RuleBook Full PDF

Rule Book Vol 1
and
Rule Book Vol 2
EMTP Rule Book Vol. 1

Table of Contents
Section 1
Introduction to the EMTP
Section 2
Computer Aspects of the EMTP
Section 3
Structure of EMTP Input Data
Section 4
Cards to Begin A New Data Case
Section 5
Lumped Linear Impedance Models
Section 6
Frequency-Dependent Transformer Model
Section 7
Distributed Parameter Transmission Line Models
Section 8
Nonlinear Elements
Section 9
Switches
Section 10
Sources
Section 11
User-Supplied Initial Conditions
Section 12
Output-Variable Specification Cards for the Time-Step Loop
Section 13
User-Supplied Source Functions (Type 1-10 Sources)
Section 14
"TACS"Transient Analysis of Control Systems
Section 15
Output Processing
Section 16
Description of Program Output
Section 17
Informative and Error Messages
Section 18
EMTP Data Modules
Section 19
High Voltage DC
Section 20
Static Var Compensator
Section 21
LDFL: Three-Phase Load Flow
Section 22
Network Reduction
Section 23
Protection Systems
Section 24
Air Gap Arc Model
Section 25
References and Bibliography
Section 1
Introduction
1.1
Program Development - Historical Notes ............................................... 1-1
1.2
Applications Of The EMTP ...................................................................... 1-2
1.3
Program Capabilities - Overview ........................................................... 1-3
1.4
Basic Elements Represented In The Program ......................................... 1-4
1.5
Program Input And Output ..................................................................... 1-8
1.6
Support Programs ................................................................................... 1-9
1.7
Supported Hardware Platforms ............................................................... 1-10
1.8
User Support And Maintenance .............................................................. 1-10
1.9
EMTP Documentation .............................................................................. 1-11
Section 1
Introduction
1.1 Program Development - Historical Notes
The Electromagnetic Transients Program (EMTP) is a computer program for simulating
electromagnetic, electromechanical, and control system transients on multiphase electric power
systems. It was first developed as a digital computer counterpart to the analog Transient Network
Analyzer (TNA). Many other capabilities have been added to the EMTP over the years and it has
become the de facto standard in the utility industry.
The EMTP was developed in the late 1960's by Dr. Hermann Dommel, who brought the program
to Bonneville Power Administration (BPA). When Professor Dommel left BPA for the University
of British Columbia in 1973, two versions of the program started to take shape: the relatively
small UBC version, used primarily for model development, and the BPA version, which expanded
to address the needs of utility engineers. The BPA version of the EMTP grew as a result of the
cooperative development effort of Dr. Scott Meyer and Dr. Tsu-huei Liu from BPA, as well as a
number of other contributors from North American power companies and universities. In order to
rationalize the development of the program and to attract funding from other utilities, the EMTP
Development Coordination Group (DCG) was founded in 1982. Original members of the DCG
included BPA, the US Bureau of reclamation, Western Area Power Administration (WAPA), the
Canadian Electrical Association (CEA), Ontario Hydro, and Hydro Quebec.
Since the inception of DCG, a number of changes have taken place in the EMTP community. In
1986, Dr. Scott Meyer left DCG (due to what at the time was described as philosophical and
political differences) to develop, and to aggressively advocate an independent version of the EMTP
which he called the ATP (Alternative Transients Program). In 1989, UBC further developed and
marketed the original version of the EMTP and concentrated on PC platforms under the trade name
MicroTran. In the mid 80's Manitoba HVDC Research Centre developed a version of the EMTP
(EMTDC) targeted primarily for the simulation of HVDC systems.
As these developments took place, DCG continued to fund EMTP research and program
development efforts. At the same time, the membership of DCG increased. At present, North
American members of DCG include WAPA, the US Bureau of Reclamation, American Electric
Power Service Corporation, Electrical Power Research Institute (EPRI), ASEA Brown Boveri ltd.,
, Canadian Electricity Association (CEA), Ontario Hydro, Hydro Quebec (BPA officially left DCG
in 1990). DCG members outside North America include CRIEPI ( Central Research Institute of
Electric Power Industry) from Japan, Eletricit de France, CEPEL (Centro de Pesquisas de Ebergia
Electrica) from Brazil, AND NEG (Nordic EMTP Group) representing Imatran Voima Oy of
Finland, Sydkdraft AB and Vattenfal AB of Sweden.
1-1
Introduction
These efforts resulted in the release of version 3 of the DCG version of the EMTP in 1996 (EMTP96).
EMTP96 represents the last version of the EMTP based on the original BPA code. This program will
be superseded by the results of a complete re-structuring of the EMTP code presently under
development by DCG, and scheduled to be released before the turn of the century. This thirdgeneration version of the EMTP will include all the functionality of EMTP96, but will also include
advanced features such as variable time step, plug-in solution modules, dynamic memory
allocation, and more.
Until the re-structured EMTP becomes a production-grade product, EMTP96 will probably be the
most comprehensive and advanced program of its kind, and it will be maintained, updated and
supported by Ontario Hydro on behalf of DCG (see Section 1.8).
1.2 Applications of The EMTP

Studies involving the use of the EMTP can be put into two general categories. One is design, which
includes insulation coordination, equipment ratings, protective device specification, control system
design, power quality assessment, harmonic studies, etc. The other is solving operating problems
such as unexplained outages or equipment failures. A partial list of typical EMTP studies follows:
Switching Surges
Deterministic
Probabilistic
Single-Pole Switching
High-Speed Reclosing
Capacitor Switching
Transient Recovery Voltages
Cable Switching Transients and sheath protection
Lightning Surges
Backflash
Induced Surges
Incoming Surges at Stations
Insulation Coordination
Overhead Lines
Outdoor Stations
Gas-Insulated Substations
Arrester Duty
Shaft Torsional Stress

Subsynchronous Resonance
Switching-Induced
1-2
Introduction
High Voltage DC (HVDC)
Controls
Electrical Transients
Harmonics
Static VAR Compensation

Controls
Overvoltages
Harmonics
Carrier Frequency Propagation

Harmonics
Ferroresonance
Series and Shunt Resonance
Motor Starting
Out-of-Phase Synchronization
Islanding or Other Disturbance Events
General Control Systems
Grounding
Asymmetrical Fault Current Evaluation
Phase Conductor Transposition
Ground Wire Losses
General Steady-State Analysis of Unbalanced Systems
Capacitor Bank Switching
Series Capacitor Protection
This is only a partial list. One of the EMTP's major advantages is its flexibility in modelling; an
experienced user can apply the program to a wide variety of studies.
1.3 Program Capabilities - Overview

The EMTP is used to solve the ordinary differential and/or algebraic equations associated with an
"arbitrary" interconnection of different electrical (power system) and control system components.
The implicit trapezoidal rule of integration is used in the discretization of the equations of most
elements which are modelled by ordinary differential equations. The result is a set of real,
simultaneous, algebraic equations which is solved at each time-step using advanced sparsity
techniques. These equations are written in nodal-admittance form (with new unknown voltages as
variables), and are solved by ordered triangular factorization. Numerical oscillations inherent to
the trapezoidal rule of integration are eliminated using a procedure called CDA (Critical Damping
Adjustment). CDA makes the simulation of power electronics devices simple and straightforward.
Initial conditions for differential equations of the various components can be determined
automatically by the program for most cases of practical interest. The calculation of initial
1-3
Introduction
conditions is normally limited to linear elements. Nonlinear resistances are always ignored during
the steady-state solution. Nonlinear reactances can either be linearized during steady state or fully
modelled to include harmonic distortion effects.
Injections of the electric network may also be specified in terms of power and voltage magnitude,
thereby providing multi-phase load flow capability.
Control system modelling (TACS or Transient Analysis of Control Systems) allows for the
superposition of an arbitrary number of linear phasor solutions of different frequencies. TACS is
also used to simulate HVDC systems, SVCs, etc. A comprehensive library of such devices,
including relay and CVT models, is included as part of EMTP96.
Large sub-networks can be reduced into compact multi-port equivalents which are valid over a
broad frequency range with the FDNE (Frequency Dependent Network Equivalent) support
program.
Large coupled RLC networks, such as the internal transformer representation used by transformer
manufacturers, can be manipulated internally without additional approximations or assumptions.
The measured response of a power transformer can be used to create frequency dependent
transformer models using the HFT (High Frequency Transformer) model.
Support programs provide additional capabilities such as the calculation of overhead line and cable
parameters, as well as the generation of more complex linear and nonlinear models for use in EMTP
simulations.
Program output consists of component variables (e.g., branch currents or voltages, machine
torques or speeds, etc.) as functions of time. Both printed and plotted output are available. Printed
plots are simple character-based plots included in the standard printed output. Built-in plotted
output is supported directly for Postscript printers. Otherwise, the EMTP has the support program
EMTPOUT that produces screen and hardcopy plots of simulation results.
1.4 Basic Elements Represented In The Program

The user defines a system to be simulated by building up component models. The types of
components that can be modelled include:
1.
Lumped resistance:
v = Ri
The specific EMTP component used to represent single-phase lumped resistances is the
series RLC branch of Section 5.1.
1-4
Introduction
2.
Lumped inductance:
di
v = L ---dt
The specific EMTP component for modelling single-phase lumped inductances is the
series RLC branch of Section 5.1.
3.
Lumped capacitance:
dv
i = C ----dt
The specific EMTP component used to represent single-phase lumped inductances is

the series RLC branch of Section 5.1.
4.
Multiphase pi-equivalents of transmission lines, where the preceding scalar R, L,

C become symmetric square matrices [R], [L], [C].
The specific EMTP component used to represent multiphase, coupled RLC circuits is
the pi-equivalent of Section 5.2. A cascaded-pi input option is available for
untransposed transmission lines, but it can be used for steady-state solution only. See
Section 5.6. There is also a special component for modelling multiphase, coupled RL
circuit which is used for power frequency transformer modelling. See the mutually
coupled RL element of Section 5.3
5.
Frequency-dependent network equivalents for approximate modelling of large

networks. The specific EMTP component used to represent frequency-dependent
network equivalents can be found in Section 5.5.
6.
Multiphase distributed-parameter transmission lines, wherein propagation time

of the line is represented. Distortionless and externally-lumped-resistance
approximations are available, as well as "exact" frequency-dependent
representations.
The specific EMTP components used to represent frequency-independent, distributedparameter transmission lines are:
(A)
Single-circuit transposed line model of Section 7.3.1.
(B)
Double-circuit transposed line model with zero-sequence coupling of

Section 7.3.4.
(C)
Untransposed line model of Section 7.3.1.
1-5
Introduction
Frequency-dependent transmission lines can be modeled with:
(A)
J. Marti line model of Section 7.4.
(B)
L. Marti cable model of Section 7.5.
(C)
Corona model of Section 7.6.
7.
High Frequency Transformer model (HFT). With the HFT model it is possible to
simulate the measured (or calculated) frequency response of a single or
multiphase transformer as a function of frequency (see Section 6).
8.
Nonlinear resistors, the v-i characteristics of which must be single-valued.

Specific EMTP components used to model nonlinear resistors are:
(A)
Type-92 true nonlinear element of Section 8.8.
(B)
Type-99 pseudo-nonlinear element of Section 8.3.
There is also a specialized EMTP component for modelling Zn0 arresters, the Type-92
resistor of Section 8.9. The Type92 nonlinear resistor can also be used to represent
multi-phase SiC arresters (see Section 8.11).
9.
Nonlinear inductors, either with the conventional single-valued characteristics

(see sketch) or including hysteresis. The specific EMTP components used to
represent single-valued (no hysteresis) nonlinear inductors are:
(A)
Type-93 true nonlinear inductance of Section 8.10.
(B)
Type-98 pseudo-nonlinear inductance of Section 8.4.
Hysteresis effect can be modeled with the Type-96 pseudo-nonlinear hysteretic reactor
of Section 8.5, or with the Type-92 hysteretic reactor model of Section 8.11.
10.
11.
Time-varying resistance. The specific EMTP components used to simulate timevarying resistances are:
(A)
Type-91 true nonlinear resistance of Section 8.7.
(B)
Type-97 staircase time-varying resistance of Section 8.2.
Switches, used to simulate circuit breakers, lightning-arrester flashover, or any

other network connection change. Diodes and dc converter valves are included.
The specific EMTP switch elements are:
(A)
Ordinary switches of Section 9.2.1 and Section 9.2.2.
1-6
Introduction
12.
(B)
Switches with randomly or systematically varied closing/opening times

of Section 9.2.3 and Section 9.2.4.
(C)
Permanently closed switches of Section 9.2.5.
(D)
TACS controlled switches (valves, diodes, etc.) of Section 9.5.
(E)
Sophisticated circuit breaker model of Section 8.10.
Voltage or current sources. In addition to standard mathematical functions

(sinusoidal, surge functions, steps, and ramps), the user may specify sources
point by point as functions of time, or in FORTRAN, or as defined by TACS (see
Point 14 below). The specific EMTP components for modelling different types
of sources are described in Section 10. The built-in modelling capabilities
include:
(A)
Type-11 step function of Section 10.3.
(B)
Type-12 and Type-13 ramp functions of Section 10.3.
(C)
Type-14 sinusoidal sources of Section 10.3.
(D)
Type-15 simplified surge function of Section 10.5.
(E)
Type-16 simplified HVDC converter model of Section 10.6.
(F)
Type-17 TACS-modulated source of Section 10.8.
(G)
Type-18 ideal transformer source of Section 10.9.
(H)
Type-60 TACS-defined source of Section 10.10.
There is also the capability to define a source with arbitrary variation as a function of
time. This capability is available with the Type 1-10 user-defined sources of
Section 10.7.
13.
Dynamic synchronous machines (3-phase balanced design only). The electrical

side is represented by Park's (Blondel's) equations, while the mechanical side is
modelled as an interconnection of masses, springs, and damping. Arbitrary
exciter and governor dynamics can be represented, by connection to TACS (see
Point 14 below). The specific EMTP component used to present the dynamic
synchronous machines is the Type59 S.M. model of Section 10.11.
14.
Unconventional rotating electromechanical energy converters of various sorts,

including induction machines and dc machines. An arbitrary number of
windings on each rotor axis is allowed, there is no restriction to 3-phase usage,
1-7
Introduction
etc. Compensation is used, and any mass-spring dynamics of the shaft are
represented by an electrical analog. Various control system connections are
possible, via TACS. The specific EMTP component used to model practically any
type of rotating machinery is the Type19 U.M. model of Section 10.12.
15.
Control system dynamics, as are normally represented on differential analyzers

(analog computers). This modelling capability goes by the name of TACS.
Nonlinear and logical operations may be represented. Input and output may be
interfaced with the electric network of the EMTP, providing a hybrid
representation. All TACS representation is user-patchable, and hence
configuration free. All control system components are described in Section 14.
The available modelling capabilities include the following:

(A)
Transfer functions in "s" (of any order) as per Section 14.3.1.
(B)
Limiters (both static and dynamic) of Section 14.3.1.
(C)
Signal sources of Section 14.2.
(D)
Supplemental variables and pseudo-FORTRAN of Section 14.5.
(E)
Special application devices such as frequency meter, relay, transport

delay, etc., of Section 14.4.
1.5 Program Input and Output

Basic data entry to the program is based on the "card image" paradigm; that is, a flat ASCII file is
created and modified with a standard editor, and column-sensitive data is entered according to the
rules described in this manual. A multi-platform graphical user environment EMTPView is
presently under development at Ontario Hydro, whereby data entry will be done using a circuit
schematic paradigm, and free-format data entry forms. EMTPView is scheduled for release in the
summer of 1997.
The input file is read by the EMTP, and an output and/or plot files are generated as output. The plot
file with the default extension "pl4" contains binary information that a support program such as
EMTPOUT can read. The output file is a plain text file reporting various results of the simulation.
The input file contains the calculation time step, length of time to be simulated, and output
requests, as well as the model data. The lumped branches are defined by resistance in ohms,
inductance in mH or in ohms at power frequency, and capacitance in F or in S (micro-Siemens)
at power frequency. The simplest travelling-wave models can be defined by surge impedances,
resistance per unit length, wave velocity, and line length for positive and zero sequence. More
complex line models require data that can only be generated with the help of support programs
such as AUX. Nonlinear elements are usually specified by current-and-voltage points for resistors,
1-8
Introduction
and current-and-flux-linkage points for inductors. Synchronous machine models use conventional
stability data for the electrical side, and parameters in English units for the mechanical side. TACS
input is specified by transferring information from the control block diagram to card images.
Most of the EMTP's input data requirements are different from, and more extensive than, other
programs such as load flow, short-circuit, and stability. This is because the program is
multiphase, it can simulate nonlinear elements, and generally uses more detailed models than the
other programs. These features are needed to accurately simulate high frequency transients which
occur during short time periods. Fortunately, there are auxiliary programs supplied with the EMTP
which assist the user in setting up the input data for transmission lines, cables, transformers, surge
arresters, and nonlinear inductors.
The primary output from a transient simulation includes plotted bus voltages, branch voltages,
branch currents, branch energy dissipation, machine variables, and control system variables.
These values can also be printed out as functions of time, but this type of output is often awkward
to use. Printed maximum values of the variables and the times at which they occurred are also
available.
A steady-state phasor solution is performed before the transient simulation to determine the initial
conditions, and this can also be a useful study tool in itself. Branch voltages and currents, bus
voltages, power loss, and power flows are determined for the entire network. A frequency scan
option is also available which systematically varies the frequency of the sources for the steadystate solution, and plots voltage magnitudes and angles as a function of frequency. This type of
output is useful for harmonic and resonance studies.
1.6 Support Programs

Under DCG development, there has been a tendency to stay away from the super-program concept,
where every feature is part of a single master program. EMTP96 now consists of the main
computational engine "EMTP" and a number of support and auxiliary programs. These programs
are:
Program
AUX
Purpose
Calculation of line and cable parameters, and EMTP
model generation.
Generation of power frequency transformer models,
such as TRELEG, BCTRAN, and TOPMAG.
Generation of data for hysteresis and eddy current
models for power transformers.
1-9
Introduction
Program
Purpose
Calculation of instantaneous flux-current
characteristics from RMS measurements.
FDNE
Generation of multi-port frequency dependent network

equivalents.
FDBFIT
Generation of frequency dependent transformer

models, primarily from measured data
3FLOW
Unbalanced three-phase load flow program.
EMTPOUT
Output processing program.
1.7 Supported Hardware Platforms

EMTP96 supports a large variety of hardware platforms and operating systems, such as
Intel-based PCs under Windows 3.1, Windows 95, Windows NT

DEC workstations under VMS
DEC ALPHA under DEC UNIX
IBM RS6000/AIX
HP9000/UX
Sun/Solaris
With the exception of VAX VMS (which lacks graphical support), all platforms have approximately
the same functionality and graphical output support. It is generally not necessary for the user to be
aware of any programming idiosyncrasies due to differences in platforms and/or operating
systems.
1.8 User Support and Maintenance

In 1996, Ontario Hydro assumed the responsibility for User Support and Maintenance of version 3
of the EMTP. This provides a centralized user support centre, in addition to program maintenance,
upgrade and development facilities. The User Support Centre also provides courses, seminars and
workshops in many aspects of EMTP usage and application.
1 - 10
Introduction
1.9 EMTP Documentation

The Rule Book is a reference document containing rules for EMTP usage. The name "Rule Book"
was chosen instead of the usual "User's Manual" because it has a minimum of user guidance and
application examples.
The Rule Book is one of many EMTP documents published, or to be published, by DCG. The
others are:
1.
EMTP Workbooks -- The workbooks are the basic introductory texts used in
most EMTP courses. They contain examples describing the usage of the basic
models available in the EMTP.
2.
EMTP Application Guide -- the Application Guide contains examples of various
kinds of applications, and how to choose the most suitable models for different
types of simulations. The reader is assumed to have some familiarity with the
EMTP.
3.
EMTP Model Verification -- this document describes cases in which the EMTP
has successfully simulated laboratory and field tests.
4.
The EMTP Theory Book, 2nd edition, by Prof. Hermann Dommel. This is the
most comprehensive compilation of the theoretical infrastructure of the EMTP.
Is is available directly from Prof. Dommel at the University of British Columbia.
1 - 11
Section 2
2.1
Availability on Different Computer Systems ......................................... 2-1
2.2
Installation-Dependent Considerations ................................................... 2-1
2.3
Input/Output Logical Units ..................................................................... 2-1
2.4
Miscellaneous Default Settings .............................................................. 2-2

2.4.1 Power System (Synchronous or Steady-State) Frequency ......... 2-2
2.4.2 Calendar Date Format ................................................................. 2-3
2.4.3 Free-Field Data Format Characters ............................................. 2-3
2.5
Re-Dimensioning The EMTP .................................................................. 2-3

2.5.1
List Sizes ..................................................................................... 2-3
2.5.2 Re-Dimensioning Procedure ....................................................... 2-7

2.5.3 Typical List Sizes ........................................................................ 2-8
Section 2
2.1 Availability on Different Computer Systems
The computer industry is continuosly changing. Operating systems and hardware platforms do
not last very long in the market. EMTP96 is supported on all the platforms used by the members of
DCG, which probably reflects a substantial cross-section of the hardware/software market. The
platforms supported by DCG in 1998 are:
IBM RS6000/AIX
HP9000/UX
Sun/Solaris
2.2 Installation-Dependent Considerations

DCG has put considerable effort into providing a uniform interface between the EMTP and the user
on all supported platforms. However, a number of aspects are inevitably dependent on the
hardware and operating system. For example, the installation procedure, how to execute a
program, assigning input/output files, plotting, etc. All these system-dependent aspects are
covered in the installation guide that accompanies the package.
2.3 Input/Output Logical Units

This information is only relevant in operating systems such as VAX/VMS, where the assignment of
input, output, and plot files is done through assignments to specific logical units. Note that in
2-1

order to present a common interface between all platforms, the logical unit assignment in the
distribution version of VAX/VMS is done through a command file.
Logical unit 5
Input data file for EMTP input data cards
Logical unit 6
EMTP printed output
Logical unit 7
EMTP punched-card output
Logical unit 4
Storage of binary data points for plotting (i.e., the "pl4"

file)
The following logical unit assignments are internal to the program. If these are explicitly assigned
by the user in a VMS environment, data loss may occur. In other operating systems, logical unit
assignments should only be apparent in the case of a system crash or some other unusual form of
execution termination.
Logical unit 1
scratch file
Logical unit 2
scratch file to store memory contents for each new

energization of a "STATISTICS" or "SYSTEMATIC" data
case.
Logical unit 3
scratch file
Logical units 8-15
scratch files
2.4 Miscellaneous Default Settings

2.4.1
Power System (Synchronous or Steady-State) Frequency
The default steady-state frequency of power system operation is assumed to be 60 Hz unless

otherwise specified with the "POWER FREQUENCY" miscellaneous data card (Section 4.1.8).
2-2
2.4.2
Calendar Date Format
The date format convention is the one used most often in the United States; that is, "MM/DD/YY",
where:
"MM"
"DD"
"YY"
two decimal digits for the month

(e.g., "03" for March);
two decimal digits for the day within
the month;
two decimal digits for the year
(e.g., "95" for 1995).
At present, the user has no control over date format.
2.4.3
Free-Field Data Format Characters
As explained in Section 4.1.8, two special characters are used in conjunction with EMTP free-field
data input. A comma is the default separator character "CSEPAR" and a dollar sign is the default
continuation character "CHCONT".
2.5 Re-Dimensioning The EMTP

The EMTP stores most data in tables or arrays, which can be sized by the user by recompiling
module "newmods" and then re-linkning it to the rest of the EMTP libraries. This means that only
users with the appropriate compiler can re-dimension the EMTP. Since re-dimensioning is
installation-dependent, specific instructions are given in the corresponding installation guide.
2.5.1
List Sizes
The contents of the list sizes that can be controlled by the user are:
List 1: LBUS
Maximum number of nodes.
List 2: LBRNCH
Maximum number of branches (e.g. a 3-phase overhead

line counts as three branches).
List 3: LDATA
Maximum number of R, L, C storage cells. This parameter

is used by devices such as pi-circuits and series R-L-C
branches).
2-3

List 4: LEXCT
Maximum number of sources. Each dynamic

Synchronous Machine (see List 17) contributes 3.
List 5 : LYMAT
Maximum number of storage cells in the admittance

matrix used in the time-step loop solution. Storage is for
the upper-triangle (including diagonal) only.
List 6: LSWTCH
The maximum number of switches. Diodes and valves

also count as switches.
List 7 : LSIZE7
Total number of characters allocated for node names.
List 8: LPAST
Maximum number of modal past-history points, for

distributed-parameter transmission lines.
List 9 : LNONL
Maximum number of nonlinear and pseudo-nonlinear

elements.
List 10: LCHAR
Maximum number of points which define the

characteristics of the nonlinear of pseudo-nonlinear
elements of List 9.
Note: Type-96 hysteretic inductors (Section 8.5), SiC
arresters (Section 8.10) and circuit breakers (Section 8.12)
are an exception.
For the first Type-96 element, 2N+8 cells are allocated,
where N is the number of data cards that define the
characteristic. Second and later elements might use the
reference-branch feature, in which case requirements drop
to six cells.
For each Type-92 SiC arrester, two additional cells are
required to store the information about the gap model.
For each dynamic circuit breaker exactly 6 cells are used.
List 11: LSMOUT
Maximum number of Type-59 synchronous machine

(S.M.) output variables.
List 12: LSIZ12
Maximum number of output quantities.
List 13: LFDEP
Reserved; set to unity.
List 14: LWT
Average number of branches per mode of the frequency

dependent network equivalents. The total size of the
floating point arrays CIKFD and RMFD is determined as a
product of List 14 and List 27. The built-in default value
is 18.
List 15: LTAILS
Reserved; set to unity.
2-4

List 16: LIMASS
Maximum total number of masses of all Type-59 S.M.

usage (total over all machines).
List 17: LSYN
Maximum number of Type-59 synchronous machines

(S.M. components) of Section 10.11.
List 18: MAXPE
Maximum number of branch or switch power and energy

requests (branch cards with column 80 having a value of
"4").
List 19: LTACST
Maximum number of floating-point cells of total storage

for all of the TACS tables. See Section 14.
List 20: LFSEM
Storage for frequency-dependent lines which rely upon

recursive convolution. For JMARTI line models
(Section 7.4), any branch requires: 5 + 7 times the number
of poles used to represent the two functions Zc and A.
For N-phase line, with NBT RL blocks in all modes and
NSHN corona branches, the CORONA model of Section 7.6
requires Q8 + 8N + 5NR + 5NBT + NSHN (3N + NBT)
storage cells.
List 21: LFD
Maximum number of cells used to store transformation

matrices for the constant-parameter and frequencydependent distributed transmission line models. For each
non-copied line section of "N" phases, 2xNxN cells are
required. Excluded from this count is the LMARTI cable
model of Section 7.5.
List 22: LHIST
For each JMARTI lines, 15 cells are required for each

coupled phase.
List 23: LSIZ23
Ignore for all computer systems which use overlaying

(such as IBM VM) -- unless the EMTP load flow ("FIX
SOURCE") is used. But for fully-virtual systems not short
of address space, and for computer systems where
COMMON blocks are not in order, this list is required. It
specifies the size in floating-point words of three giant
vectors which are used for node renumbering and the
phasor steady-state solution. Default is 4000.
List 24: NCOMP
Maximum number of phases for compensation-based

models, at peak problem size. The actual maximum
number of phases then varies inversely with the number of
nodes. Use of NCOMP=3 is most common, allowing for 3phase compensation with full-size problems, 6-phase
compensation with half-size problems, etc.
2-5

List 25: LSPCUM
Maximum number of floating-point cells of total storage

for all Type-19 universal machine (U.M.) tables. If U.M.
modeling of Section 10.12 is not of interest, set to unity.
Practical production usage allowing two 3-phase
induction machines typically takes LSPCUM = 1500.
List 26: LSIZ26
Storage for working vectors used for various tasks in the

EMTP.
List 27: LSIZ27
The number of frequency dependent network equivalents

times two.
The following lists pertain to LMARTI or FDQ frequency-dependent cable model (Section 7.5), as
per the following definitions:
List 28: LFDUNT
Maximum number of cable units (with or without

frequency-dependent ) transformation matrix [Q]).
List 29: LFDUCQ
Number of frequency dependent line/cable units which

will have a constant transformation matrix [Q]. Note that
LFDUCQ must be LFDUNT. If LFDUCQ is zero or
negative, then all frequency-dependent branches will be
assumed to have a frequency-dependent transformation
matrix.
List 30: LFDPH
Maximum number of phases allowed in a given cable unit

LFDPH must be 3.
List 31: LFDPOL
Maximum order allowed on any rational-functions

approximation (LFDPOL 20).
List 32: LFDPAV
Average order of all rational-functions approximations

(LFDPAV < LFDPOL).
List 33: LFDHAV
Average number of history terms for a given mode

(branch). This number is (roughly) the average of Tau/
Deltat for all modes.
List 34: LFDPHA
Average number of phases per unit.
The following lists pertain to the FDB model used in for the High Frequency Transformer model,
(Section 6), as per the following definitions:
List 35: LTUAV
Average number of FDB branches
List 36: LTPAV
Average number of terms in the approximation of a FDB

unit
2-6

List 37: LTUN
Total number of FDB units. Can be larger or smaller than

LTUAV.
The last list item is a global address offset.

List 38:
Addressing offset
Please note the following: An abbreviated version of these explanations appears as part
of the case-summary statistics at the end of the output of each data case being solved. See the
example of Section 17.4.
Certain primary-level non-solution overlays have giant working arrays (a maximum of one per
overlay) which are sized the same as common block /LABEL/ except for a possible built-in offset
which very crudely adjusts for the amount of code of the overlay. If the user wants to manually
apply an additional offset to this storage, he can add a fifth data card as follows:
1.
Punch an extra "9" in column one of the first data card that is read by VARDIM
(see Section 2.5.2). This is taken as a special flag (VARDIM extracts it from
LBUS before LBUS is used).
2.
Add an additional data card before the last card containing the extra desired
offsets (list 38). The format is 10I8, with one data field for each such primarylevel non-solution overlay. Currently there are two: OVER29 and OVER31.
Common usage of this feature is for virtual machines, where large list sizes are requested; unless
specially compensated for, the giant working arrays would be dimensioned far beyond any
reasonable program needs. In this case, offsets can be negative. Using the negative offsets saves
on virtual address space.
2.5.2
Re-Dimensioning Procedure
Conceptually (details may be different for each platform), the steps to re-dimesion the EMTP are as
follows:
1.
Edit the file containing the list sizes described in Section 2.5.1
2.
Run the variable dimensioning support utility, generally called VARDIM. This
utility uses the listsizes file as input, and as output it generates the FORTRAN
source module newmods.f .
3.
compile newmods.f and re-link newmods.o to the previded EMTP libraries (in
some systems it will be newmods.for and newmods.obj).
2-7

To obtain a set of list sizes which is an integer multiple of all the default sizes, punch a "9" in the
first column of the second data card and an integer N ending in column 8 of the same card. For
example, to size the table 4 times as large as the default size, punch, on the second data card, a "9"
in column 1 and a "4" in column 8.
Should the variable-dimensioning program detect illegal or absurd data as having been supplied by
the user in his attempt to redimension, it will terminate execution with an error message.
2.5.3
Typical List Sizes
A typical listsize.dat file in shown below. The "*" character in column 1 is used to indicated a
comment line.
*
* General lists:
*
*-- 1--><-- 2--><-- 3--><-- 4--><-- 5--><-- 6--><-- 7--><-- 8--><-- 9--><--10-->
90000754
2000
6000
300
10000
120
4500
10000
225
480
*--11--><--12--><--13--><--14--><--15--><--16--><--17--><--18--><--19--><--20-->
150
150
15
2500
150
30
4
15
9000
40000
*--21--><--22--><--23--><--24--><--25--><--26--><--27->
3000
3000
20000
24
1200
8101
* FDQ & FDB models lists:
*--28--><--29--><--30--><--31--><--32--><--33--><--34-->
20
18
36
50
35
500
6
* Offsets
*--35-->
70000
2-8
Section 3
Structure of EMTP Data Input
3.1
Structure of a Data Deck ........................................................................ 3-1
3.2
Preliminary Considerations In Data Preparation ................................... 3-3
3.3
Description of Electric Networks .......................................................... 3-3
3.4
Description of Control Systems ............................................................. 3-11
3.5
Cards to Begin a New Data Case ........................................................... 3-12
3.6
$-Cards ................................................................................................... 3-15
3.7
Fortran Data Format Specifications ....................................................... 3-17
Section 3
3.1 Structure Of A Data Deck
The structure of an EMTP input data file is based on a "data deck" paradigm. In the early days of
the EMTP, a data case was built by stacking punched cards in data decks, and data decks would be
separated with (sometimes coloured) blank cards. This manual (as a concession to tradition) will
refer to cards, card images, data decks, and other perhaps outdated terms, since they are still
useful to describe the structure of an EMTP data file.
Any number of cases can be stacked in an EMTP input data file:
blank card
special end-of-job termination
card, if required by system
blank card for termination
of cases
data deck for nth case
data deck for 1st case
Each EMTP simulation is described in a data deck which consists of the following parts (in this
order):
1.
Cards to begin a new data case. Included are a "BEGIN NEW DATA CASE" card,
special request cards of various types (if any), and finally the EMTP
miscellaneous data cards.
2.
Specially-requested extensions to miscellaneous data cards (optional).
3.
Cards which define any TACS (Transient Analysis of Control Systems)

representation (optional).
4.
Cards for linear and nonlinear branches, transformers, and transmission lines (or
cables), terminated by a blank card.
3-1

5.
Cards for electric-network switches, diodes, AC/DC converter valves, switchedresistance elements, and switched-inductance elements, terminated by a blank
card.
6.
Source cards for electric network (voltage sources, current sources, dynamic
synchronous machines). These are terminated by a blank card.
7.
Load flow cards for "FIX SOURCE", "FIX SOURCE1", or "LOAD FLOW" usage
(optional).
8.
Cards for overriding the internally-calculated initial conditions (optional).
9.
Node-voltage output specification cards. These are terminated by a blank card if

the specification is selective. But if all node voltages are requested by means of a
"1"-punch in column number 2, then there is to be no blank terminator card.
10.
Cards for specifying Type 1-10 EMTP source functions point by point. There is
one card for each time-step, terminated by a "9999" card (optional).
11.
Batch-mode plotting cards, either Postscript output or simple line printer plotting
(optional).
12.
Blank card.
This data structure is for a single deterministic transient simulation, which is the most common
case (see next paragraph for exceptions). Often data classes 2, 3, 7, 8, 10, and/or 11 are not
required, and can simply be omitted. Please note that the terminating blank cards for classes 4, 5,
6, 9 and 12 must be present even if elements of the data class are not present (or needed) in a given
simulation.
There are valid EMTP data cases which do not have the above structure, such as statistical studies,
and "REPLOT" (used for the batch-mode plotting of a previously-solved data case).
Blank cards are used extensively as a terminator card, for various classes of data. Any blank card
can be replaced by a card with the characters "BLANK " in the first 6 columns and any comments in
the remaining columns. For example, "BLANK CARD ENDING BRANCH CARDS". The EMTP
recognizes all such cards and blanks them out internally before processing them.
Comment cards may be added to the data deck at any point. Comment cards are identified by
characters "C " (Upper case C + space) in columns 1-2. Comment cards will appear as part of the
input data listing, but will be otherwise disregarded by the EMTP. Different levels of comments can
also be specified by using "Ci" where i is an integer; this is useful to control the amount of printout.
Blocks of data can be "commented out" by using $DISABLE and $ENABLE cards (see Section 3.6).
3-2
3.2 Preliminary Considerations In Data Preparation

Before entering data for the various elements of the simulation, the following simulation
parameters must be defined:
1.
Selection of the step width t for computing electromagnetic transients at

discrete points in time t, 2t, 3t, .... TMAX.
2.
Maximum time TMAX to which study is to be carried out. Computer costs are
approximately proportional to maximum time. Therefore, if uncertain about a
case, commence with a shorter maximum time.
3.
Units for system parameters. Resistances are always in ohms. For inductances
and capacitances, two options are available:
(A)
Inductances are either specified in mH, or as inductive reactances L in .

Mixed options are not permitted; that is, all branch data must either be L
in mH or L in at the same frequency f=XOPT (see Section 4.2.1).
(B)
Capacitances are either specified in F, or as susceptances C in S (or

mho). Mixed options are not permitted; that is, all branch data must
either be C in F or C in S at the same frequency f=COPT (see Section
4.2.1).
4.
Data scaling. It is best to use physical units per unit quantities may lead to
confusion.
5.
Network connection diagram. To minimize data errors, it is best to prepare a

network diagram showing all the relevant nodes, branches, generators and
sources prior to editing the input data file.
3.3 Description Of Electric Networks

Each non-ground node of the electric network is given a 6-character alphanumeric name. Only
identical sequences of six characters will be recognized as one and the same node. Therefore
"VOLTS " will be a different node than " VOLTS". "
" (6 blanks) is the name reserved for
ground (common reference or absolute ground). Special names which the user should avoid are
"......", "TYPE-16", "HEIGHT", "MARGIN", "SMOOTH", "TARGET", "BRANCH".
The EMTP is based on a "nodal admittance" formulation. The description of EMTP network
elements is constrained by this type of formulation. For example, node voltages are implicitly
referred to ground, and the node voltages of a "floating" delta-connected transformer winding are
3-3

undefined. The general conventions and assumptions used in an EMTP simulation are discussed
next.
Node Voltage:
Node voltage is the voltage from node to absolute ground. Its value is obtained by specifying the
node name on cards for node voltage output (Section 12). The following figure shows the voltages
at terminal A and terminal B of some device. Note that the voltages are measured with respect to
absolute ground.
terminal A
terminal B
VA
VB
Branches:
Branches are identified by their parameters and the names of the nodes to which they are
connected, or a 6-character element name (Section 3.3.1).
NODE-K
NODE-M
3-4
Switches:
Switches are identified by their parameters and the names of the nodes to which they are
connected, or a 6-character element name (Section 3.3.1).
NODE-K
NODE-M
Branch Voltage:
Branch voltage is the voltage difference across a branch. Output voltage in the EMTP consists of
an arbitrary selection by the user of these two types of voltages. To obtain a voltage difference
where no branch exists (or where such output is otherwise not permitted), a very high resistance
branch can be connected between the nodes in question, and then request the branch-voltage
output. Permissible values of R are hardware dependent. However, R = 1016 is a reasonable value
in most cases. The voltage difference output can also be obtained with the help of TACS (Section
14).
Vk-Vm
NODE-K
NODE-M
3-5
Voltage and Current Source Names:

Voltage and current sources are identified by one node name and are implicitly assumed to be
connected between the node and ground. If no ground resistance is involved, then "ground" is the
common neutral. Otherwise "ground" means local ground.
+
V
Voltage Source
Current Source
Voltage Sources in Series:

If more than one voltage source is specified at the same node, the voltages will be assumed to be in
series from ground to that node. This permits the representation of a complicated wave form as a
sum of different functions.
+
v1
vtotal = v1 + v2 + vN
+
vN
3-6
Current sources in parallel:

If more than one current source is specified at the same node, the currents will be assumed to be
connected in parallel from ground to node.
itotal = i1 + i2 + iN
Simultaneous voltage and current source at the same node:

If voltage and current sources are specified at the same node, the voltage sources override and the
current sources are ignored. Current sources do not influence the network in this case because
they are directly short-circuited through the voltage sources.
+
i
Voltage Source
3-7
Current Source
is ignored
Current source between two nodes:

A current source from node A to node B can be represented by two current sources, one out of
node A and one into node B.
node A
node A
node B
node B
represented as
Voltage source between two nodes:

If the voltage source between two nodes k and m has an internal resistance Ri, then the
corresponding Norton equivalent current source in parallel with Ri, should be used.
Ri
+
v
represented as
Ri
v
i=
Ri
If Ri = 0, then the voltage source between two nodes cannot be handled by the EMTP.
A steady-state ac voltage source (TSTART < 0) with internal inductance Li between two nodes k
and m,
v ( t ) = Vo cos ( t + o )
(3.1)
3-8

(v = vk - vm if nothing is connected to terminals k, m) can be replaced by two ac current sources in
parallel with the inductance Li, with
Vo
i = i k = i m = -------L i cos ( t + o 90 )
(3.2)
Example: Voltage source v = 100 cos (t-30o) with f = 60Hz and Li = 100mH between nodes
PLUS and MINUS can be represented with an inductance branch card for L = 100 mH between
PLUS and MINUS, and two current source cards,
at node PLUS: Amplitude = 2652.6, phase angle = -120o

at node MINUS: Amplitude = -2652.6, phase angle = -120o
Li
+
i
Li
represented as
With the case being started from an automatically-computed ac steady-state solution (which
requires TSTART < 0 on the current source cards), the initial conditions for the current iL in the
inductance are automatically correct.
If initial conditions are zero (i.e., v(t) =0 for t 0), then Laplace transform techniques can be used
to find the proper current sources. Assuming v(t) = Vo sin() for concreteness, we have
Vo
V ( s ) = ----------------2
2
s +
Vo
I ( s ) = -------L i
(3.3)
2
---------------------2
s(s + )
(3.4)
3-9

Vo
Vo
- cos t
i ( t ) = -------- -------L i Li
(3.5)
In this case, the voltage source is equivalent to one constant current source and one sinusoidal
current source at each end.
Differences Between Internal Resistance and Inductance:

Current sources with parallel resistance Ri give identical solutions as voltage sources with the same
resistance in series, and the conversion is valid for any type of source. With internal inductances,
the conversion described above is only valid for AC sources. Also, the transient solution with
current sources with parallel inductance Li may differ slightly from that with voltage sources with
the same inductance in series because the differential quotient in Li di/dt must be replaced by a
difference quotient. For example, with 60 Hz sources, the differences are relatively minor if
t = 50 s, whereas errors of about 0.3% can be obtained with t = 500 ms. AC steady-state
solutions give identical results.
6-Character Element Name for Branch and Switch Identification:

In addition to specifying node names to describe a branch, it is also possible to assign names to
branches. This feature is useful to identify output from a branch when two or more branches are
connected between a given pair of node names.
Every branch and switch is given a default name by the EMTP. If DIAGNOSTIC printout is turned
on, these names will be seen:
LIN001, LIN002, etc. for rows of the linear branch table;
NLN001, NLN002, etc. for rows of the nonlinear element table;
SWT001, SWT002, etc. for rows of the switch table.
These default names can be overriden by the user. For the naming of linear and nonlinear branches
that are not a copy of some preceding element, the following rules apply:
1.
To provide a name use BUS3 and BUS4 fields in columns 1526:

BUS3 = "NAME " ---- A request word for the naming operation
BUS4 = "??????" ---- Whatever branch name the user desires.
This is not possible for switches, however, since no such name fields exist.
2.
To copy a previously-named branch using the 6-character branch name of that

original branch (rather than the pair of terminal node names), use the two 6
character name fields of columns 15-26:
3 - 10

BUS3 = "COPY " ---- A request word for the copying operation
BUS4 = "??????" ---- Name of the branch being copied.
3.
To name a branch that is a copy of some preceding branch (so that columns 1526 are already being used), precede the branch to be named by a special, extra
card which performs this function. The extra card should contain the just-listed
BUS3 and BUS4 in columns 1526 and the following special-request text in
columns 314:
"BRANCH NAME:" ---- if following branch to be named is linear
"NONLIN NAME:" ---- if following branch to be named is nonlinear.
This also applies to switches (using "SWITCH NAME:"). Note that for switches, it is
the only way of providing a name.
3.4 Description Of Control Systems

The modelling of control system dynamics using TACS (Section 14) is separate and distinct from
the description of the electric network (resistors, capacitors, voltage sources, etc.).
Control system blocks can take various forms: Laplace transfer functions, summing junctions,
nonlinear algebraic operators (multiply, divide, square root, etc.), logical operations, etc. The
output of each control system block is given a distinct 6-character alphanumeric name, so as to
identify the signal in question. Such blocks can be defined and interconnected arbitrarily by the
user, forming a control system circuit. One such sample illustrative diagram is shown below:
IN
K1
----------------1 + T1 s
k2
-------------------------------2
A + Bs + Cs
OUT
K3
1 + T3 s
----------------1 + T2 s
Sample Control System
3 - 11

Any given 6-character alphanumeric name can be used once in TACS and also once in the electric
network. There is no confusion by the EMTP, since the TACS solution is completely separate from
the electric network solution. Such double usage is advisable in the case of interface quantities, to
remind the user of the connection.
Signal sources which serve as input can be either self-contained (e.g., a sinusoidal oscillator, or a
step function), or can be controlled by the electric network (e.g., any node voltage or any switch
current can be used as a TACS source). Likewise, any TACS variable can be passed back to the
electric network for control purposes (e.g., the status of an electric network switch can be
controlled by TACS, as can a voltage source, or the field voltage of a dynamic synchronous
machine). Variables which are passed back and forth between the electric network and TACS are
referred to as interface quantities.
Electrical
Network
Control
System
Dynamic interaction between electric network and control system
3.5 Cards To Begin A New Data Case

Each new case begins with a number of special-request and miscellaneous cards. These cards are
described briefly in the following. More information can be found in Section 4.
Simulation Control
Card
Description
BEGIN NEW DATA CASE
case-separation card
END LAST DATA CASE
optional card to end the EMTP simulations
ABORT DATA CASE
skip the current data case
CDA
critical damping adjustment
FREQUENCY SCAN
multiple-frequency steady-state solutions
INITIALIZATION WITH HARMONICS
include the effects of harmonic distortion
START AGAIN
restart a halted simulation
3 - 12

TIME STEP LOOP
transfer of control to time-step loop
ANALYTIC SOURCES USAGE
user-supplied sources
MISCELLANEOUS DATA CARDS
read miscellaneous data cards
Output
Card
Description
CHANGE PRINTOUT FREQUENCY
change frequency of solution printout
AVERAGE OUTPUT
average the values of EMTP output variables
LIMIT ON PLOT OSCILLATIONS
redefine no. of oscillations allowed before

averaging
BEGIN PEAK VALUE SEARCH
time at which extrema computation is to

begin
PEAK VOLTAGE MONITOR
request for peak node voltage
MODAL VOLTAGE OUTPUT
modal voltage output for distributed lines
KILL CODES
print context of the kill codes
DIAGNOSTIC
selective diagnostic printout
ALTERNATE DIAGNOSTIC PRINTOUT
selective printout in time-step loop
OUTPUT WIDTH 80
request for 80-column output
OUTPUT WIDTH 132
EXECUTION MONITOR
monitor selected variables while EMTP is

executing
Statistics
Card
Description
TABULATE ENERGIZATION
RESULTS
combine different runs of a statistics case
STATISTICS OUTPUT SALVAGE
save results of a statistics run
OMIT BASE CASE
skip base solution in a statistics case
TIME OF DICE ROLL
the time before which no random closing

occurs
3 - 13

user-supplied random switching times
USER SUPPLIED SWITCH TIMES
Plotting
Card
Description
REPLOT
batch-mode plot of previously-run case
USER IDENTIFICATION
user i.d. information for batch-mode plot

file
PRINTER LINES PER INCH
redefine no. of lines/inch on printer output
PLOTTER PAPER HEIGHT
redefine height of paper on batch mode plot
CUSTOM PLOT FILE
change precision of the plot file
TACS
Card
Description
ABSOLUTE TACS DIMENSIONS
allocation of storage for the 8 TACS tables
RELATIVE TACS DIMENSION
relative sizing of the 8 TACS tables
TACS EMTP SOURCES
TACS-defined, user-supplied, sources
TACS INIT
simplify initialization of TACS data cases
TACS ORDER
simplify order of TACS data cases
POSTPROCESS PLOT FILE
post-process plot file using TACS
Synchronous Machine
Card
Description
ABSOLUTE U.M. DIMENSIONS
allocation of storage for the 4 U.M. tables
RELATIVE U.M. DIMENSIONS
relative sizing of the 4 U.M. tables
Load Flow
Card
Description
FIX SOURCE
old EMTP load flow
FIX SOURCE 1
new EMTP load flow
LOAD FLOW
three-phase load flow
3 - 14
Advanced Options
Card
Description
FILE REQUEST
access to module "MIDOV1" during input
RENUMBER BYPASS
bypass transient node renumbering
FREE FORMAT
redefine free format delimiters
MODIFY SWITCH LOGIC
alternative switch logic
STEP ZERO COUPLE
subnetwork i.d. with all switches closed
REDEFINE TOLERANCE EPSILN
redefine EPSILN for cases without

miscellaneous data cards
HIGH RESISTANCE
redefine R of internally-added resistor
POWER FREQUENCY
synchronous power system frequency
ZINC OXIDE
control parameters for solution of non-linear

elements
3.6 $-cards
The dollar sign "$" is conventionally used as the continuation character CHCONT for free-format
specification (see Section 4.1.8). But when placed in column number one, and when followed by
the appropriate key word, it represents a request for a higher-level file operation. The following
are valid $ cards:
$DISABLE / $ENABLE
Ignore data block
$LISTOFF / $LISTON
Disable output for data block
$VINTAGE,M
Turn on/off high-precision input format
$INCLUDE
Include contents of an external file
$DIAGNOSTIC,M
Control diagnostic output
$STOP
Stop program execution
$WIDTH,M
Change output width
$NETRED
Data for network reduction module
3 - 15
$DISABLE/$ENABLE:
A $DISABLE card tells the EMTP to treat all following data cards as if they were comment cards,
until the subsequent appearance of a $ENABLE card. This feature is useful to temporarily remove
large blocks of data or complex components from a data case without commenting out lines one by
one.
$LISTOFF/$LISTON:
A $LISTOFF cards tells the EMTP not to print and interpret data cards which follow. This continues
until a $LISTON card is encountered. The object is to minimize the size of the output file by
omitting the listing and interpretation of blocks of data which have been tested and used before.
Unlike $ENABLE/$DISABLE, $LISTON/$LISTOFF have no effect on the actual data entered.
$VINTAGE,M:
A $VINTAGE,1 card turns on the wide data field option for a given card image. A $VINTAGE,0
turns off the wide data field option. Note that this option is not available for all branch cards.
$INCLUDE:
This is probably the most important $-card. A $INCLUDE allows the inclusion of an external file
into the current input data file. When $INCLUDE is encountered, the EMTP switches to the named
file for all input. This continues until an end-of-file is encountered, at which time reading from the
original input file is resumed. For example, $INCLUDE, TRANSF.DAT will include the contents of
file transf.dat into the current input data file.
The $INCLUDE card is also used in combination of the EMTP DATA MODULES (EDM) feature (see
Section 18). EDM is a powerful feature that allows the creation of module templates, that can
simplify and modularize the structure of an EMTP data file.
$DIAGNOSTIC,M:
A $DIAGNOSTIC,M card, is used to control the amount of diagnostics printout locally, where "M" is
the integer printout control variable IPRSUP (see Section 4.2.2, integer miscellaneous data card).
Within the portion of the program currently being executed, this allows redefinition of the
diagnostics level. When the current subroutine is left, however, the $-card definition is lost. Freeformat is used for the ",M" part, so embedded blanks are permitted here. Printout can be turned off
later in the overlay using $DIAGNOSTIC,0 (remember that level zero implies none). Use of this
$-card does not replace the "DIAGNOSTIC" special-request card of Section 4.1.2 (for diagnostic
control overlay by overlay).
$STOP:
Use this card to terminate execution immediately. This feature is useful when trying to isolate an
input data error within a large and complex simulation.
3 - 16
$WIDTH,M:
Use the $WIDTH,80 and $WIDTH,132 cards to switch between 80 and 132-column EMTP output
widths for different parts of the program output. This will override any earlier "OUTPUT WIDTH
80" or "OUTPUT WIDTH 132" card (see Section 4.1.2).
$NETRED:
The $NETRED request gives access to the network reduction feature described in Section 22. The
network reduction feature allows the simulation of a large RLC coupled network where only a few
user-defined nodes and branches are available to the rest of the electrical network. This is based
on classical Kron reduction techniques, therefore, there are no approximations involved.
3.7 Fortran Data Format Specifications

Throughout this manual, data formats are indicated in fixed-format FORTRAN notation, with
symbols like F10.3, I8, A6, or E10.3. Note that FORTRAN syntax and rules are implicitly
followed in all data card descriptions. The following examples should illustrate most instances of
FORTRAN fixed-format usage:
F10.3:
A decimal number is to be inserted in this space (field) of width 10 columns. In

general, a decimal point should be entered. Otherwise, it is assumed that digits in the
last 3 of the 10 positions are to the right of the decimal point. As long as the decimal
point is entered, the ".3" is overridden, and positioning within the field is immaterial.
Examples follow:
1 2 3 4 5 6 7 8 9
I8:
gives 39.5
gives 132.745
gives 3.14159
Integer data (no decimal point allowed) is to be inserted in the space (field) of width 8
columns. Integers should be "right-adjusted," as far to the right in the field as
possible, since blanks are interpreted as zeros.
3 - 17

Examples follow:
1 2 3 4 5 6 7 8 9
A6:
gives 39
gives 1024
gives 99000
An arbitrary collection of characters (letters; numbers; punctuation) is to be inserted in

this field of width 6. Blanks are a special unique character, remember, so "BUS A" and
"BUS A" are not the same.
Examples follow:
BUS-1A
1.3+A5
GROUND
E10.3:
A decimal number is to be inserted in the field of width 10 columns. Use of scientific

notation, with a power of 10 following the symbol "E", is added onto the previous rules
for F-formats. Omitting the E and its following exponent is interpreted as a unity
multiplier, so the E10.3 format then behaves just like F10.3. Right adjust the number
if an E is entered, since just as with the I-format, any trailing blanks are interpreted as
zeroes.
Examples follow:
1 2 3 4 5 6 7 8 9
1
E 3
gives 1.609 10300
E 3
E + 4
gives 1609
gives 3.1415926
gives 5.1
gives 0.00052
gives 52000
Free format:
There also is the option of free-format data specification, for many of the different EMTP data
structures. The rules are relatively simple. Rather than positioning data fields in fixed column
locations, there is an ordering from left to right, with a separator character (usually a comma) used
3 - 18

to delineate the field boundaries. If more than one data card is needed for the string of data items,
a continuation character (usually a dollar sign) is used rather than a separator character, which
means that another data card is to be read before extracting the next data item.
To illustrate, consider the special request word "POWER FREQUENCY" which provides for
redefinition of the steady-state frequency (see Section 4.1.8). Using free-format, this reads:
POWER FREQUENCY, 50
Rather than punching 50.0 in columns 33-40, then, ", 50" has been appended to the key word. The
comma is a separator character, dividing the key word from the parameter; the space after it is
optional, since blank characters are ignored in this mode. Note also that no decimal point has been
punched, for integers and floating point variables are interchangeable when using free-format. For
a realistic, full-scale application of free-format usage, see the data cards punched by the
transformer impedance matrix routine of the EMTP support program AUX.
In case the user wants to redefine the free-format characters CSEPAR (for field separation) and
CHCONT (for continuation), see Section 4.1.8. Please note that not all EMTP data structures can be
handled by the EMTP free-format.
For numeric data fields, blanks are ignored, no matter where they appear. For example, ",10 3," is
the same as ",103,". For floating-point numbers, either "D" or "E" can be used to indicate a power
of ten. Thus "1.E+9" and "1.D+9" are identical. Also, the plus sign preceding an exponent can be
dropped, so that "1.E9" is equally valid.
Errors with the use of EMTP free-format capability can be confusing and often unpredictable. A
missing comma can direct the program to read data from the next (and possibly unrelated data
card). The main advantage of free format is that data is not restricted to a pre-defined field width.
That is the reason why punched output from AUX is often generated in free format.
Restrictions on the usage of free-format:

Restriction 1: The branch type code ITYPE and four node names BUS1, BUS2, BUS3, and
BUS4, must be contained in columns one through 26. Note that this is the normal
fixed-format location. If the last name and its separator comma ("BUS4,") must be to
the left of column 27, then the numeric data to follow must also be to the right of
column 26.
Restriction 2: For the uncoupled series R-L-C branch, at least nine data fields are required.
Since only three floating-point fields are actually used, this suggests that six extra
commas after the third of these might be a good idea. This restriction also applies to
pi-circuits of Section 5.6
3 - 19
Restriction 3: For coupled R-L elements, at least six data fields are required. Since the first
phase only shows the use of two numbers, an extra four would seem to always provide
safety.
3 - 20
Section 4
4.1
Cards to Begin a New Data Case
Special Request Cards ........................................................................... 4-1

4.1.1
Simulation Control ..................................................................... 4-4
4.1.2
Output ........................................................................................ 4-11
4.1.3
Statistics ..................................................................................... 4-20
4.1.4
Plotting ....................................................................................... 4-23
4.1.5 TACS ........................................................................................... 4-26
4.2
4.1.6
Universal Machine ..................................................................... 4-31
4.1.7
Load Flow .................................................................................. 4-33
4.1.8
Advanced Options ...................................................................... 4-34
Miscellaneous Data Parameter Cards .................................................... 4-39

4.2.1
First Miscellaneous Data Card

(Floating-point Miscellaneous Data) ......................................... 4-39
4.2.2
Second Miscellaneous Data Card

(Integer Miscellaneous Data Card) ............................................ 4-41
4.3
Specially-Requested Extensions To Miscellaneous Data Cards ............ 4-44

4.3.1
Extra "STATISTICS" or "SYSTEMATIC"

Miscellaneous Data Card (Optional) ......................................... 4-44
4.3.2
4.4
Card for Varying the Printout Frequency (Optional) ................. 4-50
TACS Data Cards (if any) ....................................................................... 4-51
Section 4
Each new data case to be solved begins with one or more of the data types described in this
section. There can be any number (possibly zero) of special-request input card described in
Section 4.1. The special request cards with the exception of "BEGIN NEW DATA CASE", which
must come first can be in any order. The miscellaneous data cards of Sections 4.2 to 4.4 come
after the last of any such special-request input for the data case.
4.1 Special Request Cards

The special request cards can be read under free format. Take the "FREQUENCY SCAN" output of
Section 4.1.1 as an example. A valid sample card image using free format is as follows:
"FREQUENCY SCAN, 10., 5.0, 100., 0". Even further abbreviation is possible: rather than using the
full request word, an acronym made up of the first letter of each sub-word can be used. In this
case, "FREQUENCY SCAN" can be abbreviated to "FS" if desired.
Each new case begins with a number (possibly zero) of special-request and miscellaneous cards:
Section 4.1.1:
Simulation Control
Card
Description
BEGIN NEW DATA CASE
case-separation cards
ABORT DATA CASE
CDA
FREQUENCY SCAN
include effects of harmonic distortion
START AGAIN
TIME STEP LOOP
transfer of control to the time-step loop
user supplied sources
read miscellaneous data cards now
4-1
Section 4.1.2:
Output
Card
Description
AVERAGE OUTPUT

averaging

begin
KILL CODES
DIAGNOSTIC
ALTERNATE DIAG PRINTOUT
OUTPUT WIDTH
80
OUTPUT WIDTH
132
EXECUTION MONITOR
Section 4.1.3:

executing
Statistics
Card
Description
RESULTS
OMIT BASE CASE
TIME OF DICE ROLL

occurs
user supplied random switching times
4-2

Section 4.1.4:
Plotting
Card
Description
REPLOT
batch-mode plot of case solved before
USER IDENTIFICATION
user i.d. info. for batch-mode plot file
redefine height of paper on a CalComp plot
CUSTOM PLOT FILE
Section 4.1.5:
TACS
Card
Description
RELATIVE TACS DIMENSIONS
TACS EMTP SOURCES
TACS-defined,
user supplied, sources
TACS INIT
TACS ORDER
Section 4.1.6:
postprocess plot file using TACS
Universal Machine
Card
Description
Section 4.1.7:
Load Flow
Card
Description
FIX SOURCE
old EMTP load flow
FIX SOURCE1
new improved EMTP load flow
LOAD FLOW
three phase load flow
4-3

Section 4.1.8:
Advanced Options
Card
4.1.1
Description
FILE REQUEST
RENUMBER BYPASS
FREE FORMAT
MODIFY SWITCH LOGIC
STEP ZERO COUPLE
subnetwork ident. with all switches closed

HIGH RESISTANCE
redefine R of internally added resistor
POWER FREQUENCY
ZINC OXIDE
control parameters for solution of nonlinear elements
Simulation Control
BEGIN NEW DATA CASE
END LAST DATA CASE
optional card to end the EMTP simulations
ABORT DATA CASE
CDA
FREQUENCY SCAN
include effects of harmonic distortion
START AGAIN
TIME STEP LOOP
transfer of control to the time-step loop
user supplied sources
read miscellaneous data cards now
4-4
Begin New Data Case: This card should be used at the beginning of each data case. A
"BEGIN NEW DATA CASE" card should also be placed at the end of all data cases, followed by a
blank card.Blank Card to indicate the termination of EMTP execution
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
ABORT DATA CASE: When this card is used, the EMTP will skip input records until the next
"BEGIN NEW DATA CASE", is encountered. This card can be useful when more than one data case is
stacked in the same input file.
1
123456789012345
2
3
4
5
6
7
8
67890123456789012345678901234567890123456789012345678901234567890
ABORT DATA CASE
CDA: This card enables the Critical Damping Adjustment procedure (CDA).
123
1
2
3
4
5
6
7
8
45678901234567890123456789012345678901234567890123456789012345678901234567890
CDA
The solution algorithm of the EMTP is based on the solution of differential equations using the
trapezoidal rule of integration. Although the trapezoidal rule is always stable, certain type of
simulations can lead to bounded numerical oscillations. Typical examples of simulations where
numerical oscillations take place is when there is a discontinuity in the current of an inductive
circuit, or a discontinuity in the voltage of a capacitive circuit. The basic principle behind CDA is
to replace the trapezoidal rule of integration with two backward Euler integration steps every time
there is a discontinuity in the electrical network (e.g., when a switch opens or when a diode
changes state). These two integration steps use half the time step of the normal simulation, and
effectively eliminate any numerical oscillations.
In simulations where there is no cause for numerical oscillations, the presence or absence of CDA
has little or no effect, so CDA is very safe to use. However, CDA is not a substitute for poor
simulation techniques. For example, consider a simulation where a shunt reactor is switched out
of the network. In this case, there is a high frequency voltage oscillation at the terminals of the
reactor, as the energy stored in the reactor dissipates (usually through shunt and series loses). The
4-5

frequency and magnitude of this voltage oscillation depend on the inductance of the reactor and the
stray capacitances between the terminals of the reactor and ground. If this case is modelled
improperly by omitting the stray capacitances, the answers given with and without CDA are equally
incorrect: without CDA there would be oscillations of the wrong frequency and magnitude, whereas
with CDA, there would be no oscillations at all.
MISCELLANEOUS DATA CARDS: The floating-point and integer miscellaneous data cards are
described in Section 4.2. For a conventional simulation, they come after all special requests of the
type now being considered. But in special circumstances, the user may want to define one or more
of the miscellaneous data parameters ahead of time. In such case, the miscellaneous data cards of
Section 4.2 can be read ahead of time provided they are preceded by this special-request word:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890

I0E8.0 floating point miscellaneous data card
I0I8 integer miscellaneous data card
The user is advised not to do this casually.
FREQUENCY SCAN: The "FREQUENCY SCAN" feature of the EMTP allows for the repetition of
steady-state phasor solutions, as the frequency of sinusoidal sources is automatically incremented

between a beginning and an ending frequency. Rather than conventional EMTP time-response
output, the user then has available a frequency-response output. When plotted, the time axis of
conventional EMTP simulations becomes the frequency axis, with the result being a Bode plot.
Polar coordinates (magnitude and angle of the phasor variables) are used for output purposes. For
example, the following could apply to the node voltage of phase "a" somewhere in the network:
lVal
a
Va = lVal a
Va = lVal a
Frequency
Frequency
4-6

To produce a "FREQUENCY SCAN" simulation, the following card must precede input of the
floating-point miscellaneous data card of Section 4.2.1:
1
2
3
4
5
6
7
8
123456789012345678901234 56789012 34567890 12345678 90123456 7890123456789012 34567890
FREQUENCY SCAN
A24
fmin
fmax
NPDEC
MODSYM
E8.0
E8.0
E8.0
I8
A16
fmin
(2532)
Starting frequency of the scan, in Hertz. This must be a positive

number.
f
(3340)
The frequency increment for linear spacing between adjacent

frequency points. In this case, fk+1 = fk + f. If logarithmic
spacing between adjacent frequency points is desired, this field
should be left blank.
fmax
(41-48)
Maximum frequency of the scan, in Hertz. This must be greater

than or equal to fmin.
NPD
For linear spacing of the frequency points, leave this field blank.
For logarithmic spacing, NPD is the number of points per decade.
In this case, adjacent frequency points are related by:
(4956)
fK+1 = (101/NPD) . fK
MODSYM
(57-62)
Flag indicating the type of line/cable model to be used in the

frequency scan.
When MODSYM=TIME-DOMAIN, frequency scan calculations are to
be performed using the same models normally used during
transient simulations. On the other hand, when
MODSYM=FREQUENCY-DOMAIN, frequency scan calculations are
made using "correct" frequency-domain models.
In the case of transmission lines and/or cables,
MODSYM=FREQUENCY-DOMAIN implies that an exact-
representation is read from an external file supplied by AUX (field
FILENAME in the first LMARTI or FDQ branch card of Section 7.5).
If MODSYM=TIME-DOMAIN different line/cable models will have
different representations. For example:
4-7
The LMARTI model with constant Q will have an equivalent representation calculated from the approximations by rational
functions of Yc' and A', and Q will be assumed to be constant.
The distributed-parameter, lossless line model with lumped

losses (also known as Dommel's line model) will assume a
lossless -representation electrically connected to three
lumped resistances.
Note that MODSYM is a global flag in the sense that it is not

associated with any particular model. Therefore, it is not possible
to mix "FREQUENCY-DOMAIN" and "TIME-DOMAIN"
representations in the same simulation. In the future, all EMTP
models will have the provision to include correct frequencydomain representations of their normal time-domain forms.
Special attention should be paid to integer miscellaneous data parameter "KSSOUT", when using the
"FREQUENCY SCAN" option (see the Section 4.2.2 definition). If KSSOUT=1, the full steady-state
branch flow and injection printout will result, for each solution frequency of the scan. This could
produce large quantities of printed output, if the number of frequency points is large. For an
additional output option see Section 16.
INITIALIZATION WITH HARMONICS: Type-93 and Type-98 nonlinear inductances can become
saturated even in the steady-state solution. Their saturated state causes them harmonic distortion
in the fundamental frequency steady-state solution. The correct initial conditions can be obtained
only by including the distortion effects into the steady-state initialization procedure.
The solution technique uses two iterative loops: the Vrms-Irms loop to obtain a better solution at the
fundamental frequency and the Flux-Current loop to account for the harmonic distortions in the AC
steady-state network solution. The technique can handle networks containing lumped L, R, C
elements, and distributed parameters transmission lines.
To use this option the following card must precede input of the floating-point miscellaneous data
card:
1
2
3
4
5
6
7
8
12345678901234567890123456789012 34567890 12345678 90123456789012345678901234567890
A32
:
KHAMON
KNHAR
I8
I8
4-8
KHAMON
(33-40)
KNHAR
(41-48)
Flag specifying which iteration loops will be present during the

solution.
=0
both the Vrms-Irms and Flux-Current Loops are requested.
=1
only the Vrms-Irms loop is requested.
Number of odd harmonics (including the fundamental) to be

included in the solution. The value can vary between 1 and 8.
Default value is 8.
Note that if the frequency scan and initialization with harmonic distortion are requested in the
same EMTP case, the request card for the frequency scan feature must precede the request card of
this section. Also, note that Type-96 hysteretic reactors are not included in this initialization
procedure. On the other hand, type-92 hysteretic reactors can be used with this initialization
procedure.
START AGAIN: Integer miscellaneous data parameter MEMSAV will result in the dumping of
EMTP memory onto disk at the conclusion of a simulation (i.e.,
at t=TMAX). Such a simulation can
be restarted.
The request to load those memory contents from disk is of the form shown below. The "START
AGAIN" card should be followed by an arbitrary number of cards which redefine switch closing
and opening times Tclose and Topen, terminated by a 9999-card (in columns 1-4). Opening time
Topen will be redefined only if columns 25-34 are punched with a positive number input. More
commonly, it is the closing time Tclose which will be altered, for switches which are open at the
end of the preceding simulation. Columns 15-24 are always read, and Tclose is always redefined
(blank is interpreted as a request for zero closing time). Following the "9999" terminator, TMAX
can be altered with a "MISCELLANEOUS DATA CARDS" request, and then a "TIME STEP LOOP"
request actually transfers control to the time-step loop
.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
START AGAIN, file spec
4-9
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234567890123456789012345678901234567890
BUS1
BUS2
I2
A6
A6
Tclose
Topen
E10.0
E10.0
A typical example of the START AGAIN option is shown below:

BEGIN NEW DATA CASE
C
BENCHMARK DC-49
File dc32.bin contains the memory snapshot
START AGAIN, DC32.BIN,
of previous run
Re-defined switch
0FAULT
.011990
2.0
9999
.000020
.018
1
1
1
1
1
C
FOLLOWING STEP NUMBERS ARE CONTINUATION OF DC-32, WHICH WENT
C
500 STEPS (DELTAT=20 MICROSECONDS, TMAX=10 MSEC).
500
1
502
2
510
10
550
50
TIME STEP LOOP
PRINTER PLOT
BLANK CARD ENDING PLOTTING (WHICH IS ILLEGAL IN BATCH MODE).
BEGIN NEW DATA CASE
BLANK
Between the "MISCELLANEOUS DATA CARDS" and the "TIME STEP LOOP" requests, a number of
special request cards can be used. The most commonly used one is "CHANGE PRINTOUT
FREQUENCY", to alter the frequency of the time-step loop printout. If this is done, remember that
step numbers begin where the previous halted run left off (e.g., for TMAX=50 ms and DELTAT=100
s, the first printed step would be for number 500). Note that while all requests may be accepted
by the EMTP, some of the more complex ones cannot possibly be honoured in fact. Perhaps the best
example is "RENUMBER BYPASS" of Section 4.1.8: since renumbering is to be skipped (control is
transferred immediately to the time-step loop), there is no way node renumbering can be altered.
Even worse are some requests which will simply make erroneous or inconsistent certain
parameters of the memory contents to be restarted. An example of such a potential disaster is
provided by use of the "ABSOLUTE TACS DIMENSIONS" request of Section 4.1.5. Use of this feature
would alter the pointers to the TACS tables, without changing the tables themselves (which are
already built), leading to "garbage out", and most likely an operating system interrupt of some sort.
4 - 10

TIME STEP LOOP: Used in conjunction with the "START AGAIN" request, a "TIME STEP LOOP"
card will begin execution by a transfer of control to the time-step loop.
1
2
3
4
5
6
7
8
12345678901234 567890123456789012345678901234567890123456789012345678901234567890
TIME STEP LOOP
ANALYTIC SOURCE USAGE: As explained in Section 10, source types 1 through 10 are
reserved for functions which are directly defined by the user. If one or more such source functions
is to be defined in FORTRAN within a special user-supplied version of subroutine "ANALYT", then
the following special request record must precede input of the miscellaneous data cards for the
data case in question:
1
2
1234567890123456789012
3
4
5
6
7
8
3456789012345678901234567890123456789012345678901234567890
It is the user's responsibility to see to it that his own special module "ANALYT" has replaced the
dummy one which comes with the program, when actually executing a data case which contains
such an "ANALYTIC SOURCES USAGE" request. Note that to use a user-supplied "ANALYT"
subroutine, it is necessary to compile "ANALYT" and re-link it to the rest of the EMTP libraries.
4.1.2
Output
AVERAGE OUTPUT

averaging

begin
KILL CODES
DIAGNOSTIC
4 - 11

ALTERNATE DIAG PRINTOUT
OUTPUT WIDTH
80
OUTPUT WIDTH
132

executing
EXECUTION MONITOR
CHANGE PRINTOUT FREQUENCY: The frequency of the printout in the time-step loop is
controlled by the integer miscellaneous data parameter IOUT of Section 4.2.2. But this frequency
can be changed as simulation time progresses. Begin with the special-request card shown below,
and follow this with the card of (KCHG, MULT) pairs of Section 4.3.2:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 345678901234567890123456789012345678901234567890
First Change
KCHG
MULT
I8
I8
Second Change
KCHG
I8
MULT
Etc.
I8
This card specifies at what time-step numbers the printout frequency is to be changed, and what
values the frequency is to be changed to. Up to five pairs of step numbers and new printout
frequencies are permitted, as per the following definitions:
KCHGI
Time-step number at which the i-th variation in the printout frequency is

to begin. Printout for this particular step number is always provided, as a
beginning to the new frequency of output.
MULTI
Modified value of "IOUT" (see Section 4.2.2), to begin at time-step

number KCHGi.
Note that the printout frequency can also be changed setting the integer miscellaneous data
parameter IPUN = 1 (Section 4.2.2), in conjunction with the extension card in Section 4.3.2.
However, using the "CHANGE PRINTOUT FREQUENCY" request card is the recommended option.
4 - 12

AVERAGE OUTPUT: The special-request card "AVERAGE OUTPUT" is used to average
successive values of EMTP output variables before printing and plotting.
1
2
3
4
5
6
7
8
12345678901234 567890123456789012345678901234567890123456789012345678901234567890
AVERAGE OUTPUT
The mathematics of EMTP solution are not affected; this is just a massaging of the answers at the
time answers are printed and written to the LUNIT4 plot file. By averaging is meant the arithmetic
mean of the current and the preceding variable value:
V
+V
new
old
V = ----------------------------------2
This option was useful (before CDA was implemented in the EMTP) to mask cases of numerical
oscillations (see Section 9.1.2).
LIMIT ON PLOT OSCILLATIONS: Variable "NSMTH" is a limit on the number of successive
"ups" and "downs" which a curve being plotted is to be allowed, before the averaging of
successive ordinates for all later time is to be instituted. Default value is 50. This default value
can be overridden with the "LIMIT ON PLOT OSCILLATIONS" request card. As in the case of the
"AVERAGE OUTPUT" request card; this option is probably obsolete with the availability of CDA.
1
2
3
4
12345678901234567890123456789012 34567890

A32
5
6
7
8
1234567890123456789012345678901234567890
NSMTH
I8
BEGIN PEAK VALUE SEARCH: The integer miscellaneous data parameter MAXOUT (see
Section 4.2.2) provides for the calculation and output of variable extrema. The same vector of
extrema is used for "STATISTICS" and "SYSTEMATIC" output as well. Normally, extrema are
wanted over the full time span of the simulation. However, there are cases where it may be
desirable to ignore extrema in a certain initial interval of the simulation, and only consider
transients after a certain minimum time. For example, in a "STATISTICS" simulation, it is possible
that random closing follows the opening of breakers (clearing of a fault), but that the peak
simulation voltage occurs during the deterministic opening portion of the simulation. If extrema
were calculated over the whole time span of the simulation, each energization would generate
identical peaks, which would defeat the purpose of the statistical study. So, in this case, the
4 - 13

solution is to inhibit the calculation of extrema until transients of the opening phase of execution
have passed. This is possible, using a special-request card which reads as follows:
1
2
3
4
12345678901234567890123456789012 34567890
5
6
7
8
1234567890123456789012345678901234567890
BEGMAX
A32
E8.0
Here BEGMAX (columns 33-40) is the time in seconds at which the extrema computation is allowed
to begin.
Alternatively, it is possible to calculate extrema in pre-defined time intervals by setting BEGMAX=1.0, and by adding an extra card with monotonically increasing times T1, T2, etc. In this case,
extrema are to be calculated only for those portions of the simulation which intersect the subintervals (T1, T2), (T3, T4), etc. The format for this extra card is shown below, where the final
subinterval is followed by some very large beginning time of a non-existent bounding interval:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 345678901234567890123456789012345678901234567890
T1
T2
T3
T4
E8.0
E8.0
E8.0
E8.0
Etc.
At present, only two sub-intervals are allowed, so either T3 or T5 has to be a very large time.
PEAK VOLTAGE MONITOR: This option is used to obtain the highest peak node voltage in the
simulation. At the end of the simulation, in addition to the regular peak value printout for the
requested output variables, the highest peak voltage in the entire simulation will be printed. This
value is in volts, not in per unit.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
MODAL VOLTAGE OUTPUT: This request is used to obtain modal-voltage rather than phase-
voltage output at the ends of a distributed-parameter transmission line with constant transformation
matrix (see Section 7).This is a rather obscure feature, and it should be used with caution.
4 - 14

The first step to request modal output is to insert the modal voltage output request card (before the
miscellaneous data cards of Section 4.2).
1
2
3
4
12345678901234567890123456789012 34567890
MODE VOLTAGE OUTPUT

A32
5
6
7
8
1234567890123456789012345678901234567890
MODOUT
I8
Columns 1-19:
Enter the special request word "MODE VOLTAGE OUTPUT".
Columns 33-40:
Enter the number of phases (or coupled conductors) which make

up the distributed-parameter line whose modal output is desired.
A blank or zero field will be given default value equal to 3,
which is the most common case.
Such a special request card can be preceded by a "BEGIN NEW DATA CASE" card, and be either
preceded or followed by parameter-redefinition cards, if desired; comment cards can also be
mixed in anywhere (using "C " in columns 1 and 2).
Additional the following steps must be followed:
Point 1:
The distributed-parameter transmission line in question (for which the user wants
mode voltage output) must be the final branch-component of the data case; it must
immediately precede the blank card which terminates all branch cards.
Point 2:
The very first branch of the data case should consist of fictitious high-resistance
branches from each node at both ends of the line in question to ground. Equal in
number to 2MODOUT, column 80 of these uncoupled (type-zero) resistive branches
must be punched with a 2 so as to request branch-current output. Use a resistance
value which is so large that the problem is unaffected by addition of these resistors
(see Section 5.1.5 for acceptable limits on different computer systems).
After the case has been run, modal voltages will appear in the output locations which should
otherwise have been reserved for the branch currents of Point 2 above. The first "MODOUT" output
currents are in reality modal voltages to local ground at the "BUS1" end of the line, in natural
order; the next "MODOUT" output currents are mode voltages to ground at the other end ("BUS2"
end) of the line, also in natural order. A reminder message to this effect is printed immediately
below the column headings for the output variables, so the user will not forget. Note that if one is
plotting, type9 plot requests (see Section 15) must be used, since the EMTP assumes that the
modal voltages are currents in the 2MODOUT high-resistance branches which were supplied by the
user.
4 - 15

KILL CODES: It is option will generate the text of any particular fatal EMTP error message (KILL
CODE)
without actually having to set up a data case that will produce such an error condition.
1
2
3
4
5
6
7
8
12345678901234567890123456789012 34567890 12345678 90123456789012345678901234567890
KILL CODES
A32
KILL1
KILL2
I8
I8
The key-word "KILL CODES" is entered in column 1-10, and KILL1 and KILL2 are the beginning and
ending KILL numbers of the consecutive KILL codes which are to be printed. Note that KILL2 must
not be less than KILL1.
Caution: Systems which do not allow the printing of garbage may have trouble with this
feature. VAX is one such system, and the printing of a range of KILL codes will
typically end prematurely with an I/O complaint by the operating system.
DIAGNOSTIC: This option turns on diagnostics printout in specific portions of the program. This
should be regarded as a documented diagnostic tool for program maintenance purposes only.
Parameter "IPRSUP" of the integer miscellaneous data card (Section 4.2.2) turns diagnostic printout
on or off irrespective of the location of that printout within the program. That is, either all printout
is turned on, or all is turned off, at a certain threshold level. If positive, "IPRSUP" will override the
selective printout request of this present section. But if "IPRSUP" is to be punched zero or blank,
then the EMTP diagnostic printout can be controlled overlay by overlay. A special-request card
bearing the key word "DIAGNOSTIC" in columns 1-10 is used in this case. The following format is
applicable for this card which must precede the miscellaneous data cards:
A20
IPRSOV(K)
(IPRSOV(I) , I=1, 30)
21
22
23
24
25
26
27
28
29
30
19
20
16
17
18
6
7
8
9
10
11
12
13
14
15
4
5
DIAGNOSTIC
1
2
3
1
2
3
4
5
6
7
8
12345678901234567890 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90
I2 format
diagnostic printout control parameter as restricted to overlay number

"K". Within overlay "K", IPRSOV(K) has the same meaning as
"IPRSUP" as described in Section 4.2.2.
4 - 16

For overlays "K" which exceed 30, IPRSOV(K-30) is used to control
the diagnostic printout. That is, only 30 distinct codes have been
provided for, with these being reused cyclically for higher overlay
numbers.
In the selection of these diagnostic-printout codes, special care should always be taken with regard
to the time-step loop, which is overlay number 16. Remember that diagnostic printout can be
generated for each time-step, so beware of the termination time "TMAX" in relation to the step-size
"DELTAT", should such printout be requested. Another consideration is for large networks, which
will heavily impact on overlay number 7 (network node-renumbering). Here printout can be
produced for each row of the matrix.
Control of diagnostic printout for the time-step loop code of overlay 16 deserves further
explanation. Actually, it is broken into four principal modules: SUBTS1, SUBTS2, SUBTS3, and
SUBTS4. Printout control parameter IPRSOV(16) applies to the first of these (SUBTS1), 17 applies to
the second, 18 applies to the third, and 19 is for the last (SUBTS4). Since the time-step loop is
followed by overlay number 20, there is no contradiction or confusion in this usage.
If diagnostic printout is being turned off and on as a function of time (see Section 4.3.2, with a
minus sign on MULT), then (IPRSOV(J), J=16, 19) alternate with (IPS(J), J=1, 4) for control of
diagnostic printout of the time-step loop. There are two distinct cases:
Case 1.
If the user wants to start with no diagnostic (step zero or one), then IPRSOV(16)
through IPRSOV(19) should all be zero, and the "ALTERNATE DIAGNOSTIC
PRINTOUT" request should be used to define the nonzero print codes (one or
more nonzero).
Case 2.
If the user wants to start with diagnostic (step zero), then IPRSOV(16) through
IPRSOV(19) should not all be zero. In this case, no "ALTERNATE DIAGNOSTIC
PRINTOUT" request is needed, since zero values are the default.
ALTERNATE DIAGNOSTIC PRINTOUT: This option turns on diagnostics printout in specific

portions of the program. This should be regarded as a documented diagnostic tool for program
maintenance purposes only
If minus signs are applied to MULT(K) as part of the printout-frequency changes of Section 4.3.2,
then diagnostic printout codes of the time-step loop are switched. Those of the DIAGNOSTIC
request are in effect initially. Then, at the step of the first change, an alternate set of printout
control parameters (IPR(K), K=1, 4) is used instead. These four new variables replace IPRSOV(16)
through IPRSOV(19), to control diagnostic printout of the four pieces of the time-step loop. Upon
the second such change (second negative MULT), there is a switch back to the original DIAGNOSTIC
printout control codes. Etc. (for each minus sign, the two sets of control variables are switched).
4 - 17

The alternate set of printout control parameters (IPR(K), K=1, ..., 4) is initialized to zero at the
beginning of execution. Hence if this is what the user wants, he needs to do nothing special. This
corresponds to Case 2 at the end of Section 4.1.2 - DIAGNOSTIC (begin with diagnostic printout at
time zero).
But if the time-step loop is to begin without any diagnostic printout (Case 1, Section 4.1.2 then nonzero IPR(K) must be defined by the user. This is done using a specialrequest word "ALTERNATE DIAGNOSTIC PRINTOUT", as shown in the following format:
DIAGNOSTIC),
1
2
3
4
5
6
7
8
12345678901234567890123456789012 34567890 12345678 90123456 78901234 5678901234567890
A32
OUTPUT WIDTH
IPR1
IPR2
IPR3
IPR4
I8
I8
I8
I8
80: It is possible to switch to 80column output (rather than the more common
132-column output) by means of an "OUTPUT WIDTH 80" request. In the absence of such a request,
EMTP printout will normally be 132 columns.
Use of 80-column output has some serious limitations which should be understood. First, the
interpreted tabulation of input data cards, as well as the connectivity table, will be truncated at
column 80, and all output to the right of this point will simply be lost. Second, not all displays any
longer convey full information. For example, the phasor branch flows are now only in polar
coordinates, for the near end of the line. Case summary statistics now only display PRESENT SIZE
figures and computer times without the voluminous associated English language text, etc. If in
doubt, try the feature on a small problem, and see if the output is satisfactory.
The request "OUTPUT WIDTH 80" is a single-time request which then would normally remain in
effect for the entire EMTP solution. But there are times when the output width might profitably be
switched between 80 and 132 columns at different points of the EMTP execution. This is generally
possible via the universal dollar card $WIDTH as described in Section 3.6.
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
OUTPUT WIDTH 80
4 - 18
132: This special request card forces output width output to 132 columns in
those systems where 80-column output is used as default.).
OUTPUT WIDTH
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
OUTPUT WIDTH 132
EXECUTION MONITOR: This special request card activates a small plotting window in the EMTP
Launcher that shows a plot of selected EMTP variables as a function of time as the simulation
progresses (Windows and Sun Solaris versions only).
The format for this special request card is as follows:
1
2
3
4
5
6
7
8
12345678901234567890 123456789012 345678901234 567 890 123 456 789 012 345 678 901 234 567 890
KEYWORD
Ymin
Ymax
IR
I1
I2
I3
I4
I5
I6
I7
A20
Keyword
(1-20)
Set to "EXEC STATUS" for graphical display of execution

progress
Ymin
(21-32)
Minimum value of Y-axis. If left blank or if the

simulation exceeds Ymin, the plot will re-scale
automatically.
Ymax
(33-44)
Minimum value of Y-axis. If left blank or if the

simulation exceeds Ymax, the plot will re-scale
automatically.
IR
Number of points to skip in the execution display. In

other words, only every IR'th point will be plotted.
Default=100.
(45-47)
I1, I2,
...
(48-50) ...
Number of EMTP output variable to be displayed. The

numbering follows the same order as the listing of output
EMTP variables (I1=1 plots the first node voltage, if any).
Up to 10 variables can be displayed simultaneously.
4 - 19
I8
I9
I10
4.1.3
Statistics
RESULTS
OMIT BASE CASE
TIME OF DICE ROLL

occurs
user supplied random switching times
TABULATE ENERGIZATION RESULTS: Provided that the user has previously executed a
"STATISTICS" data case one or more times beforehand, and saved the essential internal tables (using
"STATISTICS OUTPUT SALVAGE" request, then it is possible to proceed with statistical tabulation.
This begins with the following special-request card
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
TABULATE ENERGIZATION RESULTS
followed by a statistics miscellaneous data card (see Section 4.3.1). Next, specify 3-digit integers
that uniquely identify all previous simulations (see parameter JFLSOS of STATISTICS OUTPUT
SALVAGE) are specified:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 56789012 34567890
JF1
JF2
JF3
JF4
JF5
JF6
JF7
JF8
JF9
JF10
I8
I8
I8
I8
I8
I8
I8
I8
I8
I8
Use as many cards as desired, skipping any of the 8-column data fields if this is convenient.
Ordering of the integers is immaterial. Signal the end of such data by punching "9999" in a field
following the last data-case specification.
STATISTICS OUTPUT SALVAGE: Because "STATISTICS" data cases can be so time-consuming,
it is often convenient to solve them in pieces. Take a really big case (either many nodes, or many
time-steps, or both), which might require 15 minutes for the base case simulation. If 100 shots of
"STATISTICS" simulation were desired, it would be an all day operation (with the probability of a
4 - 20

computer crash in the interim, which would waste the entire simulation). In such cases it might be
desirable to split a large job into a number of smaller ones (e.g., five jobs of 20 shots each), solve
the smaller jobs individually, and then combine and tabulate results later using the "TABULATE
ENERGIZATION RESULTS" feature.
The following special request card is required for the saving of "STATISTICS" results for later
tabulation:

A29
JFLSOS
1
2
3
4
5
6
7
8
12345678901234567890123456789 012 345678901234567890123456789012345678901234567890
I3
The 3-digit integer field JFLSOS allows the user to identify the disk files which result using a
decimal serialization between 001 and 999. If JFLSOS is left blank or punched with zero, the EMTP
shall itself generate such integer identification randomly. The only disadvantage with this is that
then the user does not know the file names before the case has been solved (the random
serialization will appear in the column 1-50 interpretation of the "STATISTICS OUTPUT SALVAGE"
data card).
OMIT BASE CASE: The conventional "STATISTICS" or "SYSTEMATIC" data case creates a base
case before any of the NENERG (integer miscellaneous data parameter; see Section 4.2.2)
energizations begin. This is a preliminary shot right down the middle, with all of the variances set
to zero. The user can suppress this extra, preliminary simulation by a special request card reading:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
OMIT BASE CASE
Resulting statistical tabulations are unaffected, since the base case solution is not part of the
statistical processing. A typical application for this feature is the case where one is solving one of
several small pieces of a bigger problem, using "STATISTICS OUTPUT SALVAGE", where using the
"OMIT BASE CASE" option would avoid repeating the same base case solution for each piece of the
problem that is solved.
TIME OF DICE ROLL: As explained in Section 9, a "STATISTICS" data case can model the
random opening of switches (rather than the more conventional random closing). If the user wants
4 - 21

to manually define the simulation time at which the EMTP tables are to be saved, so that each
energization need not repeat the deterministic simulation up to this point, the following specialrequest card should be used:
1
2
3
4
12345678901234567890123456789012 34567890
TIME OF DICE ROLL
5
6
7
8
1234567890123456789012345678901234567890
TENERG
A32
E8.0
Variable TENERG (columns 33-40) is the time before which no random switching can reasonably
occur. In the case of statistical switch opening, TENERG should be set to -1.
USER SUPPLIED SWITCH TIMES: This special request allows the user to specify the random
switch closing/opening times. To use this feature, the data case should be prepared the same way
as for the regular statistics case except the following two items:
1.
Input a special request card "USER SUPPLIED SWITCH TIMES" before the first
miscellaneous data card.
1
2
3
4
12345678901234567890123456789012 34567890
5
6
7
8
1234567890123456789012345678901234567890
IUNIT
A32
I8
Here, IUNIT of columns 33-40 (default equals to 24) is the unit number of the file
storing the user-supplied switch times. The connection and disconnection of this file
of switch times is installation-dependent, to be handled either externally (for VAX,
using $ASSIGN; for IBM, in the JCL), or via the installation-dependent $ATTACH function
of "CIMAGE".
2.
Specify the switch closing times (TCLOSE) and opening times (TOPEN) of all the
switches in a data file numbered IUNIT using 5E15.0 formats:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 123456789012345 678901234567890 123456789012345 67890
TCLOSE (I)
E15.0
I=1
E15.0
KSWTCH
E15.0
4 - 22
E15.0
E15.0
1
2
3
4
5
6
7
8
123456789012345 678901234567890 123456789012345 678901234567890 123456789012345 67890
TOPEN (I)
I=1
E15.0
E15.0
KSWTCH
E15.0
E15.0
E15.0
Repeat this set of switching times (closing and opening) for each energization.
4.1.4
Plotting
REPLOT
batch-mode plot of case solved before
USER IDENTIFICATION
user i.d. info. for batch-mode plot file
redefine height of paper on a CalComp plot
CUSTOM PLOT FILE
REPLOT: If a previously-solved data case had miscellaneous data parameter "ICAT" set to 1 or 2,
then the plot data points of that solution were saved on disk as a permanent file. Should the user
now, at a later time, wish to do Postscript batch-mode plotting from this data, he needs only to
precede his plot cards by the following "REPLOT" request card:
1
1234567890
2
3
12345678901234567890
REPLOT
MM/DD/YY HH.MM.SS
A10
4
5
6
7
123456789012345678901234567890123456789012
8
34567890
IPRSUP
A20
I8
Columns 1-6:
Must contain the key request word "REPLOT".
Columns 11-30:
.File name of the "pl4" plot file which contains the data to be replotted.
Columns 73-80:
Parameter controlling diagnostic printout; leave "IPRSUP" blank

(or punch a zero) for production runs. See further definition
associated with usage on second miscellaneous data card,
Section 4.2.2.
4 - 23
Warning:
Before blindly trying the just-described feature for the first time, a user
is advised to first read the installation-dependent instructions for his
own computer system. The "REPLOT" feature is quite installationdependent, and may not even be activated on some systems. A data
case in which the user wants to produce Calcomp plots by means of
the "REPLOT" feature then consists of the following components:
A data case in which the user wants to produce Calcomp plots by means of the "REPLOT" feature
then consists of the following components:
1.
First, a "BEGIN NEW DATA CASE" card, if so desired. This is optional, as per
Section 4.1.1 description.
2.
Possible special request cards to redefine built-in EMTP plot parameters, as per
Section 4.1.4. These have key request-words "PLOTTER PAPER HEIGHT" and
"PRINTER LINES PER INCH".
3.
Then the "REPLOT" card as just described.
4.
Finally, the plotting cards, as described in Section 15.
USER IDENTIFICATION: In conjunction with the batch-mode Postscript plotted output of
Section 15, it is possible to give a set of batch-mode plots a unique six-character identification
name "USERID", by using the following special-request card:
1
2
3
123456789012345678901234 567890
USER IDENTIFICATION
A24
4
5
6
7
8
12345678901234567890123456789012345678901234567890
USERID
A6
PRINTER LINES PER INCH: This is used only in conjunction with the line-printer plotting of
graphs. Variable "LNPIN" gives the number of lines per inch to which the printer has been set (e.g.
6 or 8) In any case, redefinition is by means of the above-listed special request card which carries
the text "PRINTER LINES PER INCH" in columns 1-22, and the new desired value for "LNPIN"
punched in columns 33-40.
1
2
3
4
5
6
7
8
1234567890123456789012 3456789012 34567890 1234567890123456789012345678901234567890
A22
LNPIN
I8
4 - 24

PLOTTER PAPER HEIGHT: Variable "SZPLT" is the height of the graph paper on which
Postscript plots are to be drawn. Protection against pen movements which would go off the top of
the paper (higher than "SZPLT" inches, relative to the location where the pen is initialized at the
bottom of the paper) is provided by this variable. The nominal value presently used for "SZPLT" is
10.0 inches. The user should make sure, by consulting with Program Maintenance, what built-in
value exists for his program. In any case, redefinition is by means of the above-listed special
request card which carries the text "plotter paper height" in columns 1-20 and the new desired
value for "szplt" punched in columns 33-40 as E8.0 information.
1
2
3
4
5
6
7
8
12345678901234567890 123456789012 34567890 1234567890123456789012345678901234567890
A22
SZPLT
E8.0
CUSTOM PLOT FILE: The program user will normally employ whatever plot file precision has
been decided upon by Program Maintenance for his computer system, and not think further about
the question. For plotting only, which is a low-accuracy operation, any computer precision is
adequate, so the user need never worry. But not so for other uses of the plot file like
"POSTPROCESS PLOT FILE" (where plot file variables become TACS sources of a subsequent
simulation), or "FOURIER ON" (to request Fourier series decomposition of a plot file variable), full
EMTP solution precision may be desired. For those computers that store output variables on disk
with reduced precision, a special request (the "CUSTOM PLOT FILE" request illustrated below) is
required to override the default decision favouring reduced precision.
Actually, the "CUSTOM PLOT FILE" request toggles the choice of plot file precision. If Program
Maintenance has asked for single-precision plot files with an M4PLOT=2 statement in installationdependent SUBROUTINE SYSDEP, then "CUSTOM PLOT FILE" will switch to full-precision (doubleprecision) plot file usage. The reverse is also true, although not commonly the case.
2
3
4
5
6
7
8
1
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
CUSTOM PLOT FILE
4.1.5
TACS
4 - 25

TACS-defined,
TACS EMTP SOURCES
user supplied, sources
TACS INIT
TACS ORDER
ABSOLUTE TACS DIMENSIONS:
postprocess plot file using TACS
Request for Allocation of TACS Storage:
TACS is the control system modelling feature of Section 14. Total working space for all TACS
tables is assigned in List 19, along with the other EMTP lists (see Section 2.3). But there still
remains the question of how this total working space is to be divided among the individual TACS
tables. This allocation is performed at execution time, according to the description of the present
section. The first way to allocate total TACS storage is with a request for absolute TACS table
sizing, followed by one data card giving those desired sizes:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
LT1
LT2
LT3
LT4
LT5
LT6
LT7
LT8
I8
I8
I8
I8
I8
I8
I8
I8
The format for the list sizes is (8I8), with the 8 TACS table sizes having the following meaning:
LT1:
Maximum number of TACS dynamic function blocks, having Laplace

transfer functions H(s).
LT2:
Maximum number of nonzero factors of the triangularized TACS

network matrix. This limit applies to both the steady-state solution and
the transient solution.
LT3:
Maximum total number of input variables to TACS dynamic function and

supplemental devices.
4 - 26

LT4:
Maximum number of TACS sources. This includes built-in sources (e.g.,

"TIMEX" or "UNITY"), and variables passed from the electric network to
be TACS driving functions.
LT5:
Maximum number of supplemental variables and devices (type codes

99, 98, or 88 punched in columns 1-2).
LT6:
Maximum total number of extra INTEGER pointers which are associated

with supplemental variables and devices. There is no easy, exact
formula, although sample figures can be given. Each parenthesis
requires 3 cells, as does each arithmetic operation ("+", "*", "=").
LT7:
Maximum total number of extra REAL cells which are associated with
TACS variables. Each TACS variable requires one or more such cells.
Again, a simple rule is impossible to state. Each first-order function
block requires 10 cells, intermediate variables of supplemental-variable
expression evaluation each requires one, the average device might
require between 3 and 5, and the most complex device (RMS sensor,
Type-66 (requires 1 / (f * DELTAT) where "f" is fundamental frequency.
LT8:
Maximum number of distinct TACS variables used in the TACS data

specification.
Provided the total storage represented by this user request is less than or equal to the List 19 space
available, execution will proceed normally. Otherwise, execution will be terminated with a
KILL=1 EMTP error message pointing to List 19.
RELATIVE TACS DIMENSIONS: The second way to allocate total TACS storage is with a
request for relative sizing. That is, rather than request a specific number of function blocks,
supplemental variables, etc., one can simply request a proportional allocation (e.g., 5% of the total
storage for function blocks, etc.). There first is the special-request card, then a single data card
bearing the 8 proportions:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
LT1
LT2
LT3
LT4
LT5
LT6
LT7
LT8
I8
I8
I8
I8
I8
I8
I8
I8
4 - 27

But one problem with this second option is that the user may not have a feel for the relative space
taken by different tables. To aid the user, we present an approximate correspondence between
nominal absolute dimensions and the proportions which produced them:
LIST
LT1
LT2
LT3
LT4
LT5
LT6
LT7
LT8
20
90
100
20
30
250
300
60
11
15
21
28
ABSOLUTE DIM.
RELATIVE DIM
(%)
Table 1: Approximate Correspondence between Absolute and Relative TACS dimensions

In the absence of special request card for TACS dimensioning, the EMTP sets dimensions according
to the following default absolute dimensions:
LT1
LT2
LT3
LT4
LT5
LT6
LT7
LT8
20
90
100
20
30
250
300
60
Depending upon the size of the List 19, execution may or may not be possible (i.e., even with
default dimensioning, there still might not be sufficient table space for execution).
TACS EMTP SOURCES: As explained in Section 10, the source types 1 through 10 are reserved
for functions which are directly defined by the user. If one or more such source functions is to be
defined within TACS as a signal whose value is determined by the user-defined TACS data, then the
following special request card must precede input of the miscellaneous data cards:
1
2
3
4
5
6
7
8
12345678901234567890 123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
TACS EMTP SOURCES
10
A6
A6
A6
A6
A6
A6
A6
A6
A6
A6
Names of TACS variables which control electric network sources having type-codes 1-10
First, the special text "TACS EMTP SOURCES" is to be entered in columns 1-17 of the card. Then, if
an EMTP electric-network voltage or current source of the type-code number K (for K between one
and ten) is to be controlled by TACS, the 6-character name of the controlling TACS variable is to be
entered in the K-th name-field of the card. Note that any name entered on this card must be
defined as part of the TACS data specification of Section 14. (See also simpler source type 60-99 in
Section 10).
4 - 28

TACS INIT: This option is provided to allow cross-checking between the initialization algorithms
used in versions 2 and 3 of the EMTP. If the TACS INIT request card is used, initial conditions of
TACS data cases will be similar to those computed in version 2.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
TACS INIT
or
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
TACSIN
TACS ORDER: This option is provided to allow cross-checking between the results of versions 2
and 3. With the special request "TACS ORDER", the ordering of TACS follow that of version 2.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
TACS ORDER
or
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
TACSOR
POSTPROCESS PLOT FILE: TACS can be used for the postprocessing of plot files using the
following procedure: First connect the old plot file (input to the post processor) to LUNIT2 using
$OLDFILE:
$OLDFILE,
<plot file specification> , <LUNIT2>
Note that exact formats cannot be indicated, due to the installation-dependent nature. Users
should consult the specific information for their computer systems. As for LUNIT2, this is to be the
I/O unit number that is assigned to variable LUNIT2. This will be 2 unless a system-dependent
assignment within module "SYSDEP" alters the natural assignment. Variables in this plot file will
4 - 29

be assigned to user-defined TACS sources in natural order by the EMTP, thereby providing the
connection for postprocessing by the user.
Next comes the key request, as shown below. This is a universal card. Here "IPLOT" is to be an
integer indicating frequency of the output, not unlike the miscellaneous data parameter of the same
name. For example, if IPLOT = 3 is used, then only every third point of the old plot file shall be
used. The most common (and least tricky) case is for IPLOT = 1, so that there are as many output
points as there are input ones.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
POSTPROCESS PLOT FILE, IPLOT
The data case is completed by cards for a TACS-only (TACS STAND ALONE) EMTP data case. There
are generally to be as many TACS sources as there are variables in the old data file, with 6-character
names being arbitrary. In the order defined, these are automatically (internally) connected to
variables of the old plot file. Most post-processing will involve supplemental variables, though
function blocks are also useful (e.g., 1/S for integration of voltage to give flux). This step size is
the step size used to generate the plot file, times the parameter IPLOT then used.
4.1.6
Universal Machine
ABSOLUTE U.M. DIMENSIONS: The universal machine (U.M.) component of Section 10.12 has
total working space for all tables assigned in List 25 of the overall program variable dimensioning
(see Section 2.3). However, the way in which the total working space is allocated among the four
U.M. tables can be defined by the user.
The first way to allocate total U.M. storage is with a request for absolute U.M. table sizing:
1
2
3
4
5
6
7
8
12345678901234567890123456789012 34567890 12345678 90123456 78901234 5678901234567890
A32
NCLFIX
NUMFIX
IOTFIX
IBSFIX
I8
I8
I8
I8
4 - 30
NCLFIX
Maximum total number of U.M. coils in the data case. This is the total for all
components involved.
(33-40)
U.M.
NUMFIX
Maximum number of U.M. components for the data case.
(41-48)
IOTFIX
Maximum number of U.M. output quantities (total for all machines).
(49-56)
IBSFIX
(57-64)
Maximum number of U.M. 6-character alphanumeric names (total for all

machines).
Note that the total storage represented by this user request must be less than or equal to the List 25
(see Section 2.5.1). Otherwise, an EMTP error stop will result. In either case, the minimum size
for List 25 will be printed out as part of the column 1-50 interpretation of the Type-19 request card
for U.M. modelling.
RELATIVE U.M. DIMENSIONS: The second way to allocate total U.M. storage is with a request
for relative sizing. That is, rather than request a specific number of coils, machines, etc., one can
simply request a proportional allocation (e.g., 50% of the total storage could go for coils, etc.).
The format for such an allocation is:
1
2
3
4
5
6
7
8
12345678901234567890123456789012 34567890 12345678 90123456 78901234 5678901234567890
A32
JNCL
JNUM
JIOT
JIBS
I8
I8
I8
I8
We still use integers, note; and the four data fields correspond one to one with the four tables as
described for "ABSOLUTE U.M. DIMENSIONS". But here the meaning is a fraction of total available
space for that table. In order to use this effectively, one must know the multiplicity of the four
tables:
REAL
19
INTEGER
14
ALPHANUMERIC
For a computer with half-length integers (e.g., IBM and VAX which use REAL*8 and INTEGER*4
translations), the overall byte weighting is: (6, 26, 1, 1). On the other hand, for a word machine
4 - 31

where variables all have the same length, the overall word weighting is: (8, 33, 2, 1). Using
whichever of these is appropriate, multiply the corresponding absolute dimensions by these
components in order to produce the corresponding relative dimensions.
As an example of this, suppose that absolute dimensions that are deemed to be balanced are the
default ones of the EMTP: (20, 3, 50, 60). Operating on a VAX, the weighting (6, 26, 1, 1) then
produces the associated relative dimensions (120, 78, 50, 60). If the data card "RELATIVE U.M.
DIMENSIONS, 120, 78, 50, 60" were then placed at the top of a data deck with involved U.M.
modelling, these numbers would never have to be altered, no matter how many machines were
involved. Of course, if there were too many machines for the available memory, there would be an
overflow error stop. But this could be corrected by redimensioning the EMTP with larger List 25,
without touching the "RELATIVE U.M. DIMENSIONS" card.
If the user fails to input one or the other of the above two cards which specify U.M. table sizing,
then the EMTP allocates absolute dimensions in "UMOFFS" of overlay 5 which equal (20, 3, 50, 60).
These, then, are default dimensions.
4.1.7
Load Flow
FIX SOURCE
old EMTP load flow
FIX SOURCE1
new improved EMTP load flow
LOAD FLOW
three phase load flow
FIX SOURCE: The "FIX SOURCE" feature of the EMTP is used to obtain steady-state phasor
solutions for initial conditions, while observing power constraints on one or more buses. This
option is functionally similar to that of a multi-phase "load flow" program, albeit in somewhat
restricted form (details of this feature are given in Section 10.13).
To use the load flow feature, a special request card with "FIX SOURCE" entered in columns 1-10
must precede input of the floating-point miscellaneous data card.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
FIX SOURCE
FIX SOURCE1: The "FIX SOURCE" feature of the EMTP is used to obtain steady-state phasor
solutions for initial conditions, while observing power constraints on one or more buses. This
option is functionally similar to that of a multi-phase "load flow" program, albeit in somewhat
restricted form (details of this feature are given in Section 10.14).
4 - 32

To use the load flow feature, a special request card with "FIX SOURCE1" entered in columns 1-11
must precede input of the floating-point miscellaneous data card. Full details of this feature are
given in Section 10.14.
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
FIX SOURCE1
LOAD FLOW: The "LOAD FLOW" feature of the EMTP is described in Section 21. The use can use
either:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
LOAD FLOW
or
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
LF3
4.1.8
Advanced Options
FILE REQUEST
RENUMBER BYPASS
FREE FORMAT
MODIFY SWITCH LOGIC
STEP ZERO COUPLE
subnetwork ident. with all switches closed

HIGH RESISTANCE
redefine R of internally added resistor
POWER FREQUENCY
4 - 33
File Request: The principal installation-dependent module that is called at the start of
execution of each data case is "SYSDEP". But there is yet another such module in overlay 1 which
is used for initialization (see Section 2.1): "MIDOV1". For some computer systems, for certain
special types of data cases, it may be necessary or desirable to have "MIDOV1" called at a certain
point during the data input This is possible by means of the special request card "FILE REQUEST".
Do not use this without knowledge.
1
2
3
4
5
6
7
8
123456789012 34567890123456789012345678901234567890123456789012345678901234567890
FILE REQUEST
RENUMBER BYPASS: Sparsity-oriented node renumbering is normally used on the coefficient
matrix [Y] of the time-step loop in order to preserve sparsity during the triangular factorization of
[Y]. It is possible to bypass or omit node renumbering by means of the following special-request
card (which must precede the miscellaneous data cards):
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
RENUMBER BYPASS
In such a case, nodes will remain numbered in the order that they will be encountered upon the
reading of EMTP branch data. On any one branch card, it is the left node (data field "BUS1" of
columns 3-8) which is processed before the right node (data field "BUS2" of columns 9-14).
Actually, more than just a bypass of transient network renumbering is involved. The full [Y] is
retriangularized whenever [Y] changes (e.g., when a switch changes status). Switch nodes and
pseudo-nonlinear element nodes are not necessarily eliminated after unknown voltage nodes which
are not in this class. There is no partition-a triangularization outside of the time-step loop in this
case. This is the whole rationale behind the use of the "RENUMBER BYPASS" feature there is no
artificial forcing of switch and pseudononlinear-element nodes toward the bottom of the matrix.
Use of the "RENUMBER BYPASS" feature is of arguable value and rarely use. It was originally
conceived to handle special problems. One such example is a full AC/DC converter representation
which might require 50 or 100 EMTP switches to represent the valves.
FREE FORMAT: As explained in Section 3.7, certain classes of EMTP data may be entered either
according to the conventional fixed-format specifications, or using the free-format option. This
latter mode makes use of two special alphanumeric characters: one which separates data fields,
and one which requests a continuation card. Should the user want to re-define one or both of these
4 - 34

characters, a data card in the following format should be entered, and should precede the
1
2
3
4
5
6
7
8
1234567890123456 78901234 56789012 345678901234567890123456789012345678901234567890
FREE FORMAT
A16
CSEPAR
(17-24)
CHCONT
(25-32)
Note:
CSEPAR
A8
CHCONT
A8
The single alphanumeric character which is to serve as separator

between adjacent data fields. Enter this character left-adjusted in the
field shown (i.e., in column 17). Mnemonically, the name signifies
"Character SEPARator". If left blank, a comma (",") will be
assumed; if no such redefinition card is present, a comma will be
assumed.
The single alphanumeric character which is to serve as a request
card for a continuation data card. Enter this character left-adjusted
in the field shown (i.e., in column 25). Mnemonically, the name
signifies "CHaracter CONTinuation". If left blank, a "$" sign will be
assumed. Remember that this must be a special character which the
user will never employ as part of his non-comment EMTP data cards.
If all data cards of the data case in question are to be read using fixed formats, enter
"9" (nine) as "CHCONT". By definition, this will bypass all attempts at free-format data
reading (except for the free-format FORTRAN expression in TACS), and will thus speed
up input slightly. It will also guard against possible confusion of the fixed-format data
with free-format recognition rules.
The free-format option is available for most EMTP data structures. But there are important
exceptions, so it should be used with caution.
The appearance of either "CSEPAR" or "CHCONT" characters (which have default values "," and
"$", respectively) on a data card determines whether or not the card uses free format. Therefore, it
is important that these two characters not be buried in bus names, or TACS variable names.
MODIFY SWITCH LOGIC: This is one of those EMTP options of questionable value which
remain in the program because it is probably easier to leave in rather than remove it. The purpose
of this feature is to use a non-standard way to handle switch logic internally.
4 - 35
1
2
3
4
5
6
7
8
12345678901234567890 123456789012345678901234567890123456789012345678901234567890
MODIFY SWITCH LOGIC
STEP ZERO COUPLED: With this special request card "STEP ZERO COUPLED", the program will
carry out automatic subnetwork identification with all the non-grounding switches closed at t=0;
and, no further subnetwork identification work would be done during later switching operations in
the time-step loop. Without this special request, the program will determine the subnetwork
identification dynamically at each time-step that a switching operation has occurred. More
information on subnetwork identification can be found in Reference 21, Vol. XIV, pagination
CUMA-30 to 33.
1
12345678901234567
2
3
4
5
6
7
8
890123456789012345678901234567890123456789012345678901234567890
STEP ZERO COUPLED
REDEFINE TOLERANCE EPSILN:
1
2
3
4
12345678901234567890123456789012 34567890
A32
5
6
7
8
1234567890123456789012345678901234567890
EPSILN
E8.0
The familiar floating-point miscellaneous data parameter "EPSILN" (see Section 4.2.1) is used far
more than for the stated purpose of checking matrix singularity. Often the EMTP wants to know
whether a floating-point result is getting small, and "EPSILN" (possibly scaled by a power of ten) is
usually used as the standard of comparison. The default value is set in installation-dependent
module "SYSDEP" (see Section 2.1), and typically has a value of 1.E-8 for REAL*8 versions.
HIGH RESISTANCE:
1
2
3
4
12345678901234567890123456789012 34567890
HIGH RESISTANCE
A32
5
6
7
8
1234567890123456789012345678901234567890
KPARTB
I8
4 - 36

Type-99 pseudo-nonlinear elements (see Section 8.2) have large resistance internally added,
always. The resistance is normally based on the near-zero round-off tolerance "FLZERO" which is
defined within the installation-dependent module "SYSDEP" (see Section 2.1). For example, the
REAL*8 versions for IBM and VAX will have FLZERO = 1.E-12. The EMTP uses resistance value
RINF = .01/FLZERO ohms, or 1.E+10 for REAL*8 machines. By means of the "HIGH RESISTANCE"
request card shown above, the default value of RINF can be replaced by 10**KPARTB. The user
thus has control within a power of ten.
POWER FREQUENCY:
1
2
3
4
5
6
7
8
123456789012345 67890123456789012 34567890 1234567890123456789012345678901234567890
STATFR
POWER FREQUENCY
A15
E8.0
Variable "STATFR" is the synchronous power system frequency, in Hertz. The default is set within
the installation-dependent module "SYSDEP" of overlay 1, and it is normally 60 Hz. Yet special
cases can be of interest (e.g., the 400Hz usage of the aircraft industry, or the usage of both 50Hz
and 60Hz within Japan). In any case, if the user wants to change variable "STATFR", he needs only
punch a card according to the above format; the text "POWER FREQUENCY" is to be punched in
columns 1-15, and the desired new power frequency goes in columns 33-40 as E8.0 information.
If the user chooses to alter two or more such parameters, the order of the requests is immaterial. If
the data case in question is a "REPLOT" case, the above-described parameter redefinition cards (if
any) must precede the "REPLOT" card, since immediately thereafter, program control will be
transferred to plotting.
ZINC OXIDE: Control parameters of the Newton-Raphson iterative solution are specified via a
special request word to be read in arbitrary order prior to the miscellaneous data cards. The
request word "ZINC OXIDE" (or just "ZO" in abbreviated form) is used.
1
2
3
4
5
6
7
8
1234567890123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
ZINC OXIDE
A16
MAXZNO
I8
EPSZNO EPWARN EPSTOP ZNOLIM1 ZNOLIM2

E8.0
E8.0
E8.0
4 - 37
E8.0
E8.0

MAXZNO
(17-24)
EPSZNO
(25-32)
EPWARN
(33-40)
EPSTOP
(41-48)
ZNOLIM(1)
(49-56)
ZNOLIM(2)
(57-64)
Maximum number of Newton iterations allowed for the

solution of the surge arresters in each sub-network. The
Newton iteration will go this long unless tolerance EPSZNO
is attained first.
Convergence tolerance for the Newton iteration. All
voltage corrections of the coupled elements must be
smaller than this in absolute value for the equations to be
judged solved. A blank (zero) value is taken as a request
for the value EPSILN Vref, where EPSILN is the familiar
floating-point miscellaneous data parameter of
Section 4.2.1. For a comment about VREF see the
description of ZNOLIM(1) below.
Voltage convergence tolerance for non-fatal warning
message about a "sloppy solution." A blank (zero) data
field is taken as a request for the value Vref / 1000
representing one tenth of one percent of rated voltage,
which is about the limit of graphical resolution.
Voltage convergence tolerance for a fatal error stop, complaining about a non-converged iteration. A blank (zero)
data field is taken as a request for the value Vref / 10 or
10% of rated voltage. The current error in such a case
could be very high due to the extreme nonlinearity (e.g.,
1.126 = 11.9).
The maximum per unit (based on VREF) voltage correction

at each iteration step. A blank or zero is given the default
value of 1.0. Note that the control of this parameter may
help resolve difficult convergence cases. For Type-91
(3333.) and Type-92 (4444.) elements, VREF is generated
by the program.
The maximum per unit based on VREF arrester voltage
during the iterations. A blank or zero is given the default
value of 1.5. This parameter applies to the ZnO arrester
model only and may be manipulated by the program if it
causes an oscillatory solution.
Note that data in this card can also be specified in free format (i.e., with commas separating all data
fields). See also Section 8.6.7.
4 - 38
4.2 Miscellaneous Data Parameter Cards

4.2.1
First Miscellaneous Data Card (Floating-point Miscellaneous

Data)
The first non-comment card which fails to be recognized as any of the preceding special request
types will be taken to be the beginning of "regular data" for a conventional EMTP data case. This
regular data begins with two miscellaneous data cards: first one card for floating-point parameters,
and then one for integers, as follows:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 789012345678901234567890
DELTAT
TMAX
XOPT
COPT
EPSILN
TOLMAT
TSTART
E8.0
E8.0
E8.0
E8.0
E8.0
E8.0
E8.0
where:
DELTAT
(1-8)
TMAX
(9-16)
XOPT
(17-24)
COPT
(25-32)
The time-step size t of the numerical integration, in sec. The solution

will be calculated at time instants with this spacing: t=0, t, 2t, etc. in
seconds.
The termination time Tmax of the study, in seconds. The simulation
covers time interval O < K t < Tmax.
Parameter indicating whether inductance or inductive-reactance is to be
inputted on all linear branch cards:
=0
All inductance values on branch cards are interpreted as L in

mH (millihenries).
=f
All inductance values on branch cards are interpreted as

L = 2fL in ohms at frequency f = XOPT.
Parameter indicating whether capacitance or capacitive reactance is to

be inputted on all linear branch cards:
=0
All capacitance values on branch cards are interpreted as C in

microfarads.
=f
All capacitance values on branch cards are interpreted as

C = 2fC in micromhos (S) at frequency f = COPT.
4 - 39

EPSILN
(33-40)
Near-zero tolerance, used for checking singularity of the real

coefficient matrix [G] within the time-step loop, where
[G]vnode(t) = inode(t)
A blank or zero field is given a default value which is dependent upon

the translation being used. The 64bit (REAL*8) versions are
commonly given default values of 1.E8.
TOLMAT
(41-48)
TSTART
(49-56)
4.2.2
A second near-zero tolerance, used for checking singularity of the

steady-state complex admittance matrix [Y]. A blank or zero field is
given default value equal to 'EPSILN'.
The beginning simulation time. This will normally be zero (the data
field can be left blank). If a previously-solved data case is to be
restarted, however, then this field must be contain the terminal (ending)
time of that preceding partial simulation.
Second Miscellaneous Data Card (Integer Miscellaneous Data

Card)
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 56789012 34567890
IOUT
IPLOT
IDOUBL
I8
I8
I8
IOUT
KSSOUT
I8
MAXOUT
IPUN
I8
I8
MEMSAV
I8
ICAT
I8
NENERG
IPRSUP
I8
I8
Parameter providing control over the frequency of the printed output:
(1-8)
= 0 or 1
Output quantities are printed for every time-step.
=K
For K > 1, results are printed for every K-th step only, that is,
for times t=0, (Kt), (2Kt), (3Kt), etc. If ICAT = 3, then
IOUT controls the frequency at which points are dumped into
the MATLAB post-processing file (see Section 15.4).
4 - 40

IPLOT
IDOUBL
(17-24)
KSSOUT
(25-32)
Parameter providing control over the frequency of the plotted

transients simulation output:
= 0 or 1
Every computed point will be used for plotting.
=M
For M > 1, only every M-th computed point will be used for
plotting. This saves plotting time and reduces storage
requirements if, for some reason, a smaller step-size is
needed for solution than for graphical output. If the user
should punch an even value for M, the program will
automatically increment it by one, to make it odd. This is
because even values will produce deceptive results for an
oscillating curve, graphing only the upper or lower
envelope.
= 1
No plotting is possible, as no permanent file of plot values is

created. Use this option if no plots are desired, in order to
speed execution.
Parameter controlling the printing of a network connectivity table,

showing how branches and switches topologically interconnect the
network busses.
=0
No such output is provided.
=1
A printout of network connectivity is provided. For each

node, a list is given of the nodes to which it is connected
through physical branches and switches. Mutual coupling
between phases of multiphase elements is ignored in this
output, as is the capacitance to ground of pi-circuits and
distributed-parameter lines. Since the node name " " (6
blank characters) which normally signifies ground would
not show up in print, the word "TERRA" is used in the
connection list.
Parameter controlling the printout of all linear-branch line flows,

switch flows, and all voltage-source injections, for the steady-state
phasor network solution:
=0
No steady-state solution printout.
=1
Print complete steady-state solution (branch flows, switch

flows, source injections).
=2
Print switch flows, source injections; but not branch flows.
=3
Print branch flows requested by column 80 punches, switch

flows and source injections.
4 - 41

Parameter controlling the printout of maximum absolute
values of all output variables, the maximum which occurred
during the simulation:
MAXOUT
(33-40)
=0
No such output.
=1
Print the maxima, as calculated using every time step.
=2
Print the maxima, as calculated using only the time steps ti

for which the output solution vector must be formed for
printing or plotting. If the user's branch cards carry a power
or energy-output request, every time-step is automatically
used though.
Parameter controlling the punching of terminal conditions of
the simulation (at time t = TMAX):
IPUN
(41-48)
MEMSAV
(49-56)
=0
No such punching.
=1
Node voltages and branch currents (plus capacitor voltage

on series R-L-C branches) at the last time step are punched
on cards; these can be used as initial conditions for other
studies starting from this state of the system. The present
version of the program does not punch the branch currents
for branches with distributed parameters; in such cases a
warning message is printed:
= 1
An extra, special card will immediately follow this second

miscellaneous data card, to vary the printout frequency as
described in Section 4.3.2. See also Section 4.1.2.
Parameter controlling the dumping of memory onto disk at the end of

the simulation (time TMAX), for subsequent use with the "START
AGAIN" request of Section 4.1.1.
=0
No such memory saving will occur.
=1
Yes, save the memory contents for later use.
Restriction: The "START AGAIN" reactivation of a previouslyhalted simulation can only be guaranteed if the same EMTP version
with the same dimensions is used.
ICAT
(57-64)
Parameter which controls the permanent disk storage of plot data

points which are accumulated on logical unit number 4 during the
simulation:
=0
Plot (pl4) file will not be created, though conventional

batch-mode plotting at the conclusion of the solution
process is possible.
4 - 42
NENERG
(65-72)
=1
Plot (pl4) file will be created. No conventional batch-mode

plotting of the results in this case will be allowed (the EMTP
will skip over plot cards, discarding them until it finds the
blank card which terminates plot cards).
=2
Combination of the above. That is, the plot data file is saved
on disk for later reuse, plus, conventional batch-mode
plotting requests are honoured.
=3
Export data in MATLAB form (see Section 15.4).
Control parameter related to the running of multiple-energization

studies (either "STATISTICS" or "SYSTEMATIC").
=0
Conventional, single, deterministic EMTP simulation is

desired; no dice rolling!
=K
For K > 0, this is a request for a statistical study of this data

case, in which K energizations (switch closings) will be
performed.
For K < 0, this is a request for a systematic study of this data
case, in which K energizations will be performed.
See Section 9 for a complete discussion of such studies
(both "STATISTICS" and "SYSTEMATIC"). Should "NENERG "
be punched nonzero, an added special statistics or
systematic miscellaneous data card must follow, giving
necessary scalar parameters for the study (see
Section 4.3.1).
IPRSUP
(73-80)
Parameter controlling the output of program diagnostic printout.

Production runs should always leave this field blank (or punch a zero).
=0
There will be no diagnostic printout, unless such was

selectively requested using the "DIAGNOSTIC" key-word card
of Section 4.1.2.
=M
For increasingly-positive "M", more and more diagnostic

printout will be generated. Such a positive "M" overrides
any selectively-requested diagnostic printout (key-word
"DIAGNOSTIC").
4 - 43
4.3 Specially-Requested Extensions To Miscellaneous Data

Cards
There are two fields on the integer miscellaneous data card which can require the Input of
additional data, data which is to immediately follow the integer (2nd) miscellaneous data card:
(A)
If 'NENERG' of columns 65-72 is nonzero, a special miscellaneous data card

must immediately follow the integer miscellaneous data card, for this
"STATISTICS" or "SYSTEMATIC" case.
(B)
If 'IPUN' of columns 41-48 has value -1, the preceding card (if any) is to be
followed by a card which varies the printout frequency.
The format and meaning of these additional cards is as described in the subsections which follow.
4.3.1
Extra "STATISTICS" or "SYSTEMATIC" Miscellaneous Data Card

(Optional)
If and only if parameter 'NENERG' of the integer miscellaneous data card has a positive value, the
following additional special "STATISTICS" miscellaneous data card is required:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 56789012 34567890
ISW
ITEST
I8
IDIST
AINCR
XMAXMX
DEGMIN
DEGMAX
STATFR
SIGMAX
I8
F8.0
F8.0
F8.0
F8.0
F8.0
F8.0
I8
where:
ISW
=1
Request for the printed output of all variable switch closingopening times, for every energization.
=0
No printed output of variable switching times will occur.
=0
An extra, randomly-selected offset time, calculated using the

parameters "DEGMIN", "DEGMAX", and "STATFR", is to be added
to the randomly-generated switching times, for each
energization.
=1
No such additional offset is added.
(1-8)
ITEST
(9-16)
4 - 44
NSEED
I8
IDIST
=2
This extra offset is added to the randomly-generated switch

closing times for each energization, but not to the random switch
opening times.
=3
This extra offset is added to the randomly-generated switch

opening times for each energization, but not to the random
switch closing times.
=0
All randomly-generated switch closing times (for "STATISTICS"

switches) will have Gaussian (normal) distribution.
=1
All randomly-generated switch closing times will have a

uniform distribution.
(17-24)
AINCR
(25-32)
The post-solution statistical tabulation of overvoltages, following the

solution of all "NENERG" energizations, will use a voltage
discretization increment of "AINCR" in per unit. A zero or blank field
will be given default value AINCR=0.05 per unit.
Note: Variable "AINCR" is also used as a flag for the request of

additional printout by the EMTP. Add a bias of "55." in order to also
produce the familiar base-case printout of minima and maxima (along
with their associated times of occurrence) for each energization. This
printout is in addition to, and not a replacement for, the regular
printout that occurs for each energization.
XMAXMX
(33-40)
DEGMIN
(41-48)
DEGMAX
(49-56)
The maximum per unit overvoltage that the counting algorithm will
consider. This is a per unit bound on the tabulated overvoltage
distributions which will be outputted, with a default value of 2.0
assumed if the user should input a zero or blank value. A "-" sign
on XMAXMX means that the built-in random number generator in the
module "SANDMN" is going to be used.
If statistics parameter "ITEST" of columns 9-16 of the same card is
zero or blank, the additional random time which is to be added to the
randomly-generated switching times is produced from these three
parameters for each energization using the following formula:
STATFR
(57-64)
offset
1 ( STATFR )
- { ( 1 ) ( DEGMIN + DEGMAX ) }
= -----------------------------------360
where is a random number which is uniformly distributed over the

unit interval (0.0, 1.0). "DEGMIN" and "DEGMAX" are the minimum
and maximum angles in degrees which define a "window" of a sine
wave having frequency "STATFR" Hz., from which the added time is
randomly selected according to a uniform distribution. As an
illustrative sample sketch, consider the following example:
4 - 45
window
DEGMIN = 30
DEGMAX = 120
STATFR = 60 Hz
30
120
360 16.667 ms
Blank or zero values for fields "DEGMIN" and "DEGMAX" are given
default values of zero and 360 degrees, respectively; the corresponding
default value for a blank or zero "STATFR" is 60.
The same options are used for ITEST = 2 and/or ITEST = 3.
SIGMAX
(65-72)
NSEED
(73-80)
For a Gaussian (normal) distribution, "SIGMAX" gives the point at

which the "tails" of the distribution will be truncated, for calculating
the random switch times. "SIGMAX" is entered as a multiple of the
standard deviation, and is symmetric about the mean. A zero or blank
field will be given a default value of 4.0, which means that no random
switch times further removed than 4 from the mean will be allowed.
Parameter which controls whether or not the random switch times
will be identical, should a given "STATISTICS" data case be rerun a
second time:
=1
a re-run of the data case will produce identical answers; the

randomnumber generating algorithm uses a constant seed
which does not depend on the time of day.
=0
a re-run of the data case will produce different answers; the seed
does depend on the time of day.
The re-initialization of the random number generator is done for each

energization.
If and only if parameter "NENERG" of the integer miscellaneous data card has a negative value, the
following additional special "SYSTEMATIC" miscellaneous data card is required:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 1234567890123456789012345678901234567890
ISW
I8
ITEST
AINCR
XMAXMX
I8
F8.0
F8.0
4 - 46

Same meaning as for "STATISTICS" immediately above.
ISW
(1-8)
ITEST
=0
The inputted switch-time TCLOSE, which is read from a

"SYSTEMATIC" switch card, is understood to be the mean switch
time. The variation will be around such centre values then.
=1
The inputted switch-time TCLOSE, which is read from a

"SYSTEMATIC" switch card, is understood to be the minimum or
beginning switch time. The variation will be for times greater
than this then.
(9-16)
AINCR
(25-32)
XMAXMX
(33-40)
If and only if parameter "NENERG" of the integer miscellaneous data card has a positive value, and
field "ISW" of the "STATISTICS" miscellaneous data card is punched with the integer 4444, then a
special test of the random switch times is produced. In this case, no simulations are actually
produced. Instead, the random switch closing times are just generated, and subsequently tabulated
statistically, so as to compare the observed distribution with the expected theoretical distribution.
For this case where "ISW" is punched with value 4444, two other fields of the "STATISTICS"
miscellaneous data card take on special meaning:
1
2
3
4
5
6
7
8
12345678 9012345678901234 56789012 34567890123456789012345678901234 56789012 34567890
ISW
I8
ISW
AINCR
SIGMAX
F8.0
F8.0
Punched with special flag 4444, in this case.
(1-8)
AINCR
(25-32)
SIGMAX
(65-72)
The sample and theoretical cumulative distribution functions will be

tabulated at a number of uniformly-spaced time instants. The timespacing of this tabulation is "AINCR", as a multiple of the standard
deviation of the switch in question.
The range of the just-described tabulated comparison is "SIGMAX" on
either side of the mean closing time, as a multiple of the standard
deviation.
4 - 47

A number of comments might be appropriate, as to usage of this diagnostic feature:
Comment 1: Characteristic printout of this feature is a series of tabulations for different pairs
of switches. Recall that if closing times for switch "A" and switch "B" are both of Gaussian
(normal) distribution, then the difference will also be Gaussian and will have mean and standard
deviation given by:
T = TA TB
and
2
2
= A + B
It is this difference of closing times which is checked, and this is not the same as just looking at the
closing times of any one switch. Recall that no random number generating algorithm is truly
random; the K+1st number is a function of the Kth number, somehow (usually). Since there is only
a finite number of different bit permutations within a computer word, every generating algorithm
must eventually repeat. Presumably the period is so large that it is of no practical interest for EMTP
usage.
For M Gaussian switches, there will be M(M-1)/2 such pairings of switchings. For each, there will
be one tabulated output, of which the following (extracted from the benchmark DC50) are typical:
SWITCH PAIR 2
C1
TO CSW1
AND
A1
TO ASW1
TIME
-3.5000 -3.0000 -2.5000 -2.0000 -1.5000 -1.0000 -0.5000 0.0000 0.5000
2.5000 3.0000
SAMPLE
0.0000 0.0020 0.0120 0.0340 0.0740 0.1680 0.2960 0.4960 0.6720
0.9960 0.9980
THEORETICAL
0.0000 0.0013 0.0062 0.0228 0.0668 0.1587 0.3085 0.5000 0.6915
0.9938 0.9987
TIME
SAMPLE
THEORETICAL
3.5000
1.0000
0.9998
3.5000
1.0000
0.9998
1.5000
2.0000
0.8380
0.9240
0.9800
0.8413
0.9332
0.9772
1.0000
1.5000
2.0000
0.8700
0.9100
0.9160
0.8413
0.9332
0.9772
4.0000
1.0000
1.0000
SWITCH PAIR 3
C1
TO CSW1
AND
B1
TO BSW1
TIME
-3.5000 -3.0000 -2.5000 -2.0000 -1.5000 -1.0000 -0.5000 0.0000 0.5000
2.5000 3.0000
SAMPLE
0.0000 0.0000 0.2000 0.0140 0.5000 0.2140 0.3380 0.4960 0.7320
0.9900 1.0000
THEORETICAL
0.0000 0.0013 0.0062 0.0228 0.0668 0.1587 0.3085 0.5000 0.6915
0.9938 0.9987
TIME
SAMPLE
THEORETICAL
1.0000
4.0000
1.0000
1.0000
Concerning units, it will be noted that everything has been normalized. The printed "TIME" row
gives the number of standard deviations from the mean. The corresponding actual time in seconds
could be found from:
t k = T + TIME ( K )
Entries in the "TIME" row should normally vary from -SIGMAX to +SIGMAX. Note that one
compartment is missing on the left, since tabulations correspond to the right edge of the
compartment.
4 - 48
Comment 2: The tabulation described above applies only to Gaussian (normal) switchclosing time. If one or more switches happen to be either deterministic (non-STATISTICS) or
uniformly-distributed, they are ignored for purposes of this special tabulation.
Comment 3: Recall that miniature printer plots of individual switch closing times are
outputted at the end of the printout for a conventional "STATISTICS" solution. See Section 12.4 for
an example. Such plots of switch closing times will also end the run being considered here. All
parameters and assumptions related to this output remain unchanged.
Comment 4: A complete EMTP data case must be used, at least through the blank card which
terminates switch cards (see Section 3 for the structure of EMTP data cases). EMTP source cards
and all later cards of a conventional case are optional; if so supplied by the user, they will be
skipped by the EMTP as it searches for a "BEGIN NEW DATA CASE" record (just as if an EMTP error
stop had occurred).
4.3.2
Card for Varying the Printout Frequency (Optional)
If and only if parameter 'IPUN' of the second miscellaneous data card is set to 1, the following
card should follow immediately after the second miscellaneous data card:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 345678901234567890123456789012345678901234567890
First Change
KCHG
MULT
I8
I8
Second Change
KCHG
I8
MULT
Etc.
I8
This card specifies at what time-step numbers the printout frequency is to be changed, and what
values the frequency is to be changed to. Up to five pairs of step numbers and new printout
frequencies are permitted, as per the following definitions:
KCHGI
Time-step number at which the i-th variation in the printout frequency

is to begin. Printout for this particular step number is always provided,
as a beginning to the new frequency of output.
MULTI
Modified value of "IOUT" (see second miscellaneous data card,

Section 4.2.2), to begin at time-step number KCHGi.
4 - 49

Pairs must be ordered corresponding to increasing time (increasing "KCHG"). No bounding entry is
needed, as the last nonzero pair (if less than the full five pairs) simply remains in effect from the
time it commences until the end of the study. The first "KCHG" value, read from columns 1-8, must
be positive.
Special Time-Control of Diagnostic Printout:

Minus signs can be appended to the MULTi values as defined above, if the user wants to control the
diagnostic printout of the time-step loop as a function of time, rather than as a function of time step
number. Recall that diagnostic printout will produce output for every time step; if a large number
of steps are taken, astronomical quantities of paper will be consumed -- unless such printout is
selectively controlled as a function of time. The following scheme is used:
(A)
If a minus sign is appended to MULT(J), then at step number KCHG(J) the

diagnostic printout status will be changed. That is, there will be a switching of
the two sets of diagnostic printout control parameters:
1.
(IPRSOV(K), K=16, 19) of Section 4.1.2, DIAGNOSTIC;
2.
(IPR(K), k=1, 4) of Section 4.1.2.,

(B)
If there is no minus sign, there will be no change in status of the diagnostic

printout at the associated time step.
If one or more minus signs are used as just described, it is only the absolute value of MULTi which
controls the production (non-diagnostic) printout frequency as originally described. That is, the
use of minus signs does not interfere with the original usage.
4.4
TACS
Data Cards (if any)
stands for "Transient Analysis of Control Systems". In very general terms, it directly
provides the user with modelling capability, as fully documented in Section 14. If such modelling
is involved in a data case, this present mention is just a reminder that such data is to be physically
positioned at this point in the data deck before the first EMTP branch card, but after the
miscellaneous data cards (and any extensions).
TACS
The beginning of TACS data is flagged by means of one of the two following special request cards:
4 - 50
IFLAG
1
2
3
4
5
6
7
8
123456789012345678 90 123456789012345678901234567890123456789012345678901234567890
TACS STAND ALONE
A18
Columns 21 - 80 available for additional comments
I2
This card introduces the TACS data cards describing a system modelled in TACS without any
associated electric network components. The data case will look as follows:
TACS STAND ALONE
(See
TACS cards
Section 14)
plotting cards (See Section 15)
BEGIN NEW DATA CARDS
Variable IFLAG of columns 19-20 specifies the ordering the case is solved:
IFLAG=
1 case is solved according to the user supplied ordering
IFLAG=
0 case is solved based on the ordering given in Section 14.2
IFLAG=-1
case is solved opposite to the ordering user inputted
TACS HYBRID
A18
IFLAG
1
2
3
4
5
6
7
8
123456789012345678 90 123456789012345678901234567890123456789012345678901234567890
Columns 21 - 80 available for additional comments
I2
This card introduces the TACS data cards describing a system modelled in TACS and including
components interfaced with associated components of the electric network. The data case will
look as follows:
TACS HYBRID
TACS cards
(See Section 14)

electric network cards (See Sections 6-10)
plotting cards
BEGIN NEW DATA CASE
4 - 51

Variable IFLAG of columns 19-20 specifies the ordering the case is solved:
IFLAG=
1 case is solved according to the user supplied ordering
IFLAG=
0 case is solved based on the ordering given in Section 14.2
IFLAG=-1
case is solved opposite to the ordering user inputted
NOTE: The keyword "TACS OUTPUTS'" are also recognized as introducing a TACS hybrid case,
for convenience to users of older data cases.
4 - 52
Section 5
5.1
5.2
Lumped, Uncoupled, Series R-L-C Branch ............................................ 5-1

5.1.1
Applications ............................................................................... 5-1
5.1.2
Branch Card ............................................................................... 5-2
5.1.3
Examples .................................................................................... 5-3
5.1.4
Alternate High-Precision Format ............................................... 5-4
5.1.5
Maximum Resistance ................................................................. 5-4
5.1.6
Minimum Resistance ................................................................. 5-5
Pi-circuit Branch .................................................................................... 5-6

5.2.1
Applications ............................................................................... 5-6
5.2.2
Branch Cards .............................................................................. 5-9
5.2.3
Example ..................................................................................... 5-10
5.2.4
Alternate High Precision Format ............................................... 5-12
5.2.5
Alternative Performance Equation for Series

Segment of Pi-Circuit ................................................................ 5-12
5.3
Mutually-Coupled R-L Branches ........................................................... 5-13

5.3.1
Applications ............................................................................... 5-13
5.3.2
Branch Cards .............................................................................. 5-14
5.3.3
Optional Use of Z0 and Z1 for 3-Phase

Coupled R-L Branches ............................................................... 5-16
5.3.4
5.4
5.5
Alternate Performance Equations .............................................. 5-17
Saturable Transformer Component ........................................................ 5-18

5.4.1
Description of a Single-Phase Transformer ............................... 5-18
5.4.2
Branch Cards .............................................................................. 5-19
5.4.3
Comments About Delta Transformer Connections ................... 5-22
Frequency Dependent Network Equivalents ......................................... 5-24

5.5.1
Applications ............................................................................... 5-24
5.5.2
Branch Cards .............................................................................. 5-25
5.5.3
Example ..................................................................................... 5-26

5.6
"CASCADED PI" Feature ........................................................................ 5-27

5.6.1
General Explanations ................................................................ 5-27
5.6.2
Some More-Specific Comments About "CASCADED PI" .......... 5-29
5.6.3
Format and Meaning of Different

Classes of Data for "CASCADED PI" .......................................... 5-30
5.6.4
Sample Usage of "CASCADED PI" Feature ................................ 5-35
2-2
Section 5
This section describes the rules for linear impedance models, which include:
5.1
Lumped, Uncoupled, Series R, L, C Branch
5.2
Pi-circuit Branch
5.3
Mutually-Coupled R-L Branches
5.4
Saturable Transformer Component
5.5
Frequency Dependent Network Equivalents
5.6
CASCADED PI Feature
The first five devices can be used for both steady-state and transient calculations, while the use of
the last one is limited to the steady-state only.
5.1 Lumped, Uncoupled, Series R-L-C Branch

5.1.1
Applications
The lumped linear impedance models are used for the following types of uncoupled branches:
5-1
5.1.2
Branch Card
R
BUS1
BUS2
A single data card is needed to describe the series R-L-C branch:
I
T
Y
P
E
I2
Node Names
Node Names
Of Reference
Branch
BUS 1
BUS 2
BUS 3 BUS 4
A6
A6
A6
A6
R()
L(mH)
or
L()
C(F)
or
C
(mho)
E6.2
E6.2
E6.2
IOUT
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 56789012345678901234567890123456789 0
I1
Rule 1:
Branch type is 0 (punch 0 in column two or leave blank.
Rule 2:
Specify the terminal nodes by name (fields BUS1 and BUS2, columns 3-8 and 9-14,
respectively).
Rule 3:
Specify the desired R, L, C data in the appropriate fields (fields R, L and C, columns
27-32, 33-38 and 39-44, respectively).
Rule 4:
At least one of the R,L,C values of the branch must be non-zero. For zero values,
either punch 0.0 or leave the appropriate columns blank.
Rule 5:
The numerical values for parameters R, L, and C are in the following units. Recall that
variables "XOPT" and "COPT" come from the floating-point (first) miscellaneous data
card of Section 4.2.1.
(A)
Specify R in ohms.
(B)
Specify inductance L as:

(i)
inductance L in mH if XOPT = 0.
(ii)
reactance L in ohms with = 2 * XOPT
(iii)
(rad/s) if XOPT > 0.0.
5-2

(C)
Rule 6:
Specify capacitance C as:

(i)
capacitance C in F if COPT = 0.
(ii)
susceptance C in mho with = 2 * COPT (rad/s) if COPT >0.0.
Should the branch R, L, C data be identical to that of a preceding branch, the following storage-saving option can be used:
Specify the node names of that preceding branch in fields BUS3 and BUS4 (columns
15-20 and 21-26, respectively) and leave R, L, C data fields blank.
CAUTION:
If the reference branch has other branches in parallel, it is not clear

which of them should be the reference branch. Therefore, the first branch among
parallel branches with identical node-name pairs shall always be the reference branch.
Note that two branches 'NODE-A' to 'NODE-B' and 'NODE-B' to 'NODE-A' do not have
identical node-name pairs (order is reversed) and can therefore be used as two distinct
reference branches.
Rule 7:
Output options for printing and/or plotting:

Specify "1" in column 80 to get branch current;
Specify "2" in column 80 to get branch voltage;
Specify "3" in column 80 to get both branch voltage and current;
Specify "4" in column 80 to get branch power and energy consumption (see also
description of steady-state and transient output in Section 16.2 - 16.3).
5.1.3
Examples
1
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4-><----R<----L<----C
VOLT-A 10-A
1.
123. .21E-3
3
VOLT-B 10-BVOLT-A 10-A
1
PORT
300.
Above examples (assuming XOPT=0 and COPT=60) represent the following branches:
1)
VOLT-A
10-A
123mH 0.21 103 mho (f=60H z)
5-3

Current and voltage across branch will be included in output:.
2)
VOLT-B
10-B
123mH 0.21 103 mho (f=60H z)
Current through this branch will be included in output:

PORT
3)
300
5.1.4
Alternate High-Precision Format
The $VINTAGE card (see Section 3.6) provides for an alternate high-precision format. Specifically,
the R, L, C fields can be switched to 3E16.0 (columns 27-74 in this case) if $VINTAGE, 1 precedes
such a group of new branch cards, and $VINTAGE, 0 follows the grouping. Picircuits (Section
5.2), coupled RL Branches (Section 5.3), and the original distributed line (Section 7.3) are other
components that allow wider formats; they can also be included in the grouping between the
$VINTAGE cards, then:
$VINTAGE, 1
Any mixture of series R-L-C, picircuit, coupled RL, and distributed line cards, as long
as all use the wide formats.
$VINTAGE, 0,
ITYPE
BUS1
BUS2
L (L)
C (C)
IOUT
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 7890123456789012 3456789012345678 9012345678901234 56789 0
I2
A6
A6
E16.0
E16.0
E16.0
I2
5.1.5
Maximum Resistance
A high resistance branch can be added to monitor the voltage difference between two nodes. The
largest value permissible is system dependent and should not exceed the square root of the largest
number which the computer system being used can represent in floating-point form. In VAX VMS
5-4

the maximum R permissible would be 1.E18 ohms. In other systems this number can be higher.
However, for most practical purposes 1.E10 is a sufficiently large number to represent an
"infinite" resistance.
5.1.6
Minimum Resistance
Ideally, a zero-impedance branch could be used to monitor branch currents. However, this is not
possible for the EMTP (an attempt to input a zero-impedance branch will yield an EMTP error
message). Instead, a "small" impedance value must be used, where the minimum permissible or
desirable value depends on several considerations:
1.
The fundamental limitation is due to floating-point word length of the computer

system being used. Since R will be inverted to build the nodal admittance
matrix, limitations similar to those of the largest number permitted also apply
here.
2.
Singularity tolerances "EPSILN" and "TOLMAT" (floating-point miscellaneous

data parameters; see Section 4.2.1) are actually used by the EMTP to protect the
user against a too-small impedance value. Parameter "EPSILN" is used for the
real, all-resistive transient equivalent-network solution, while "TOLMAT" applies
to the complex (phasor) admittance-matrix solution for steady-state initial
conditions. The general idea is that all impedances which terminate on a given
node should not differ drastically in value, as measured by the just-mentioned
singularity tolerances. Specific relevant points related hereto include the
following:
(A)
Considering "EPSILN", the impedances in question are equivalent Norton

resistances which result from trapezoidal-rule integration (see Reference
7). For a resistor, this is just the value of the resistance in ohms; for an
inductor of L Henries, the impedance in ohms is 2L/t; for a capacitor of
C Farads, the impedance is t/2C.
For distributed-parameter transmission lines or cables, treat them like
resistors having resistance equal to the characteristic impedance. In
the just-stated formulas, t is the time-step-size "DELTAT", in seconds
(see Section 4.2.1).
(B)
3.
Considering "TOLMAT", the impedances in question are phasor

impedances, depending not only upon the element, but also upon the
frequency of the sinusoidal excitation being used. The impedance of a
resistor is just the value of the resistance in ohms; for an inductor, use L
for a capacitor, use l/(C).
A permanently-closed switch can sometimes be advantageously used in place of

a low-impedance branch. See the write-up of Section 9.2, Class 5 switch
5-5

("measuring" switch). If a "MEASURING" switch is used, all concern over the
question of "how small is small" is avoided. The solution time may either
increase or decrease, it all depends on the case.
5.2 Pi-circuit Branch

5.2.1
Applications
This class of branches provides for the representation of lumped-element resistance, inductance,
and capacitance matrices. For N conductors, the associated differential equations are:
di km
v k v m = L ----------dt + R i km
dv k
1
-i k = 2 C -------dt + i km
(6.2)
dv m
1
i m = --2- C --------dt i km
(6.3)
ik
(6.1)
ikm
+
im
R , L
vk
vm
C
-------- 2
Figure 5.1:
C
-------2
Schematic Representation of a Pi-Circuit
All matrices are assumed to be symmetric, and it will be noted that [C] is split in two, with half of
the total on each end of the branch.
5-6
end
k
1
2
R + j L
Coupling
given by
matrices
R L,
COUPLED
1
--- C
2
end
m
1
1
--- C
2
Principal use of this model is to represent short, untransposed sections of transmission line steadystate or quasi steady-state simulations. By connecting many such short sections in series, keeping
track of actual transpositions (if any), a model for a long line can be made. For such use, the
matrices [R], [L], and [C] can be automatically calculated by the EMTP support program AUX.
While [C] = [0] (no capacitance) is a legal input option, this special situation represents only
mutually-coupled RL branches, for which the separate input format of Section 5.3 has been
provided.
To specify a multiphase (N > 1) capacitance matrix to ground, then input a near-infinite series
impedance branch, and ground all conductors at the far end. In particular, set
Rii = 10 for all i
Rij = 0 for all i j
[L] = [0]
[C] = twice the desired final matrix, since it will be halved.
Here is a "large" exponent, ideally infinite. In practice it is limited by the computer installation
being used. Refer to discussion of high-impedance branches, Section 5.1.5
5-7
R11=10
R22=10
R33=10
Earth
Figure 5.2:
Multiphase Coupled Capacitance
Elements of the matrices [R], [L], and [C] have the following meaning in the sinusoidal steadystate at frequency f (=2f):
Diagonal Rii + jLii =
self impedance of branch i (impedance of loop "branch iground return");
Off-diagonal Rik + jLik =
mutual impedance between branches i and k. (Rik = 0

with non-zero ground resistivity).
Diagonal Cii =
sum of all capacitances connected to the nodes at both

ends of branch i;
Off-diagonal Cik =
negative value of capacitance from branch i to branch k.
If the values of a 3-phase -circuit are given as Z1, C1 (positive sequence) and Z0, C0 (zero
sequence), then convert to matrix elements as follows::
Z11 = Z22 = Z33 = 1/3 (Z0 + 2Z1)
Z12 = Z13 = Z23 = 1/3 (Z0 - Z1)
Identical relationships hold for the capacitances C.
5-8
(6.4)
5.2.2
Branch Cards
To specify an N-conductor pi-equivalent, at total of N(N+1)/2 cards must be used:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
I2
Node Names
Ref. Branch
BUS1
BUS2
BUS3
A6
A6
A6
elements (k, m)
elements (k, m+1)
elements (k, m+2)
BUS4
A6
E6.2
E6.2
E6.2
E6.2
E6.2
E6.2
E6.2
L
E6.2
C
E6.2
Rule 1:
Number the phases 1, 2,..., N. Make out one branch card (plus possible continuation
cards; see Rule 3) for each phase, and stack them in this sequence. Indicate this
sequence by punching 1, 2, ..., N in columns 1-2 of these cards (field ITYPE). Limits
on N are 0 < N < 41.
Rule 2:
Specify each of the phases 1, 2,..., N by the names of the nodes at both ends field BUS1
and BUS2, columns 3-8 and 9-14, respectively). Nodes may be grounded (indicated
by blank field name) if desired.
Rule 3:
Matrices [R], [L], and [C] are symmetric, so only need be specified on and below the
diagonal. The following format applies:
1
2
3
4
5
6
7
8
12 345678901234 567890123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
Node Names of
1st Branch
R11
L11
C11
Node Names of
2nd Branch
R21
L21
C21
R22
L22
C22
Node Names of
3rd Branch
R31
L31
C31
R32
L32
C32
R33
L33
C33
Node Names of
4th Branch
R41
L41
C41
R42
L42
C42
R43
L43
C43
R44
L44
C44
Rule 4:
When one card is not sufficient for all required R-L-C values (for the fourth and later
phases), then "continuation cards" are used, with columns 1-26 left blank, it will be
noted.
5-9

Rule 5:
At least one of the matrices [R], [L] must be non-zero. Matrix [C] may be zero, though
then the Section 5.3 input should be used.
Rule 6:
The numerical values for [R], [L] and [C] are in the same units as in Rule 5 of
Section 5.1. If branch data are identical with those on a preceding set of N branch
cards, then the following storage-saving option may be used: Repeat node names of
the first branch of that preceding set of branch cards in the provided columns 15-20
and 21-26 (fields BUS3 and BUS4) of the first branch in the same sequence and leave R,
L, C blank. On the 2nd, 3rd, ... , Nth branch card only the information in columns 1-2,
3-8, and 9-14 (fields ITYPE, BUS1, and BUS2) is used.
CAUTION:
If the reference branch has other branches in parallel, it is not clear

reference branches.
Rule 7:
No branch current output is possible for this branch type. However, the branch voltage
can be obtained on the first two phases (where column 80 of the card is not being used)
by punching a "2" in column 80.
5.2.3
Example
Consider the modelling of the two 3-phase circuits on the same right of way, so that they are
mutually coupled. Together these two circuits constitute six coupled conductors (N=6).
10-A
10-B
10-C
11-A
11-B
11-C
20-A
20-B
20-C
21-A
21-B
21-C
Figure 5.3:
Cascaded Pi-Equivalents
5 - 10
30-A
30-B
30-C
31-A
31-B
31-C

Note that each circuit is represented by 2 cascaded pi-equivalents. The data for the pi-equivalents
are shown below.
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4-><----R<----L<----C<----R<----L<----C<----R<----L<---C
1 10-A 20-A
1.41 8.334 .164
2 10-B 20-B
0.97 2.8586-.0289 1.23 9.1392 .1626
3 10-C 20-C
0.91 2.2823-.0088 0.86 3.4318-.0273 1.16 9.518 .1662
4 11-A 21-A
1.00 2.0986-.0180 0.94 2.1248-.0101 0.90 2.0398-.0053
1.41 8.334 .164
5 11-B 21-B
0.94 2.1248-.0101 0.89 2.5064-.0104 0.86 2.6454-.0087
0.97 2.8586-.0289 1.23 9.1392 .1626
6 11-C 21-C
0.90 2.0398-.0053 0.86 2.6454-.0087 0.83 3.1597-.0142
0.91 2.2823-.0088 0.86 3.4218-.0273 1.16 9.518 .1662
1 20-A 30-A 10-A 20-A
2 20-B 30-B
3 20-C 30-C
4 21-A 31-A
5 21-B 31-B
6 21-C 31-C
[R] =
1.41
0.97
0.91
1.00
0.94
0.90
[L] =
1.23
0.86
0.94
0.89
0.86
symmetric!
1.16
0.90 1.41
0.86 0.97 1.23
0.83 0.91 0.86
8.3340
2.8586
2.2823
2.0984
2.1248
2.0398
[C] =
9.1392
3.4218
2.1248
2.5064
2.6454
0.1640
-0.0289
-0.0088
-0.0180
-0.0101
-0.0053
1.16
symmetric!
9.5180
2.0398
8.3340
2.6454
2.8586
3.1597
2.2823
0.1621
-0.0273
-0.0101
-0.0104
-0.0087
at 60
9.1392
3.4218
symmetric!
-0.1062
-0.0053
0.1640
-0.0087 -0.0289
-0.0142 -0.0088
5 - 11
9.5180
F
0.1626
-0.0273
0.1662
5.2.4
Alternate High Precision Format
The $VINTAGE card (see Section 3.6) provides for an alternate high-precision format. Specifically,
the R,L,C fields can be switched to 3E16.0 (columns 27-74 in this case) if $VINTAGE, 1 precedes
such a group of new branch cards, and $VINTAGE, 0 follows the grouping. Here only one triplet of
(R,L,C) values is allowed on each data card. For the second and later rows, ordering is from left to
right (e.g., the card for column 2 follows that for column 1, etc.). Series R-L-C branches
(Section 5.1), coupled RL branches (Section 5.3) and the original distributed line (Section 7.3) are
other components that allow such new, wider formats; they can also be included in the grouping
between the $VINTAGE cards as in Section 5.1.4.
To remove any ambiguity as to the format, the following diagram corresponds to the case of
narrow formats as illustrated under Rule 3 of Section 5.2.2.
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 7890123456789012 3456789012345678 9012345678901234567890
1
SA
RA
R11
L11
C11
SB
RB
R21
L21
C21
R22
L22
C22
R31
L31
C31
R32
L32
C32
SC
5.2.5
RC
Alternative Performance Equation for Series Segment of PiCircuit
Instead of describing the series branch of multi-conductor (N > 1) pi-circuit by means of matrices
[R] and [L], it may be preferable to use arrays [R] and [B] = [L]-1, defined by the alternative
performance equation
( vk vm ) = L
di km
R i km + ----------dt
(6.5)
If this is the case, use the following procedure:

Rule 1:
Immediately preceding all mutually coupled branches (N > 1) for which the RB-option
is desired, insert an extra data card with the code "USE RB" punched in what is normally the BUS1-field (columns 3-8).
5 - 12
12 345678
1
2
3
4
5
6
7
8
901234567890123456789012345678901234567890123456789012345678901234567890
USE RB
Rule 2:
For all multi-conductor coupled circuits in question, punch the elements of [B] where
those of [L] would normally appear.
[L] becomes [B]
Rule 3:
To return to the use of [R] and [L] after having used [A] and [R], insert an extra data
card with the code "USE RL" punched in what is normally the BUS1-field. For all data
following this card, then, use of the R-L formulation is assumed -- until another "USE
RB" card should be encountered.
12 345678
1
2
3
4
5
6
7
8
901234567890123456789012345678901234567890123456789012345678901234567890
USE RL
The numerical values of [B] are to be specified in the following units:

(A)
Specify [B] in 1/mH if XOPT = 0.0.
(B)
Specify [B] in 1/ohms with = 2 * XOPT if XOPT > 0.0
5.3 Mutually-Coupled R-L Branches

5.3.1
Applications
This class of branches provides for the representation of lumped-element, mutually-coupled RL

branches. Associated with these branches are matrices [L] and [R] having performance equations
di km
v k v m = L ----------dt + R i km
5 - 13
(6.6)

Matrices [R] and [L] are assumed to be symmetric.
End
m
End
k
vk
vm
Mutual Coupling
Figure 5.4:
Schematic Representation of Coupled RL Elements
Note that this branch type is identical in performance to the pi-equivalent of Section 5.2, if the
capacitance matrix [C] is set to zero. But in the present section the inductance data fields (see
below) have 12 columns rather than just 6, allowing for greater precision. This option is intended
to be used primarily for coupled impedances representing transformers, where it is necessary to
have highly-accurate values for self and mutual impedances; otherwise the leakage impedances get
lost in the magnetizing impedance.
5.3.2
Branch Cards
To specify N mutually-coupled R-L elements, the following rules and format are to be observed:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234 567890 123456789012 345678 901234567890
Node Names
Reference BR
BUS1
BUS2
BUS3
BUS4
I2
A6
A6
A6
A6
elements (k, m)
elements (k, m+2)
elements (k, m+2)
E6.2
E12.2
E6.2
E12.2
E6.2
E12.2
Rule 1:
Number the phases 51, 52,..., (50+N). Make out one branch card (plus possible continuation cards; see Rule 3) for each phase, and stack them in this sequence. Indicate this
sequence by punching 51, 52,..., (50+N) in columns 1-2 of these cards (field ITYPE).
Limits on N are 0 < N < 41.
Rule 2:
Specify each of the phases 51, 52, ...., (50+N) by the names of the nodes at both ends
(field names BUS1 and BUS2, columns 3-8 and 9-14, respectively). Nodes may be
grounded (indicated by blank field names) if desired.
Rule 3:
Matrices [R] and [L] are symmetric, so only need be specified on and below the diagonal. The following format applies:
5 - 14
1
2
3
4
5
6
7
8
12 345678901234 567890123456 789012 345678901234 567890 123456789012 345678 901234567890
51
Node Names of
1st Branch
R11
L11
52
Node Names of
2nd Branch
R21
L21
R22
L22
53
Node Names of
3rd Branch
R31
L31
R32
L32
R33
L33
54
Node Names of
4th Branch
R41
L41
R42
L42
R43
L43
R44
L44
When one card is not sufficient for all required RL values (for the fourth and later
elements), then "continuation" cards are used, with columns 1-26 left blank, it will be
noted.
Rule 4:
At least one of the matrices [R], [L] must be non-zero.
Rule 5:
The numerical values for [R] and [L] are in the same units as per Rule 5 of Section 5.1.
Rule 6:
If branch data are identical with those on a preceding set of N branch cards, then the
following storage-saving option may be used: Repeat the node names of the first
branch of that preceding set of branch cards in columns 15-20 and 21-26 (fields BUS3
and BUS4) of the first branch in the same sequence of this new group, and leave R and
L fields blank. On the 2nd, 3rd,..., Nth branch card, only the information in columns
1-2, 3-8, and 9-14, respectively (ITYPE, BUS1, BUS2) is used.
CAUTION: If the reference branch has other branches in parallel, it is not clear
reference branches.
Rule 7:
No branch current output is possible for this branch type. However, the branch voltage
can be obtained on the first two phases (where column 80 of the card is not being used)
by punching a "2" in column 80.
5 - 15
5.3.3
Optional Use of Z0 and Z1 for 3-Phase Coupled R-L Branches
Suppose that the user has exactly three coupled R-L branches, with the associated [R] and [L]
matrices having common diagonal values Zs and common off-diagonal values Zm, as shown in the
sketch. Then, rather than inputting [R] and [L], it may be more convenient to input the associated
zero and positive-sequence values, (R0, L0) and (R1, L1), respectively.
Zs Zm Zm
Z m Zs Z m
Zm Zm Zs
, ,
transformation
Z0 0 0
0 Z1 0
0 0 Z1
The preceding format specification is modified slightly, giving the following:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 789012 345678901234 567890123456789012345678901234567890
I2
Node Names
Sequence Values
BUS1
BUS2
A6
A6
E6.2
E12.2
Rule 1:
Number the 3 phases 51, 52 and 53. Make out a branch card for each, in this sequence.
Indicate this sequence by punching 51, 52, and 53 in columns 1-2 of these cards (field
ITYPE).
Rule 2:
Specify the phases by means of the names of the nodes at both ends (use columns 3-8
and 9-14; field names BUS1 and BUS2). Nodes may be grounded (indicated by blank
field name) if desired.
Rule 3:
Resistance and inductance values are punched in columns 27-32 and 33-44 (fields R
and L). Zero sequence values R0, L0 go on the first (phase 51) card, while positivesequence values R1, L1 go on the second card; columns 15-80 of the third card are to
be left blank. The numerical values for [R] and [L] are in the same units as per Rule 5
of Section 5.1.
Rule 4:
Once a 3-phase branch has been so inputted, it can be used as a reference branch for
identical 3-phase branches to follow. See Rule 6 of Section 5.3.2.
WARNING! If columns 15-80 of that third branch card are not left blank, the EMTP
will not recognize the coupled branches as being specified with sequence values. In
5 - 16

that case, the phase-domain format will be assumed, with the result being one very
unbalanced, erroneous representation. Sometimes something as innocent as an illegal
column-80 punch for branch current will cause big trouble.
Example: For the bus names as specified in the diagram and sequence parameters as tabulated
below, the associated input data cards should be as shown on the data form below.
N1A
N2A
N1B
N2B
N1C
N2C
Mutual Coupling
R0 = 4.5 ohm
L0 = 87.6 mH
R1 = 3.1 ohm
L1 = 66.4 mH
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->
<--R--<-----L----51
N1A
N2A
4.5
87.6
52
N1B
N2B
3.1
66.4
53
N1C
N2C
5.3.4
Alternate Performance Equations
The use of [R] and [B] instead of [R] and [L] is exactly as described for multi-conductor pi
circuits (see end of Section 5.2). This option is particularly useful for representing transformers
which draw no magnetizing current; in this case, [R] and [L] do not exist, though [R] and [B] do.
For marginal cases of very-high magnetizing impedance, one avoids the near-singularity problem,
with its associated need for high precision on input data. The saturable transformer component of
Section 5.4 is based on this approach, with N-1 2x2 matrices [A] and [B] used to represent the N1 ideal transformers and associated non-primary leakage impedances.
5 - 17
5.4 Saturable Transformer Component

5.4.1
Description of a Single-Phase Transformer
The use of saturation is not mandatory. If the flux-current magnetization characteristic consists of
a single finite-slope segment, then all-linear components are used in the model, and numerically
the resulting solution will be identical (except for roundoff differences) to that of Section 5.3
where matrices [R] and [L] or [R] and [B] are used. Thus even in the linear case, the present
section may be used.
The saturable N-winding single-phase transformer in question is modelled as per the figure below.
Pertinent points include the following:
N-1 single-phase, 2-winding ideal transformers are involved, providing the
correct transformation ratios of windings 2, 3,...,N with respect to winding 1.
2.
Each winding k has an associated leakage-impedance branch, characterized by

resistance Rk and inductance Lk. All leakage inductances with the exception of
L1 must be non-zero.
3.
Saturation and magnetizing-current effects are confined to a single nonlinear

reactor in the winding-1 circuit. This is internally modelled as a type-98 pseudononlinear reactor, should the saturation characteristic have two or more
segments; for a single segment, a simple, constant linear inductance element is
used in which case the transformer is completely linear.
4.
Core losses are confined to the constant, linear resistance Rmag which is in
parallel with the saturation branch.
BUS11
BUS21
L1
N 1 N2
Rmag
ideal
N2 Nn

ideal
Figure 5.5:
BUS12
BUS22
BUS1n
BUS2n
winding n
winding 1
R1
winding 2
1.
Equivalent Circuit Used to Represent Saturable, N-winding, Single-Phase

Transformer
5 - 18
5.4.2
Branch Cards
To input a single-phase saturable transformer component, the following rules apply:

Rule 1:
The first data card for the component is to be punched according to the following format:
isteady
steady
E6.2
E6.2
Bus3
TRANSFORMER
A6
Rmag
I OUTMG
Ref.
Name
Special
Request
Word
BUSTOP
1
2
3
4
5
6
7
8
12 345678901234 567890 123456 789012 345678 901234 567890 12345678901234567890123456789 0
A6
E6.2
I1
Special request word field is punched with the 12 characters, "TRANSFORMER" in

columns 3-14.
BUS3
(15-20)
Reference-component name. Leave blank unless using the referencecomponent procedure of Rule 4.
isteady
(27-32)
Components of point in the flux-current plane used to define the linear

inductance which represents the magnetizing branch during the phasor steadystate solution.
steady
(33-38)
Meaning as in Section 8.4.2.
i
BUSTOP
(39-44)
A 6-character alphanumeric name for the internal bus at the top of the
magnetizing branch. This name uniquely identifies the transformer.
Rmag
(45-50)
The constant, linear resistance which parallels the magnetizing reactance,

accounting for core loss. The specification is in ohms, with a value of zero or
blank to mean Rmag =
IOUTMG
Output specification for magnetizing-reactance branch.

Punch "1" for branch current, "2" for branch voltage, or "3" for both.
(80)
5 - 19

Rule 2:
Assuming that the reference name BUS3 is left blank, the card of Rule 1 is followed by
cards which define the saturation characteristic of the magnetizing branch. Format is
exactly like for the type-98 pseudo-nonlinear reactor of Section 8.4.
infinite
extension
i
(A)
The origin (i=0, =0) is an implied point, not to be inputted explicitly.
(B)
Current and flux pairs for the breakpoints are punched in fields CUR and FLUX
(columns 1-16 and 17-32, respectively). Punch one pair of values per card,
inputted in monotone-increasing order (movement away from the origin). Both
coordinates must be strictly monotonically increasing.
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
CUR
FLUX
E16.0
E16.0
(C)
The final point on the characteristic merely defines the slope of the final
segment, which is assumed to extend to infinity. The last point is followed by a
terminating card with "9999" punched in columns 13-16.
Normally the first point of the characteristic will equal (isteady, ysteady) in order to
provide continuity between steady-state and transient solutions at time zero, though
this is not mandatory. Note that if only one i- point is inputted, a linear transformer
results; with no i- date points preceding the 9999-card, the magnetizing reactance is
assumed not to exist.
Rule 3:
After the above saturation cards, if any, come cards specifying the transformer characteristic. For each transformer winding k (k=1, 2,..., N), the user inputs a winding card
in the following format
5 - 20
I2
Node Names
Leakage Z
BUS1
BUS2
Rk
A6
A6
E6.2
ITYPE
(1-2)
BUS1
(3-8)
BUS2
(9-14)
Lk
Volts
(Nk)
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 789012 345678 901234 56789012345678901234567890123456789 0
E6.2
E6.2
I1
Winding number (1, 2, ...). Cards must be placed in natural order, with winding 1
first, then winding 2, etc.
Six-character node names of the buses to which the winding in question is
connected. As usual, a blank field is taken to mean ground.
Rk
Leakage values associated with the winding in question. Rk is resistance in
(27-32) ohms, while Lk is inductance in mH.
Lk
(unless XOPT parameter of first miscellaneous data card is non-zero, in which
(33-38) case Lk is in ohms at frequency XOPT Hz). Rk may be zero, but Lk must be nonzero (with the exception of winding 1, where L1=0 is permitted if R1 0.0).
Volts
A number proportional to the number of turns of the winding in question. It is
(39-44) convenient simply to use the rated winding voltage in kV.
IOUT
(80)
Rule 4:
Only for winding 1 (the primary), a 1-punch in column 80 will make branch
current i1 an output variable. Identifying node names for the current output will
be "BUS1" of the primary terminals, and internal node "BUSTOP".
Should it be desired to specify a transformer having parameters identical to those of a
previously-inputted unit, use the following reference-component procedure:
(A)
On the card of rule 1, punch on the fields shown below. Field BUS3
(columns 15-20) now carries the name which was punched as variable
BUSTOP (columns 39-44 for the reference component of which a copy is
desired).
Special
Request
Word
Ref.
Word
BUSTOP
1
2
3
4
5
6
7
8
12 345678901234 567890 123456789012345678 901234 567890123456789012345678901234567890
5 - 21
5.4.3
(B)
Omit the cards of Rule 2 (the i curve).
(C)
For the cards of Rule 3 (the winding cards), punch only the fields ITYPE
(columns 1-2), BUS1 (columns 3-8), and BUS3 (columns 9-14).
Comments About Delta Transformer Connections
Delta transformer connections can sometimes be a little tricky. They also are not uncommon, so
that a little specific explanation of this configuration is in order.
First, there is the equivalent circuit when modelled using three single-phase saturable transformer
components. If it is the primary which is delta-connected, and if the secondary is grounded-Wye,
then Figure 5.6 shows the resulting interconnection of elementary components:
TA
R1
L1
T1
N1
N2
N1
T3
R2
L2
R2
L2
SA
SB
N2
TB
L1
L1
R1
T2
R1
N2
N1
R2
TC
Figure 5.6:
L2
SC
Interconnection of Elementary Components
Here R1 and L1 are the primary leakage-impedance (assumed to be the same for all three
transformers), and R2 and L2 are those for the secondary. Bus "T1" is the internal node "BUSTOP"
for the first single-phase transformer whose primary terminals are "TA" and "TB", and whose
secondary terminals are "SA" and ground. Et cetera for the other two single-phase transformers as
per the following EMTP data-setup listing:
5 - 22

C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C --keyword-->-ref->
<-is--<flxs-<bstp-<RmagTRANSFORMER
2.0
1137. T1
3.00E5
C ----- CUR ----<---- FLUX ----2.0
1137.
50.
1165.
10080.
1478.
9999
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->
<---Rk<---Lk<---Nk
1
TA
TB
.25
25.
3.03E5
2
SA
.25
25.
3.03E5
C --keyword-->-ref->
<-is--<flxs-<bstp-<RmagTRANSFORMER T1
T2
1
TB
TC
2
SB
TRANSFORMER T1
T3
1
TC
TA
2
SC
A second important point concerns the need for a path to ground on the delta side of the bank: a
floating delta is not allowed, since mathematically the voltages there would then only be defined
within an arbitrary constant. Of course if a transmission line is connected to the delta side, there is
no problem, since line capacitance provides the connection to ground. Likewise a connection to
one or more voltage sources, either direct or through other "simple" elements, will satisfy the
requirement. But the common, isolated, delta-connected tertiary of 3-winding transformers
requires special attention. There are two common, general approaches:
Unbalanced Solution: If the user is not interested in voltages on the delta side, it is
convenient to simply ground one corner of the delta (see Figure 5.7). This provides the needed
ground connection without altering the problem solution elsewhere (since no current can flow in
the connection to ground anyway, by Kirchhoff's current law).
Figure 5.7:
Grounding One Corner of the Delta Winding
5 - 23
Balanced Solution: If preserving the inherent balance on the delta side is important, then
stray capacitance can be added. The simplest procedure is to connect three equal capacitors from
the corners of the delta to ground, as in Figure 5.8. Caution must be exercised to use realistic
values, however; as such capacitors go to zero, a matrix singularity will be approached, and the
user will get into trouble with miscellaneous data parameter singularity tolerances "EPSILN" and
"TOLMAT" (see Section 4.2.1). Values of 3000 pF to 8000 pF are more or less typical values for
stray capacitances of power transformer windings..
Figure 5.8:
Addition of Stray Capacitances
5.5 Frequency Dependent Network Equivalents

5.5.1
Applications
This class of branches provides for the representation of lumped-element equivalents of a network
as seen from a set of terminal ports. The branch cards for the equivalents are automatically
produced by a dedicated support program, the Frequency Dependent Network Equivalent (FDNE)
program.
The equivalents represent the network in the modal domain. For any N-phase network for which
an M-port equivalent is to be used, there are:
(A)
M positive and zero sequence modal equivalents of the network from the port
in question to ground (corresponding to the self impedance/admittance term).
(B)
M(M-1)/2 positive and zero sequence equivalents describing the coupling

(mutual impedance) between the different ports.
5 - 24
RPI
RI
CI
LI
Figure 5.9:
Structure of Individual Branch
The mathematics of the network equivalents are documented in Reference 37.

The frequency dependent network equivalents can be used for both AC steady-state and time-step
solutions.
5.5.2
Branch Cards
To specify a single-port N-phase frequency dependent network equivalent, the following rules and
format are to be observed:
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234 567890123456789012345678901234567890
ITYPE
Node Names
BUS1
I2
A6
BUS2
A6
Ref. Branch
BUS3
BUS4
A6
A6
Code1
Code2
E6.2
E12.2
Rule 1:
Number the phases 51, 52, ..., (50+N). Make out one branch card for each phase, and
stack them in this sequence. Indicate this sequence by punching 51, 52, ..., (50+N) in
columns 1-2 of these cards (field ITYPE). Limits on N are 0 < N < 41.
Rule 2:
Specify each of the phases 51, 52, ..., (50+N) by the names of the nodes to which the
equivalent is connected (use columns 3-8 and 9-14, respectively, fields BUS1 and
BUS2). Nodes may be grounded (indicated by blank field names) if desired.
Rule 3:
Specify "-6666." in columns 27-32 and 33-44. This indicates to the program the use
of the frequency dependent network equivalents.
5 - 25

Rule 4:
Terminate bus cards by a special card with the word "BRANCHES" punched in columns
3-10. This will switch the program logic to reading in the specification of the individual branches in the ground and aerial modes.
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678 9012345678901234 5678901234567890
RI
LI
E16.0
E16.0
CI
RPI
E16.0
E16.0
Rule 5:
Specify the values of the elements Ri, Li, Ci and Rpi of the individual branches of the
ground mode.
Rule 6:
At least one of the elements Ri, Li and Ci must be non-zero. A zero (blank) value of
Rpi is interpreted by the program as an open circuit, i.e., Rpi = [blank].
Rule 7:
The numerical values for Ri, Rpi, Li and Ci are the same units as per Rule 5 of Section
5.1.
Rule 8:
Terminate the specification of branch elements by a card with 9999 punched in columns 13-16.
Rule 9:
Specify the branch elements of the aerial mode according to Rules 5-8.
Note that the branch cards described above are normally generated automatically by the FDNE
support program, and included into the EMTP data case with a text editor or by using $INCLUDE
statements.
5.5.3
Example
The following is a sample data listing representing an actual data case provided by Ontario Hydro.
The parameter XOPT and COPT (first miscellaneous data card, Section 4.2.1) are: XOPT = 60.0 and
COPT = 60.
C
ADMITTANCE FROM PINRDA TO PORCQA
ZERO SEQUENCE
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4-><CODE1<---CODE2--51PINRDAPORCQA
-6666. -6666.
52PINRDBPORCQB
-6666. -6666.
53PINRDCPORCQC
-6666. -6666.
BRANCHES
C
6 BRANCHES
C <-----R-------<------LI-------<------CI-------<------RPI-----0.44825188E+02 0.18483540E+03 0.00000000E+00 0.58593750E+06
5 - 26

-0.17391516E+03 -0.58667725E+02 -0.10715895E+03
0.23261044E+03 0.54195309E+02 0.27190245E+02
-0.29840744E+03 -0.55958225E+02 -0.11419177E+02
0.36792017E+03 0.58396465E+02 0.61198645E+01
0.25000001E+16 0.00000000E+00 0.75809151E+00
9999
C -------------------------------------------------------C
ADMITTANCE FROM PINRDA TO PORCQA
POS SEQUENCE
C
6 BRANCHES
C <-----R-------<------LI-------<------CI-------<------RPI-----0.39113164E+01 0.52014275E+02 0.00000000E+00 0.11319939E+09
-0.50570450E+01 -0.25848560E+02 -0.15892096E+03
0.71209798E+01 0.25758076E+02 0.39726288E+02
-0.88492546E+01 -0.25793795E+02 -0.17605963E+02
0.10501760E+02 0.25707409E+02 0.99280367E+01
0.11264092E+05 0.00000000E+00 0.35342913E-01
9999
5.6 "CASCADED PI" Feature

5.6.1
General Explanations
The "CASCADED PI" option can be used only for runs which stop after the steady-state phasor
solution is complete (Tmax 0, meaning that no transient simulation is to follow). For such
studies, numerous pi-circuits (see Section 5.2) are often cascaded to approximate the distributed
nature of a transmission line, with possible conductor transposition occurring at the points of
interconnection of the pi-circuits; series or shunt elements may sometimes be present at these
interconnection points also. The "CASCADED PI" option can be efficiently used in such cases,
provided the user is only interested in the solution at the line terminals. Use of this feature makes
it impossible to find what is going on, i.e., the solution variables, at the internal interconnection
nodes and adjacent branches.
The mathematical modelling associated "CASCADED PI" is worth summarizing, since a general
understanding aids efficient usage of this feature. Components representing the line are defined
sequentially as a chain, from one end (the sending end) to the other (the receiving end). This data
is processed as it is read sequentially so as to produce at any stage of the process a
mathematical equivalent for all components the sending end and the last component read. This is
shown schematically in Figure 5.11.
5 - 27
RA1
RC1
RB1
1B1
1C1
1A1
2B1
2C1
open 2C1
2A1
3A1
3B1
6C1
6A1
Figure 5.10:
Equivalent
for
Components
1, 2, , K-1
Figure 5.11:
5B1
GB1
GC1
7B1
7C1
7A1
6B1
5C1
5A1
4A1
4B1
4C1
GA1
Sample Problem to illustrate usage of the "CASCADED PI" feature. Bus

(RA1, RB1, RC1) is the sending end of the line, and Bus (GA1, GB1, GC1) is
the receiving end.
Component
K
Equivalent
for
Components
1, 2, , K-1
Schematic Illustration of Cascading Operation for Kth Component.
The form of mathematical equivalent used for this is simply the nodal admittance matrix [Y].
Hence when input processing of the last component of the chain is complete, an admittance matrix
among the terminal nodes of the line exists, as shown at the right. This is an exact mathematical
equivalent for all of the interconnected components of the chain, as seen from the terminal nodes.
Matrix [Y] is symmetric and complex, i.e., admittance element Ykm = Gkm +jBkm.
5 - 28
YSS YSR
YRS YRR
From the aforementioned description, several points concerning speed and storage requirements
associated with this feature may be deduced:
Point 1:
The storage requirement (for [Y]) varies as the square of the number of transmission
line conductors; it is independent of the number of components which are cascaded
together.
Point 2:
Computational effort to produce the equivalent is proportional to the number of components which are cascaded together. The computer time associated with this effort
adds to the "data input" time figure of the summary case-termination statistics, not to
the time figure for the steady-state solution.
5.6.2
Some More-Specific Comments About "CASCADED PI"
Before detailing the specific format of data input for the "CASCADED PI" feature, a few additional
comments about restrictions and/or assumptions might be mentioned. Figure 5.10 shows a sample
problem, the Section 5.6.4 setup of which illustrates some of these points:
1.
The first line segment of the chain elements which are to be cascaded may not
have any series or shunt connections. That is, the sending end must begin with a
pi-circuit.
2.
Second and later sections can each consist of up to four types of subcomponents, as follows, in the order shown:
(A)
Series uncoupled R-L-C branches, if any.
(B)
Shunt uncoupled R-L-C branches, if any.
(C)
New line-position (thought of conceptually as transposition)

specification, if any.
(D)
Specification of new (i.e., changed) pi-circuit parameter matrices [R],

[L], and [C].
In order to utilize the "CASCADED PI" modelling features as part of an EMTP data case, the
following classes of data are involved, in the order indicated:
5 - 29

Class 1:
"CASCADED PI" header card, i.e., special request card which begins the definition process.
Class 2:
Cards which specify the sending-end and receiving-end bus names, as well as the [R],
[L], and [C] matrices which apply to the first pi-circuit.
Class 3:
Line position card for the first pi-circuit, the one at the sending end of the line.
Class 4:
Line position card, if any.
Class 5:
Cards defining Series R-L-C branches, if any.
Class 6:
Cards defining shunt R-L-C branches, if any.
Class 7:
Cards specifying new [R], [L] and [C] parameter matrices, if any.
Class 8:
"STOP CASCADE" card, to close the definition process.
Classes 1, 2, 3, and 8 are used only once, for the sending and the receiving end of the line in
question. Classes 4, 5, 6, and 7 apply to the interior of the line, to be repeated over and over again,
once for each station.
5.6.3
Format and Meaning of Different Classes of Data for "CASCADED PI"
Class 1: "CASCADED PI" Header Card: One begins the definition process for a line by
means of the special request card having "CASCADED PI" punched in columns 3-14, as per the
following format:
1
2
3
12 345678 901234 5678901234567 89012
BUS1
BUS2
Cascaded Pi
4
5
6
7
8
345678 901234567890123456789012345678901234567890
NPCHAS FREQCS
I6
E6.2
Cols 3-14
Punched with the key request word "CASCADED PI".
Cols 27-32
Variable "NPHCAS", the number of phases or conductors in the circuits

to be cascaded. The maximum permissible number is dependent upon
user EMTP dimensioning (see Section 2.5.1); execution should
terminate with an overflow error message, if the user attempts to
represent a line having too many conductors.
Cols 33-38
Variable "FREQCS", the frequency in Hz of the sinusoidal steady-state

phasor solution which is to be performed.
5 - 30
Class 2: Specification of Bus Names, Matrices [R], [L], [C]: These data
cards are for inputting the sending-end and receiving-end bus names of the line, plus the [R], [L],
and [C] matrices which characterize the line geometry of the first section. The format is identical
to that which is used for a conventional pi-circuit (see Section 5.2.2 for format).
Rule 1:
Number the phases 1, 2, ... NPHCAS. Make out one branch card plus possible continuation cards (see Rule 6) for each phase and stack them in this sequence. Indicate this
sequence by punching 1, 2, ... NPHCAS in columns 1-2 of these cards (field ITYPE).
These numbers will be referred to later as the line position number for this base
matrix.
Rule 2:
Specify the circuit consisting of NPHCAS phases by the names of the nodes at both
ends field names BUS1 and BUS2, columns 3-8 and 9-14, respectively). Nodes may be
grounded (indicated by blank field name) if desired. These names do not necessarily
correspond to the R-L-C matrix values on the same card. The line position card takes
care of the relation between bus name and R-L-C matrix values.
Rule 3:
Matrices [R], [L] and [C] are symmetric, so only need be specified on and below the
diagonal. (See Section 5.2.2 for format).
Rule 4:
At least one of the matrices [R], [L] must be non-zero. Matrix [C] may be zero. Specifically, [R] + j[L] must be non-singular.
Rule 5:
The numerical values for [R], [L], and [C] are to be specified according to the Rule 5
of Section 5.1.
Rule 6:
If this cascaded pi-circuit is identical with another cascaded pi-circuit then the following storage-saving option may be used:
Repeat node names of the first branch of that preceding set of branch cards in the
provided columns 15-26 of the first branch in the same sequence and leave R, L, C
blank. On the 2nd, 3rd, ... NPHCAS-th branch card only the information in columns 114 is used.
CAUTION:
1.
Same as in Rule 5, of Section 5.1.
2.
Rule 1 and 2 of this section do not necessarily hold. Rules 1 and 2 of Section 5.3
are more pertinent.
NOTE: The program will properly process the reference data with or without the CASCADED
PI header card and STOP CASCADE terminator card. Care must be taken that either
both or none of these cards should appear.
5 - 31

Rule 7:
No branch voltage output, or branch current output, is possible for this component, so
don't put any of those special punches in columns 80. Of course the complete steadystate solution will show all branch flows.
Class 3: Line Position Card for Sending-End Section: The line position card
is used to specify control parameters pertaining to the associated line section and lumped elements.
As used for Class 3 data, it applies only to the very first segment, the beginning of the cascading
process. The general definition is contained in the Class 4 data explanation which follows.
Class 4: Line Position Card: The line position card is used to specify control
parameters pertaining to the associated line section and lumped elements. The following format
applies:
1
2
3
4
5
6
7
8
12 345678 9012 3456 7890 1234 5678 9012 3456 78901234567890123456789012345678901234567890
DSECTJ
MULTIP
MSER
MBR
MSECT
Phase 1
Phase 2
Phase 3
Phase-location indicator MAPCAS (I)
E6.2
I4
I4
I4
I4
I4
I4
I4
Rule 1:
Etc. for all phases

(MAPCAS ( I ), I 1, NPHCAS

Length of segments is specified by DSECTJ. This is a per unit length where the base
length is determined by the length of line represented by the equivalent pi-matrix.
That is, the length of the line represented by this matrix is 1.0.
Example: If a 5 mile length of line is needed for a section and the R-L-C equivalent pi-matrix is
per mile, then DSECTJ=5.0. However, if the R-L-C equivalent pi matrix is for a 5 mile
length of line, then DSETJ=1.0.
Rule 2:
If N identical sections are to be cascaded together without any transpositions, then

MULTIP can be set to N rather than include N line position cards. There may be series
and shunt connections between these sections, but these connections must be the same
for all sections handled by this multiplicity feature. Default value of 1 if MULTIP is left
blank.
Rule 3:
Specifications of series connections:
MSER=1
This indicates that there is a R-L-C series connection for this section(s) and
data for this is to follow.
MSER=0
This indicates that there is no R-L-C series connection for this section(s).
5 - 32

MSER= -1 This indicates that there is a R-L-C series connection for this section(s). The
data from the last R-L-C series connection is to be used.
Comment:
Should the user desire to put the same R-L-C connection between every other
section this can be easily accomplished as follows:
The first section preceded by the R-L-C series connection would have
MSER=1. The next section would have MSER=0. The third section, which is
to be preceded by a R-L-C series connection would have MSER = -1.
Rule 4:
Specification of R-L-C shunt connections MBR=1, 0, -1 has the same meaning for
shunt R-L-C connections as MSER has for series connection.
Rule 5:
Specifications of line-section parameters R, L and C:
Rule 6:
MSECT=0
means old R-L-C values are used for following or blank sections.
MSECT=1
a new R-L-C matrix is to be read in.
Specifications of line position. These numbers are a map of the position of the phases
of the cascaded circuit. The fields across the card starting in column 25 correspond to
the phases in the order their names appear in the BUS1, BUS2 fields on the NPHCAS
equivalent-pi cards. The number entered in these fields is the row number of the R-LC equivalent pi-matrix. Thus if the conductor connecting the nodes specified by the
first equivalent-pi cards BUS1-BUS2 field has electrical properties which are specified
by the third row of the R-L-C matrix then a 3 would be put in column 28.
For 14 or fewer conductors, columns 25-80 of the data card suffice, as shown. But for
15 or more conductors, the MAPCAS(I) data spills over onto as many extra cards of the
format (24X, 14,I4) as are required to complete the data input.
Class 5: Series R-L-C Branch Specification: Class 5 data consists of series R-L-C
branches, connected in series with any particular conductor (phase) of the line. The following
format applies:
ITYPE
1
2
3
4
5
6
7
8
12 345678901234567890123456 789012 345678 901234 567890123456789012345678901234567890
I2
Series
R
Series
L
Series
C
E6.2
E6.2
E6.2
5 - 33

Rule 1:
Fields ITYPE of columns 1-2 is to be punched with the conductor number which the RL-C branch being defined is to be placed in series with.
Rule 2:
If a short-circuit is desired, simply do not enter a data card of this class for that conductor.
Rule 3:
If an open circuit is desired, punch field R of columns 27-32 with "999999" to be

interpreted by the EMTP as a special flag meaning "open circuit".
Rule 4:
At least one of the parameter values R, L or C of the card must be non-zero.
Rule 5:
If the desired branch has no resistance, set R equal to zero (or leave blank). If the
branch has no inductance, set L equal to zero (or leave blank). If the branch has no
capacitance, set C equal to zero (or leave blank), which is interpreted by the EMTP as
though series capacitance C were actually infinite, a capacitive short-circuit.
Rule 6:
The numerical values for R, L and C are to be specified according to Rule 5 of

Section 5.1.
Rule 7:
The end of such series R-L-C branch cards is signalled by a blank card.
Class 6: Shunt R-L-C Branch Specification: Series R-L-C branches may be

connected in shunt, thereby being categorized as shunt R-L-C branches of data class number 6.
This is illustrated by the Figure 5.10 schematic diagram, at buses 7A1, 7B1, and 7C1. The shunt
branches in question can be connected between any two nodes which are either conductor nodes or
intermediate extra nodes, including local ground. The applicable format and rules for such shunt
R-L-C branches are as follows:
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 789012 345678 901234 567890123456789012345678901234567890
Terminal
Node Numbers
Branch
Parameter Values
N1
N2
I6
I6
E6.2
E6.2
E6.2
Rule 1:
Fields N1 and N2 (columns 3-8 and 9-14, respectively) are to be punched with node
numbers. If one wants a conductor, the associated node number is the positive integer
(row or column number) assigned to it in the line-parameter matrices R, L, and C. Zero
or blank is reserved for local ground. Any other new, intermediate nodes are given
consecutive negative integer values -1, -2, etc. for as many as are needed. Such numbering are local to the line section in question, with one beginning over again with -1
the next time. A limit of "NPHCAS", such intermediate nodes for each section exists
(i.e., not more than one for each conductor).
Rule 2:
At least one of the parameter values R, L or C of each branch must be non-zero.
5 - 34

Rule 3:
If the desired branch has no resistance, set R equal to zero (or leave blank). If the
branch has no inductance, set L equal to zero (or leave blank). If the branch has no
capacitance, set C equal to zero (or leave blank), which is interpreted by the EMTP as
though series capacitance C were actually infinite, a capacitive short circuit.
Rule 4:
The numerical values for R, L and C are to be specified according to Rule 5 of Section
5.1.
Rule 5:
The end of such shunt R-L-C branch cards is signalled by blank card.
Class 7: Change of Line Parameters: During the cascading process, the line
geometry can be altered if desired, which necessitates the redefinition of matrices [R], [L], and
[C]. If this is the case, field "MSECT" of the associated line-position card (Class 4 data) must have
been punched with value "1". The general data format for Class 7 input then is like that of Class 2,
with the following exceptions:
Rules 2 and 6 of the Class 2 data description do not apply. There are no bus names "BUS1" and
"BUS2" to be inputted, so leave columns 3-14 blank. No storage-saving option is possible, so
likewise enter nothing in columns 15-26.
Class 8: Termination of CASCADED PI Data: The last card in the definition

sequence for "CASCADED PI" line is a record reading "STOP CASCADE", as per the following
format:
1
2
3
4
5
6
7
8
12 345678901234 567890123456789012345678901234567890123456789012345678901234567890
Stop Cascade
This card tells the EMTP that it has now processed all sections, making the line which connects the
sending-end and receiving-end buses which were defined on the Class 2 data input.
5.6.4
Sample Usage of "CASCADED PI" Feature
Figure 5.10 displays a sample network for which the steady-state sinusoidal phasor solution can be
found in two different ways: first, by using the conventional, manual EMTP data setup and, second,
by using the "CASCADED PI" modelling capability just described. The listing of the "CASCADED
PI" setup is shown below.
BEGIN
C
C
C
C
C
NEW DATA CASE

BENCHMARK DC-9
UTPF TEST CASE NO. 18
SAMPLE SMALL STEADY-STATE PROBLEM WHICH ILLUSTRATES MANY FEATURES OF THE
'CASCADED PI' CAPABILITY OF THE T.P. ----- TRANSPOSITION, CASCADING
OF PI-CIRCUITS, THE CONNECTION OF SERIES BRANCHES, THE CONNECTING OF
5 - 35

C
SHUNT BRANCHES.
C
C ----dt<---Tmax<---Xopt<---Copt<-Epsiln<-Tolmat<-Tstart
0
0
60.
C -Iprnt<--Iplot<-Idoubl<-Kssout<-Maxout<---Ipun<-Memsav<---Icat<-Nenerg<Iprsup
1
1
C
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
CASCADED PI
3 60.0
1RA1
GA1
.877 8.40 .1628
2RB1
GB1
.747 4.14-.0252 .852 8.43 .1559
3RC1
GC1
.735 3.47-.0067 .723 4.17-.0277 .829 8.46 .1571
1.0
2
0
0
1
1
2
3
1.0
1
1
0
0
2
3
1
2
999999
3
13.1449.071
BLANK CARD ENDING 1ST CLASS 5 (SERIES R-L-C) SET OF DATA
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
1.0
1
1
1
0
3
1
2
1
13.14
2
13.14
3
13.1449.071
BLANK CARD ENDING 2ND CLASS 5 (SERIES R-L-C) SET OF DATA
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
1
-1
13.1449.071
2
-1
13.1449.071
3
-1
13.1449.071
-1
5.0 13.14
BLANK CARD ENDING FIRST CLASS 6 (SHUNT R-L-C) SET OF DATA
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
2.0
1 -1 -1
1
1
2
3
1
.829 8.46 .1571
2
.723 4.17-.0277 .852 8.43 .1559
3
.735 3.47-.0067 .747 4.14-.0252 .877 8.40 .1628
STOP CASCADE
BLANK CARD ENDING BRANCH CARDS
BLANK CARD ENDING SWITCH CARDS
14GA1
424.35
60.
0.0
-.1
14RA1
424.35
60.
10.0
-.1
14GB1
424.35
60.
-120.0
-.1
14RB1
424.35
60.
-110.0
-.1
14GC1
424.35
60.
120.0
-.1
14RC1
424.35
60.
130.0
-.1
BLANK CARD ENDING SOURCE CARDS
BLANK CARD ENDING SELECTIVE NODE VOLTAGE OUTPUTS
PRINTER PLOT
BLANK CARD ENDING PLOT CARDS
BEGIN NEW DATA CASE
BLANK
5 - 36
Section 6 Frequency-Dependent Transformer Model

6.1
6.2
Background Information ........................................................................ 6-1

6.1.1
Power Frequency Module .......................................................... 6-2
6.1.2
Hysteresis/Saturation Module .................................................... 6-3
6.1.3
Eddy Current Module ................................................................ 6-4
6.1.4
High Frequency Transformer Module ....................................... 6-4
FDB Model ............................................................................................. 6-7
6.2.1
General Considerations .............................................................. 6-7
6.2.2
Input Data Format ...................................................................... 6-8
6.2.3
FDB Branch Card Format ........................................................... 6-9
6.2.4 FDB Component Branch Input Format ....................................... 6-10

6.2.5
Examples .................................................................................... 6-11
Section 6
The Frequency-Dependent Transformer Model consists of four stand-alone components.
Separately, these modules simulate power frequency behaviour, eddy current effects, hysteresis/
saturation, and the high frequency characteristics of a transformer. This section describes how to
assemble these components to create a comprehensive model that simulates the linear as well as
the nonlinear behaviour of a power transformer.
6.1 Background Information

The Frequency-Dependent Transformer Model consists of several components which can be
combined to obtain the desired simulation capabilities. These components or building blocks are:
6.1.1
Power frequency module
6.1.2
Hysteresis/Saturation module
6.1.3
Eddy current module
6.1.4
High Frequency Transformer (HFT) module
Hysteresis Model
HFT Model
Eddy Current Model
Power Frequency Model
Figure 6.1:

6-1
6.1.1
Power Frequency Module
The power frequency module can be any of the following EMTP transformer models: TRELEG,
BCTRAN, or TOPMAG. These are the most basic transformer models available in the EMTP, and they
consist of coupled RL branches that reproduce the standard open and short circuit zero and positive
sequence tests at power frequency (e.g., 60 Hz). These three models have very similar behaviour
at power frequency, and they give reasonably good, but underdamped answers up to 1 kHz or so.
Except for TRELEG, (see Figure 6.2), [Y()] for all power frequency models degenerate to a simple
diagonal resistive matrix at DC (see Rrence 17).
When data for the frequency-dependent transformer model is obtained from field measurements, it
can be difficult to obtain accurate measurements at both high (hundreds of kHz) and low
frequencies (e.g., around 60 Hz) with the same instrumentation. Using different instruments for
different frequency ranges can create continuity and consistency problems in the measurements.
Therefore, it is good practice to rely on the power frequency module for the low frequency portion
of the spectrum, and on the High Frequency Transformer (HFT) module for mid to high
frequencies. This is done by subtracting the frequency behaviour of the power frequency module
from the measured data, and then producing an HFT model of the difference. This subtraction is
done automatically with the FDBFIT support program. Please note that the punch file obtained after
running FDBFIT contains the EMTP data cards for both the power frequency module and the HFT
module.
+0
10
-1
10
-2
10
TRELEG
MEASUREMENTS
-3
10
Real part of Y11
-4
10
becomes negative
here
-5
10
-6
10
-7
10
-8
10
-1
10
Figure 6.2:
+0
10
+1
10
+2
+3
10
10
Frequency
+4
10
Real Part (absolute value) of Y11 zero-sequence.

response
6-2
+5
10
TRELEG
+6
10
and measured
+0
10
-1
10
TRELEG
-2
10
MEASUREMENTS
-3
10
-4
10
-5
10
-6
10
-1
10
+0
10
+1
10
+2
10
F
+3
10
+4
10
+5
10
+6
10
Figure 6.1: Imaginary Part (absolute value) of Y11 zero-sequence. TRELEG and measured
response
6.1.2
The hysteresis/saturation of a power transformer can be modelled using the type 96 pseudononlinear hysteretic reactor (see Section 8.5), or the type 92 hysteretic reactor (see Section 8.11)
generated using the FITSAT module of AUX. The type 96 hysteretic element is based on switched
inductances and a piecewise linear description of the flux-current characteristic of the transformer.
The type 92 hysteresis model, on the other hand, is more flexible, it does not involve switched
inductances, and it is defined by closed-form third-order analytical expressions for saturation and
hysteresis. The parameters for the type 92 hysteresis model can be obtained from the flux-current
characteristic of the transformer with the FITSAT support routine.
When specifying the nonlinear module, it should be remembered that a portion of the magnetizing
branch can be specified in the power frequency module. In fact, in the case of TRELEG, a fraction
of the magnetizing branch must be specified to produce a legitimate model. This is not the case
with BCTRAN or TOPMAG, where a zero magnetizing impedance can be used.
6-3
6.1.3
Eddy current module
Eddy currents are induced in transformer core laminations by the alternating flux in the core. As
frequency changes, flux distribution in the iron core lamination changes. For high frequencies, the
flux will be confined to a thin layer close to the lamination surface, whose effective thickness
decreases as the frequency increases.
There are a number of possible representations for eddy current effects, ranging from a single RL
branch to high-order ladder networks. These representations are generally based on the physical
characteristics of the core, and core laminations. The AUX module EDDYC, produces a ladder
network (see Figure 6.1) which provides a good compromise between computational speed and
accuracy [see Reference 34]. The resulting ladder network should be connected in parallel with
the nonlinear branch representing saturation and/or hysteresis.
Eddy current effects are not as important at power frequency as they are at high frequencies.
Strictly speaking, though, the effect of the eddy current network evaluated at power frequency
should be taken into account when specifying the magnetizing branch for the power frequency
module.
6.1.4
High Frequency Transformer module
The High Frequency Transformer (HFT) module reproduces the behaviour of a transformer over a
wide frequency range [See Reference 24]. Raw input data for the HFT model can be in the form of
field measurements of [Y()]; that is, frequency scan measurements of the nodal admittance
matrix of the transformer. The raw data must be manipulated and approximated with rational
functions, which can later be represented as FDb branches in the EMTP. The FDBFIT support routine
handles the data acquisition, manipulation and approximation process. On output, FDBFIT
produces a punch file that contains a multi-port, multi-phase pi-circuit. Each branch in the picircuit contains a multi-phase FDB model.
For example, a two-winding, three-phase transformer would be represented with a two-port threephase pi-circuit, as in Figure 6.4:
Y12
Y22 + Y21
Y11 + Y12
Figure 6.2: HFT model for a two-winding three-phase transformer
6-4

The nodal admittance matrix [Y} for this transformer can be partitioned in 3 x 3 blocks
Y
Y =
Y
11
21
Y
Y
12
22
where [Yij] is an 3 x 3 block matrix of the form

y
y
11 12
Y
y y
ij = 21 22
y y
31 32
y
y
y
13
23
33
If [Yij] is assumed to be balanced (or if all the diagonal elements are averaged to produce ys and
all the off-diagonal elements are averaged to produce ym) then,
ys ym
Y ij = y m y s
ym ym
ym
ym
ys
Note that in the pi-circuit of Figure 6.4, we have that
shunt1
= Y 11 + Y 12
shunt2
= Y 22 + Y 12
= Y 12
series
Since the elements of the pi-circuit are described by balanced matrices, they can be modelled
using zero and positive sequence parameters. The FDB branches produced by FDBFIT represent the
approximations by rational functions of Yshunt and Yseries in zero and positive sequence.
6-5
+0
10
-1
10
TRELEG
-2
10
MEASUREMENTS
-3
10
-4
10
DIFFERENCE FUNCTION
-5
10
-6
10
-1
10
+0
10
+1
10
+2
+3
10
10
Frequency
+4
10
+5
10
+6
10
Figure 6.3: Imaginary Part (absolute value) of Y11 positive sequence. TRELEG, raw data and
difference function
When the frequency response of the transformer is obtained from field measurements, it is often
found that due to instrumentation limitations, measured data that is accurate at high frequencies
(e.g., 1 kHz to 500 kHz) loses accuracy in the power frequency range, especially the real part of the
elements of [Y]. Therefore, it is good practice to rely on the power frequency module for low
frequency response and on the HFT module for high frequency response. This is achieved by
subtracting the frequency behaviour of the power frequency module from the measured data, and
the producing a fit of the difference using FDB models. This subtraction is done automatically with
the FDBFIT support program. The punch file obtained after running FDBFIT contains the EMTP data
cards of both the power frequency module and the HFT module.
Note:
From the EMTP point of view, the HFT module is just a combination of FDB branches
generated by the support program FDBFIT. The user need only be concerned with the
specification of the terminal node names and output requests.
6-6
6.2
FDB
6.2.1
Model
General Considerations
The FDB (Frequency Dependent Branch) model consists of a number of parallel RLC branches of
the general form shown in Figure 6.6. In a given FDB branch there are N component branches, and
the discrete R, RL, RC, and C elements may or may not be present in every component branch.
This model can be used to represent the behaviour of any frequency dependent device which can
be synthesized with the type of branches shown in Figure 6.6, or with a subset thereof.
RL
RC
VL
VC
(a) Single Component
(b) N component branch
Figure 6.4: Single-phase FDB branches
The FDB model was originally implemented to represent the high frequency behaviour of the
impedances of a transformer, but it can be used for practically any transfer function that can be
realized with a rational function that has a finite number of poles and zeroes and no time delay.
For example, the modal characteristic impedance of the JMARTI line model can be represented
with FDB branches where only RC components are present. An FDNE equivalent network can also
be modelled with FDB branches (there is an option to accept FDNE input format, as described in
Section 6.2.2
6-7

While single-phase FDB branches can be simulated in the EMTP by combining existing series RLC
branches, the manual specification of a large number of parallel RLC branches would be awkward
and time consuming. For example, to specify five FDB branches (each with 15 component
branches) manually, it would be necessary to create up to 150 intermediate node names in addition
to the 10 node names to which the FDB branch would be connected. Additionally the EMTP would
require 75 entries in the branch tables. By using the FDB model only 10 node names and 5
branches have to be specified. Multiphase FDB branches, on the other hand cannot be modelled
directly with standard EMTP components. The computational effort required for manual
specification is also larger. For example, a transient simulation with 5 FDB branches is
approximately 50% faster than its manual-entry counterpart.
For the HFT model, a more realistic number of FDB branches would be 45 with an average 10 to 15
component branches each. In other words, if it were to be modelled manually, over a thousand
intermediate nodes would have to be specified.
6.2.2
Input Data Format
To specify an FDB branch, two or more data cards are required. The data structure is the following:
Input Option 1:
[NPHASE FDB branch cards]

[filename containing FDB data}
name of file containing FDB data

NPHASE FDB
Input Option 2:
branch cards
[NPHASE FDB branch cards]

FDB_DATA (keyword)
[contents of filename in input option 1]
contents of file with FDB data

FDB_DATA
NPHASE FDB
6-8
keyword
branch cards
6.2.3
FDB
branch card format
FDB
Branch Cards:
BUS1
BUS2
BUS3
BUS4
I2
A6
A6
A6
A6
ITYPE
KEYWORD
A28
IBAL
I
T
Y
P
E
NPHASE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 7890123456789012345678901234 56 78 901234567890123456789 0
I
O
U
T
I1
I2
I1
Set ITYPE = -1 for the first card. Ignored for cards 2 to NPHASE.
(1-2)
BUS1
First node of the FDB branch
(3-8)
BUS2
Second node of FDB branch
(9-14)
BUS3
First node for reference branch option.
(15-20)
BUS4
(21-26)
Second node for reference branch option. BUS3 and BUS4 are the
names of another FDB branch from which data is to be duplicated
(some memory saved when this option is used).
KEYWOR
D
Set KEYWORD to "FDB_BRANCH". Left justification is not

necessary. Uppercase characters must be used.
(27-54)
(55-56)
Number of phases in a multiphase FDB branch. If NPHASE is

negative, it is assumed that subsequent data is in FDNE-compatible
format.
IBAL
IBAL
=0
Assume o transformation matrix. This means that

only FDB_DATA cards for zero and positive sequence
have to be provided
IBAL
>0
Assume a user-defined modal transformation matrix.

This means that NPHASE sets of FDB_DATA cards must
be provided.
NPHASE
(57-58)
IOUT
(80)
Set IOUT=1 to produce branch current output at the sending and

receiving ends of the cable. Set IOUT=2 for branch voltage, and
IOUT=3 for both branch currents and branch voltages.
6-9
6.2.4
FDB
Component Branch Input Format
Following NPHASE branch cards, FDB component data must be specified.
Input option 1:
Enter the filename where the values for the RLC component branches
are stored. The same filename can be used by more than one FDB branch.
Input option 2:
Alternatively, the component branch data can be appended to the branch

card specification by replacing the filename with the keyword
"FDB_DATA" in uppercase characters.
FDB component branches are usually generated with an auxiliary routine such as FDBFIT, so the
average user should not really have to know the specific format used for the FDB component branch
specification. However, this information would be useful for users who wish to use the FDB model
to simulate other system components.
-1BUSA BUSB
FDB_BRANCH
1
1
FDB_DATA
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
Number of RLC branches I3 format
2
Followed by R,L,C data in 3E25.0 format
C
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.13448552156843914D+05 -0.28194088449482568D+00 0.00000000000000000D+00
End of branch 1
-0.30322071295568220D-09 -0.58969169265007891D+05
-0.64808676108169777D+03 0.72305124403379425D-01 0.00000000000000000D+00
End of Branch 2
0.11889682080970598D-08 0.41557999166208247D+05
The values of R, RL, RC, L and C follow the following conventions:

L is expressed in Henrys and C in Farads (regardless of the value of XOPT and COPT.
If R=0, its value is assumed to be zero.
If RL=0 or RC=0, then they are assumed to be absent. That is, pure inductance or pure
capacitance.
If L=0 or C=0, then they are assumed to be absent and RL or RC are added to the value of R.
If RL=L=0 or RC=C=0 then the RL or RC branches are assumed to absent (short-circuited).
If IBAL =1, and NPHASE > 1, then a modal transformation matrix [Q] can be specified immediately
following the last FDB data card using the 3E25.0 format. Note that [Q] must be entered as [Qt]
(i.e., as the transpose of [Q]).
6 - 10

For example for
1 0 1
1 1
1 --- --2Q =
2
1 1
1 --2- --2
the corresponding data cards would as shown below:

C
C
C
C
C
Modal transformation matrix Q follows, entered as Q transpose

Q(1,1) Q(2,1) Q(3,1)
Q(1,2)
Q(2,1) Q(3,2)
Q(1,3) Q(2,3) Q(3,3)
1.00000000000000000D+00 1.00000000000000000D+00 1.00000000000000000D+00
0.00000000000000000D-00 -0.50000000000000000D+00 0.50000000000000000D+00
1.00000000000000000D+00 -0.50000000000000000D-00 -0.50000000000000000D+00
6.2.5
Examples
The following examples were obtained using FDBFIT to calculate an HFT model for a three-winding
wye-wye-delta three-phase transformer. Both HV and LV windings are solidly grounded. The
power frequency model subtracted was a TRELEG model. The tertiary delta winding is buried; that
is, the [Y] matrix frequency scan measurements were only made on the HV and LV sides of the
transformer. in order to obtain a two-winding equivalent model. Note that in the case of the
TRELEG model, the buried delta winding is included. Nonlinear and eddy currents models are not
included.
Example 1: FDB data included in the EMTP file:

C
First the FDB branch from HV to ground
C HFT MODEL FOLLOWS
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
-1H1-A
FDB_BRANCH
3
1
H1-B
1
H1-C
1
FDB_DATAb
C
This is the FDB branch from HV to ground
C
C Approximation of branch [ 1, 1]
C ZERO SEQUENCE: Block-balanced Y matrix, Fitting by elements. No LF branch
48
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
6 - 11

0.13448552156843914D+05 -0.28194088449482568D+00
-0.30322071295568220D-09 -0.58969169265007891D+05
-0.64808676108169777D+03 0.72305124403379425D-01
0.11889682080970598D-08 0.41557999166208247D+05
0.00000000000000000D+00
0.00000000000000000D+00
.
[rest of the data ommited]
.
.
C
C POSITIVE SEQUENCE: Block-balanced Y matrix, Fitting by elements. No LF branch
37
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.13242460171497542D+04 0.98588171160729765D-01 0.00000000000000000D+00
0.10481860198387582D-08 0.53326716265625320D+06
0.71283061010485053D+04 0.18836361872649557D+00 0.00000000000000000D+00
0.12166962499451844D-09 -0.57089791253081849D+06
.
.
.
C
C NEGATIVE SEQUENCE: Block-balanced Y matrix, Fitting by elements. No LF branch
37
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.13242460171497542D+04 0.98588171160729765D-01 0.00000000000000000D+00
0.10481860198387582D-08 0.53326716265625320D+06
.
.
.
C
This is the FDB branch from HV to LV
C
-1H1-A X1-A
FDB_BRANCH
3
1
H1-B X1-B
1
H1-C X1-C
1
FDB_DATA
C
26
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.25400149807135404D+04 0.27646048915195331D+00 0.00000000000000000D+00
0.31597543514350458D-09 0.11106467526595997D+08
.
.
.
C
17
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.94896271633808840D+03 0.36916545872172923D+00 0.00000000000000000D+00
6 - 12

0.30590651686364647D-09 0.48937290659215435D+06
.
.
.
C
17
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.94896271633808840D+03 0.36916545872172923D+00 0.00000000000000000D+00
0.30590651686364647D-09 0.48937290659215435D+06
.
.
.
C
This is the FDB branch from LV to ground
C
-1X1-A
FDB_BRANCH
3
1
X1-B
1
X1-C
1
FDB_DATA
C
42
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
-0.25400149807135404D+04 -0.27646048915195331D+00 -0.00000000000000000D+00
-0.31597543514350458D-09 -0.11106467526595997D+08
.
.
.
C
31
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
-0.94896271633808840D+03 -0.36916545872172923D+00 -0.00000000000000000D+00
-0.30590651686364647D-09 -0.48937290659215435D+06
.
.
.
C
31
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
-0.94896271633808840D+03 -0.36916545872172923D+00 -0.00000000000000000D+00
-0.30590651686364647D-09 -0.48937290659215435D+06
.
.
.
6 - 13

Now comes the TRELEG low frequency model
C
C
C Power frequency TRELEG model follows
51,H1-A ,
,,,
0.563200000000E+00, 0.360838777035E+06
52,X1-A ,
,,,
0.164891784765E-01, 0.736626325226E+05
0.268000000000E-01, 0.150410276706E+05
53,T1-A ,T1-B ,,,
0.207556476682E-02, 0.400149116732E+05
-0.646357321451E-02, 0.817057471861E+04
0.690000000000E-02, 0.443869934400E+04
54,H1-B ,
,,,
0.000000000000E+00, -0.180109786766E+06
0.450729231370E-01, -0.367680781658E+05
0.166576311163E-01, -0.199731463549E+05
0.563200000000E+00, 0.360838777035E+06
55,X1-B ,
,,,
0.450729231370E-01, -0.367680781658E+05
0.000000000000E+00, -0.750592977487E+04
-0.333695525588E-02, -0.407734858919E+04
0.164891784765E-01, 0.736626325226E+05
0.268000000000E-01, 0.150410276706E+05
56,T1-B ,
,,,
0.166576311163E-01, -0.199731463549E+05
-0.333695525588E-02, -0.407734858919E+04
0.000000000000E+00, -0.221489337600E+04
0.207556476682E-02, 0.400149116732E+05
-0.646357321451E-02, 0.817057471861E+04
0.690000000000E-02, 0.443869934400E+04
57,H1-C ,
,,,
0.000000000000E+00, -0.180109786766E+06
0.450729231370E-01, -0.367680781658E+05
0.166576311163E-01, -0.199731463549E+05
0.000000000000E+00, -0.180109786766E+06
0.450729231370E-01, -0.367680781658E+05
0.166576311163E-01, -0.199731463549E+05
0.563200000000E+00, 0.360838777035E+06
58,X1-C ,
,,,
0.450729231370E-01, -0.367680781658E+05
0.000000000000E+00, -0.750592977487E+04
-0.333695525588E-02, -0.407734858919E+04
0.450729231370E-01, -0.367680781658E+05
0.000000000000E+00, -0.750592977487E+04
-0.333695525588E-02, -0.407734858919E+04
0.164891784765E-01, 0.736626325226E+05
0.268000000000E-01, 0.150410276706E+05
59,
,T1-A ,,,
0.166576311163E-01, -0.199731463549E+05
-0.333695525588E-02, -0.407734858919E+04
0.000000000000E+00, -0.221489337600E+04
0.166576311163E-01, -0.199731463549E+05
-0.333695525588E-02, -0.407734858919E+04
0.000000000000E+00, -0.221489337600E+04
0.207556476682E-02, 0.400149116732E+05
-0.646357321451E-02, 0.817057471861E+04
0.690000000000E-02, 0.443869934400E+04
C
Magnetizing branches follow
C
C 3456789 123456789 123456789 123456789 123456789
T1-A
4295.6
T1-B
4295.6
T1-A T1-B
4295.6
C
6 - 14
,,,,,
$
,,,,,
$
$
,,,,,
$
$
,,,,,
,,,,,
$
$
,,,,,
$
,,,,,
$
$
,,,,,
$
$
,,,,,
$
$
,,,,,
$
$
,,,,,
,,,,,
$
$
,,,,,
$
$
,,,,,
$
,,,,,
$
$
,,,,,
$
$
,,,,,
$
$
,,,,,

Type 92 hysteresis model follows
C
C ----- Slope 1 ---><------ Slope 2 ---><------ Curvat ----><------ Coerc
-->
C ----- Sslop 1 ---><------ Sslop 2 ---><------ Cursat ----><------ Fslxs
-->
C ----Vref
--><------ Remflx ----><------ Eps1 ------><------ Eps2 ----->
92T1-A
8888.
1
28.9321308136
-5.80931901932
41.3372077942
1.21254003048
2.40782117844
0.651428490528E-04
2464.86596680
136.873901367
500000.000000
92T1-B
8888.
1
28.9321308136
-5.80931901932
41.3372077942
1.21254003048
2.40782117844
0.651428490528E-04
2464.86596680
136.873901367
500000.000000
92T1-B T2-A
8888.
1
28.9321308136
-5.80931901932
41.3372077942
1.21254003048
2.40782117844
0.651428490528E-04
2464.86596680
136.873901367
500000.000000
Example 2: FDB data stored in a separate file:

First the FDB branch from HV to ground
C HFT MODEL FOLLOWS
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
-1H1-A
FDB_BRANCH
3
1
H1-B
1
H1-C
1
filename for FDB data zero, positive and negative seq.
yhg_fit.pch
C
This is the FDB branch from HV to LV
C
-1H1-A X1-A
FDB_BRANCH
3
1
H1-B X1-B
1
H1-C X1-C
1
yhx_fit.pch
C
This is the FDB branch from LV to ground
C
-1X1-A
FDB_BRANCH
3
1
X1-B
1
X1-C
1
yxg_fit.pch
C
Now comes the TRELEG low frequency model
C
C
C Power frequency TRELEG model follows
51,H1-A ,
,,,
0.563200000000E+00, 0.360838777035E+06 ,,,,,
52,X1-A ,
,,,
0.164891784765E-01, 0.736626325226E+05 $
0.268000000000E-01, 0.150410276706E+05 ,,,,,
53,T1-A ,T1-B ,,,
0.207556476682E-02, 0.400149116732E+05 $
-0.646357321451E-02, 0.817057471861E+04 $
6 - 15

.
Example 3 FDB data with user-defined modal transformation matrix:

-1BUS1A BUS2A
FDB_BRANCH
3
-2BUS1B BUS2B
-3BUS1C BUS2C
FDB_DATA
2
Followed by R,L,C data in E25.0 format
C
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.13448552156843914D+05 -0.28194088449482568D+00 0.00000000000000000D+00
-0.30322071295568220D-09 -0.58969169265007891D+05
-0.64808676108169777D+03 0.72305124403379425D-01 0.00000000000000000D+00
0.11889682080970598D-08 0.41557999166208247D+05
2
C
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.13448552156843914D+05 -0.28194088449482568D+00 0.00000000000000000D+00
-0.30322071295568220D-09 -0.58969169265007891D+05
-0.64808676108169777D+03 0.72305124403379425D-01 0.00000000000000000D+00
0.11889682080970598D-08 0.41557999166208247D+05
2
C
C R(I)
L(I)
RL(I)
C C(I)
RC(I)
0.13448552156843914D+05 -0.28194088449482568D+00 0.00000000000000000D+00
-0.30322071295568220D-09 -0.58969169265007891D+05
-0.64808676108169777D+03 0.72305124403379425D-01 0.00000000000000000D+00
0.11889682080970598D-08 0.41557999166208247D+05
C
C Modal transformation matrix Q follows, entered as Q transpose
C Q(1,1) Q(2,1) Q(3,1)
C Q(1,2) Q(2,1) Q(3,2)
C Q(1,3) Q(2,3) Q(3,3)
1.00000000000000000D+00 1.00000000000000000D+00 1.00000000000000000D+00
0.00000000000000000D-00 -0.50000000000000000D+00 0.50000000000000000D+00
1.00000000000000000D+00 -0.50000000000000000D-00 -0.50000000000000000D+00
6 - 16
Section 7
Distributed-Parameter Transmission Lines
7.1
General Considerations .......................................................................... 7-1
7.2
Considerations Regarding Constant-Parameter Modelling ................... 7-3
7.3
7.2.1
Distortionless Line Modeling .................................................... 7-3
7.2.2
Lumped Resistance Line Modeling ........................................... 7-4
Constant-Parameter Transmission Line Model ..................................... 7-5

7.3.1
Branch Cards .............................................................................. 7-5
7.3.2
Alternate High-Precision Format ............................................... 7-8
7.3.3
Examples .................................................................................... 7-9
7.3.4
Special Double-Circuit Distributed Line,

with Zero-Sequence Coupling ................................................... 7-11
7.3.5
7.4
7.5
7.6
7.7
Examples of Special Double Circuit Lines ................................ 7-13
JMARTI Line Model ............................................................................... 7-16
7.4.1
Data Requirements ..................................................................... 7-16
7.4.2
Example ..................................................................................... 7-19
LMARTI Cable Model ............................................................................. 7-20
7.5.1
Data Requirements ..................................................................... 7-21
7.5.2
Examples .................................................................................... 7-24
EXACT-PI Model for Frequency Scan Calculations .............................. 7-27
7.6.1
Input Data Requirements ........................................................... 7-28
7.6.2
Example ..................................................................................... 7-30
Corona Model ........................................................................................ 7-31

7.7.1
Introduction ................................................................................ 7-31
7.7.2
The Suliciu model ...................................................................... 7-32
7.7.3
Data input rules .......................................................................... 7-39
7.7.4
Examples .................................................................................... 7-41
Section 7
Distributed-Parameter Transmission Line Models
7.1 General Considerations
The voltages and currents on a set of transmission lines can be described by partial differential
equations. For N phases (conductors) these are:
i
v
R' i
------ = L' ---t +
x
(7.1)
v
i
G' v
------ = C' ----t +
x
(6.2)
A sketch of an incremental section of single phase transmission line is shown below:
Rx
Lx
Cx
Gx
Figure 7.1: Incremental section, length = x

In general, the line equations (7.1 and 7.2) are vector-matrix equations, for which the matrices
[R], [L] and [C] can be determined from the EMTP support program AUX.
If the transmission line is assumed to be transposed (or just balanced if 2phase), then all matrices
have only two distinct element values: all diagonals have a common value "S", and all offdiagonals have a common value "M".
7-1
S
M
Figure 7.2:
Self
Mutual
Structure of Matrices for Transposed Lines
Such matrices can be diagonalized by a family of modal transformation matrices such as o,

Karrenbauer, and symmetrical components. But if such a transposition assumption is not realistic
for the study contemplated, the following two options are available for constant-parameter
representations:
(A)
the option of cascading lumped-parameter sections (Section 5.6) each of which

approximately represents the actual construction over the may be 20 or so miles
of the section. Yet such a lumped-parameter solution is costly in computer
time and storage requirements, and has a very poor frequency response, so
should be avoided if practical.
(B)
the option of using a constant and real modal transformation matrix for the
diagonalization process evaluated at the frequency at which the line parameters
are calculated. This type of model can be obtained using the support program
AUX.
In the above discussion it was assumed that all line parameters do not change with frequency.
Such an assumption is not realistic for many types of studies. For frequency dependent modelling,
the following alternatives are available:
(A)
JMARTI model for both transposed and untransposed lines (constant

transformation matrix).
(B)
LMARTI model for cables, where the modal transformation matrix can either be
assumed to be constant or frequency dependent. This model also includes an

option to use EXACT-PI line or models for frequency scan simulations.
(C)
Specialized line model designed to represent the corona effects.
The modelling of frequency dependence for untransposed lines is not finalized yet. For more
details about the issues discussed above, see Reference 8.
7-2
7.2 Considerations Regarding Constant-Parameter

Modelling
Constant parameter models in the EMTP are based on the assumption that R, L, and C are constant
(evaluated at a given frequency). There are two ways to account for resistances:
7.2.1
Distortionless Line Modelling
7.2.2
Lumped Resistance Line Modelling
7.2.1
Distortionless Line Modeling
A transmission-line mode is known to be distortionless if parameters for that mode satisfy the
relationship R'/L' = G'/C' where R', L', G', and C' are the resistance, inductance, conductance, and
capacitance per unit length (see figure 7.1). Of course, actual leakage conductance for an
overhead line is very nearly zero, and no provision for inputting a non-zero value is provided by
the EMTP. If the user wants distortionless-line modelling, it is presumed that the input value of R'
is meant to be a measure of the total line losses. Therefore, the program splits the losses into
series and shunt losses by computing the series resistance R'series and the leakage conductance G'leak
from the relation:
R' series
-------------------- =
L'
G' leak
R'
----------------- = -------C'
2L'
(6.3)
With this formula, the AC steady-state results are practically identical whether the line is modelled
as distortionless or with a lumped resistance (see Section 7.2.2). Transient responses differ mainly
in the initial rise of voltage pulses. The attenuation constant a is found from the relation:
R' L'C '
R'
= -------------------- = ------2L'
2Z
(6.4)
where
Z =
L'
----C'
(6.5)
The corresponding attenuation factor is e-l where l is the line length. Such distortionless
modelling is generally used only for the positive sequence, if at all.
7-3
7.2.2
Lumped Resistance Line Modeling
In this case, the series resistance is pulled outside of the distributed line, and represented as a
lumped element. This is the default representation for constant-parameter line modelling (see
variable IPUNCH, cols. 53-54). The program automatically cuts the line in two, inserting half of the
resistance in the middle, and one fourth at each end:
distributed L, C
Name of
1st Node
distributed L, C
Name of
2nd Node
Ground
Figure 7.3:
Lumped Resistance modelling
The half-length distributed sections then become lossless (hence distortionless, with attenuation
= 0). Assuming a lumped R produces acceptable results as long as
Equation is needed here
R l << Zc (surge impedance)
(6.6)
This condition is normally satisfied for most transmission lines. It may, however, not be valid for
some cable models.
7-4
7.3 Constant-Parameter Transmission Line Model

7.3.1
Branch Cards
To specify an N-conductor, distributed-parameter transmission line, conform to the following

rules and format:
(left end)
Node Names
for phases
Gen-A
End-A
Gen-B
Gen-C
Line Parameters
End-B
in the
End-C
Modal Domain
Node Names
for phases
(right end)
I2
A6
Rule 1:
A6
BUS3
A6
BUS4
A6
E6.2
A
E6.2
B
E6.2
E6.2
IPOSE
BUS2
Branch
ILINE
BUS1
Ref.
IPUNCH
N ode Names
Line Length
I
T
Y
P
E
Resistance in
R [ / length]
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 567890 12 34 56 78901234567890123456789 0
I2 I2 I2
I
O
U
T
I1
Number the phases -1,-2,-3,...,-N. Make out one branch card for each phase, and stack
them in this sequence. Indicate this sequence by punching -1,...,-N in columns 1-2 of
these cards (field ITYPE).
The above scheme can be used only if N is less than 10. If N is greater than 9, then use
the following scheme:
N
USE
10
-A
11
-B
12
-C
13
-D
14
-E
15
-F
16
-G
17
-H
18
-I
Note:
N cannot exceed 18.

7-5

Rule 2:
Specify each of the phases -1,...,-N by the names of the nodes at both ends (columns 3
14; field names BUS1 and BUS2, columns 3-8 and 9-14, respectively). Nodes may be
grounded (indicated by blank name field), if desired.
Rule 3:
If the parameters for this line are identical to those of another Nphase line previously
inputted, then the reference card option of Rule 6 in Section 5.2.2 may be used, except
that no storage is saved here. Columns 1520 and 21-26 with field names BUS3 and
BUS4 are used in this case. Otherwise, they are left blank.
Rule 4:
For an Nphase continuously-transposed line, there exist only two modes. Specify the
zero-sequence parameters on the first card (the card for the first phase), and the positive-sequence parameters on the second card (the card for the second phase). Leave
the modal parameters blank on the third and later branch cards. For an Nphase
untransposed line, there exist N different modes. Specify the modal parameters for the
first mode on the first card, and those for the second mode on the second card, etc. for
the third and later modes.
The four basic pieces of modal information needed are resistance, inductance,
capacitance, and length; yet three forms for this input are possible;
(A)
In all cases, punch resistance R' in ohms per unit length in columns 2732, and
line length l in consistent units in columns 4550.
(B)
Parameter ILINE (columns 5152) determines how L' and C' may be disguised.
Remember that variables XOPT and COPT come from the floating-point (first)
miscellaneous data card of Section 4.2.1.
ILINE
"A" columns 33 - 38
"B" columns 39 - 44
L in mH / length if XOPT = 0
C in mF / length if COPT = 0
Reactance L in W / length at frequency XOPT if XOPT 0.
Susceptance C in mho / length at

frequency COPT if COPT 0.
Propagation velocity in length / sec
Surge impedance zs in ohms

1
zs =
l
L'C'
v = ----------------
L'
-----C'
Surge impedance zs in ohms

2
zs =
Travel time of line, in sec
L'
-----C'
= --- =
v
L'C' l
Table 1: Summary of ILINE Input Options

Rule 5:
Output options for printing and/or plotting use a column-80 punch (variable IOUT).
The branch voltage output request, by setting IOUT=2, is now honoured for multi7-6

phase lines only. No mode voltage or branch current output is available for singlephase lines because of the removal of the Karrenbauer transformation. A blank or
zero will give no such output, of course.
Rule 6:
All modes of a line must have travel time t in excess of the time-step size t
(Section 4.2.1). If not, the program will stop with an error message (see Section 17.5).
Rule 7:
Both L' and C' must be non-zero for each mode.
Rule 8:
Variable IPUNCH of columns 5354 specifies the type of modelling to be used on the
mode of the card in question:
IPUNCH = 0
lumped-resistive modelling (the usual case)
IPUNCH = 1
distortionless-mode modelling
It is not obligatory to model the two modes of a multi-phase line the same way.
IPOSE = 0 or blank:
Rule 9:
IPOSE = N
Rule 10:
Line is assumed to be transposed.
Number of phases of an untransposed line.
Only when IPOSE is non-zero, the current transformation matrix [Ti] is read immediately after the N-th branch card. The elements of the matrix [Ti] are read in by rows
(row 1 first, then row 2, etc.). For a given row, the real part of matrix element for all
columns come first; then the imaginary part follows on a new card. NOTE THAT THE
IMAGINARY PART MUST BE ENTERED EVEN THOUGH IT IS IGNORED BY THE PROGRAM.
Within each row, elements are read in order of increasing column number. Each row
begins with a new card, and there are 6 or fewer elements per card. The following format is applicable to row 'k' of a 6phase line.
1
2
3
4
5
6
7
8
123456789012 345678901234 567890123456 789012345678 901234567890 123456789012 34567890
TI (J,K)
TI (J,K+1)
TI (J,K+2)
TI (J,K+3)
TI (J,K+4)
TI (J,K+5)
E12.0
E12.0
E12.0
E12.0
E12.0
E12.0
If the transmission line had only 3 phases, the final three fields (columns 3772)
would not be used. For N 6 phases, exactly 2N data cards are involved; for 7 N
12 phases, exactly 4N data cards are involved, etc. Note that there must be exactly
2N2 elements for [Ti].
7-7

Theoretically, the transformation matrix [Ti] is complex as well as frequencydependent. For many practical cases, however, the matrix [Ti] can be approximated by
a frequency-independent matrix with acceptable accuracy.
7.3.2
Alternate High-Precision Format
The $VINTAGE card (see Section 3.6), provides for an alternate high-precision format.
Specifically, the R,A,B,L fields can be switched to 4E12.0 (columns 2774 in this case) if
"$VINTAGE, 1" precedes such a group of new branch cards, and "$VINTAGE, 0" follows the
grouping. Note that picircuits (Section 5.2), coupled R-L branches (Section 5.3), and the simple
series R-L-C branch (Section 5.1) are other components which allow such new, wider formats; they
can also be included in the grouping between the $VINTAGE cards, then:
$VINTAGE, 1
Any mixture of series R-L-C, pi-circuit, coupled R-L, and distributed
line cards, as long as all use the new wide formats.
$VINTAGE, 0
Line Length
The rule for inputting IPOSE is different from that described earlier, because only one column
(column 79) is available for IPOSE on this alternate high-precision format:
IPOSE = 0
or blank
Line is assumed to be transposed.

For an N-phase untransposed line, IPOSE is non-zero. If N is less than
10, IPOSE = N. If N is greater than 9 the assignment of IPOSE is:
N
IPOSE
10
11
12
13
14
15
16
17
18
Note:
A
B
C
D
E
F
G
H
I
N cannot exceed 18.

7-8
8
0
IOUT
Resistance
R [ /length]
IPOSE
BUS2
IPUNCH
BUS1
ILINE
ITYPE
1
2
3
4
5
6
7
12 345678 901234 567890123456 789012345678 901234567890 123456789012 345678901234 56 78 9

This high precision format is used when the branch cards for an untransposed line are
automatically generated by the EMTP support program AUX.
Note:
7.3.3
1.
A negative sign on the length of the line is used as a flag for EMTP to recognize
an untransposed line when this format is used.
2.
Complex-valued transformation matrix [Ti] elements use 6E12.5 format.
Examples
(A)
Consider the 3phase continuously-transposed line shown in the sketch in

Section 7.3.1 the (preceding format rules). Suppose that line length is 180
miles, with parameters as follows:
R [ohm / mile]
Zero sequence
Positive sequence
X [ohm / mile]
at 3000 Hz
10.9
77.4
.0484
29.4
C [F / mile]
.0107
.0192
Corresponding data cards for input to the EMTP then are (assuming XOPT = 3000 and COPT = 0):
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus-->Bus-->Bus-->Bus--><---R<---L<---C<--len 0 0 0
-1 GEN-A END-A
10.9 77.4 .0107 180.
0
-2 GEN-B END-B
.084 29.4 .0192 180.
0
-3 GEN-C END-C
(B)
Consider four separate, uncoupled single-phase lines, as per Figure 7.4.

Corresponding input cards might then be as follows (assuming XOPT = 60 and
COPT = 60):
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
-1 LEFT RIGHT
.3
.4
12.6 0.1
0
0
-1 10-A 10-B
260. 0.66
2
0
-1 10-A 10-B
260. 0.66
2
0
-1PT
XY
0.2
300. 3.0E08 180. 1
0
-1 L1
R1
LEFT RIGHT
7-9
LEFT
10-A
PT
10-B
R = 0
Zs = 260
= 660ms
XY
2
Figure 7.4:
(C)
R = 0.3 / mile
L = 0.4 / mile
C = 12.6 mho / mile
length = 0.1 miles
RIGHT
R = 0.2 / m
Zs = 300
v = 300106 m / s
length = 200 m
Single-phase Lines
Consider an example of branch cards and transformation matrix generated by

the "LINE CONSTANTS" routine of the EMTP support program AUX for a 138mile long untransposed line.
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus-->Bus-->
<---R------<-----------<-----------<----------- 0 000
$VINTAGE,1
-1JDA
LMA
0.56097E+00 0.62519E+03 0.13038E+06-0.13800E+03 1 3
-2JDB
LMB
0.28195E-01 0.29090E+03 0.18276E+06-0.13800E+03 1 3
-3JDC
LMC
0.32016E-01 0.27623E+03 0.18181E+06-0.13800E+03 1 3
$VINTAGE,0
0.58810448 0.70710678 -0.40330771
0.0
0.0
0.0
0.55521730 0.0
0.82139258
0.0
0.0
0.0
0.58810448 -0.70710678 -0.40330771
0.0
0.0
0.0
The high-precision format of Section 7.3.2 was used in this example.
Note:
As explained in Section 7.3.2, the length of the line has an extra minus sign appended
to it to signal the presence of an untransposed line.
7 - 10
7.3.4
Special Double-Circuit Distributed Line, with Zero-Sequence

Coupling
Applicability: The double-circuit configuration, where both circuits share the same right of
way, is common enough so as to deserve special attention. Here the restriction to two 3
conductor circuits is made, for a total of six phases or conductors for the components. The
modelling option of this section assumes that the circuits are themselves individually continuously
transposed, but that there is inter-circuit zero-sequence coupling. As such, it is a special 6conductor extension to the continuously-transposed distributed-parameter transmission line
modelling of Section 7.3.1.
Zs Zm Zm
Zm Zm Zm
Zm Zs Zm
Zm Zm Zs
Zm Zm Zm
Zm Zm Zm
Zm Zm Zm
Zs Zm Zm
Zm Zm Zm
Zm Zs Zm
Zm Zm Zs
Zm Zm Zm
000
000
000
2b m
2c
Zs Zm Zm
Zp Zp Zp
Zm Zm Zs
Zm Zs Zm
Figure 7.6: Structure of Matrices for

Double-Circuit Line
Having Zero-Sequence
Coupling Between Circuits
1a
1b
1c
2a
2b
000
000
000
000
000
1a
000
000
2a m
Matrix Structure Assuming

Continuous Transposition
Zp Zp Zp
Zp Zp Zp
000
000
Figure 7.5:
Zp Zp Zp
Zm Zs Zm Zp Zp Zp
Zm Zm Zs Zp Zp Zp
Zs Zm Zm
2c
1b
1c
Figure 7.7: Double-Circuit Line

Supported by a
Single Tower
Figure 7.8:
7 - 11
Double-Circuit Line where Each

Circuit is Supported by its Own Tower

This special double-circuit representation is an approximation, but the idea is not too different
from approximating a transposed single-circuit line as a balanced line, rather than representing it as
three untransposed sections.
In general terms, the special double-circuit representation of this section is to be preferred if the
two circuits are physically separated, as in Figure 7.8; if both circuits are supported by the same
tower as in Figure 7.7, it is questionable if this has any advantage over the conventional
continuously-transposed representation of Section 7.3.1.
Mathematics of Double-Circuit Model: Mathematically, the impedance matrix has

the structure of Figure 7.6, with three distinct impedances: Zs, Zm, and Zp. The capacitance
matrix has the same structure with Cs, Cm, and Cp. This is to be contrasted with the continuouslytransposed model of Figure 7.5, where there are only two distinct impedances Zs, Zm. Several
specific points concerning this difference follow:
Point 1:
The two circuits of the double-circuit component representation are assumed to be

identical. That is, rows and columns 1, 2, and 3 of Figure 7.5 can be interchanged
(permuted) with those of 4, 5, and 6 without altering the matrix.
Point 2:
Zs and Zm are the self and mutual impedance associated with either one circuit taken
separately.
Point 3:
Zp is the mutual impedance between any one of the three conductors in one circuit and
any one of the three conductors of the other circuit.
Point 4:
The three distinct phase impedances Zs, Zm and Zp can be indirectly specified by
means of the three distinct modal coupling parameters ZG, ZL, and ZIL. The relationship is:
ZG = Zs + 2Zm + 3Zp
ZL = Zs - Zm
(6.7)
ZIL = Zs + 2Zm - 3Zp

Identical relationships hold true for capacitances. Equation (7.7) describes modal
parameters of a six-phase line with the matrix structure of Figure 7.6. These
parameters are related to the positive and zero sequence parameters of two parallel
three-phase lines, as customarily used in short-circuit programs:
ZG = Z 0 + Z 0-0
ZL = Z 1
(6.8)
ZIL = Z 0 - Z 0-0
7 - 12

where Zo and Z1 are the zero and positive sequence impedances of each (identical)
three-phase circuit, and Z0-0 the zero sequence coupling between the two circuits.
Since the parameters Z0, Z1, and Z0-0 can be directly obtained from the "LINE
CONSTANTS" routine of the EMTP support program AUX by asking for symmetrical
components, users may prefer Eq. (7.8) to obtain the input data, rather than do the
averaging discussed in Section 7.3.5.
Branch Data Format for Double-Circuit Distributed Line: Data preparation

and input for the double-circuit distributed-parameter line component is the same as that for a
continuously-transposed line (see Section 7.3.1) of six conductors, except for the following:
1.
The three conductors of one of the circuits must be numbered -1, -2, and -3;
conductors of the other circuit are then numbered -4, -5 and -6.
2.
Modal parameter values are to be punched on the first three data cards
(corresponding to the conductors -1, -2, and -3) in the following order:
(A)
The "G" mode goes on the card of conductor 1, and the "L" mode goes on
the card of conductor number 2. Note that for the special case of Zp and
Zs being equal, this data input is absolutely identical to what would have
been done for the continuously-transposed line model.
(B)
The "IL" mode goes on card number 3 (that for the third conductor).
(C)
Cards 4 through 6 are to be left blank, for columns 27-80.
The fields and units used for such data of columns 27-80 are identical to those for the
constant distributed parameter line (see example in Section 7.3.5).
7.3.5
Examples of Special Double Circuit Lines
Consider a 180-mile double-circuit line having the one-line diagram shown below. Note that
termination of both lines is on the same 3phase bus "SEND" at one end, while on different busses
("REC1" and "REC2") at the other end.
7 - 13
(1)
REC 1
(2)
SEND
Figure 7.9:
REC 2
Double-Circuit Line
Assumed modal parameters for this line are taken to be:

R [ohm / mile]
L [H / mile]
C [F / mile]
Mode G
0.58
0.00340
0 .0091
Mode L
0.034
0.00068
0 .0166
Mode IL
0.035
0.00110
0 .0116
Data cards for input to the EMTP then might appear as follows:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
-1SENDA REC1A
.58
3.4
.0091 180.
0
-2SENDB REC1B
.034 .68
.0166 180.
0
-3SENDC REC1C
.035 1.1
.0116 180.
0
-4SENDA REC2A
-5SENDB REC2B
-6SENDC REC2C
The assumed structure of the line-constants matrices of Figure 7.6 is an approximation which
makes the problem tractable for easy computer solution; it is not really the result which one finds
from running the Line Constants routine for any particular double-circuit line. Yet it does provide
a reasonably good model for the case of identical circuits supported by different towers, as the
following example shows.
Consider two 3-phase 500kV circuits which are separated by 125 feet (distance between centre
lines). The resulting impedance matrix [Z] for 50Hz is found by the support routine "LINE
CONSTANTS" to be:
7 - 14
1.03245E-01
8.95662E-01
7.59838E-02 1.01856E-01
4.53672E-01 8.97228E-01
7.66662E-02 7.59844E-02 1.03246E-01

4.31747E-01 4.53669E-01 8.95656E-01
7.84993E-02 7.58673E-02 7.66138E-02 1.03246E-01

3.26215E-01 3.42918E-01 3.70026E-01 8.95656E-01
7.57433E-02 7.51370E-02 7.58673E-02 7.59844E-02 1.01856E-01

3.98823E-01 3.27779E-01 3.42918E-01 4.53668E-01 8.97228E-01
7.63482E-02 7.57433E-02 7.64993E-02 7.66662E-02 7.59838E-02 1.03245E-01

2.85776E-01 3.08823E-01 3.26215E-01 4.31747E-01 4.53672E-01 8.95662E-01
An arithmetic averaging of these elements so as to fit the Figure 7.6 model produces the following
mean values:
Zs = 1.02782E-01 + j8.96182E-01
Zm = 7.62115E-02 + j4.46363E-01
Zp = 7.60354E-02 + j3.27722E-01
Deviations of the actual values from these mean values then are summarized by the following
statistics:
Quantity
Maximum Fractional Deviation

from the Sample Mean
Sample Standard Deviation as

Fraction of the Sample Mean
Rs , Ls
0.90% , 0.12%
0.63% , 0.08%
Rm , Lm
0.60% , 3.27%
0.42% , 2.32%
Rp , Lp
1.18% , 9.75%
0.60% , 6.43%
It is the average values just stated which are next converted to modal values, using the formulas of
Point 4 (Section 7.3.4). One finds .483311+j2.772074, 0.0265705+j.449819 and
.0270988+j.805742 for modes "G", "L", and "IL", respectively.
An analogous calculation would of course have to be performed for the capacitance matrix.
7 - 15
7.4 JMARTI Line Model

7.4.1
Data Requirements
The first step in the specification of an N-conductor distributed-parameter transmission circuit

using the JMARTI line model, is to obtain the punched-card output of the EMTP support program
AUX which corresponds to the geometry (including the line length and ground resistivity) of
interest.
SENDA
SENDB
SENDC
Parameters
for JMART
FrequencyDependent
Representation
RECA
RECB
RECC
The resulting cards should conform to the following rules:

Rule 1:
There is to be a branch card for each branch (mode), punched according to the following format:
I2
A6
BUS2
BUS3
BUS4
SKIP
A6
A6
A6
E6.2
I2 I2
ITYPE =
IPOSE
BUS1
IPUNCH
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 34567890123456789012 34 56 789012345678901234567890
(1-2)
-1, -2, -3,..., -N until N < 10. If N is greater than 9. Use -A, B, -C,-D, ..., -I for values of itype between -10 to -18.
BUS1, BUS2
The names of the nodes at both ends of the branch.
(3-8, 9-14)
BUS3, BUS4
The names of the nodes of the reference branch.
(15-20, 21-26)
SKIP
=0
All the data input of poles and residues of the characteristic

impedance Zc and propagation function A1 for the present
mode (described in the following Rule 2 through Rule 7) will
be printed out as part of the output.
=1
Only the data inputted under Rule 2 and Rule 5 will be

included in the output listing.
(27-32)
7 - 16
IPUNCH
=2
All the data inputted under Rule 2 through Rule 7 will be

skipped in the output listing.
= 2
serves as the flag for a JMARTI branch.
=0
or blank - Line is assumed to be transposed.
=N
Number of phases of an untransposed line.
(53-54)
IPOSE
(55-56)
The just-described branch (mode) is to be immediately followed by the data cards explained in
Rules 2 through 7.
Rule 2:
Then comes a card with the number of poles of Zc and its value at infinite frequency.
1
2
3
4
5
6
7
8
12345678 90123456789012345678901234567890 1234567890123456789012345678901234567890
NPZ
AKOZC
I8
E32.20
The number of poles of the characteristic impedance Zc.Flk
NPZ
(1-8)
AKOZC
The value of Zc at infinite frequency.
(9-40)
Rule 3:
Then come a number of cards which carry the values of the residues of NPZ poles of
Zc. Each card can carry up to 3 residues in the format of E26.0.
1
2
3
4
5
6
7
8
12345678901234567890123456 78901234567890123456789012 34567890123456789012345678 90
VOLTBC( II)
E26.0
Rule 4:
VOLTBC(I +1)
E26.0
VOLTBC(I +2)
E26.0
The values of all the poles of Zc are inputted in the format of E26.0. The corresponding residues are inputted under Rule 3.
7 - 17
1
2
3
4
5
6
7
8
12345678901234567890123456 78901234567890123456789012 34567890123456789012345678 90
VOLTBC( II)
VOLTBC(I +1)
E26.0
Rule 5:
VOLTBC(I +2)
E26.0
E26.0
Then comes a card with the number of poles of A1 and the travel time of the branch.
1
2
3
4
5
6
7
8
12345678 90123456789012345678901234567890 1234567890123456789012345678901234567890
NPA
TAU
I8
E32.20
(1-8)
The number of poles of the propagation function A1 of the

branch.
TAU
The travel time of the branch.
NPA
(9-40)
Rule 6:
Same format as described in Rule 3. Only now the residues are for the poles of the
propagation function A1.
Rule 7:
Same format as described in Rule 4. Only now the poles are for A1.
Rule 8:
Repeat the data cards from Rule 1 to Rule 7 for each of the JMARTI branches.
Rule 9:
Only when IPOSE is non-zero on JMARTI branch card, the transformation matrix [Ti] is
read in after all the JMARTI branch cards are inputted according to Rules 1 to 8. The
elements of [Ti] are read in by rows (row 1 first, then row 2, etc.). For a given row, the
real part of the element for all columns come first; then the imaginary follows by starting on a new card. NOTE THAT THE IMAGINARY PART OF [TI] MUST ALWAYS BE
ENTERED, EVEN THOUGH IT IS IGNORED BY THE EMTP. Within each row, elements are
read in the order of increasing column number. Each row begins with a new card, and
there are 6 or fewer elements per card.
1
2
3
4
5
6
7
8
123456789012 345678901234 567890123456 789012345678 901234567890 123456789012 34567890
TI (J,K)
E12.0
TI (J,K+1)
E12.0
TI (J,K+2)
TI (J,K+3)
TI (J,K+4)
TI (J,K+5)
E12.0
E12.0
E12.0
E12.0
7 - 18

If the transmission line has only 3 phases, the final three fields (columns 3772) are
not used. For N less than or equal to 6, exactly 2N data cards for [Ti] are involved; for
N greater than or equal to 7, and less than or equal to 12, exactly 4N data cards are
involved, etc. Note that there must be exactly 2NN elements for [Ti].
Note:
Theoretically, the current transformation matrix [Ti] is complex as well as frequency

dependent. In the time-step loop, the EMTP can handle only constant and real
transformation matrices. By default, the modal transformation matrix automatically
selected by AUX, is calculated at 1.2 kHz or so (depending on the line configuration).
Keep in mind that the constant transformation matrix assumption is an approximation.
The accuracy of this approximation depends on the line configuration, and the type of
simulation. In general, higher errors can be expected at frequencies below 100 Hz
(for details see Reference 8).
7.4.2
Example
On the following page is an example of a set of Marti's branch cards which are generated on the
punched output by running the EMTP support program AUX for an untransposed line:
C
PUNCHED CARD OUTPUT OF JMARTI SETUP WHICH BEGAN AT 10.43.27 10/28/87
C ENGLISH
C
1.3157 .156800 4
.9500
0.00 103.08
80.67
20.
45.
C
2.3157 .156800 4
.9500 -21.00
67.08
44.67
20.
45.
C
3.3157 .156800 4
.9500 +21.00
67.08
44.67
20.
45.
C
0.5000 3.47000 4
.4950 -22.50 121.50 104.08
ST1X2
C
0.5000 3.47000 4
.4950 +22.50 121.50 104.08
ST1X2
C
C
200.0
500.
1
250.25
1
C
200.0
60.
1
250.25
1
C
200.0
.06
1
250.25
7 10
1
C
-1
2PA
1LA
1.
-2 3
21
0.24152036054498543294E+03
0.166158032894826892E+03 0.105654039484767281E+04 0.222963352743750246E+03
-0.459016029960812135E+02 -0.240015385272436621E+03 0.205282472635053614E+01
0.266488083740615355E+04 -0.126743913527252300E+04 -0.775861791755235348E+02
0.292668577030444279E+02 0.489493248727771357E+02 0.233192096884617662E+03
0.258054061325210291E+03 0.181145307088604071E+03 0.327204892486895631E+03
0.782146076957718250E+03 0.700926154520921420E+04 -0.188897791957832190E+03
0.152367356049145806E+04 0.345788242449950729E+04 0.797700079602043043E+05
0.187101360693176031E-02 0.759421838374094521E+00 0.993615936464874089E+00
0.124469700766352670E+01 0.127876130284928663E+01 0.164881136494811220E+01
0.553854381163242249E+01 0.589948683075975588E+01 0.640822744216896567E+01
0.733914881928147100E+01 0.860678522066533414E+01 0.147469097348116158E+02
0.190833145267801694E+02 0.258060123204732781E+02 0.409354195297442205E+02
0.155493338922723456E+03 0.635287585050358800E+03 0.643898459236923031E+03
0.736580325616889397E+03 0.278654503811752801E+04 0.636505096224125591E+05
16
0.13680807949881993823E-02
0.377661839973843442E-04 0.201502016119700501E-01 0.187033510649276629E+00
0.792022679405346974E+01 0.237906450182637039E+02 0.201568059293244108E+03
7 - 19

0.324374055857230837E+04 0.440770715729592858E+03 0.836111853682627020E+04
0.779930376633985338E+04 0.138611853787623463E+05 0.360896386026898053E+05
0.109543531511294401E+06 -0.878133434579960681E+04 0.476176633848973637E+06
-0.253444688665194190E+06
0.119025767042912384E-01 0.227638448681862932E+01 0.213206937687088875E+02
0.432912766911082684E+03 0.719739621874220106E+03 0.600489056217948871E+04
0.350217033032571953E+05 0.388859746140533252E+05 0.100456852658521037E+06
0.133265979587640650E+06 0.204382133582062819E+06 0.302429577277157587E+06
0.535192724776360323E+06 0.606416620048329802E+06 0.123498140513805603E+07
0.221425590999286057E+07
-2
2PB
1LB
1.
-2 3
26
0.40354347682829263988E+03
0.183176390786193512E+03 0.180043758935006417E+04 0.756735921432968750E+01
-0.839978151040693248E+03 0.156075581237605395E+03 0.780688436741060929E+02
Immediately after the last branch card follow the cards with the transformation matrix [Ti]:
0.87620967
0.00000000
-0.34077604
0.00000000
-0.34077604
0.00000000
0.39921404 0.00000000
0.00000000 0.00000000
0.64831634 -0.70710678
0.00000000 0.00000000
0.64831634 0.70710678
0.00000000 0.00000000
Note the use of a real transformation matrix (all imaginary parts are 0.0).
7.5 LMARTI Cable Model

The LMARTI or FDQ (Frequency Dependent Q matrix) underground cable model takes into account
the frequency dependence of the cable parameters as well as the frequency dependence of the
modal transformation matrix. This model was specifically developed to model single-core directburied multi-phase cables. For more details, please refer to Section 11 of the Rule Book part II.
To create the LMARTI model from cable configuration data, it is necessary to approximate (with
rational functions) the characteristic admittance Yc, and the propagation function A for each mode
of the cable, as well as the modal transformation matrix [Q] in the frequency domain. The
"CABLE MODEL" module of the support porgram AUX, is used to create the model.7.5.2 Examples
7 - 20
7.5.1
Data Requirements
The first step in the specification of an N-conductor distributed-parameter underground cable

using the LMARTI cable model, is to obtain the punched-card output of the EMTP support program
AUX which corresponds to the geometry (including the line length and ground resistivity) of
interest.
SENDA
SENDB
SENDC
Parameters
for LMART
FrequencyDependent
Representation
RECA
RECB
RECC
The resulting cards should conform to the following rules:

Rule 1:
There is to be exactly NPHS cards (where NPHS is the number of phases/modes)

punched according to the following format:
I2
A6
A6
A6
A6
ITYPE
Filename
A28
I2 I2
I2 I2
Filename (continued)
IOUT
BUS4
IDEBUG
BUS3
ICHOP
BUS2
NPHS
BUS1
IMODEL
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 78901234567890123456789012 34 56 78 90 1234567890123456789 0
A13
I1
ITYPE = -N where "N" is the N-th branch of a given line/cable unit
and 1 < N < 61. If 10 < N < 35, then upper case letters (in ascending
alphabetical order) are used; that is:
A ITYPE = 10
B ITYPE = 11
Y ITYPE = 34
Z ITYPE = 35
If 36 < N < 61, then lower case letter (in ascending alphabetical
order) are used; that is:
a ITYPE = 36
7 - 21

b ITYPE = 37
y ITYPE = 60
z ITYPE = 61
BUS1
(3-8)
BUS2
(9-14)
BUS3
(3-8)
BUS4
(9-14)
FILENAME
(25-52,
61-74)
The names of the nodes of the reference branch (the node names of
the first branch of a previously defined LMARTI cable model. Note
that IMODEL and NPHS in the current branch "must" match those of
the reference branch. Since no storage space is saved using this
option, and since the same filename can be used in the field
FILENAME as many times as necessary, there is no real need to use
this option.
Name of the file which contains the cable data in rational-functions
form. This file is obtained directly from the punched output of the
EMTP support program AUX.
If the field within columns 27 to 52 is not sufficient for the name of
the file, then columns 61 to 64 can be used as a continuation field
(blank characters within the file name are not allowed). If
FILENAME is left blank, then the fitted-functions data will be read
immediately after the last branch card of the current line/cable unit.
This is equivalent to using $INCLUDE "filename" immediately after
the last branch card. Note that FILENAME is ignored from the second
to last branch cards.
IMODEL
(53-54)
NPHS
(55-56)
ICHOP
(57-58)
This field determines which subset of the LMARTI model is to be

used:
IMODEL= 3
The modal transformation matrix [Q] or [Ti] is

assumed to be constant and real.
IMODEL= 4
The modal transformation matrix [Q] or [Ti] is

assumed to be frequency-dependent.
Number of phases of the current line/cable unit. This field is ignored

from the second to last branch cards.
This flag controls the option to reduce the order of the rational
functions approximations of Yc' and A' according to the time-step of
the simulation.
7 - 22

ICHOP = 0
(default) The order of the approximations of Yc' and

A' will be reduced according to the Nyquist
frequency of the simulation (Nyquist frequency = 1/
(2*t)).
ICHOP = 1
No truncation occurs.
Note that if the simulation stops after the steady-state calculations,

then no truncation occurs. ICHOP only affects the approximations of
Yc and A'. The order of the approximation by rational functions
approximation of the elements of transformation matrix is never
reduced. This field is ignored from the second to the last branch
cards.
IDEBUG
(59-60)
This flag controls the amount of printed output pertaining to the

processing of LMARTI lines/cables at input time.
IDEBUG = 0
No output is produced.
IDEBUG = 1
The modal transformation matrix (for IMODEL=-3).

Modal time displacement constant (numerically
similar to the modal travel time). Average number
of terms truncated when ICHOP=0.
IDEBUG = 2
Number of terms truncated on the approximation of

each element of Yc' and A', as well as the Nyquist
frequency when ICHOP=0.
IDEBUG = 3
Verbose record of the approximations by rational

functions of the elements of Yc' and A' (after
reduction if ICHOP=0) and modal transformation
matrix.
IDEBUG = 4
This is the highest level of output production. At the

moment it produces debugging information for the
order-reduction procedure (ICHOP=0).
Note that IDEBUG=3 also prints the information indicated for

IDBEUG=2 and IDEBUG=1. Similarly, IDEBUG=2 also includes the
information printed when IDEBUG=1. This field is ignored from the
second to the last branch cards.
IOUT
(80)
Has the same meaning as IOUT in the lumped RLC elements of

Section 5.1.
IOUT = 1
Branch currents will be included on output. Note

that sending and receiving end currents are printed
following the sign convention shown in Figure 7.10.
7 - 23

IOUT = 2
Branch voltage vk(t)-vm(t) will be included on

output.
IOUT = 3
Branch currents and branch voltages will be included

on output.
Im
Ik
node k
node m
Vk
Vm
Figure 7.10: Currents at the sending and receiving ends
Note:
Columns 75-79 must be left blank.
Rule 2:
The remaining cards which describe the approximations of Yc', A' and the modal transformation matrix are to be contained in a file whose name has been specified in the
field FILENAME. This data is punched automatically and its format will not be discussed here.
7.5.2
Examples
C <-bus1<-bus2<-bus3<-bus4<--- filename ------------<-<-<-<-<--filename cont.--<

-1core1acx001a
400pplplm1.pch
-4 6
1
-2grx001sx001a
-3core1bcx001b
-4grx001sx001b
-5core1ccx001c
-6grx001sx001c
C
-1cx001acx002acore1acx001a
-4 6
-2sx001csx002c
-3cx001bcx002b
-4sx001asx002a
-5cx001ccx002c
-6sx001bsx002b
Note : The reference branch option has been used in the second cable. Also note that for a threephase direct buried cable, the conductor sequence is core1, sheath1, core2, sheath2, core3, sheath3.
7 - 24

File 400pplplm1.pch contains the model data generated by the CABLE MODEL module of AUX. A
portion of this file is shown below:
Approximation of element
1 of Yc
17
0.3339555118542865D-01
0.2223677279208564D-08 -0.6684726628398290D-04
0.7384409391633254D-03 -0.3056914502439972D-02
-0.7298927389063723D-02 -0.2858774571664257D-01
0.8442824307049253D-01 -0.2881790434770883D-01
-0.1690571980956414D+01 -0.2365588209355900D+01
-0.7946774587083623D+01 -0.9785979113856933D+02
0.1986512447084994D-02
0.5871791759167237D+00
0.1274630805149068D+01
0.1498229131117012D+01
0.5089053892138579D+01
0.7570728553905202D+01
0.9603847609306016D+02
0.1009163238193162D+03
0.2903809497895774D+03
0.9004557218803037D+03
0.1226173154995020D+05
0.1513500158403253D+06
C
2 of Yc
15
0.9550240017317235D-02
-0.3054120340852447D-07 -0.9321185244931170D-04
-0.5650469156851982D-03 -0.3239571844278076D-01
.
.The rest of the elements of Yc follow...
.
.C
1 of A
27
0.1844849658552810D-04
0.1786600986956503D-09
0.1498895887849891D-06
0.6540243805399673D-06
0.2049320091295968D-05
0.1572154297917118D-04
0.3384480312905494D-04
0.2943090227818116D-03
0.1464468845799371D-02
0.1862921017770613D-01
0.1686836598109972D+00
0.8612468825196593D+00
0.7261158030490222D+01
.
.The rest of the elements of A follow...
.
C
1 1 of Q
6
-0.5243117677914846D+00
0.4368676581007832D-05
0.8439369552329936D-03
0.1100164044052870D+01
0.2473131884090338D+02
0.1991207479843018D-02
0.4693995043383073D+00
0.3033031962201489D+03
0.6777886767395064D+04
C
2 1 of Q
4
0.5243044625844220D+00
0.2890451978418303D-06 -0.3261833325015998D+03
-0.1083757134952251D+03
0.1986455815471463D-02
0.8022169617706264D+02
7 - 25
-0.2647276426251857D-04
0.5127772262033974D-02
-0.4412886474055768D-01
-0.1442501860944369D+01
-0.2659464793002834D+01
0.9856600971551481D+00
0.4763668332710630D+01
0.1101518064677846D+02
0.1392947075436179D+03
0.1899621522766753D+04
-0.1373833807181917D-03
0.3949144684554902D-01
0.2202117214337431D-06
0.7314225510421220D-05
0.1358009200545962D-03
0.7980331544034139D-02
0.1424096799671657D+01
0.1095428521156367D+03
0.1956723254606404D-01
0.1539419540773927D+03
0.1084467378093935D+02
0.4176351539378856D+05
0.3658882418768325D+03
0.8077328248000710D+02

0.1060551710254020D+03
1 of Q
1
0.1000000000000000D+01
0.0000000000000000D+00
0.1000000000000000D+01
.
.The rest of the elements of Q follow...
.
To simulate a cable assuming that the modal transformation matrix Q is constant and real set model
= -3.
C <-bus1<-bus2<-bus3<-bus4<--- filename ------------<-<-<-<-<--filename cont.-<
-1core1acx001a
400pplplm2.pch
-3 3
1
-2core1bcx001b
1
-3core1ccx001c
1
C
-1cx001acx002acore1acx001a
-3 3
-2cx001bcx002b
-3cx001ccx002c
Where file 400pplplm2.pch contains the rational function approximations of Yc and A, and a
constant Q, as illustrated below:
Partial contents of file 400pplplm2.pch:
APPROXIMATION OF YC, MODE
25
2.5722103746361303D-02
-9.1398897276766932D-06
-2.5510582177052298D-05
-1.8725863256199870D-04
-1.2911588871813812D-03
4.6599760683497870D-02
-3.4026640618059707D-02
-3.5497987809727067D-01
-6.0950166147299063D+02
-5.5155764731978083D+05
6.5924871369958270D-03
1.8394084249973127D-01
5.3482764962293816D-01
1.3559561913443330D+00
5.2661746991350418D+00
9.4169625411677313D+00
7.2495738480539536D+02
4.8923322467790282D+06
4.7947903261815631D+08
-4.7732419642510682D-06
4.5182017907377946D-05
-3.5487865421470267D-04
-2.8587520169833231D-03
-5.4609788012209141D-02
-4.5358605682249719D-02
-2.8320054298228014D+00
-3.8099092340431350D+04
2.4202419850084102D-05
-3.9818936513439709D-04
-4.3932154938838136D-04
-4.5595739951930064D-03
-3.7652946736098552D-02
-1.6930689205273065D-02
-7.9897666513396406D+00
-2.5081032355653864D+05
1.2793111735678223D-01
2.2755119316504574D-01
7.4808301945931244D-01
1.8647732694238314D+00
5.3910101049294138D+00
1.4250926240921896D+01
1.1476564331069469D+04
7.7330708753132582D+07
1.6868480586342925D-01
2.6383476691742730D-01
1.0459708285568725D+00
2.8949523353067189D+00
7.1799372131564200D+00
2.4437958645343056D+01
6.4138070958513677D+04
2.4903807389847651D+08
.
.The rest of the elements of Yc follow...
.
7 - 26

C
APPROXIMATION OF AP, MODE
27
1.4204032678206035D-04
2.2332138559952998D-07
7.6264241520928741D-05
6.8482753782447720D-04
1.1476096108158689D-02
4.6994552886982888D-01
3.1208726249524719D+02
4.9100303877194619D+03
6.1584551958969116D+04
8.2119854208426476D+08
1.9868059829050395D-03
5.4573666067375422D-01
4.8959451674433011D+00
8.4460006221681988D+01
7.6924045903434384D+02
1.6388509551340768D+04
1.2714138220962210D+05
6.6450369054580760D+05
3.4930759989122635D+06
1.3701856289354453D-05
1.4242504228592440D-04
1.9076840502814071D-03
6.2401121797168410D-02
8.2199132206362844D-01
1.9069946110165149D+02
9.4925383820301521D+03
2.1904793609980655D+05
-8.2737807923962355D+08
1.9603027928962263D-01
1.0209950917730483D+00
6.8339896171366584D+00
2.2566844476789343D+02
7.7953601437854052D+02
2.1496676261269975D+04
2.3262423149958560D+05
1.0989689270439534D+06
3.4965362466765195D+06
1.9020297585217422D-05
9.5674980316514301D-04
1.6963197003467395D-03
7.5571060506626976D-01
1.8668880613207570D+01
1.2590033397309805D+03
3.9133442105124277D+04
5.4875784844188547D+06
3.5600758605173695D+05
2.7221187863356757D-01
3.4232567262692291D+00
1.2304901876088393D+01
6.6275546410586476D+02
4.1273408717570293D+03
3.6052040439918047D+04
4.2823307527454871D+05
2.7790446910967064D+06
1.4074761811830206D+07
.
.The rest of the elements of A follow...
.
C
REAL TRANSFORMATION MATRIX Q AT 1.0000E+03 HZ
C
LISTED BY COLUMNS: Q(1,1), Q(2,1), ..., Q(1,2), Q(2,2), ...
7.0710678118660831E-01 -1.2070099351709118E-13 -7.0710678118648696E-01
5.7732216739095044E-01
5.7740646868391633E-01
5.7732216739095080E-01
4.0828802950727069E-01 -8.1645683898291242E-01
4.0828802950747961E-01
C
7.6
EXACT-PI Model for Frequency Scan Calculations
The EXACT-PI model provides an exact single-frequency representation of an overhead line or an

underground cable in terms of a multiphase pi-equivalent. The model can be used by the EMTP
when frequency scans or single-frequency steady-state solutions are requested. The model is
punched out as a Y-matrix that includes the series and shunt branches of the multiphase EXACT-PI
circuit. This Y matrix is produced for each frequency point in a logarithmic or linear frequency
interval.
It is important to realize that the EXACT-PI model is only a frequency domain representation of a
cable at a given frequency. It is not an EMTP time domain model and it can only be used in
frequency scans or steady-state calculations. On the other hand, the EXACT-PI model is correct
(inasmuch as line or cable parameters are correct) at each frequency since there are no
approximations involved.
7 - 27
7.6.1
Input Data Requirements
To run a frequency scan with the exact-pi model, the special request card FREQUENCY SCAN must
be included in the EMTP data deck. The format and usage of this card is described in Section 4.1.1
and is summarized below.
1
2
3
4
5
6
7
8
123456789012345678901234 56789012 34567890 12345678 90123456 7890123456789012 34567890
FREQUENCY SCAN
A24
fmin
fmax
NPDEC
MODSYM
E8.0
E8.0
E8.0
I8
A16
fmin
(2532)
Starting frequency of the scan, in Hertz. This must be a positive

number.
f
(3340)
The frequency increment for linear spacing between adjacent

frequency points. In this case, fk+1 = fk + f. If logarithmic
spacing between adjacent frequency points is desired, this field
should be left blank.
fmax
(41-48)
Maximum frequency of the scan, in Hertz. This must be greater

than or equal to fmin.
NPDEC
For linear spacing of the frequency points, leave this field blank.
For logarithmic spacing, NPD is the number of points per decade.
In this case, adjacent frequency points are related by:
(4956)
fK+1 = (101/NPD) . fK
MODSYM
Set MODSYM=FREQUENCY-DOMAIN for EXACT-PI modelling.
(57-62)
Note that MODSYM is a global flag in the sense that it is not
associated with any particular model. Therefore, it is not possible
to mix "FREQUENCY-DOMAIN" and "TIME-DOMAIN"
representations in the same simulation. In the future, all EMTP
models will have the provision to include correct frequencydomain representations of their normal time-domain forms.
It is important to match the frequency points generated in AUX (see Pi-Exact Line Model in
Section 9.6.1, Rulebook II) with the frequency points expected in the EMTP (i.e., fmin, fmax, f
and NPDEC in columns 25 to 56 above). Otherwise, an error condition will occur.
Additionally, the appropriate cards to describe the cable itself in the EMTP must be included. For
the EXACT-PI model the FDQ or LMARTI cable model cards should be used, as explained in Section
7.5.1 of the EMTP Reference Manual. A brief description is also included below:
7 - 28
I2
A6
A6
A6
A6
Filename
A28
I2 I2
I2 I2
IOUT
BUS4
IDEBUG
BUS3
ICHOP
BUS2
NPHS
BUS1
IMODEL
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 78901234567890123456789012 34 56 78 90 1234567890123456789 0
A13
I1
(1-2)
Set ITYPE to -1, -2, -3, ...-N for conductors 1, 2, 3, ...,N. If

10 N 35 use -A, -B, -C, ..., -Z. If 36 N 61 use -a, -b, c,...,-z.
BUS1
Sending end bus name.
ITYPE
(3-8)
BUS2
Receiving end bus name
(9-14)
BUS3
Reference bus, sending-end.
(15-20)
BUS4
Reference bus, receiving end.
(21-26)
FILENAME
(25-52, 61-74)
IMODEL
(53-54)
Name of the file that contains punched output from AUX. If this
field is blank, punched output from AUX must be appended
immediately following the last conductor card.
Set to -4 for EXACT-PI or for FDQ model options. Set to -3 if [Q]
is constant.
(55-56)
Number of conductors. For example, if a 3-phase cable has two

conductors per phase (e.g., core and sheath) then set NPHS to 6.
ICHOP
Does not apply to EXACT-PI model.
NPHS
(57-58)
IDEBUG
Does not apply to EXACT-PI model.
(59-60)
IOUT
(80)
7.6.2
Set IOUT=1 to produce branch current output at the sending and

receiving ends of the cable. Set IOUT=2 for branch voltage, and
IOUT=3 for both branch currents and branch voltages.
Example
Frequency scan request card:

7 - 29

C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C
<-fmin--<-deltaf<-fmax--<-npdec-<--MODsYM------FREQUENCY SCAN
10.
1000.
10FREQUENCY-DOMAIN
The branch cards for the cable follow

C <-bus1<-bus2<-bus3<-bus4<--- filename ------------<-<-<-<-<--filename cont.-<
-1core1acore2a
400exactpi.pch
-3 6
1
-2core1bcore2b
1
-3core1ccore2c
1
Partial contents of file 400exactpi.pch follow:

C
1.00000E+01
0.73711E+02
-0.67119E+01
-0.26081E+02
-0.17029E+01
-0.33013E+02
0.80545E+00
0.63200E+01
0.19200E-02
0.44822E-07
-0.26814E-19
0.28837E-18
0.56029E-14
0.11095E-05
-0.95506E-19
0.19467E-18
0.45823E-13
Hz. ---> Series-Real: First frequency point

-0.10005E+02 0.27102E+02 -0.26081E+02 -0.67119E+01
-0.17029E+01 -0.91537E+01 0.27280E+02 -0.20270E+02
-0.67119E+01 0.73711E+02 -0.53771E+01 -0.14038E+01
-0.10005E+02 0.27102E+02
---> Series-Imag:
-0.60751E+01 -0.18084E+01 0.63200E+01 0.80545E+00
0.19200E-02 -0.48842E+01 -0.14852E+01 0.10316E+02
0.80545E+00 -0.33013E+02 0.19778E+01 0.33487E+00
-0.60751E+01 -0.18084E+01
---> Shunt -Real:
-0.44822E-07 0.14118E-06 0.28803E-18 -0.27180E-19
0.56029E-14 -0.44822E-07 0.14118E-06 0.13189E-18
-0.27397E-19 0.44822E-07 -0.50008E-19 0.54695E-14
-0.44822E-07 0.14118E-06
---> Shunt -Imag:
-0.11095E-05 0.58798E-05 0.19486E-18 -0.96023E-19
0.45823E-13 -0.11095E-05 0.58798E-05 0.74764E-19
-0.96056E-19 0.11095E-05 -0.10567E-18 0.42521E-13
-0.11095E-05 0.58798E-05
0.77624E+02
-0.53771E+01
-0.67119E+01
-0.28718E+02
0.19778E+01
0.80545E+00
0.44822E-07
-0.50161E-19
-0.26774E-19
0.11095E-05
-0.10450E-18
-0.95247E-19
C
Hz. ---> Series-Real: Second frequency point
-0.10572E+02 0.26988E+02 -0.25320E+02 -0.65953E+01 0.75481E+02
-0.16870E+01 -0.95341E+01 0.27208E+02 -0.18758E+02 -0.50774E+01
-0.65953E+01 0.70782E+02 -0.50774E+01 -0.13456E+01 -0.65953E+01
-0.10572E+02 0.26988E+02
---> Series-Imag:
-0.34336E+02 -0.62381E+01 -0.19035E+01 0.78283E+01 0.11137E+01 -0.30212E+02
0.11137E+01 0.58583E-01 -0.50496E+01 -0.15716E+01 0.11540E+02 0.22610E+01
1.12202E+00
0.70782E+02
-0.65953E+01
-0.25320E+02
-0.16870E+01
.
.The rest of frequency points follow...
.
7 - 30
7.7 Corona Model

7.7.1
Introduction
Corona modelling is used for an accurate knowledge of the magnitudes of lightning and switching
overvoltages in an optimized design of transmission lines. Corona has a significant effect on
overvoltage and on wave propagation. The currently available methodology for corona modelling
relies heavily on experimental testing of transmission lines for extracting the charge-voltage
characteristic.
The physical phenomenon of corona is very complex (see Reference 12 and Reference 28). It
includes ionization, effects of mobility, diffusion, deionization and the mutual effect of space
charges and electric field. Sophisticated transmission line models are currently available in the
EMTP (see Reference 8).These models can accurately represent the distributed nature and the
frequency dependence of transmission line parameters. The representation of corona, however,
involves a distributed nonlinear hysteresis behaviour and is difficult to combine with the EMTP
transmission line mathematics.
Most EMTP type corona models are based on the representation of the macroscopic effects of the
corona phenomenon (see Reference 3). The charge-voltage (q-v) response of a conductor can be
used to characterize the corona phenomenon. Extensive experimental studies on the corona
characteristics of single and bundled conductors under impulse voltage conditions in a large cage
are available in Reference 19. Four line conductors were examined: 2 single 1.2 and 1.823
diameter conductors, a bundle of 4x1.2 diameter conductors and a bundle of 6x1.1823 diameter
conductors. The applied impulse waveforms of both polarities were covering the range of
switching and lightning surges: 260x2500s, 75x2500s, 15x1000s and 2.5x60s. A typical q-v
curve is shown in Figure 7.11
In Figure 7.11 the corona onset voltage is given by V0, E0 is the corona onset voltage gradient and
Vcr is the applied impulse crest voltage. Corona occurs when the strength of the electric field on
the surface of a conductor becomes sufficient to ionize the surrounding air. The electric field
intensity at the onset of corona can be predicted from the Peeks formula:
0,308
E 0 = 31m 1 + ------------r
(kV/cm)
(7.1)
where m is the irregularity factor (0.7 for fair weather and 0.5 for rain, see Reference 19), d is the
relative air density factor (3.92xatmospheric pressure (in cmHg) over temperature in Kelvin) and r
is the conductors radius in cm.
7 - 31
q (mC)
V0
Vcr1
Vcr2
conductor voltage v (kV)
Figure 7.11:
A typical q-v characteristic of a conductor
If the geometrical capacitance of a conductor (or bundle) is given by C0 and C is the apparent
capacitance after corona onset, then the following experimental observations are found in
Reference 19:
rain conditions lower the dissipated energy (the area of the q-v curve)
E0 is lower under rain conditions
the slope after corona onset is less steep and the area enclosed by the q-v curve is smaller for a
negative polarity impulse compared to a positive polarity impulse
E0 is lower for impulses of positive polarity
E0 decreases as the conductor diameter increases
E0 is lower for a single conductor compared to a bundle of conductors
the corona onset level increases for impulses with steeper fronts
C/C0 is a voltage independent ratio and it is found to be higher for shorter impulses, for conductors of larger diameter, fair weather conditions and for a decreasing number of conductors
in a bundle
the q-v characteristic for fast impulses (or high frequency ac) is narrower than for slow
impulses
Peeks formula is in good agreement for switching impulse measurements, but for faster
impulses the measured corona onset gradient is 10 to 15% higher.
7.7.2
The Suliciu model
The nonlinear q-v characteristic can be simulated directly using RC circuits and diodes or can be
described by analytical expressions. Such models are static models where a fixed q-v characteristic
is assumed. But, as indicated earlier, experimental data shows that q-v curves are significantly
affected by the rate of rise of the applied surge. The only model that has been so far able to account
for this dynamic behaviour is called the Suliciu model (see Reference 32).
7 - 32

This report presents the implementation of the Suliciu corona model in the EMTP line models (see
Reference 41). The major difficulty for corona modelling is the availability of field tests for
practical transmission line cases. The parameters needed for the Suliciu model are particularly
difficult to establish. The fitting method remains a trial and error method based on q-v
characteristics measured in cages. The module proposed in Reference 16 can be used for
simulating the q-v characteristic of a line with the Suliciu model and comparing with q-v data
from field tests. It requires an initially known set of Suliciu model parameters and can simulate
only a single line section connected to a voltage source and an external resistance to account for
losses.
Basic Assumptions:
To approximately account for the distributed nature of corona, EMTP multiphase transmission line
models must be subdivided into a large number of sections to insert corona branches connected
from node to ground. The choice of a section length must be such that its travel time is a fraction
of the surge rise-time. It is assumed that at any given simulation time-point, the voltage along each
line section is uniform.
The basic corona branch model equations are presented in Reference 16, they are conveniently
recalled and modified according to the proposed solution context.
If x is the radius of a cylinder on which space charge is concentrated when conductor voltage falls
to zero, then for a multiphase system (vectors and matrices at time t):
1
(6.2)
V = Vx + Cr Q
(6.3)
Q = Cx Vx + Qc
1
Cr
(6.4)
= Co Cx
where V is the line end (or section end) voltage, Vx is the voltage inside the cylinder, Q is the total
line charge, Cr is the capacitance of the cylinder to ground, Cx is the capacitance of the line
conductor to cylinder boundary, Qc is the corona charge inside the cylinder and Co is the
1
geometric capacitance of the line. If P r = C r , P x = C x and P 0 = C 0 then the phase voltage

of phase a in a 3-phase system can be found from equation 7.4:
v a = v x + p 0 p x q a + p 0 q b + p 0 q c
a
11
11
12
13
7 - 33
(6.5)

where:
V = va vb vc
Vx =
vx
vx
vx
Q = qa qb qc
p 0 P 0 and
px Px
The first two terms of this equation, represent voltage induced on phase a by its own charge and
the last two terms represent mutual capacitive coupling from other phases. Equations 7.2 and 7.3
can be rewritten using equation 7.4:
V = P 0 C x Vx + Pr Qc
(6.6)
Q = C0V + C0Px Qc
(6.7)
It follows that the corona charge is given by:

Q cor = C 0 P x Q c
(6.8)
and the corona current (the corona branch current) is found from:
I cor = C 0 P x I c
(6.9)
Ic is the corona current vector inside the cylinder and its members can be found from the Suliciu
equation:
7 - 34
0
g2
ic =
d
q =
dt c
g1 + g
0
g4
g3 + g
if g 2 0
state 6
if g 1 0 < g 2
state 2
if g 1 > 0
state 1
if g 4 0
state 5
if g 4 < 0 g 3
state 4
if g 3 < 0
state 3
gj = kj [ ( cj cx ) ( vx vj ) ( qc ) ]
vx > 0
vx 0
(6.10)
j=14
where kj, cj and vj are model parameters, c x C x , v x V x and q c Q c .
Solution method:
The corona branch appears as a nonlinear function connected at intermediate line section nodes.
The compensation algorithm is the most appropriate for solving such a branch. If Vth and Rth are
the Thevenin voltage and resistance respectively, then the circuit equation at the corona branch
nodes is given by:
V th R th I cor = V
(6.11)
The combination of the above equation with equations 7.6 and 7.9 results in:
C x P 0 V x + P r Q c + R th C 0 P x I c = V th
(6.12)
The trapezoidal rule of integration relates charge to current at each solution time-point:
t
Q c = ---2 I c + Q hist
(6.13)
t
Q hist = ---2 I c, t t + Q c, t t
7 - 35
(6.14)

Equation 7.12 now becomes:
t
V x + C 0 P x P r ---2 + C 0 P x R th C 0 P x Ic = C 0 P x V th C 0 P x P r Q hist
(6.15)
Equation 7.15 is solved at each time-point for Vx after finding the corona state and replacing for Ic
according to equation 7.10.
Validation:
This section provides a single case of field test comparison for the previously described EMTP
corona model. Further currently undertaken validation work will be presented in a future report.
The EMTP corona model usage rules are presented in the Section 7.7.3.
The field tests reported in Reference 11 and Reference 13 (Garys line) are measurements of surge
propagation in an actual 12.5km 220kV 3-phase line and laboratory measurements of the conductor
q-v curves. Only phase a is energized with a surge function:
u ( t ) = U m [ 0,988e
0,123t
1,064e
4,1t
sin ( 12,3t + 70 ) ]
(6.16)
Waveforms were originally measured for Um = 850kV and Um = 995kV (see Reference 3), but
only data for the 850kV case is available in this report.
The Suliciu model parameters (for equation 7.10) are:
cx = 8.41 pF/m,
c1 = c3 = 16.8pF/m,
c2 = c4 = 35pF/m,
k1 = k3 = 4MHz,
k2 = k4 = 0.8kHz,
v1 = v3 = 320kV and
v2 = v4 = 220kV
Sample data files are shown in Section 7.7.4.
7 - 36
Validation with the constant parameter line model:

The original line is subdivided into 500 sections of 25 meters. Physical line data with the constant
parameter option is given in the EMTP data file cp25aux.dat. The created line model is stored in
cp25aux.pun. Model parameters are evaluated at 100kHz. The actual simulation data case is found
in cp25.dat. The simulation results with corona are shown in Figure 7.12. It is concluded that the
model can reproduce the field test with a reasonable accuracy. The simulation results without
corona modelling are shown in Figure 7.13. This illustrates that corona modelling provides a
much more realistic estimate of the attained maximum voltages and the rate of voltage rise.
700
1km
___field data
------EM TP
600
3km
Voltage (kV)
500
400
7km
10km
300
200
100
0
Figure 7.12:
10
15
20
25
30
time (seconds*1E-06)
35
40
45
Garys field test, simulation with cp-line (100kHz data) and the
Suliciu corona model
7 - 37
800
___field data
------E M TP
1k m
700
3k m
V oltage (k V )
600
7k m
500
10km
400
300
200
100
0
10
15
20
25
30
tim e (s ec onds*1E -06)
35
40
45
Figure 7.13: Garys field test, simulation with cp-line (100kHz data), no
corona model
Validation with the frequency dependent line model:

The original line is again subdivided into 500 sections of 25 meters. Physical line data with the
frequency dependent parameter option is given in the EMTP data file fd25aux.dat. The created line
model is stored in fd25aux.pun. The Ti matrix is evaluated at 100kHz. The actual simulation data
case is found in fd25.dat. The simulation results with corona are shown in Figure 7.14. The
simulation results without corona modelling are shown in Figure 7.15.
The overall effect of frequency dependence is less important when corona is modelled. When
corona is not modelled, then the parameters of the cp-line model are only correct at one frequency,
but as the surge propagates along the line its frequency content changes and thus a frequency
dependent model becomes more appropriate.
7 - 38
700
1km
___field data
------EMTP
600
3km
Voltage (kV)
500
7km
400
10km
300
200
100
0
10
15
20
25
30
time (seconds*1E-06)
35
40
45
Figure 7.14: Garys field test, simulation with fd-line (100kHz data) and
the Suliciu corona model
___field data
------E M TP
800
1k m
700
3k m
V oltage (k V )
600
7k m
500
10k m
400
300
200
100
0
10
15
20
25
30
tim e (s ec onds *1E -06)
35
40
45
Figure 7.15: Garys field test, simulation with fd-line (100kHz data), no
corona model
7 - 39
7.7.3
Data input rules
The EMTP corona model is based on the insertion of the nonlinear Suliciu corona branch model in
a standard constant parameter (CP-LINE) or frequency dependent (FD-LINE) line model. The
standard line model is subdivided into short sections of equal length. It is assumed that during
transient analysis the voltage along a short section is approximately constant. The length of a
section should be chosen according to this criteria.
The corona line definition has two parts: the usual line model data (multiphase CP-LINE or FDLINE) and the corona line declaration. The line model is the target line for corona branch insertion.
Corona line declaration:
1
2
3
4
5
6
7
8
12 345678 9 0123 4 567890 123456 789012345678901234567890123456789012345678901234567890
CORON
A
LINE
A7
A4
BUSKa BUSMa
A6
A6
CORONA LINE:
keyword for corona line declaration
BUSKa:
identifies the k node of "phase a" of the target line
BUSMa:
identifies the m node of "phase a" of the target line
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
NSECT
NPH
I8
I8
SECTION_LEN
E25.0
NSECT:
number of line sections of the complete line
NPH:
number of phases of the target line
SECTION_LEN:
length of the target line
7 - 40
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890123456789012345678901234567890
Cx (pF/m)
E25.0
Cx:
pF/m, capacitance of the corona cylinder, the

diagonal elements of a multiphase line are identical
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
Vnode_out1
Vnode_out2
Vnode_out3
Vnode_out4
Vnode_out5
Vnode_out6
I8
I8
I8
Vnode_out#
is an intermediate node number for node voltage

output request,
default: leave blank
Node numbering: from 1 to Nsect+1 for phase a,
from Nsect+2 to Nsect+Nsect+2 for phase b, ...
A plot file with the extension .pc4 is created during
the simulation and can hold a maximum of 6 output
requests.
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
c1 (pF/m)
k1 (Hz)
v1 (volts)
c2 (pF/m)
k2 (Hz)
v2 (volts)
c3 (pF/m)
k3 (Hz)
v3 (volts)
c4 (pF/m)
k4 (Hz)
v4 (volts)
E25.0
E25.0
E25.0
7 - 41
c1 to c4:
pF/m, Suliciu corona branch parameters
k1 to k4:
Hz, Suliciu corona branch parameters
v1 to v4:
V, Suliciu corona branch parameters
The program automatically calculates the geometric capacitance C0 (matrix diagonal in the
multiphase case) and verifies the following conditions: c2 > c1 > C0, c4 > c3 > C0, ki > 0, v1 >
v2 > 0 and |v3| > |v4|.
Line model data:

The line model data (the target line) follows the standard EMTP rules, specified parameters are
those of a single line section which has a length SECTION_LEN. The program will automatically
interconnect a total of NSECT identical sections and insert corona branches.
7.7.4
Examples
Example 1: cp25aux.dat:
BEGIN NEW DATA CASE
C Gary's line
LINE CONSTANTS
C 3456789 12345678><--------><--------><--------><--------><--------><-------->
C
LINE-MODEL
CP-LINE
QREAL
METRIC
2
C CONDUCTOR CARDS
C ><---><DC-res><><perm--><diam--><Horiz-><------><Vmid--><------><Alph>
001 0.5.0896209 4
1.0
2.64
0.
12.
002 0.5.0896209 4
1.0
2.64
7.
12.
003 0.5.0896209 4
1.0
2.64
14.
12.
BLANK CARD ENDING CONDUCTORS CARDS
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789
C There is only one frequency card in the LINE-MODEL module.
C This is a section of 25meters
C -----><-------->
<--km-->
>iseg
<
100.
100000.
0.025
0
.CTLFIT
50 1 0 0 0
.OUTFIT
1 1 1
C 3456789 123456789<--------><--------><--------><--------><--------><-------->
.NODES
BUS1A
BUS2A
BUS1B
BUS2B
BUS1C
BUS2C
BLANK CARD ENDING FREQUENCY CARDS
BLANK
BEGIN NEW DATA CASE
BLANK
7 - 42
Example 2: cp25.dat:
BEGIN NEW DATA CASE
C To compare with Gary's line, field results are in kV
C
C Rising time = 0.3e-06 sec
C 0.3e-06*speed_light=0.9e+02 meters
C Length of section must be a fraction of 90 m.
C
C ----dt<---tmax<-------------------------------------------------------------->
0.04E-06 4.7E-05
C --IOUT<--Iplot<-Idoubl<-KssOut<-MaxOut
<---Icat
1
1
0
0
0
0
0
3
0
0
TACS HYBRID
88TIMEXX =TIMEX*1.0E+06
88VAR1
=-0.123*TIMEXX
88VAR2
=-4.1*TIMEXX
88VAR3
=12.3*TIMEXX+PI*7/18
88VAR4
=0.988*EXP(VAR1)-1.064*EXP(VAR2)*SIN(VAR3)
88BUS1A
=850000.*VAR4
33BUS1A
BLANK
C <----><----><----><---->
CORONA LINE BUS1A BUS2A
C 3456789 123456789 1234><23456789 123456789 1234><23456789 123456789 1234>
C
NSECT
NPH
Length of a section in meters
500.
3
25.
C
Cx in PF/m
8.41
C output at: 1km 3km 7km 10km 12.5km phase A, 1km phase B
41
121.
281.
401.
501.
503.
C
C
C in PF/m
K in Hz
V in volts
16.8
4.0E+06
320.0E+03
35.0
0.8E+03
220.0E+03
16.8
4.0E+06
-320.0E+03
35.0
0.8E+03
-220.0E+03
$INCLUDE cp25aux.pun
BLANK End of circuit data......................................................|
BLANK End of switch data.......................................................|
C Bus--><I<Amplitude<Frequency<--T0|Phi0<---A1
<---Tstart<----Tstop
60BUS1A
0.0
10.0
BLANK End of source data.......................................................|
C Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->Bus-->
BUS1A
BLANK End of output requests............................................... ...|
BLANK End of Plot Request
BEGIN NEW DATA CASE
BLANK
7 - 43
Example 3: fd25aux.dat:
BEGIN NEW DATA CASE
C Gary's line
LINE CONSTANTS
C 3456789 12345678><--------><--------><--------><--------><--------><-------->
LINE-MODEL
FD-LINE
QREAL
LOG
0.01
10
16
METRIC
2
C CONDUCTOR CARDS
C ><---><DC-res><><perm--><diam--><Horiz-><------><Vmid--><------><Alph>
001 0.5.0896209 4
1.0
2.64
0.
12.
002 0.5.0896209 4
1.0
2.64
7.
12.
003 0.5.0896209 4
1.0
2.64
14.
12.
BLANK CARD ENDING CONDUCTORS CARDS
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789
C There is only one frequency card in the LINE-MODEL module.
C This is a section of 25meters
C -----><-------->
<--km-->
>iseg
<
100.
100000.
0.025
0
.CTLFIT
50 1 0 0 0
.OUTFIT
1 1 1
C 3456789 123456789<--------><--------><--------><--------><--------><-------->
.NODES
BUS1A
BUS2A
BUS1B
BUS2B
BUS1C
BUS2C
BLANK
BEGIN NEW DATA CASE
BLANK
Example 4: fd25.dat:
BEGIN NEW DATA CASE
C To compare with Gary's line, field results are in kV
C
C Rising time = 0.3e-06 sec
C 0.3e-06*speed_light=0.9e+02 meters
C Length of section must be a fraction of 90 m.
C
C ----dt<---tmax<-------------------------------------------------------------->
0.04E-06 4.7E-05
C --IOUT<--Iplot<-Idoubl<-KssOut<-MaxOut
<---Icat
1
1
0
0
0
0
0
3
0
0
TACS HYBRID
88TIMEXX =TIMEX*1.0E+06
88VAR1
=-0.123*TIMEXX
88VAR2
=-4.1*TIMEXX
88VAR3
=12.3*TIMEXX+PI*7/18
88VAR4
=0.988*EXP(VAR1)-1.064*EXP(VAR2)*SIN(VAR3)
88BUS1A
=850000.*VAR4
33BUS1A
BLANK
C <----><----><----><---->
CORONA LINE BUS1A BUS2A
C 3456789 123456789 1234><23456789 123456789 1234><23456789 123456789 1234>
C
NSECT
NPH
Length of a section in meters
500.
3
25.
7 - 44

C
Cx in PF/m
8.41
C
41.
401.
121.
501.
281.
503
C
C
C in PF/m
K in Hz
V in volts
16.8
4.0E+06
320.0E+03
35.0
0.8E+03
220.0E+03
16.8
4.0E+06
-320.0E+03
35.0
0.8E+03
-220.0E+03
$INCLUDE fd25aux.pun
BLANK End of circuit data......................................................|
BLANK End of switch data.......................................................|
C Bus--><I<Amplitude<Frequency<--T0|Phi0<---A1
<---Tstart<----Tstop
60BUS1A
0.0
10.0
BLANK End of source data.......................................................|
BUS1A
BLANK End of output requests............................................... ...|
BLANK End of Plot Request
BEGIN NEW DATA CASE
BLANK
7 - 45
Section 8
Non-Linear Elements
8.1
General Comments About The Pseudononlinear Models ...................... 8-1
8.2
Staircase Time-Varying Resistance (Type-97 Element) ....................... 8-2
8.3
8.4
8.5
8.6
8.2.1
Applications .......................................................................... 8-2
8.2.2
Format and Rules .................................................................. 8-3
Pseudononlinear Resistance (Type-99 Element) ................................... 8-5

8.3.1
Applications .......................................................................... 8-5
8.3.2
Format and Rules .................................................................. 8-6
8.3.3
Example ................................................................................ 8-8
Pseudononlinear Reactor (Type-98 Element) ........................................ 8-9

8.4.1
Applications .......................................................................... 8-9
8.4.2
Format and Rules .................................................................. 8-9
8.4.3
Example ................................................................................ 8-11
8.4.4
Possible Discontinuity at Time Zero .................................... 8-12
Pseudononlinear Hysteretic Reactor (Type-96 Element) ...................... 8-13

8.5.1
Applications .......................................................................... 8-13
8.5.2
Format and Rules .................................................................. 8-14
8.5.3
Example ................................................................................ 8-17
8.5.4
Considerations Regarding Type-96 Usage ........................... 8-18
General Comments About True Nonlinear Modelling .......................... 8-21

8.6.1
Solution Method ................................................................... 8-21
8.6.2
Effects of Number and Location of NonLinear and

Time-Varying Branches ....................................................... 8-22
8.6.3
Network Connectivity Requirement ..................................... 8-22
8.6.4
Singularity Check ................................................................. 8-23
8.6.5
Singular Matrices .................................................................. 8-25
8.6.6
Treatment in Steady-State Solution to Set

Initial Conditions .................................................................. 8-25
8.6.7
8.7
Newton-Raphson Iteration Control Parameters .................... 8-25
Multiphase Time-varying Linear Resistance ......................................... 8-26

8.7.1
Applications .......................................................................... 8-26
Section 8
8.8
8.9
8.10
8.11
8.12
8.13
Non-Linear Elements
8.7.2
Format and Rules ................................................................. 8-27
8.7.3
Sample Data Listing ............................................................. 8-29
Multi-phase Piecewise Linear Resistance with Flashover .................... 8-29

8.8.1
Applications ......................................................................... 8-29
8.8.2
Format and Rules ................................................................. 8-30
8.8.3
Sample Data Listing ............................................................. 8-32
Multi-phase ZnO Surge Arrester ........................................................... 8-33

8.9.1
Applications ......................................................................... 8-33
8.9.2
Format and Rules ................................................................. 8-35
8.9.3
Use of Gaps for Practical Arrester Applications ................. 8-37
8.9.4
Sample ZnO Data Structures ............................................... 8-38
8.9.5
Remedies for Problems with Convergence .......................... 8-39
Multi-Phase Active Gap Surge Arrester Model .................................... 8-40

8.10.1
Applications ......................................................................... 8-40
8.10.2
Format and Rules ................................................................. 8-41
8.10.3
Example ............................................................................... 8-44
Nonlinear Hysteretic Reactor (Type 92) ............................................... 8-45

8.11.1
Introduction .......................................................................... 8-46
8.11.2
Initialization ......................................................................... 8-48
8.11.3
Time-step loop solution ....................................................... 8-49
8.11.4
Data structure ....................................................................... 8-49
8.11.5
Example ............................................................................... 8-52
Dynamic Circuit Breaker Model ........................................................... 8-52

8.12.1
Background .......................................................................... 8-52
8.12.2
Data Format and Rules ........................................................ 8-53
8.12.3
Arc Models - Brief Explanation ........................................... 8-57
8.12.4
Usage Note ........................................................................... 8-58
8.12.5
Sample Data Cases ............................................................... 8-58
Single-phase Nonlinear Inductance (Type-93 Element) ....................... 8-61

8.13.1
Applications ......................................................................... 8-61
8.13.2
Format and Rules ................................................................. 8-62
8.13.3
Example ............................................................................... 8-65
Section 8
Non-Linear Elements
The EMTP supports the following two classes of nonlinear elements:
(A)
pseudononlinear model
(B)
true nonlinear model
Each of these two models uses a different solution technique. The pseudononlinear model
assumes a piecewise linear representation of the nonlinear characteristics of the device, and it
relies on data calculated in the previous time-step; consequently, this model does not always
follow properly fast changes in the system. The true nonlinear model provides a correct
instantaneous solution by solving iteratively the combination of nonlinear equations and an
appropriate equivalent of the linear portion of the system using the compensation method.
8.1 General Comments About The Pseudononlinear Models

Conceptually, the idea behind pseudononlinear models is very simple. Since any segment of the
piecewise linear vi or -i curve is a straight line (Figure 8.1), such a segment can be presented in
the program by a resistor in parallel with an appropriate current source (Figure 8.2).
R
i
Figure 8.1: Nonlinear v-i characteristic.
8-1
Figure 8.2: Representation of a

segment of the nonlinear
characteristic
Non-Linear Elements
The only problem is with limits which are not observed by the linear representation as shown in
Figures 8.3 and 8.4. We do not have a true (simultaneous) nonlinearity because the program relies
on history (previous time-step results) to decide on what segment to operate. It is, therefore,
important to use a small enough time-step so that one moves up and down the nonlinearity slowly
and smoothly. The program changes segments only after having illegally operated outside the
range of the current segment for one time-step (see Figure 8.4).
infinite extension
illegal operation
x
segment being
modelled
infinite extension
Figure 8.3: Linear approximation

used in program
Figure 8.4: Sample movement to a

higher segment
The following elements use the pseudononlinear modeling:

(A)
Type-97 Staircase Time-Varying Resistance (Section 8.2)
(B)
Type-99 Pseudononlinear Resistance (Section 8.3)
(C)
Type-98 Pseudononlinear Reactor (Section 8.4)
(D)
Type-96 Pseudononlinear Hysteretic Reactor (Section 8.5)
8.2 Staircase Time-Varying Resistance (Type-97 Element)

8.2.1
Applications
This element can be used to represent a time-varying resistance R(t) is assumed to change in a
staircase manner as shown in Figure 8.5.
8-2
Non-Linear Elements
R2
R5
R3
t1=0 t2
R4
last value
continues
to tR =
tr
t3 t4
t5
Figure 8.5: Time-varying resistance
8.2.2
Format and Rules
Data cards for the staircase time-varying resistance begin with a single branch card for a Type-97
pseudononlinear resistance:
ITYPE
BUS1
BUS2
BUS3
BUS4
VFLASH
TDELAY
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 90123456789012345678901234567890123456789 0
I
O
U
T
I2
A6
A6
A6
A6
E6.0
E6.0
I1
Rule 1:
Branch type is 97 (field ITYPE; columns 1-2).
Rule 2:
Specify the terminal nodes of the branch by name (fields BUS1 and BUS2, columns 3-8
and 9-14, respectively). One node may be grounded if desired (indicated by bank
field for node name in question).
Rule 3:
Data fields VFLASH and TDELAY determine at what time during the simulation the element is to be connected. TDELAY is a fixed time delay before which the branch is not
conducting. VFLASH is the minimum flashover voltage.
8-3
Non-Linear Elements
VFLASH
=0
no flashover delay
(27-32)
= VF
with VF > 0, the element is not connected until the

branch terminal voltage v satisfies | v | > VF.
TDELAY
= 1
The first-step resistance value R1 is assumed to be

present for all time t < 0, as part of the sinusoidal
phasor solution. Element time tR and simulation
time t are identical; field VFLASH is ignored.
=0
No special fixed time delay (there may be a delay

due to positive VFLASH, though). Element is not
connected for the steady-state phasor solution.
= TD
With TD > 0, the element will not be connected

until the simulation time t reaches this delay value.
Positive VFLASH may provide added delay beyond
this point, until such time as the branch voltage
then exceeds the flashover value.
(33-38)
Rule 4:
If the R(tR) characteristic (exclusive of VFLASH and TDELAY) is identical to that of a

preceding type-97 element, the following storage-saving option may be used: enter the
node names of the preceding reference branch in fields BUS3 and BUS4 (columns 1520 and 21-26, respectively), and omit the cards defining the R(tR) characteristics as
mentioned in Rule 6 below.
Rule 5:
Output options for printing and/or plotting use field IOUT of column 80, as follows:
IOUT = 1
=2
=3
=4
Rule 6:
produces branch-current output

produces branch-voltage output
produces both branch-current and branch-voltage output
produces branch-power and energy consumption output
The R(tR) characteristic is defined point by point, by cards which immediately follow
the first branch card; these points are terminated by a 9999-card (entered in
columns 13-16).
(A)
The beginning of the characteristic, at time tR=0, is the first point to be entered.
(B)
Time and resistance pairs defining each step of the staircase characteristic are
entered in field TIME and RESIS (columns 1-16 and 17-29, respectively), one
pair of values per card; these points are terminated by a 9999-card (entered in
columns 1316).
8-4
Non-Linear Elements
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
TIME
RESIS
E16.0
E16.0
(C)
The order of input for points of the characteristic must be as per the numbering
on Figure 8.5, with time in seconds being monotonically increasing. All
resistance values must be positive, in units of ohms [voltage/current].
8.3 Pseudononlinear Resistance (Type-99 Element)

8.3.1
Applications
The pseudononlinear resistance model can be used to approximately model a nonlinear v-i
characteristic (see Section 8.8 for true nonlinear modelling). Please note the following:
(A)
Upon flashover, there is no relevant history; the user must specify which
segment it is appropriate to jump into (see Figure 8.6).
(B)
The user must use a small enough time-step so that one moves up and down
the nonlinearity slowly, smoothly. The program will only change segments
after having illegally operated outside the range of the current segment for one
time step, note (see Figure 8.4).
Vflash
p
m
Ju
in
1
i
Figure 8.6: Nonlinear v-i characteristic, with flashover

8-5
Non-Linear Elements
To avoid connectivity difficulties (e.g., a floating branch), the program automatically inserts a
very-high-impedance resistive branch in parallel with a Type-99 element, if no parallel branch
exists in the input data. This branch has negligible effect on the answers. The user must exercise
care so as to always set up a physically-realistic problem. For example, the series connection of
two Type-99 elements is absurd, since mathematically, there is no way to determine the voltage
split between the two perfect gaps; the problem is not even defined. The same holds true of the
series connection of a Type-99 element and a flashover switch.
99
flashover switch
8.3.2
99
99
Format and Rules
ITYPE
BUS1
BUS2
BUS3
BUS4
VFLASH
TDELAY
JUMP
VSEAL
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 567890 12345678901234567890123456789 0
I
O
U
T
I2
A6
A6
A6
A6
E6.0
E6.0
E6.0
E6.0
I1
Rule 1:
Rule 2:
Specify the terminal nodes by name (fields BUS1 and BUS2; columns 38 and 9-14,
respectively). One node may be grounded, if desired (blank for node name).
Rule 3:
If the vi characteristic (exclusive of VFLASH, TDELAY, and JUMP) is identical with a

preceding type-99 element, use the following storage saving option: enter the node
names of that preceding reference branch in fields BUS3 and BUS4 (columns 15-20 and
21-26, respectively), and omit the cards defining the vi characteristic as mentioned in
Rule 9 below.
Rule 4:
Field VFLASH (columns 27-32) specifies the breakdown voltage for the branch. Until
terminal voltage exceeds |VFLASH|, a Type-99 element is an open circuit. Such an
open circuit is assumed for the steady-state solution also.
8-6
Non-Linear Elements
Rule 5:
The element will open up again after having been conducting when a current zero
occurs, provided a time interval of TDELAY seconds has elapsed since the most recent
firing (initiation of conduction as per Rule 4). It is assumed that such a polarity
change will occur only while operation is on segment 1 (that passing through the origin); if the polarity change occurs while operating on higher segments, an appropriate
warning message is printed out (see Section 17, Message 14).
Rule 6:
Field JUMP (columns 39-44) gives the segment number that is to be jumped into upon
flashover. Segment L is defined as the segment that has point L at its upper end (see
numbering on Figure 8.6). A zero (or blank) field is automatically converted to a
default value of unity.
Se
gm
en
tL
Point L
Rule 7:
If data field VSEAL (columns 45-50) is left blank (normal operation), then current
interruption occurs on current zero. To operate the circuit before a zero current crossing, set VSEAL to a positive, indicating a voltage threshold (sealing-off voltage) below
which conduction is not allowed.
Rule 8:
Output options for printing and/or plotting use field IOUT in column 80:
IOUT
Rule 9:
=1
=2
=3
=4

The vi characteristic is defined point by point on cards which immediately follow the
first branch card; these points are terminated by a 9999-card (entered in columns 1316).
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
CUR
VOLT
E16.0
E16.0
(A)
The origin of the characteristic (i=0, v=0) is an implied point, which must not
be entered explicitly.
8-7
Non-Linear Elements
(B)
Current and voltage pairs of the breakpoints are entered in fields CUR and
VOLT (columns 116 and 17-32, respectively) one pair of values per card.
(C)
The order of input of points is as per Figure 8.6, where both current and voltage
values must be monotonically increasing. Segments with negative, zero, or
infinite slope are not allowed.
Note:
Rule 10:
The program assumes that the final two points define a segment which
extends to infinity. Care must be taken if "noisy" data is used to define
the characteristic, since the last two points may define absurd
asymptotic values.
To allow a Type-99 element to flash over only once, the "SINGLE FLASH" option can
be used. This is governed by the following special rules and restrictions:
(A)
The keyword "SINGLE FLASH" is entered in columns 3344 of the 9999-card

which terminates the i-v characteristic of the element under consideration:
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234 567890123456789012345678901234567890
9999
(B)
8.3.3
SINGLE FLASH
Within any given data case, the "SINGLE FLASH" request only has to be entered
once on the 9999-card of the first type-99 element. Any other Type-99
elements whose v-i characteristic has the same first breakpoint voltage v1 as the
first Type-99 element will also be allowed to flash just. Therefore, any number
of identical Type-99 elements can all be treated as single-flash units. Except
for the prohibition against flashing a second time, "SINGLE FLASH" Type-99
elements are identical to the conventional Type-99 elements.
Example
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4-><VFLSH<TDEL-<JUMP-<VSEAL
99ALPHA BETA
1.2E5 .065
3
3
C ----CUR-------<-------VOLT---1.0
50000.0
3.0
100000.0
10.0
300000.0
30.0
500000.0
9999
99TESLA
ALPHA BETA
1.1E5
2
3
8-8
Non-Linear Elements
8.4 Pseudononlinear Reactor (Type-98 Element)

8.4.1
Applications
The pseudononlinear reactor can be used to approximately model a nonlinear flux-current

characteristic. Unlike Type-93 single-phase nonlinear reactors of Section 8.13, Type-98 reactors
discussed here do not have to be separated by travel time.
The pseudononlinear reactor is the inductive analog of the pseudononlinear resistance of
Section 8.3, using a parallel connection of resistor and current source for internal representation.
The user must use a small enough time-step so that one moves up and down the nonlinearity
slowly, smoothly. The program will only change segments after having illegally operated outside
the range of the current segment for one time step.
infinite extension
illegal operation
x
segment being
modelled
infinite extension
8.4.2
Format and Rules
ITYPE
BUS1
BUS2
BUS3
BUS4
isteady
steady
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 90123456789012345678901234567890123456789 0
I
O
U
T
I2
A6
A6
A6
A6
E6.2
E6.2
I1
Rule 1:
Branch type is 98 (field ITYPE: columns 1-2).
Rule 2:
Specify the terminal nodes by name (fields BUS1 and BUS2; columns 38 and 9-14,
respectively). One node may be grounded, if desired (blank field for node name).
8-9
Non-Linear Elements
Rule 3:
Fields isteady and steady (columns 27-32 and 33-38, respectively) define the constant
linear inductance to be used during the sinusoidal, phasor steady-state solution. While
only the ratio (L=/i) is actually required, steady is taken to be the limit on the linear
region of operation; if the initial flux exceeds this value, the EMTP prints out a warning
message after the steady-state solution (see Section 17.1, Message 12).
Rule 4:
If the i- characteristic proper (exclusive of isteady and steady is identical with a preceding type-98 element, use the following storage-saving option: enter the node
names of the preceding reference branch in fields BUS3 and BUS4 (columns 15-20 and
21-26, respectively), and omit the cards defining the characteristic as mentioned in
Rule 6 below.
Rule 5:
Output options for printing and/or plotting use field IOUT of column 80:
IOUT = 1
=2
=3
=4
Rule 6:

The i characteristic is defined point by point, on cards which immediately follow

the first branch card; these points are terminated by a 9999-card (entered in columns
13-16).
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
CUR
FLUX
E16.0
E16.0
(A)
The origin of the characteristic (i=0, =0) is an implied point, which must not
be entered explicitly.
8 - 10
Non-Linear Elements
(B)
Current and flux pairs of the branch points are punched in fields CUR and
FLUX (columns 1-16 and 17-32, respectively), one pair of values per card.
(C)
The order of input of points is as per numbering shown below, where both
current and flux values must be strictly monotonically increasing.
Vflash
p
m
Ju
in
1
i
(D)
Usually the first point of the characteristic will equal (isteady, steady) in order
to provide continuity between the steady-state and transient solutions at time
zero, though such matching is not mandatory.
Note:
8.4.3
The program assumes that the final two points define a segment which
extends to infinity. Care must be taken if "noisy" data is used to define
the characteristic, since the last two points may define absurd
asymptotic values.
Example
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4-><Isted<Flxss
98ALPHA BETA
21.3 249.1
3
C ----CUR-------<-------FLUX---21.3
249.1
99.8
274.0
236.9
288.9
1602.0
326.3
9999
98TESLA
ALPHA BETA
1.1E5
2
3
8 - 11
Non-Linear Elements
8.4.4
Possible Discontinuity at Time Zero
The sinusoidal steady-state solution involves the linear magnetizing inductance given by
L=steady/isteady. At the first time-step, the EMTP shifts to using the user-defined nonlinear
characteristic, always starting out on the first segment. If the initial-condition point does not lie on
the characteristic, there will be a discontinuity which may produce a spurious transient
immediately after time zero. Flux is always continuous, so that adjustment is accomplished by a
horizontal jump having current discontinuity i, as shown below. If only one segment change is
required as shown, this occurs in one time-step. Therefore, an inductor of value LS could have a
corresponding voltage change in the order of LSi/t, which may be substantial for small timesteps. Present EMTP logic only allows one segment change per time-step, so if the ultimate
operating point really is on the K-th segment, K-1 time-steps would be required by the adjustment
process, during which time a spurious transient of unpredictable shape might be observed.
solution at time zero
solution at
first time step
=L
slop
e
coil flux
Jump
i
coil current
Figure 8.7: Discontinuity at Time Zero

For balanced 3-phase initial conditions, the user can always shift the time reference (rotate the
phase angles of all sinusoidal sources by some fixed phase angle) so that none of the three initial
fluxes (or alternatively, currents) exceeds 3 2 = 0,866 times its peak value. In terms of
quantities that can be read from the phasor steady-state solution printout of branch flows, this
corresponds to shifting the time origin so that current in one of the three components (for the 3phase representation) is exactly zero at time zero.
8 - 12
Non-Linear Elements
The user should always check the initial flux in Type98 pseudononlinear inductors, and be wary
of the aforementioned possible discontinuity. The initial flux in all data-case nonlinear or
pseudononlinear inductors is printed out after the complete steady-state solution, and before the
time-step-loop column headings, with a sample being:
INITIAL FLUX IN COIL 'JOHNDA' TO 'INTA
' = E13.5
Assuming that the user has chosen isteady and steady equal to the coordinates of the first point of
the characteristic, then the aforementioned discontinuity trouble will be signalled by the following
additional message:
Warning: ASSUMPTION THAT AC STEADY STATE HAS FUNDAMENTAL FREQUENCY ONLY

IS QUESTIONABLE WITH PRECEDING FLUX OUTSIDE LINEAR REGION."
Note that the EMTP provides the user with an option to obtain a steady-state solution which will
include the effects of harmonic distortion. This option is activated by the special request word
"INITIALIZATION WITH HARMONICS" (see Section 4.1.1). In this case the entire characteristic is
used to compute steady-state conditions rather than just the first linear segment.
8.5 Pseudononlinear Hysteretic Reactor (Type-96 Element)

8.5.1
Applications
The pseudononlinear hysteretic inductor is very similar to the pseudononlinear inductor described
in Section 8.4. The major difference is that this element allows the hysteretic behaviour of the
core material to be represented. The element is represented internally by a resistor in parallel with
a current source (see Figure 8.8). The resistance R is changed only when operation moves from
one segment to another, whereas the value of the current source is updated at each time step. The
qualification "pseudo" has been appended because operation moves from one segment to another
only after having illegally operated outside the range of the present segment for one time step
(Figure 8.4, Section 8.1).
8 - 13
Non-Linear Elements
Figure 8.8: Representation of Type-96 Inductance in the Time-Step

The Type-96 pseudononlinear hysteretic inductor is also similar in functionality to the Type-92
hysteretic reactor. The main differences are that the Type-92 reactor is modelled as a true
nonlinear element, the shape and behaviour of minor loops are modelled more accurately, and it
can be used conjunction with INITIALIZATION WITH HARMONICS to produce a correct harmonic
steady-state initialization (see Section 8.10.1).
8.5.2
Format and Rules
ITYPE
BUS1
BUS2
BUS3
BUS4
isteady
steady
res
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 56789012345678901234567890123456789 0
I
O
U
T
I2
A6
A6
A6
A6
E6.2
E6.2
E6.2
I1
Rule 1:
Rule 2:
Specify the terminal nodes by name (fields BUS1 and BUS2; columns 3-8 and 9-14,
Rule 3:
Fields isteady and steady (columns 2732 and 33-38, respectively) define the constant
linear inductance to be used during the sinusoidal, phasor steady-state solution. While
only the ratio (L=steady/isteady) is actually required, steady is taken to be the limit on
8 - 14
Non-Linear Elements
the linear region of operation. If initial flux exceeds this value, the EMTP prints out a
warning message after the steady-state solution (see Section 17.1, Message 12). In
addition, the point (isteady, steady) must lie within the major hysteresis loop. If it does
not, the EMTP will halt this run, providing an error message indicating this mistake.
steady
i
isteady
Figure 8.9: Automatic Selection of the

Stead-State Inductance
Optionally, the EMTP can calculate isteady and steady automatically. This option is
chosen by specifying isteady = 8888. and leaving steady blank. The EMTP chooses the
steady-state point as follows: first the trajectory from the origin to the positive
saturation point is created, then the point on that trajectory where the flux is equal to
70% of the saturation flux is chosen as the steady-state point (see Figure 8.9).
steady
i
isteady
Figure 8.10: Definition of the Steady-State

Inductance
8 - 15
Non-Linear Elements
Rule 4:
Field res (columns 3944) specifies the value of residual or remnant flux in the core.
This will be used as the initial value of flux ((0)) in the coil provided a non-zero
value is not calculated for time zero by a sinusoidal, phasor steady-state solution.
Therefore, if this Type-96 element is not connected during a steady-state solution, the
value of flux in the coil at time zero will be taken as the value specified in the field
res. If this Type96 element is connected during the steady-state solution and only if
the value of flux at time zero determined by the phasor solution is zero, the value of
flux specified in the field res is used as the initial flux at time zero. (Note that if res
=0, the initial flux at time zero will be identical to that calculated by the phasor
solution). Otherwise the initial flux in this Type96 element will be taken as the value
calculated from the steady-state solution. The value of res which is specified must lie
within the major loop.
Rule 5:
If the i characteristic proper (exclusive of isteady, steady and res) is identical with
one of the preceding Type96 elements, use the following storage saving
option: Enter the node names of the preceding reference branch in fields BUS3 and
BUS4 (columns 15-21 and 21-26, respectively) and omit the card defining the i
characteristic as mentioned in Rule 7 below.
Rule 6:
Output options for printing and/or plotting use field IOUT of column 80.
IOUT = 1
=2
=3
=4
Rule 7:

The i- characteristic is defined point by point, on cards which immediately follow the
first branch card; these points are terminated by a 9999-card (entered in columns 13
16).
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
CUR
FLUX
E16.0
E16.0
(A)
The bottom half of the loop must be defined - the top half is then determined by
symmetry.
8 - 16
Non-Linear Elements
Positive Saturation
Negative Saturation
Figure 8.11: Definition of the Hysteresis Loop

(B)
Current and flux pairs of the breakpoints are entered in fields CUR and FLUX
(columns 116 and 17-32, respectively), one pair of values per card.
(C)
The order of input of the break points must be as follows: the first point
specified must be the first point following the negative saturation point; the
points are then specified in order, up to and including the first point after the
positive saturation point. The next to last point specified must always be the
positive saturation point. The negative saturation point is defined from the
positive saturation point by symmetry in EMTP.
All points must be strictly monotonically increasing in both flux and current.
8.5.3
Example
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4-><Isted<Flxss<-Fres96
GEN
8888.
1.E-9
1
C ----CUR-------<-------FLUX---1.0
-0.7
2.0
0.9
3.5
1.0
9999
8 - 17
Non-Linear Elements
8.5.4
Considerations Regarding Type96 Usage
Whenever a Type96 element is being used in a run, there must be at least one source connected
during steady-state. The sub-network containing the Type96 element may or may not be
connected in steady-state. Such a sub-network can easily be created as shown in Figure 8.12,
where the voltage source at node DUMMY is connected during steady-state. Failure to observe this
restriction will result in the run being terminated, with an appropriate error message.
Dummy
1.0 cos t
1.0
Figure 8.12: Definition of a Dummy Subnetwork

Care should be exercised when specifying values of isteady and steady which are used to define the
linear magnetizing inductance (L=steady/isteady) that represents the Type96 element during a
phasor solution. The only restriction on this point is that it lies within the major loop. As the
sketch below indicates, there are many points that lie within the major loop but yield poor linear
representations of the loop. A handy rule of thumb might be to only pick points in the first
quadrant and never pick a point which yields an inductance smaller than the inductance obtained
by using the positive saturation point as the steady-state point.
8 - 18
Non-Linear Elements
Type96 will switch from being represented as a linear inductance to being represented by the
nonlinear, multi-valued hysteresis characteristic (which is then internally represented linearly at
each time step). Herein lies a potential problem. If the initial point lies within the major loop
there is no problem, operation simply begins at that point and proceeds from there as determined
by the system.
However, if the initial point as determined by a phasor solution lies outside of the major loop, the
problem arises because operation outside of the major hysteresis loop cannot occur in a real core.
It happens here only because of the linear representation used in the steady-state solution. The
EMTP therefore will move the initial point within the major loop rather than terminating the run at
this point (which is the only other option available). The scheme for moving the initial point into
the major loop is as follows: draw a line of constant current through the initial point, determine the
two points where it intersects, the top and bottom half of the major loop (these will be the same
points if i(0) >ipositive saturation or i(0) <inegative saturation) and take the average of these two points
as the new initial point. Whenever this happens the following message is printed:
NOTE ---- NONLINEAR ELEMENT NUMBER 1 IS A TYPE-96 HYSTERETIC INDUCTOR WHICH IS

CONNECTED BETWEEN BUSES BUS-K AND BUS-M THE INITIAL FLUX-CURRENT POINT AS FOUND
BY THE PHASOR STEADY-STATE SOLUTION HAS BEEN OBSERVED TO LIE OUTSIDE THE USERDEFINED MAJOR HYSTERESIS LOOP, HOWEVER, THE INITIAL FLUX IS 0.62642E+02 AND THE
INITIAL CURRENT IS 0.1500E+04. THE EMTP SHALL NOW ALTER THIS JUST-PRINTED FLUX SO
AS TO MAKE IT LEGAL, WHILE HOLDING THE CURRENT CONSTANT.
THE LINE OF CONSTANT
CURRENT INTERSECTS THE USER-SUPPLIED MAJOR HYSTERESIS LOOP AT TWO POINTS
(POSSIBLY EQUAL, IF THE CURRENT IS LARGE ENOUGH).
THE 'UPPER' IS CUT AT FLUX
VALUE 0.60000E+02, AND THE 'LOWER' AT FLUX VALUE, 0.58000E+02. THE INITIAL
FLUX SHALL BE TAKEN BY THE EMTP TO BE THE AVERAGE OF THESE, WHICH HAS FLUX VALUE,
0.59000E+02.
8 - 19
Non-Linear Elements
It is the user's responsibility to determine if the change made is acceptable or not. It should be
realized that transients, due to sudden changes in flux (like those being made here,) may die out
very slowly. It is strongly suggested that the user make legitimate alterations which cause all such
messages to disappear.
One potential cause for this type of problem could be a poor choice of isteady and steady as
discussed above. Consider the example in which a relatively low value of inductance (by means of
point 1) was chosen to represent the loop during the phasor solution. Notice the required change in
flux as discussed above. If point 2 is used as the steady-state point rather than point 1, it is quite
likely that no such problem with initial point outside of the major loop will occur. Any time the
EMTP creates a new initial point whose flux coordinate is larger (in absolute value) than the initial
point originally calculated from the phasor solution, one should check for the situation discussed in
this example.
Figure 8.13: Problems in the Definition of the

Steady-State Inductance
If such steps as these do not eliminate problems of initial points outside of the major hysteresis
loop, the user must decide if the changes made by EMTP are of small enough magnitude to be
accepted or not.
The user should always check the initial flux in each Type-96 element in addition to checking for
the message discussed above. The initial flux in all nonlinear or pseudononlinear coils is printed
out after the complete steady-state solution, and before the time-step-loop column headings, with a
sample being:
INITIAL FLUX IN COIL 'NODEVA' TO 'NODEVB' = E13.5
8 - 20
Non-Linear Elements
By checking this, the user will be able to verify that he is correctly starting at the specified value of
residual flux as desired. One will also be able to determine if the execution inadvertently begins at
the user specified value of residual flux when, in fact, the user wants the initial flux to be
calculated from a steady-state phasor solution.
8.6 General Comments About True Nonlinear Modelling

8.6.1
Solution Method
The solution of (true) nonlinear elements in the EMTP is done using the compensation method,
where the linear portion of the system is represented with a multi-terminal Thevenin equivalent
circuit, and the nonlinear portion is represented with the correct (non-linearized) equations that
describe the nonlinearities in the system (see Figure 8.14).
ZThv
ikm
VThv
nonlinearity
ikm
m
Figure 8.14: Solution Scheme for True Nonlinear Elements
In the compensation method, the nonlinear elements are simulated as current injections, which are
superimposed on the linear network after a solution without nonlinear elements has been found
(Thevenin equivalent). This can be illustrated with the network in Figure 8.14. Current ikm must
fulfill two equations. The first equation is for the linear part
v km = v kmo ( R thev i km )
(where subscript vkmo is the voltage across k and m without the nonlinear branch). The second
equation is the relationship of the nonlinear branch itself.
di km
, t,
v km = f i km,
dt
8 - 21
Non-Linear Elements
These two equations are then solved simultaneously using Newtons method. For more detail on
the implementation of the compensation method in the EMTP, please refer to Section 12.1.2 of the
EMTP Theory Book, second edition.
8.6.2
Effects of Number and Location of NonLinear and Time-Varying Branches
There are certain restrictions regarding the location of nonlinear elements within the network.
These restrictions are:
Certain nonlinear elements should not be in the same sub-network as a Type-59 Synchronous
machine, and the Universal machine.
Any number of multiphase nonlinear elements can co-exist in the same sub-network
Single-phase Type-93 nonlinear reactor does not co-exist with any other nonlinear element in
the same sub-network.
In the context of nonlinear elements a sub-network is defined as a topologically-connected region.

Note that branches with distributed parameters do not connect their end nodes topologically
because they are separated by the travel time of the line. Therefore, distributed parameter branches
often break the network into toplogically disconnected sub-networks.
Hint for introducing disconnections: A lumped inductance L can be approximated

by a lossless line of travel time t = t and surge impedance Z = L/t. Since such a branch with
distributed parameters is erased in the connectivity check, it will help in further disconnecting the
network. This approximation of a lumped inductance by a short, lossless line is known in the
literature as a "stub-line representation" (see Reference 2). The stub-line representation must be
used with caution; it works only if Z is large compared with the surge impedance of distributed
lines (see Reference 35).
8.6.3
Network Connectivity Requirement
For the EMTP to be able to solve a problem, the equivalent resistive network which is solved at
each time step must be "connected." That is, it should be possible to pass from any one network
node to any other network node along network branches of finite resistance. Equivalently, there
must be a network path from every node to ground. This connectivity check is done in every
simulation, and the printing of the connectivity tree is controlled with IDOUBL in the second
miscellaneous data card. However, nonlinear and time-varying elements affect this requirement in
that they must be ignored while making the connectivity check. No such branch can be used as
part of the paths satisfying the connectivity criterion.
8 - 22
Non-Linear Elements
Example:
The network sketched in Figure 8.15 below cannot be solved by the EMTP in the form shown.
With switch S open as shown, ignoring the time varying element R(t) leaves the right-hand switch
node without a finite-resistance path to ground, and hence disconnected from the rest of the
network. This problem would be rejected by the program, and the case would be terminated.
R
+
L
C
R(t)
Figure 8.15: A Network that does not Satisfy the

Connectivity Requirement
Remedy:
In such cases connectivity can be provided by inserting high-resistance paths where needed. By
making such a resistance several orders of magnitude larger than other typical problem
resistances, the transient solution is unaffected for engineering purposes. However this high
resistance cannot be arbitrarily high (see Section 5.1.5). Typically 1.E+10 is a reasonable value.
In terms of floating-point miscellaneous data parameter "EPSILN", a resistance of 1.E-2/EPSILN
times other representative problem impedances might be reasonable.
8.6.4
Singularity Check
Every nonlinear or time-varying branch can have a characteristic with a region where R= (see
Figure 8.19); therefore, the solution method must account for the eventuality that such a branch is
blocked (open). This blocking might exclude a solution, which will show up mathematically in a
matrix singularity. Figure 8.20 shows such a case, where no solution exists whenever the
nonlinear branch is blocked.
8 - 23
Non-Linear Elements
R=
Figure 8.16: A v-i

Characteristic with
Infinite Resistance at i=0
nonlinear
branch
i(t)=given
Figure 8.17: A Network with a Nonlinear

Resistor can Potentially Block
The program does not recognize whether a specific characteristic may or may not block. To make
a case with a non-blocking characteristic solvable that would be unsolvable in the blocked state, it
is necessary to split the nonlinear branch into two parallel branches one linear and one nonlinear
thus providing a branch of finite resistance. This is done by splitting the current in the nonlinear
characteristic (see Figure 8.21).
one branch with potentiality of blocking
Two equivalent parallel branches

R linear = -2
V
2
R nonlinear
R = 2
nonlinear part
linear part
i
2 4
characteristic of
nonlinear part
i
Figure 8.18: A Nonlinear Resistor Replaced by Two Equivalent Resistors
8 - 24
Non-Linear Elements
8.6.5
Singular Matrices
An aspect of the compensation based solution which deserves comment is a possible singularity of
the equations being solved. The solution of each coupled element group is performed within
"ZINCOX" using the general purpose linear equation solver "DGELG". Pivoting is involved, along
with miscellaneous data parameter "EPSILN" to measure possible singularity. There has yet to be
found a situation where this tolerance falsely caused trouble, when solutions were proceeding
normally. But, it is not uncommon for an EMTP error stop (KILL=209, LSTAT(19)=3501) to occur
if the iteration is diverging. Typically, this is the result of excess voltage, which then is fed into
the nonlinearity to produce astronomical currents; the elements of the Jacobian matrix thus are
quite abnormal.
8.6.6
Treatment in Steady-State Solution to Set Initial Conditions
All nonlinear and time-varying resistances are excluded from the network in the AC steady-state
solution. Nonlinear inductances are included with their linear region as specified on the branch
card (see Section 8.4, Section 8.5, and Section 8.13). In the case of the Type-92 hysteretic reactor,
a zero-width hysteresis loop is assumed during initialization with harmonics.
8.6.7
Newton-Raphson Iteration Control Parameters
Control parameters of the Newton-Raphson iterative solution are specified via a special request
word to be read in arbitrary order prior to the miscellaneous data cards. The request word "ZINC
OXIDE" (or just "ZO" in abbreviated form) is used. (see also Section 4.1.8)
1
2
3
4
5
6
7
8
1234567890123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
ZINC OXIDE
MAXZNO
A16
MAXZNO
(17-24)
I8
EPSZNO EPWARN EPSTOP ZNOLIM1 ZNOLIM2

E8.0
E8.0
E8.0
E8.0
E8.0
Maximum number of Newton iterations allowed for the

solution of the surge arresters in each sub-network. The
Newton iteration will go this long unless tolerance EPSZNO
is attained first.
8 - 25
Non-Linear Elements
EPSZNO
(25-32)
EPWARN
(33-40)
EPSTOP
(41-48)
ZNOLIM(1)
(49-56)
ZNOLIM(2)
(57-64)
Convergence tolerance for the Newton iteration. All

voltage corrections of the coupled elements must be
smaller than this in absolute value for the equations to be
judged solved. A blank (zero) value is taken as a request
for the value EPSILN Vref, where EPSILN is the familiar
floating-point miscellaneous data parameter of
Section 4.2.1. For a comment about VREF see the
description of ZNOLIM(1) below.
Voltage convergence tolerance for non-fatal warning
message about a "sloppy solution." A blank (zero) data
field is taken as a request for the value Vref / 1000 -representing one tenth of one percent of rated voltage,
which is about the limit of graphical resolution.
Voltage convergence tolerance for a fatal error stop, complaining about a non-converged iteration. A blank (zero)
data field is taken as a request for the value Vref / 10 -- or
10% of rated voltage. The current error in such a case
could be very high due to the extreme nonlinearity (e.g.,
1.126 = 11.9).
The maximum per unit (based on VREF) voltage correction
at each iteration step. A blank or zero is given the default
value of 1.0. Note that the control of this parameter may
help resolve difficult convergence cases. For Type-91
(3333.) and Type-92 (4444.) elements, VREF is generated
by the program.
The maximum per unit based on VREF arrester voltage
during the iterations. A blank or zero is given the default
value of 1.5. This parameter applies to the ZnO arrester
model only and may be manipulated by the program if it
causes an oscillatory solution.
Note that data in this card can also be specified in free format (i.e., with commas separating all data
fields).
8.7 Multiphase Time-varying Linear Resistance

8.7.1
Applications
The model of this section provides for true nonlinear (as opposed to pseudononlinear)
representation of time-varying resistances. The model is compensation-based (see Section 8.6.1).
8 - 26
Non-Linear Elements
The time-varying resistance R(tR) is specified point-by-point as a piecewise linear characteristic.
Linear interpolation is used between the data points.
Name of
1st Node
R(tR)
Name of
R(t) 2nd Node
R=
for tR < 0
tR
Figure 8.19: Definition of a Time Dependent Resistance

The time count tR does not have to be identical with the time of the transient study. The time
count tR is started as soon as the absolute value of the voltage v across the blocked branch
(R=infinity) reaches VSTART.
8.7.2
Format and Rules
Data cards for a time-varying resistance begin with a single branch card for a Type-91 nonlinear
resistance:
BUS1
BUS2
BUS3
BUS4
I2
A6
A6
A6
A6
3333.
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012345678 901234 56789012345678901234567890123456789 0
I1
Rule 1:
Rule 2:
Rule 3:
Enter "3333." in columns 39-44 to indicate time-varying resistance modelling.
Rule 4:
8 - 27
Non-Linear Elements
IOUT = 1
=2
=3
=4

The second data card is to be entered with variable VSTART according to the following format:
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890123456789012345678901234567890
VSTART
E25.0
VSTART
(1-25)
Breakdown voltage in V (volts). The time count tR starts as

soon as |v| is > VSTART. This value cannot be left unspecified
(blank field) or set to zero. To start the time count at the start
of the transient study, set VSTART to a small value, e.g.,
VSTART = .01 V.
Next come cards which specify the time-varying characteristic point-by-point from left to right
with one card for each pair of values tR, R(tR) (columns 1-25 and 26-50, respectively). The format
is:
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 123456789012345678901234567890
R(tR) in
tR in sec
E25.0
E25.0
Terminate the grouping of these data cards with a card having 9999 entered in columns 22-25.
Control parameters of the Newton-Raphson's iteration are specified via a special request word
"ZINC OXIDE", to be read in arbitrary order prior to the miscellaneous data cards. For details, see
Section 8.6.7.
8 - 28
Non-Linear Elements
8.7.3
Sample Data Listing
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4->
<FLAG
91LEFT RIGHT
3333.
1
C VSTART ------300000.
C ----R(t)-------<------- t ---300.0
0.0
200.0
0.3
150.0
0.6
150.0
1000.
9999
8.8 Multi-phase Piecewise Linear Resistance with Flashover

8.8.1
Applications
The model of this section provides for true nonlinear modelling (as opposed to pseudononlinear)
of an arbitrary number of piecewise linear resistances.
The nonlinear resistance R(i) in series with a linear resistance RLIN is specified point-by-point.
Linear interpolation is used between the data points.
Name of
1st Node
i
R(i)
RLIN
V nonlinear
Name of
2nd Node
6
V nonlinear
V
4
This model can be used to model two types of arresters: old style lightning arresters and modern
gapless arresters.
8 - 29
Non-Linear Elements
Old Style Lightning Arrester:

R will be assumed infinite until |V| is > VFLASH; thereafter, R=R(i)+RLIN according to specified
nonlinear characteristic. The linear resistance RLIN can be used to represent the grounding
resistance of the lightning arrester; a value of zero is permissible, if desired.
Name of
1st Node
Gap flashes
over when
v VFLASH
R(i)
RLIN
Name of
2nd Node
The following two options are available:

1.
No clearing after the sparkover (discharge), i.e., the gap remains shorted once it
has fired.
2.
Clearing after the sparkover, i.e., following a discharge, the gap will open (clear)
at the next current zero crossing.
Nonlinear Resistance, Modern Gapless Arrester:

The gap is omitted by setting VFLASH 0.0. RLIN=0.0 if all resistance is in the piecewise-linear
characteristic.
8.8.2
Format and Rules
Data cards for a piecewise linear resistance begin with a single branch card for a Type-92 nonlinear
element (nonlinear resistance):
BUS1
BUS2
BUS3
I2
A6
A6
A6
Rule 1:
BUS4 NFLASH
A6
4444.
I6
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 56789012345678901234567890123456789 0
I1
Branch type is 92 (punch in columns 1 and 2).
8 - 30
Non-Linear Elements
Rule 2:
Rule 3:
Enter "4444." in columns 39-44 to indicate piecewise linear resistance modelling.
Rule 4:
The variable NFLASH (columns 27-32) defines whether the arrester will clear after the
gap discharge:
NFLASH > 0:
Single flash, i.e., the gap will discharge and clear once. It will stay
permanently open thereafter.
NFLASH = 0:
The gap will discharge and clear as many times as required by the
network conditions.
NFLASH < 0:
The gap will discharge once and stay closed thereafter.
The second data card contains variables RLIN, VFLASH and VZERO according to the
following format:
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
RLIN
VFLASH
VZERO
E25.0
E25.0
E25.0
Linear resistance in series with the piecewise linear one in units

of ohms.
RLIN
(1-25)
Gap flashover voltage in volts. If the resistor is gapless, enter

any negative number.
VFLASH
(26-50)
Arrester voltage (voltage drop across it) in the same units as

VFLASH. In almost all cases, leave it blank to start Newton's
iteration with zero current.
VZERO
(51-75)
Rule 5:
IOUT
Rule 6:
=1
=2
=3
=4

If the nonlinear characteristic is identical to that of a preceding branch, the referencebranch feature can be used to save memory. Enter "4444-" in columns 33-44.
8 - 31
Non-Linear Elements
Next come cards which specify the nonlinear characteristic point by point from left to right with
one card for each pair of values i, Vnonlinear. The format is:
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 123456789012345678901234567890
CUR
VOLT
E25.0
E25.0
Terminate the grouping of these data cards with a card having 9999 entered in columns 22-25.
If the characteristic is symmetrical with respect to the origin, the following storage saving option
can be used: Specify only the positive part of the characteristic (above the origin). Do not specify
the origin (0.0, 0.0) which will be generated automatically by the program.
Section 8.6.7.
8.8.3
Sample Data Listing
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus1->Bus2->Bus3->Bus4->NFLASH
<FLAG
92LEFT RIGHT
4444.
1
C --------- RLIN --------<------- VFLASH ---------<--------- VZERO -------600000.
C ---- CUR -----<----- VOLT ----10000.0
-600000.0
-4200.0
-500000.0
-1630.0
-400000.0
-630.0
-300000.0
-200.0
-200000.0
-50.0
-100000.0
50.0
100000.0
200.0
200000.0
630.0
300000.0
1630.0
400000.0
4200.0
500000.0
10000.0
600000.0
9999
8 - 32
Non-Linear Elements
8.9 Multi-phase ZnO Surge Arrester

8.9.1
Applications
representation of an arbitrary number of ZnO surge arresters. The basic constraint equation is
resistive, and highly nonlinear, as can be seen in the example below for q=30:
v q
i = p ---------- Vref
i/p
v/VREF
(8.1)
0.0012
0.0424
1.0
17.45
0.8
0.9
1.0
1.1
237.4
2620.0
24201.0
1.2
1.3
1.3
Typically, the reference voltage Vref is selected to be in the order of twice the rated voltage. In
theory, the choice is arbitrary (it is an extra parameter). Vref normalizes the equation, and
prevents numerical overflow. Then constants "p" and "q" are unique characteristics of the device.
The nonlinear v-i arrester characteristics can be approximated by an arbitrary number of
exponential segments. Both gapped and gapless arresters can be represented. The gapped arrester
requires the specification of two sets of exponential segments, while the gapless arrester requires
only one.
Due to the extreme nonlinearity of ZnO characteristics, very little current is drawn for voltages
that are substantially below rated voltage Vref (e.g., 0.530=9.E-10). To avoid the possibility of
exponentiation underflow, and also to speed the solution, linear representation is actually used for
low voltages (see sketch). In physical terms, the solution is unaffected by this representation.
voltage V
Vmin
exponential
characteristic
linear resistance
current i
8 - 33
Non-Linear Elements
Data cards for this component can be generated automatically from nonlinear v-i characteristic
data points using the EMTP support program AUX.
As indicated in Section 8.6.2, the use of multi-phase compensation has some limitations. The only
nonlinear elements that can belong in the same subnetwork as Type-92 ZnO arresters are:
time dependent resistance (Type-91, 3333.)
piecewise linear resistance with flashover (Type-92, 4444.)
active gap arrester (Type-92, 6666.)
Type-92 hysteretic reactor (Type-92, 8888.)
Multiphase nonlinear elements can be used in any number and configuration, in parallel or in
series, with the exception of networks where one nonlinear element's voltage is equal to the
combination of others, which will result in a singular Jacobian matrix in the Newton solution;
insertion of small resistance branches is then required. Other nonlinear elements, such as the
universal machines, cannot coexist with other types of nonlinearities in the same subnetwork.
Subdivision into subnetworks occurs due to time delay on transmission lines (see Section 8.6.2).
3-
Subnetwork
A
U.M.
line
Induction
Motor
ZnO
Subnetwork
B
ZnO
ZnO
3-
ZnO
Usage
lin
e
3-
line
Subnetwork
C
ZnO
1-
ZnO
Usage
S.M. Type 59 S.M.
In some cases a six-phase grouping of arresters can profitably be used, when both the high and the
low sides of a 3-phase transformer bank are protected, and one does not want to worry about
introducing stub lines to isolate the two sides (as described in Section 8.6.2). But the user should
approach such larger groupings with caution, since the convergence of Newton's method may be
adversely affected.
8 - 34
Non-Linear Elements
8.9.2
Format and Rules
Data cards for a ZnO surge arrester begin with a single branch card for a Type-92 nonlinear
element (nonlinear resistance)
BUS1
BUS2
BUS3
BUS4
I2
A6
A6
A6
A6
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012345678 901234 56789012345678901234567890123456789 0
5555.
I1
Rule 1:
Rule 2:
Rule 3:
Enter "5555." in columns 39-44 as an indication of ZnO modelling. No other data is

actually required on this branch card.
Rule 4:
IOUT
Rule 5:
=1
=2
=3
=4

If the nonlinear characteristic be identical to that of a preceding branch, the referencebranch feature can be used to save memory. Enter "5555." in columns 39-44.
Next come data cards for the actual ZnO characteristics (there are two characteristics, if the
arrester has a flashover gap). These ZnO characteristics are described by a number of exponential
segments (sections) of the form:
V
-
i = p --------------- VREF
(8.2)
The second data card variables VREF, VFLASH and VZERO according to the following format:
8 - 35
Non-Linear Elements
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
VREF
VFLASH
VZERO
E25.0
E25.0
E25.0
VREF
(1-25
VFLASH
(26-50)
VZERO
(51-75)
The reference voltage of the ZnO constraint equation (1)

in volts.
Gap flashover voltage in p.u. based on VREF. If the
arrester is gapless, enter any negative number.
Arrester voltage (voltage drop across it) in the same units
as VREF. In almost all cases, leave it blank so that the
Newton iteration will start with zero current.
Next come cards for specifying the exponential segments. These begin with the characteristic
before flashover (or, the only characteristic if there is no gap). Each characteristic is to be
terminated by a special card having "9999" in columns 22-25. The following format is used for
each exponential segment of each characteristic, in natural order (of increasing current and
voltage):
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
RLIN
EXPON
E25.0
E25.0
COEF
VMIN
E25.0
Coefficient "p" of ZnO constraint.
(1-25)
EXPON
Exponent "q" of ZnO constraint.
(26-50)
VMIN
(51-75)
Minimum voltage for usage of the just-started characteristic,

in per unit based on VREF. This value must be greater than
zero for the first exponential segment.
All such cards describing the exponential segments are to be in their natural order. Terminate each
grouping with a "9999" card. If the arrester is equipped with a gap, the post-flashover data must
follow that or the pre-flashover characteristic.
8 - 36
Non-Linear Elements
Section 8.6.7.
8.9.3
Use of Gaps for Practical Arrester Applications
At any point of time, the arrester is represented by a block with a variable conductivity
(resistance). The value of this conductivity varies as the operating point of the arrester changes.
The variation may be smooth or discontinuous. Consider, for example, an arrester equipped with
a shunt, passive gap shown in Figure 8.20 below. Similar diagrams can be drawn for an arrester
with a series passive gap, with the gap represented as a very high resistance.
modelled
modelled
as
a) before sparkover
as
b) after sparkover
Figure 8.20: Representation of an Arrester with a Shunt Gap

Assume that two exponential segments are used to represent the arrester before and after gap
sparkover. An idealized voltage-current characteristic of an arrester equipped with a shunt gap is
shown in Figure 8.21 below. The numbers in circles correspond to the numbering sequence of the
segments within the program. They also correspond to the data input order.
8 - 37
Voltage (Crest, P.U.)
Non-Linear Elements
Gap Sparkover
1.5
2
before sparkover
1
4
3
after sparkover
1.0
1.
10.
100.
1000
10000
Current (Crest, A)
Figure 8.21: Protective Characteristic of an arrester with a Shunt Gap

The gap, when open, permits the flow of a very small leakage current. This behaviour can be
modelled by either a linear or nonlinear (exponential) resistor. Remember, a linear segment has an
exponent a = 1.0.
Good results have been obtained with a linear resistor of 1.E8 to 1.E9 ohm. Remember, the
program uses admittance formulation and the gap characteristic is to be specified accordingly.
8.9.4
Sample ZnO Data Structures
As an example, consider the following gapless, single-phase ZnO arrester. The schematic is as
shown in Figure 8.22 below, with just a single exponential used to represent the ZnO arrester
characteristic:
p=2500 amps, q=26, Vref=778 kV.
t=0
200 miles
408 kV
ZnO
Figure 8.22: An Example with a ZnO Arrester
8 - 38
Non-Linear Elements
A listing of the data case follows:
BEGIN NEW DATA CASE
ZO,,,,.9,,
.000050
.020
C .....^.......^.......^.......^.......^.......^.......^
C -Iprnt<--Iplot<-Idoubl<-Kssout<-Maxout<---Ipun<-Memsav<---Icat<-Nenerg<-Iprsup
1
1
1
0
1
-1
0
2
C .....^.......^.......^.......^.......^.......^.......^.......^.......^.......^
C --Kchg<---Mult<---Kchg<---Mult<---Kchg<---Mult<---Kchg<---Mult<---Kchg<---Mult
2
10
33
1
40
10
100
50
C .....^.......^.......^.......^.......^.......^.......^.......^.......^.......^
C Bus-->Bus-->Bus-->Bus--><--R-><-Zs-><-v--><-l-->-|-|-|
-1SEND REC
.306 5.82 0.12 200.
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
<FLAG
92LEFT RIGHT
5555.
1
C --------- VREF --------<------- VFLASH ---------<--------- VZERO -------778000.0
-100.0
C --------- RLIN --------<------- EXPON ---------<--------- VMIN -------2500.0
26.0
0.5
9999
BLANK CARD ENDING BRANCH DATA
BLANK CARD ENDING SWITCH DATA
C Bus--><I<Amplitude<-FREQ---><-PHI -->---A1---->----------<---Tstart<----Tstop
14END
408000.0
60.
8.9.5
Remedies for Problems with Convergence
Convergence problems while using the ZnO model (or other multiphase true nonlinear elements)
are typically caused by:
1.
The wrong characteristic data set was used in the case.
2.
The time-step was too large.
3.
Some spurious voltage oscillations occurred which can be taken care of by

connecting the leakage capacitance or the bus capacitance in series with the
chosen resistance, and then in parallel with the arrester. Alternatively,
numerical oscillations can be removed by enabling the CDA procedure.
8 - 39
Non-Linear Elements
8.10
8.10.1
Multi-Phase Active Gap Surge Arrester Model

Applications
This section provides a true nonlinear active gap surge arrester model. Active-gap surge arresters
for HV systems are fairly complex to model. Dynamic mathematical models of these devices have
been developed in the past but have been difficult to use, suffering complex data acquisition
problems. The model shown here is simple, with adjustable parameters that can be used to
represent any active gap arrester, consisting of an active gap unit attached to a nonlinear SiC
(silicon-carbide) valve resistor (Figure 8.23 below) having the equation:
v = R (i)
(8.3)
To remain general, specific nonlinear v-i characteristics can be approximated by an arbitrary

number of segments, each with its own "R" and "q", which is similar to the ZnO modeling
described in Section 8.9
.
V
R
Gap
Z1
S1
Z2
S2
PE
Z3
S3
Z4
S4
Figure 8.23: A Typical Gap Arrester Circuit with Voltage-Dividing

Impedances (Zk), Active Spark Gaps (Sk) and a Magnetic
Blowout Coil L with a Protective Shunting Element PE
8 - 40
Non-Linear Elements
The active gap voltage is modelled using three parameters:
1.
a time delay (TDELAY) to represent the period during which the blowout coils
are shunted;
2.
a capacitance (CAPAG) to simulate gap voltage buildup;
3.
a peak gap voltage (VGAPM)
This model, which can be used in the same sub-network as other multi-phase nonlinear elements
(such as the ZnO surge arrester).
Section 8.6.7.
8.10.2
Format and Rules
Data cards begin with branch identification:
BUS1
BUS2
BUS3
BUS4
I2
A6
A6
A6
A6
6666.
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012345678 901234 56789012345678901234567890123456789 0
I1
Rule 1:
Rule 2:
Specify terminal nodes by name (fields BUS1 and BUS2, columns 3-8 and 9-14,
Rule 3:
Enter "6666." in columns 39-44 as an indicator of SiC active gap surge arrester modelling. No other data actually required on this branch card.
Rule 4:
IOUT
=1
=2
=3
=4

8 - 41
Non-Linear Elements
Rule 5:
If the nonlinear characteristic is identical to that of a preceding branch, the reference

branch feature can be used to save space. Enter "6666." in columns 39-44.
Next come data cards for the basic arrester parameters VREF, VFLASH and VZERO.
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
VREF
VFLASH
VZERO
E25.0
E25.0
E25.0
VREF
(1-25)
VFLASH
(26-50)
VZERO
(51-75)
The reference voltage of the arrester in volts. The rated

voltage of the arrester can be used.
Gap flashover voltage in p.u., based on VREF. This value
must be greater than zero.
Estimated voltage drop across the arrester in volts, when
connected to the network. This value is used to start the
Newton iterations. In almost all cases, it can be left blank.
Next comes data card for specifying the active gap parameters
1
2
3
4
5
6
7
8
12345678901234 56789012345678 90123456789012 34567890123456 789012345678901234567890
TDELAY
CAPAG
VGAPM
CURTOL
E14.0
E14.0
E14.0
E14.0
TDELAY
(1-14)
Time delay in seconds before the active gap voltage starts

to build up after flashover.
Note that during sparkover the blowout coils are shunted to
reduce the arrester impedance and protect the low
insulation level of the coil, so TDELAY must be greater than
0.0.
CAPAG
(15-28)
VGAPM
Equivalent capacitance in F used to model the active gap.

Peak active gap voltage in p.u. based on VREF.
(29-42)
8 - 42
Non-Linear Elements
CURTOL
(43-56)
Current margin in amps to specify gap extinction before

zero-crossing; can be left blank.
Next come cards for specifying the arrester nonlinear v-i characteristic, which may be described
by any number of nonlinear segments in natural order of increasing current and voltage.
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012345678901234567890 1234567890123456789012345 67890
COEF
EXPON
E25.0
E25.0
COEF
VMIN
E25.0
Coefficient "R" of the SiC valve resistor characteristic.
(1-25)
EXPON
Exponent "q" of the SiC valve resistor characteristic.
(26-50)
VMIN
51-75)
Minimum voltage for the usage of the just-started segment,

in per unit based on VREF. This value must be greater than
zero for the first segment. For voltages lower than VMIN
the arrester is considered linear.
Terminate segments with a "9999" card in columns 22-25.
8 - 43
Non-Linear Elements
8.10.3
Example
0.180mH
9.6mH
+
72kV
SIC
+
0.120F
Figure 8.24: An Example with an Active Gap Arrester

The test circuit of Figure 8.24 is used to flashover the active gap arrester and observe its follow
current. A capacitance precharged to 150 kv is used to flashover the 72 kV (rms) arrester at a
switch closing angle of 30 degrees of the cosine wave of the applied voltage. The arrester has a
flashover voltage of 145 kV and its nominal voltage ( 72 2 = 101,82338 kV) is used as the
reference voltage VREF. The nonlinear resistor is modelled with a single exponential segment:
v = 9780 ( i )
(8.4)
0,3
For 72 kV arresters typical coefficient "R" values range from 9300 to 12000, the value of "q" may
vary from 0.15 to 0.3. The remaining parameters are:
possible range (manufacturer)
TDELAY = 250MS
25 - 250 ms
CAPAG = 5.0 MF
5 - 12 mF
VGAPM = 70 KV
40 - 70 kV
Note that for an arrester with a nominal voltage of N kV, parameters "R" and "VGAPM" must be
multiplied by N/72, "CAPAG" by 72/N and "TDELAY" is unchanged. The EMTP data cards needed
8 - 44
Non-Linear Elements
to run this test case are as follows (the two 1.E09 ohms resistances are used for network
connectivity):
BEGIN NEW DATA CASE

C -------------<MAXZNO><ESPZNO><EPWARN><EPSTOP><ZNOLIM><ZNOLIM>
ZO
20
2.5E-6
.040
C .....^.......^.......^.......^.......^.......^.......^
C -Iprnt<--Iplot<-Idoubl<-Kssout<-Maxout<---Ipun<-Memsav<---Icat<-Nenerg<-Iprsup
2
1
1
1
1
0
0
2
C .....^.......^.......^.......^.......^.......^.......^.......^.......^.......^
C Bus-->Bus-->Bus-->Bus--><--R-><-L -><-C ->
SOURCETOP
9.6
CAP
0.12
TOP
CAP
1.E9
TOP
ARR
0.18
ARR
1.E9
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
<FLAG
92ARR
6666.
1
C --------- VREF --------<------- VFLASH ---------<--------- VZERO -------101823.38
1.42403
0.0
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C --- TDELAY--<--- CAPAG ---<--- VGAPM ---<--- CURTOL -25.E-6
5.0
0.6875
C --------- COEF --------<------- EXPON ---------<--------- VMIN -------9780.0
0.3
0.00001
9999
C Bus-->Bus--><--Tclose>---Topen-><-CURRENT><--VFLASH>------------------------>O
TOP
CAP
0.0
100000.
2CAP
150000.
3CAP
150000.
TOP
CAP
ARR
SOURCE
8.11
Nonlinear Hysteretic Reactor (Type 92)
The type 92 hysteretic reactor is a true nonlinear device designed to simulate the saturation and
hysteresis phenomena that take place in the steel core of a power transformer. It is described as a
true nonlinear element because, unlike the Type-96 pseudononlinear hysteretic reactor, it is not
modelled as a set of switched inductances or piecewise linear approximations, but rather as a
closed-form nonlinear function solved using compensation.
8 - 45
Non-Linear Elements
8.11.1
Introduction
The hysteretic reactor is modelled by a closed-form function that relates instantaneous flux vs.
current in two steps:.
sat(i) = F1(unsat)
unsat = F2(i)
where F1 and F2 are nonlinear characteristics or functions. Therefore, the hysteretic reactor
consists of two components:
A saturation function relating "saturated" flux sat to "unsaturated" flux unsat.
A hysteresis function relating "unsaturated" flux unsat to current.
The saturation function is described in closed form with a second order equation of the form
Csat = (unsat - sat ms1 - bs1) (ms2 unsat - sat - bs2)

This equation represents a hyperbola whose vertical asymptote is defined by the straight line
unsat - sat ms1 - bs1 = 0

and whose horizontal asymptote is defined by the straight line
ms2 unsat - sat - bs2 = 0

where,
unsat
Unsaturated flux
sat
Saturated flux
ms1
Inverse of the slope of the vertical asymptote
bs1
Y-axis (sat) intercept of the vertical asymptote
ms2
Slope of the horizontal asymptote
bs2
X-axis ( unsat) intercept of the horizontal asymptote
csat
Curvature of the knee point of the hyperbola

8 - 46
Non-Linear Elements
sat
vertical asymptote
horizontal asymptote
unsat
Saturation Function
The second function models the hysteresis effect between unsaturated flux and current and it also
defined by a second order equation
Chyst = (i - mh1 unsat - bh1) (i mh2 - unsat - bh2)

This equation represents a hyperbola whose vertical asymptote is defined by the straight line
i - mh1 unsat - bh1 = 0

and whose horizontal asymptote is defined by the straight line
i mh2 - unsat - bh2 = 0
8 - 47
Non-Linear Elements
unsat
horizontal asymptote
vertical asymptote
i
coercive current
Hysteresis Function
where,
8.11.2
unsat
Unsaturated flux
Current
mh1
Inverse of the slope of the vertical asymptote
bh1
Y-axis ( unsat) intercept of the vertical asymptote
mh2
Slope of the horizontal asymptote
bh2
X-axis (current) intercept of the horizontal asymptote
c hyst
Curvature of the knee point of the hyperbola
Initialization
The flux-current trajectory assumed at the start of a transient simulation depends on the past
history of the device. If the simulation starts from zero initial conditions (i.e., does not start from
steady-state), the remnant flux specified by the user (REMFLX in Card 4, Section 8.11.4) is taken
into account. The initial trajectory is automatically determined based on the following
assumptions:
1.
The initial direction of the flux trajectory is chosen to be positive (ascending) if

the remnant flux is negative. Conversely, if the remnant flux is positive, the
initial trajectory is assumed to be negative (descending) This behaviour would
be consistent with the de-energization of an isolated transformer.
8 - 48
Non-Linear Elements
2.
If remnant flux is zero, a descending trajectory is arbitrarily chosen.
If the simulation starts from steady-state initial conditions at a single frequency, the Type-92
hysteretic reactor is seen by the rest of the network as a simple linear reactance, and the past
history of the device itself is calculated from the resulting steady-state voltage at its terminals. If
the operating point during steady-state does not lie sufficiently close to the linear region of the
characteristics, a transient will take place in the first time step of the simulation due the mismatch
in initial conditions.
To obtain better initial conditions, the INITIALIZATION WITH HARMONICS option should be used.
When this option is enabled, the characteristic of the device is assumed to have no losses (or zero
width), but otherwise the correct characteristic is used. The initial transient due to mismatch in
initial conditions is usually negligible in this case.
8.11.3
Time-step loop solution
In the time-step loop of the EMTP the type 92 hysteretic reactor model simply consists of one
multi-valued function that receives instantaneous flux (and past history) as input, and produces
instantaneous current as output; that is, i = F(). The compensation method is used to calculate
the correct network solution. Since flux is not readily available, it is calculated from the node
voltage by simple integration using the trapezoidal rule. Once the correct operating point is
reached, past history arrays that keep track of earlier flux reversals and operating points are
updated.
8.11.4
Data structure
To specify a type 92 hysteretic reactor, a total of four data cards are needed to describe the
characteristics for the saturation and hysteresis components of the model. Note that these cards
can also be generated using the support routine FITSAT (see Section 7.3 of Rulebook 2) from fluxcurrent characteristics, rather than from the trajectory characteristics.
Card 1:
I2
BUS1 BUS2
A6
A6
8888.
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 34568 901234 567890123456789012345678 90123 456789012345678901234567890123456789 0
E5.0
I1
8 - 49
Non-Linear Elements
ITYPE
Set ITYPE to 92
(1-2)
BUS1
Sending end node to which the reactor is connected.
(3-8)
BUS2
Receiving end node to which the reactor is connected.
(9-14)
(38-43)
Set flag to 8888. to indicate that a type 92 hysteretic reactor is

being specified.
IOUT
Standard IOUT parameter to control branch output.
FLAG
(80)
IOUT
=1
=2
=3
=4

produces both branch-current and
branch-voltage output
produces branch-power and
energy consumption output
Card 2:
1
2
3
4
5
6
7
8
12345678901234567890 12345678901234567890 12345678901234567890 12345678901234567890
SLOPE1
SLOPE2
CURVAT
COERC
E20.0
E20.0
E20.0
E20.0
SLOPE1
(1-20)
SLOPE2
(21-40)
Slope of the vertical asymptote of the hyperbola that

describes an unsaturated hysteresis loop.
Slope of the horizontal asymptote of the hyperbola that
describes an unsaturated hysteresis loop.
(41-60)
Curvature of the hyperbola that describes an unsaturated

hysteresis loop.
COERC
Coercive current or width of the hysteresis loop for zero flux.
CURVAT
(61-80)
8 - 50
Non-Linear Elements
Card 3:
1
2
3
4
5
6
7
8
12345678901234567890 12345678901234567890 12345678901234567890 12345678901234567890
SSLOPE1
E20.0
SSLOPE1
(1-20)
SSLOPE2
(21-40)
CURSAT
(41-60)
FSLXS
(61-80)
SSLOPE2
CURSAT
FSLXS
E20.0
E20.0
E20.0
Slope of the vertical asymptote of the hyperbola that

describes the saturation curve.
Slope of the horizontal asymptote of the hyperbola that
describes the saturation curve
Curvature of the hyperbola that describes the saturation
curve
Intersection of the horizontal asymptote of the hyperbola that
describes the saturation curve and the (saturated) flux axis in
a static -i curve.
Card 4:
1
2
3
4
5
6
7
8
12345678901234567890 12345678901234567890 12345678901234567890 12345678901234567890
VREF
REMFLX
EPSLN1
EPSLN2
E20.0
E20.0
E20.0
E20.0
VREF
Reference value used to scale the saturation curve.
(1-20)
REMFLX
(21-40)
EPSLN1
(41-60)
EPSLN2
(61-80)
User-supplied remnant flux in vs (volt seconds). Only used

when the EMTP starts from zero initial conditions.
Tolerance that controls the overtaking of a reversal point.
Default: EPSLN1=10-8.
Tolerance that controls the decision that there has been a
flux reversal. Default: EPSLN2=10-8.
8 - 51
Non-Linear Elements
8.11.5
Example
The following example shows a type 92 hysteretic reactor generated using the auxiliary routine
FITSAT. The flux current characteristic used as input to FITSAT is given below. FITSAT reads data
in the same format as the pseudo-nonlinear type 96 hysteretic reactor (static -i curve) and
produces an optimal fit using the two quadratic equations described in Section 8.11.1.
1.21254
0.516416E+01
0.115702E+02
0.205235E+02
0.287663E+02
0.433500E+02
0.744448E+02
0.124409E+03
0.174762E+03
0.225144E+03
0.119298E+05
0.0
0.113744E+03
0.119479E+03
0.125214E+03
0.128014E+03
0.130813E+03
0.133134E+03
0.135729E+03
0.136548E+03
0.136958E+03
0.159351E+03
The resulting punch file obtained from FITSAT has been used to create the example shown below:
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 1234567890
C
C
8888.
--->
--->
C ----Vref
--><------ Remflx ----><------ Eps1 ------><------ Eps2 ------>
92one-a grnd
8888.
1
28.9321308136
-5.80931901932
41.3372077942
1.21254003048
2.40782117844
0.651428490528E-04
2464.86596680
136.873901367
500000.000000
92one-b grnd
8888.
1
28.9321308136
-5.80931901932
41.3372077942
1.21254003048
2.40782117844
0.651428490528E-04
2464.86596680
136.873901367
500000.000000
92one-a grnd
8888.
1
28.9321308136
-5.80931901932
41.3372077942
1.21254003048
2.40782117844
0.651428490528E-04
2464.86596680
136.873901367
500000.000000
8.12
Dynamic Circuit Breaker Model
8.12.1
Background
The models in this section provide for the representation of the current interruption process in a
power circuit breaker. More details about these models can be found in Reference 25. Throughout
8 - 52
Non-Linear Elements
this section it is assumed that all input parameters are specified in volts,ohms, and amperes: not
on a per unit basis.
Unless the EMTP user is familiar with arc physics in circuit breakers, the dynamic arc models
should be used with caution. The parameters are not always well known, and one set of
parameters may not reflect the true behaviour of the arc for all possible interruption duties. Close
consultation with the manufacturer may be advisable. For an industry consensus, see Reference 4.
The time sequence of events during the opening process in a circuit breaker is shown below.
initial voltage ramp
Varc
nonlinear arc equation zone
Vmax
t
Tpart
Tmax
Tarc
Figure 8.25: Opening sequence in a Circuit Breaker

where Vmax, Tpart, Tmax, and Tarc must be supplied by the user, as explained in Section 8.11.2.
8.12.2
Data Format and Rules
The data cards for a circuit breaker begin with a single branch card for a Type-92 nonlinear
element.
BUS1
BUS2
I2
A6
A6
7777.
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 7890123456 78 901234 56789012345678901234567890123456789 0
I1
8 - 53
Non-Linear Elements
Rule 1:
Branch Type is 92 (in field ITYPE, columns 1-2).
Rule 2:
Specify terminal nodes by name (fields BUS1 and BUS2, columns 3-8 and 9-14, respectively). One node may be ground, by leaving the node name blank. Note that the reference branch option is not available in this model
Rule 3:
Set flag to "7777." request breaker modelling
Rule 4:
IOUT = 1
=2
=3
=4

1
2
3
4
5
6
7
8
1234567890 1234567890 123456789012345678901234567890123456789012345678901234567890
IMODEL
I10
Rule 1:
CONTROL
I10
Specify the model type to be used with IMODEL (1-10) according to the table below .
Model Name / Breaker Type
X1
X2
X3
X4
VMAX
Avdonin, user-supplied data.
Vmax
Avdonin, air blast breaker.
6e-06
1.6e07
0.2
0.5
4200
Avdonin, oil-filled breaker.
6e-06
10.e07
0.15
0.6
12000
Avdonin, SF6 breaker.
1.3e-06
0.1e-07
0.15
0.28
500
Urbanek, user-supplied data.
ud
Vmax
Urbanek, air-blast breaker.
2.e-06
30.e03
450e03
8000
7000
Kopplin
Kp
Vmax
IMODEL
Rule 2:
Set CONTROL to 1 to indicate that an additional data card will be used to modify the
default solution settings. Set CONTROL = 0 to use the default settings for the solution
algorithm.
8 - 54
Non-Linear Elements
1
2
3
4
5
6
7
8
123456789012345 678901234567890 123456789012345 678901234567890 12345678901234567890
TPART
E15.6
TPART
(1-15)
TMAX
E15.6
TARC
Nch
E15.6
E15.6
TPART 0, or
any valid TACS variable name starting with a letter, used to set
TPART in TACS statements, TPART is set to the simulation timepoint at which the TACS variable bedomes greater than zero.
TMAX
(16-30)
TMAX TPART
becomes an offset when TPART is a TACS variable.

TARC
(31-45)
TARC TMAX
becomes an offset when TPART is a TACS variable.

Nch
number of breaks, default is 1, or

any valid TACS variable name starting with a letter, must be set in
TACS initial conditions and can be changed through TACS
statements; maximum 6 characters.
If the model type, IMODEL was chosen such that the default values are used (IMODEL=1., 6. or 7.)
then the card Breaker Parameters should be omitted. Instead, the timing card is followed by the
Termination Card as detailed below.
Breaker Parameters Card (conditional):
1
2
3
4
5
6
7
8
123456789012345 678901234567890 123456789012345 678901234567890 123456789012345 67890
X1
X2
X3
X4
Vmax
E15.6
E15.6
E15.6
E15.6
E15.6
The meaning of the first 4 parameters on this card, X1,X2,X3 and X4 depends on the model being
used, as explained below.
8 - 55
Non-Linear Elements
Model Name / Breaker Type
X1
X2
X3
X4
VMAX
Avdonin, user-supplied data.
Vmax
Avdonin, air blast breaker.
6e-06
1.6e07
0.2
0.5
4200
Avdonin, oil-filled breaker.
6e-06
10.e07
0.15
0.6
12000
Avdonin, SF6 breaker.
1.3e-06
0.1e-07
0.15
0.28
500
Urbanek, user-supplied data.
ud
Vmax
Urbanek, air-blast breaker.
2.e-06
30.e03
450e03
8000
7000
Kopplin
Kp
Vmax
IMODEL
Vmax
(61-75)
Vmax 0
sign adjusted by the program.
Optional control card:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 123456789012345 678901234567890 123456789012345 67890
iter_max
epsilon
corr_lim
r_arc_max
drdt_max
I15.6
E15.6
E15.6
E15.6
E15.6
when control = 1:
iter_max
(1-15)
maximum number of iterations for the solution of the nonlinear arc equation,
default is 150, integer
epsilon
(16-30)
convergence tolerance for the iterative solution of the nonlinear arc equation,
default is 1.e-07
corr_lim
(31-45)
the maximum per unit (based on arc resistance) correction at each iteration
step, default is 1.0
r_arc_max
(46-60)
maximum arc resistance allowed before the breaker is treated as an open

circuit, default is 1.e07
drdt_max
(61-75)
maximum arc resistance differential allowed before the breaker is treated as an

open circuit, default is 1.e11
8 - 56
Non-Linear Elements
8.12.3
Arc Models - Brief Explanation

(A)
Avdonin Model: This model is a derivative of the Mayr model with the time
constant q replaced by Ara and the power constant P replaced by Brb. The
model is capable of representing arc interruption and thermal failure, and has
been used for modelling current chopping. It cannot simulate dielectric
breakdown or multiple restrikes, but has the advantage of being
computationally simple and robust.
The Advonin arc equation (see Reference 31) is given by:
1
1
r
dr
r
--------------------------- = --------------- vi
AB
dt
A
which is derived from the modified Mayr model:

dr
vi
--- ----- = 1 ----r dt
P0
where:
= Ar
P0 = r
The variables r, v, and i are respectively the arc resistance (ohms), voltage
(volts) and current (amps). is the arc time constant (seconds) and PO is
the breaker cooling power (watts).
(B)
Urbanek Model: This model is the most complex of the three, and can be used
to represent arc interruption and both thermal and dielectric failure. In
addition to these, it can model current chopping and re-ignition.
The Urbanek arc equation (see Reference 31) is given by:
2
v dv
1 vi
P
dg
v
--------------- = ----- g 2 1 ------ 2 2- ----dt
dt
2
2
ud
e
e
ud
wherre g is the arc conductance (mhos), P is the minimum power input

(watts) to maintain the arc, e is the arc voltage (volts) for high arc currents,
ud is the dielectric breakdown voltage (volts) for cold arc channel.
8 - 57
Non-Linear Elements
(C)
Kopplin Model: This model is also derived from the Mayr model, and can be
used with larger time-steps than the Avdonin model or the Urbanek model. It
simulates thermal breakdown
The Kopplin arc equation (see Reference 36) is given by:
1 dg
1
--- ------ = --- vi
- 1
---g dt
P
where:
= K ( g + 0,0005 )
0,25
P = K p ( g + 0,0005 )
0,6
constant K (seconds) and Kp (watts) are model parameters.

(D)
The default breaker parameters for X1, X2, X3, and X4 are taken from
Reference 31. The solution method is a Newton-Raphson algorithm based on a
high-order predictor of the arc resistance at each time step. The non-linear arc
dr
equation zone of Figure 8.?? is entered only when ---dt >0. The breaker is
dr
assumed to be open when ---dt >1e11 or r>1e07. These parameters are user
adjustable.
8.12.4
Usage Note
The breaker model is a compensation-based true nonlinear model, but it cannot coexist with any
other nonlinear model of breaker in the same sub-network. Therefore, the restrictions mentioned
in Section 8.9.1 apply to the use of the breaker model as well
When steady-state calculations are requested in the EMTP, the breaker is automatically replaced
by a switch that is closed at t < 0 (Tclose = -1) and open at t = 0- (Topen = -1) with an infinite
current margin.
8.12.5
Sample Data Cases
Two cases are shown here, one for AC breakers and the other for HVDC breakers. In case 1, the
default Avdonin air-blast breaker is used, and for the data case listed, interruption should occur. If
the source voltage is raised to 378 kV however, thermal failure should occur. In case 2, the
Kopplin model is used to model interruption in a HVDC system.
8 - 58
Non-Linear Elements
Case 1 - Air Blast Breaker:

SRC
BUS1
+
vs
Type 92
arc model
BEGIN NEW DATA CASE

C
C Type-92 arc model testing
C Section : Nonlinear elements, breaker models
C
C ........................... Miscelaneous data ..............................
C DeltaT<---TMax<---XOpt<---COpt<-Epsiln<-TolMat<-TStart
1.0E-6 14.e-3
C --IOut<--IPlot<-IDoubl<-KSSOut<-MaxOut<---IPun<-MemSav<---ICat<-NEnerg<-IPrSup
10
10
1
C
C ........................... Circuit data ...................................
C <-Bus1<-Bus2<-Bus3<-Bus4<----R<----L<----C
SRC
BUS1
6.9
3
BUS1
57.38
1.055
C
C arc model
C <
><
>
<><
> model flag
92BUS1
7777.
3
C
>
2
select model
C TPART
><
TMAX
>< TARC
>< Nchamb
>
4.e-03
5.0e-03
6.e-03
1.
C
BLANK card terminates circuit data
C
C ........................... Switch data ....................................
C
1
2
3
4
5
6
7
C 3456789012345678901234567890123456789012345678901234567890123456789012
BLANK card terminates switch data
C
C ...........................Source data ......................................
C Bus--><I<Amplitude<Frequency<--T0|Phi0<---0=Phi0
<----Tstart<----Tstop
14SRC
134.78E3
60.
0.
-1.
9999.
BLANK card terminates source data
C
C ........................... Output requests .................................
8 - 59
Non-Linear Elements
BUS1 SRC
BLANK
BLANK
BLANK
Case 2 - HVDC Breaker:

1
14
Type 92
arc model
vs
BEGIN NEW DATA CASE

C
1.E-7 2.00E-3
C
C -IPRNT<--IPLOT<-IDOUBL<-KSSOUT<-MAXOUT<---IPUN<-MEMSAV<---ICAT<-NENERG<-IPRSUP
3
1
1
1
C >BAR1<>BAR2<>BAR3<>BAR4<> R <> L <> C <
NODEA1NODEB1
21.0 50.0
NODED1NODEC1
21.0 50.0
NODED1NODEA1
0.10 1.0
NODEA1
900.
0.15
NODEB1
900.
0.15
NODEC1
900.
0.15
C >BAR1<>BAR2<
CIRCUIT BREAKER
92NODEA1NODED1
7777
3
C 34567890<23456789><23456789><23456789><23456789><23456789><23456789><23456789>
7
1
C TPART
><
TMAX
>< TARC
>< Nchamb >
0.0
19.5e-06
40e-06
C
Ktau
><
Kp
><
><
><
Vmax
>
15.0e-06
4.0e06
1000.
C ITERMAX><EPS
>< CORRLIM
>< DRDTMAX
>< RMAX
>
950
1.0e-07
1.0000
C
C 12345><
><23456789><23456789><23456789>
BLANK CARD ENDING SWITCH CARDS
C
C Current controlled DC voltage
C
2 means source initially at max. voltage
C < CAT><><2gain789>< V or I ><234T1 89><23T2 789><23T3 789><23Idc789>
16NODEA1 2 1.390
1000.
4.4
.04
.0103
900.0
8 - 60
Non-Linear Elements
C < AN >
< k1
>< k2
><Vdc min ><Vdc max >< R isol >
-148000.0 14800.00
-70000.0 70000.0
C
C
1 means source initially on constant current control
C <
><><2gain789>< V or I ><234T1 89><23T2 789><23T3 789><23Idc789>
16
1 1.390
-60420.0
4.6
.03
.0125
450.0
C
< k1
>< k2
NODEB1 -148000.0 14800.00
-120000. -10000.0
C
C <
><><2gain789>< V or I ><234T1 89><23T2 789><23T3 789><23Idc789>
16
1 1.390
-60420.0
4.6
.03
.0125
450.0
C
< k1
>< k2
NODEC1 -148000.0 14800.00
-120000. -10000.0
NODEA1NODEB1NODEC1
BLANK CARD ENDING SELECTIVE NODE VOLTAGE OUTPUT REQUESTS
BEGIN NEW DATA CASE
BLANK
8.13
Single-phase Nonlinear Inductance (Type-93 Element)
8.13.1
Applications
representation of nonlinear reactors (inductances). As explained in Section 8.6.2, only a single
reactor is permitted per subnetwork, i.e., this model cannot be used for representation of multiphase nonlinear reactors.
The nonlinear inductance L(i) is specified point by point as a piecewise linear characteristic,
=f(i), where is the total flux linkage. Linear interpolation is used between points.
L(i)
V
1
Figure 8.26: Nonlinear Inductance

8 - 61
Non-Linear Elements
Note:
= max cos t for sinusoidal flux with max = Vrms/(4.44f).
The EMTP support program AUX has a routine that will convert a Vrms vs. Irms saturation curve into
the vs. i curve which is needed here, under the assumption that there is no hysteresis (i.e., no
losses). See routine CONVERT in Section 7.1 of Rulebook 2 for details.
8.13.2
Format and Rules
Data cards for a nonlinear inductance begin with a single branch card for a Type-93 nonlinear
element.
BUS1
BUS2
BUS3
BUS4
I2
A6
A6
A6
A6
IOUT
ITYPE
1
2
3
5
6
7
8
12 345678 901234 567890 123456 789012 345678 90123456789012345678901234567890123456789 0
isteady steady
E6.2
E6.2
I1
Rule 1:
Rule 2:
respectively). One node may be grounded, if desired (blank for node name).
Rule 3:
Specify the linear region as pairs of values isteady, steady in columns 27-32 and 33-38,
respectively.
steady
i
isteady
This linear region is used in the calculation of ac steady-state conditions (no

harmonics).
8 - 62
Non-Linear Elements
Special Cases:
(A)
isteady = 0 or blank
steady = 0 or blank
To be used if no ac steady-state solution is

asked for or if non-linear inductances are to
be totally ignored during the steady-state
solution.
(B)
isteady = 0 or blank
steady > 0
nonlinear inductance will be nonconducting in steady-state solution, but

will, in general, have a non-zero initial flux
(is computed automatically).
(C)
Is not permitted (would be a short-circuit in

steady-state). This case leads to the
following error message:
isteady > 0
steady = 0
"NO INTERSECTION WITH

NONLINEAR CHARACTERISTIC NO.
XXX".
The linear region to be used in the ac steady state solution is defined by the slope L=steady/isteady.
The initial flux (0) will be listed in the output. If the solution leads to a flux (0) > steady, then
the following warning is printed:
"*** WARNING. ASSUMPTION THAT AC STEADY STATE HAS FUNDAMENTAL
FREQUENCY ONLY IS QUESTIONABLE WITH PRECEDING FLUX OUTSIDE LINEAR
REGION."
8 - 63
Non-Linear Elements
Rule 4:
IOUT = 1
=2
=3
=4
Rule 5:

If the nonlinear characteristic as well as the linear region are identical with that of a
preceding branch, then the following storage-saving option may be used: Repeat node
names of that preceding reference branch in the provided columns 15-26 in the same
sequence and do not add any additional card for the nonlinear characteristic.
Caution: If the reference branch has other branches in parallel, it is not clear
reference branches.
Next come cards which specify the nonlinear characteristic from the left to the right, with one card
for each pair of values i, nonlinear. The format is:
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
CUR
FLUX
E16.0
E16.0
The nonlinear characteristic must be monotonically increasing, that is, the values i, must never
decrease when read in from left to right. Terminate the grouping of these data cards with a card
having 9999 specified in columns 13-16.
8 - 64
Non-Linear Elements
8.13.3
Example
L(i)
LEFT
L(i)
NODE-1
RIGHT
NODE-2
both have
the same
characteristics
C
-20
20
-1
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
93LEFT RIGHT
10.
0.7
1
C ----CUR-------<-------FLUX----20.0
-1.0
-10.0
-0.7
0.0
0.0
10.0
0.7
20.0
1.0
9999
93NODE-1NODE-2LEFT RIGHT
1
8 - 65
Section 9
9.1
9.2
9.3
Switches
Introduction to EMTP Switches .............................................................. 9-1

9.1.1
Types of EMTP Switches ....................................................... 9-1
9.1.2
Switch Opening Considerations ............................................ 9-3
9.1.3
Switch Closing Considerations ............................................. 9-6
Ordinary Switches .................................................................................. 9-7

9.2.1
Class 1: Time-Controlled Switch (Circuit Breaker Pole) ..... 9-8
9.2.2
Class 2: Flashover-Controlled Switch (Gap) ........................ 9-8
9.2.3
Class 3: Random "STATISTICS" Switch Closing .................. 9-9
9.2.4
Class 4: "SYSTEMATIC" Switch ............................................ 9-12
9.2.5
"MEASURING" Switch ........................................................... 9-14
Format and Rules For Ordinary Switches .............................................. 9-14

9.3.1
Time-Controlled Switch ....................................................... 9-14
9.3.2
Voltage-Controlled Flashover Switch ................................. 9-15
9.3.3
Statistics Switch .................................................................... 9-16
9.3.4
Systematic Switch ................................................................. 9-18
9.3.5
Measuring Switch ................................................................. 9-20
9.4
Samples Of Ordinary Switch Specification ........................................... 9-21
9.5
TACS-Controlled Switches ..................................................................... 9-23
9.5.1
TACS Controlled Switch type-11 .......................................... 9-23
9.5.2
Format and Rules for type-11 Switch ................................... 9-24
9.5.3
Samples of type-11 Switch Specification ............................. 9-26
9.5.4
TACS-Controlled Switch type-12 .......................................... 9-27
9.5.5
9.5.6
TACS-Controlled Switch type-13 .......................................... 9-29
9.5.7
9.5.8
Sample of type-13 Switch Specifications ............................. 9-30
Section 9
Switches
9.1 Introduction to EMTP Switches
9.1.1
Types of EMTP Switches
The EMTP supports two types of switches:

1.
"Ordinary" switches, for which the voltage drop is zero when closed, and the
current is zero when open. For such switches ITYPE=0 (columns 1-2 of data
card), or possibly ITYPE=76 for exceptional "STATISTICS" switches.
ITYPE
2.
=0
TACS-controlled
TACS-controlled
switch, where the switch status is controlled by TACS. The

switches have ITYPE=11, 12 or 13 (columns 1-2 of data card).
Important special cases of this general component capability are diodes and
valves (as used for AC/DC conversion).
TACS
Dynamics and logic

of switch control
status control
Valve
Diode
Gap
Electric Network Switch
9-1
Switches
Data cards for all the different types of switches are collectively referred to as switch cards, since
in all cases the computer model involves a "perfect" or "ordinary" switch. Switch cards follow the
regular EMTP branch cards, and precede the EMTP source cards.
If a network involves no such switch components, then the blank card ending switches
immediately follows the blank card ending branches.
Excluded from the above description is the dynamic circuit breaker model which is internally
represented as a "true" nonlinear element. For details see Section 8.12.
Switches can be connected quite arbitrarily, although there are a few restrictions. One is that
switch currents must be unique: there must be no loop involving closed switches. A more obvious
restriction is that Kirchhoff's voltage law must not be violated for a loop involving a voltage source
and a switch. For example, a closed switch must not be connected to two nodes of known voltage
(including ground). If a case starts from zero initial conditions, no switches are necessary for
connecting voltage and current sources to the network at time zero, provided the sources remain
connected throughout the study.
Phasor switch flows (both current and power) are an optional printout. The integer miscellaneous
data parameter KSSOUT (Section 4.2.2) continues to control such steady-state output, of which the
following is a representative illustration:
OUTPUT FOR STEADY STATE SWITCH CURRENT
NODE-K
NODE-M
I-REAL
REACTIVE
AC6
AC6Y
-0.17537736E-02
0.16848494E+00
BC6
BC6Y
0.74585229E-03
0.16848494E+00
CC6
CC6Y
0.10079213E-02
0.16848494E+00
VI-1
CTH6C
OPEN
VI-3
CTH6C
OPEN
VI-5
CTH6C
OPEN
I-IMAG
I-MAGN
DEGREES
POWER
-0.15130561E-03
0.17602884E-02
-175.0691
0.14551287E-01
0.15944653E-02
0.17602884E-02
64.9309
0.14551286E-01
-0.14431597E-02
0.17602884E-02
-55.0691
0.14551287E-01
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
OPEN
Note that there are seven switches (actually, one switch and six diodes), with four of them open
during the phasor steady-state solution. If such phasor switch output is requested (KSSOUT
positive), it will follow the phasor branch flows, and precede the injections at nodes of known
voltage.
The switch logic provides for a correct assignment of switch currents at time zero for purposes of
printing and/or plotting. For any switch which is closed in the steady-state, printed and plotted
values will be correctly assigned for step zero. Also, all such switch currents at time zero can be
seen from the printout which immediately follows the variable heading of the time-step loop. As
an illustration, consider the printout for step zero for the above example:
_____TIME-STEP LOOP begins_____
__Column headings for the
17
EMTP output variables follow.
These are ordered according to the output-variable classes:
first
8 output variables are electric-network node voltages (with respect to local ground)
next
4 output variables are branch voltages (voltage of upper node minus voltage of lower node)
next
5 output variables are branch currents (flowing from the upper node to the lower)
STEP
VI-5
TIME
S4-
4-282
4-564
R4-
9-2
A3+
3-282
3-564
R3+
Switches
CTH6C
II-3
V-1
4-4
VI-5
II-3
V-1
4-4
CEL4
CTH2C
CTH5C
CTH4S
CTH6C
CTH2C
CTH5C
CTH4S
S4***
PHASOR I(0) = -0.1753774E-02
SWITCH "AC6
" TO "AC6Y " CLOSED AFTER
0.00000E+00 SEC.
***
PHASOR I(0) = 0.7458523E-03
SWITCH "BC6
" TO "BC6Y " CLOSED AFTER
0.00000E+00 SEC.
***
PHASOR I(0) = 0.1007921E-02
SWITCH "CC6
" TO "CC6Y " CLOSED AFTER
0.00000E+00 SEC.
0 0.000000-0.980868E-11-0.567554E-10-0.488658E-10-0.149247E-10-0.767030E-10-0.570429E-10 0.175139E-10
0.962480E-10 0.924472E+02
-0.803920E+02-0.462336E+02-0.603248E-02 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
0.236073E-13
The printed "PHASOR I(0)" equals the real part of the complex phasor switch current of the steadystate display, of course. Such output will be found for any switch which was closed during the
phasor solution (three of them in this example). Unlike earlier program versions, the closing times
for these switches are zero. A nonzero "PHASOR I(0)" value indicates that the switch has closed
during the steady-state phasor solution.
9.1.2
Switch Opening Considerations
The circuit below is used to indicate the difficulties which can result during the opening of a
switch in a physically improper model:
i(t)
0.18
0.8mH
V=sint (f=60Hz)
II
22100F
Figure 9.1: Test System

Assume that both switches I and II are closed at t=0, and the capacitor is charged by a transient
current i(t). Switch II is designed to open when the current is approximately zero. If the opening
of Switch II occurs at a time when there is a residual current i in the reactor, there is no
mechanism left for dissipating the rest energy L(i)2/2. The result is bounded voltage oscillations
across the reactor from one time step to another as shown in Figure 9.2 below. The frequency of
these numerical oscillations 1/2t, the magnitude of the numerical oscillations generally straddle
or bracket the "correct solution".
9-3
Switches
Vb
Va
Figure 9.2:
a) Simulated Reactor Voltage without CDA
Vb
Va
Figure 9.2: b) Simulated Reactor Voltage with CDA

9-4
Switches
Voltage oscillations occur even if the switch opening occurred exactly at current zero. The
problem is inherent to the trapezoidal rule of integration. In general, voltage oscillations occur
when the current through a hanging inductor is interrupted. Decreasing "DELTAT" does not solve
this problem. Numerical oscillations also occur in other situations, such as current chopping in
reactors or voltage chopping in capacitors. In power electronic simulations, the most common
source of numerical oscillations is the firing of diodes.
The EMTP has an option to eliminate these numerical oscillations call CDA (Critical Damping
Adjustment). To enable CDA, simply enter the keyword "CDA" immediately after the CASE
IDENTIFICATION CARD (see Section 4.1.1). When CDA is enabled, two Backward Euler integration
steps of length t/2 replace the trapezoidal rule integration step that would have created numerical
oscillations. Since each half time-step Backward Euler solution does not entail changes in the
conductance matrix of the system being solved, CDA is numerically efficient. These two time
steps take place every time there is switching action (including slope changes in pseudononlinear
elements), and in the first time step of the simulation. For more information on CDA, please see
Reference 17 and Reference 18.
Figure 9.2 (b) shows the same simulation in Figure 9.2 (a), with but with CDA enabled. It shows
that the voltages Va and Vb are the same after the switch is opened and no current is flowing
through the reactor, which is the "best" answer for this simulation. It should be noted, however,
that from a physical point of view, the current across the reactor in Figure 9.1 cannot change
instantaneously, and that a transient should take place after the switch opens. For instance,
modelling of the dynamic arc resistance across the switch, and adding parasitic capacitances
would reflect the physics of the problem more accurately. In such case, the presence or absence of
CDA would make little if any difference.
For example, the circuit in Figure 9.1 can be modified by placing a very small fraction of the
capacitance over onto the left side of the switch (to model leakage capacitance), as shown in
Figure 9.4. This provides a path for the switching transient. (damping will be incorrect because
the arc resistance has not been included). The results of the simulation with and without CDA are
indistinguishable (see Figure 9.5).
0.18
0.8mH
10F
Figure 9.3: Modified Test Network
9-5
22000F
Switches
Vb
Va
Figure 9.4:
9.1.3
Modified Simulation Results
Switch Closing Considerations
When the EMTP prints a message that a switch is closed after T seconds (see Figure 9.5), it means
that the closing occurs precisely at time t = T. For most cases, this is simple.
switch closed
switch open
t
time
Figure 9.5: Closing of switch at time instant t = T

But a one time-step delay may actually occur in certain cases. A switch closing represents a
change in the network; certain conditions (voltages, for example) immediately following the
closure may be quite different than those immediately preceding the closure. In Figure 9.6, we
9-6
Switches
really should (ideally) get a network solution for t = T+. Among other things, this would clearly
equalize the node voltages across the switch in question, and hence perhaps thereby initiate other
flashovers which should also really be performed at time instant t = T+. But the program does not
presently perform such extra solutions in the same time-step. Equalization of node voltages for
the just-closed switch will only occur as part of the network solution at time T+t, thereby
introducing a delay of t.
Example:
For the circuit of Figure 9.6, with both the switch and Type99 pseudononlinear element open, no
current flows. Hence source voltage vs(t) all appears across the switch, which will close when vs
exceeds flashover value Vswflash. At that instant the Type99 element should flashover too,
assuming V99flash < Vswflash. But with present program logic, one time step would separate
these two events.
Vsw-flash
vs(t)
V99-flash
99
Figure 9.6: Example Illustrating Delay
9.2 Ordinary Switches

Switches which are not diodes, valves, or gaps belong to one of five different basic types or
classes, as follows:
Class 1:
Conventional (i.e., deterministic) time-controlled switch.
Class 2:
Voltage-controlled (flashover) switch.
Class 3:
"STATISTICS" switch (a time-controlled switch for which the closing or opening time is
a random variable).
Class 4:
"SYSTEMATIC" switch (a time-controlled switch for which the closing time is systematically [regularly] varied).
9-7
Switches
Class 5:
"MEASURING" switch, which is permanently closed at all times. It is used just for current or power and energy monitoring.
These five classes of switches are described below.
9.2.1
Class 1: Time-Controlled Switch (Circuit Breaker Pole)
This switch is assumed to be initially open, closes at t > TCLOSE and tries to open at t > TOPEN. In
option A below, the opening is successful as soon as the current iSWITCH has gone through zero
(detected by a sign change in iSWITCH). In option B the opening is successful as soon as |iSWITCH|
< CURRENT MARGIN or as soon as iSWITCH has gone through zero. After a successful opening, the
switch will stay open.
Option A
Option B
Opening when iSWITCH goes

through zero
Opening when |iSWITCH| < CURRENT MARGIN

or
(set CURRENT MARGIN = 0)
iSWITCH goes through zero

(set CURRENT MARGIN > 0)
iSWITCH
iSWITCH
opening signalled
(takes effect in next step)
9.2.2
} current margin
t } current margin
opening signalled
(takes effect in next step)
Class 2: Flashover-Controlled Switch (Gap)
This switch is initially open, and it tries to close after t > TCLOSE. The closing is successful as soon
as the absolute value of the voltage across the switch is > FLASHOVER VOLTAGE. After the switch
has closed, it waits until the time delay TDELAY has elapsed and then tries to open again with the
same current criteria as in option A and B above. The sequence (1) flashover-controlled closing,
(2) elapse of time TDELAY, (3) current-controlled opening and back to (1) remains effective
throughout the study (see Figure 9.7).
9-8
Switches
Voltage across switch
opens
opens
opens
Flashover
Voltage
TDELAY
TCLOSE
TDELAY
TDELAY
closes
Figure 9.7:
closes
Flashover
Voltage
closes
Sample Illustrative Operation of Flashover-controlled Switch
There are cases where the voltage across the gap oscillates around the true value (i.e., numerical
oscillations). An example of this is the case of an inductance "hanging in the air", as per the
sketch below.
To avoid such problems, the gap voltage for checking flashover is computed as the average of the
last two time-step figures: This voltage averaging is also used when CDA is invoked.
V ( t ) + V ( t t )
Vgap ( t ) = ---------------------------------------2
9.2.3
(9.1)
Class 3: Random "STATISTICS" Switch Closing
A statistical overvoltage study consists of "NENERG" (miscellaneous data parameter; see

Section 4.2.2) separate internally-generated simulations, the peak voltages of which are then
processed statistically. A data case must contain one or more Class3 switches. Such switches
are called "STATISTICS" switches, after the key request word "STATISTICS" which must be punched
in columns 5564 of the switch card in question. "STATISTICS" switches are always initially open
(in particular, they are always treated as being open for the phasor steady-state solution); they then
close once at the appropriate random times (as determined by their probabilistic specifications),
never to open again.
9-9
Switches
The closing time Tclose for each "STATISTICS" switch is randomly varied according to either a
Gaussian (normal) distribution or a uniform distribution, as illustrated in Figure 9.8 below. The
distribution is controlled by an appropriate selection of variable "IDIST" (statistics miscellaneous
data parameter; see Section 4.3) by the user. When all switch type codes "ITYPE" (Columns 1-2 of
the switch card) are zero or blank and "IDIST" is zero, all "STATISTICS" switches will use the
Gaussian distribution. If "IDIST" is unity, the switches will use the uniform distribution. In order
to use both distribution types within the same data case, the user makes "IDIST" equal to zero, and
leaves the Gaussian-distributed switches with type code "ITYPE" equal to zero or blank; the
exceptional switches which are to have uniform distributions are then flagged by using the type
code "ITYPE" equal to 76. Figure 9.8 illustrates the two available distribution types for
"STATISTICS" switches. The mean closing time T and associated standard deviation are specified by
the user on the one or more "STATISTICS" switch cards.
f(T)
0,3
------
f (T )
1
-------------2 3
0,2
------
0,1
------
T
T + 2
T 2
T T+
T 3
F(T)
T + 3
F(T)
1.0
1.0
0.5
0.5
a) Gaussian distribution IDIST = 0 and

ITYPE 76.
b) Uniform distribution IDIST = 1 or

ITYPE = 76
Figure 9.8: Available probability distributions for the closing time Tclose of a
"STATISTICS" switch. Density functions f(T) are shown above, and
cumulative distribution functions F(T) below.
In addition to switch closing-time variation caused by each switch's own distribution, there is an
added random delay which is the same for all switches, provided the user has given variable
9 - 10
Switches
"ITEST" (statistics miscellaneous data parameter; see Section 4.3) a value of zero or blank. This
added random delay which is applied equally to all "STATISTICS" switches is referred to by the
term "reference angle"; it follows a uniform distribution always, the probability distribution for
which is specified using parameters "DEGMIN", "DEGMAX", and "STATFR" of the statistics
miscellaneous data card (see Section 4.3.1). Printed output for a statistical over-voltage study will
show the reference angle for each energization, in approximately the columns which are normally
used for the step number and time value (marked "STEP" and "TIME") of a conventional
deterministic (non-"STATISTICS") data case. The numerical value printed out will be in degrees,
and will lie between "DEGMIN" and "DEGMAX". If the user sets "ITEST" equal to unity, the
reference angle will simply be constrained to always be zero --- which is mathematically
equivalent to not having any added random delay. There is also a capability for dependent
switches i.e., the closing time of one switch depends on that of second switch with both switches
being "STATISTICS" switches.
"A" (master)
"B" (slave)
The closing time of a switch labelled "B" can be made equal to the sum of two components:
TBclose = TAclose + TBrandom
(9.2)
where:
TAclose
Previously-determined closing time of another "STATISTICS"

switch, labelled here as switch "A". This is a random variable,
note, not a constant.
TBrandom
Random time delay having mean TB and standard deviation sB

(parameters as entered on the data card for switch "B"). If
desired, the mean TB of this component can be negative. The
only constraint which must be kept in mind is that the resulting
actual switch closing times must be less than the beginning time
of the simulation very infrequently.
But switch "A" might, in turn, depend on a third switch "Z". The EMTP allows an arbitrary number
of such dependencies. In terms of the above example, switch "B" is called a dependent (or slave)
switch, while switch "A" is the reference (or master) switch. An arbitrary number of levels of
dependence are permitted.
The identification or specification of the reference switch "A" in the above example is actually
made using the two 6-character alphanumeric bus names of the switch. The names of this
reference switch "A" are to be entered using 2A6 format in columns 6576 of the data card for
dependent switch "B". Where no closing time dependence exists, switch cards are completely
9 - 11
Switches
independent, and the parameters affecting any one switch are totally contained on the switch card
for that switch. This is perhaps the most common usage.
Random "STATISTICS" Switch Opening:

A switch which is to randomly open starts out as being closed in the steady-state, and it is
subsequently inhibited from opening until the time which is determined by rolling the dice. But
switch current is not generally zero at this instant of time. The actual opening is delayed until the
switch current goes through zero. If this never happens, the switch will never open.
As indicated in Section 4.1.3, variable TENERG (columns 33-40) in "TIME OF DICE ROLL" special
request card, is the time before which no random switching can reasonably occur. In the case of
statistical switch opening, TENERG should be set to -1.
A positive current margin I (columns 3544) can be used for random opening switches. When this
is done, the random opening switch would open after T > TOPEN and as soon as |ISWITCH| < I or as
soon as ISWITCH = 0.
It is possible to mix random opening and random closing within the dependency chain explained in
the description of random closing.
9.2.4
Class 4: "SYSTEMATIC" Switch
An EMTP study which involves a "SYSTEMATIC" switch consists of |NENERG| (miscellaneous data
parameter; see Section 4.2.2) separate internally-generated simulations, the peak voltages of which
are then processed and tabulated statistically. Each "SYSTEMATIC" switch has its closing time
systematically (regularly) varied between given beginning and ending times, in steps which are
uniformly spaced. Such systematically-varied switches are named "SYSTEMATIC" switches after
the key request word "SYSTEMATIC" which must be specified in columns 5564 of the switch card
in question. "SYSTEMATIC" switches are always initially open (in particular, they are always
treated as being open for the phasor steady-state solution); they can close once at the appropriate
systematic time, never to open again. Note that "NENERG" should be negative for "SYSTEMATIC"
data cases and positive for "STATISTICS" data cases.
The user should be aware of the potentially large number of energizations which can be involved in
a "SYSTEMATIC" data case. As an example, consider a data case which has three independent
"SYSTEMATIC" switches. The internally-generated energizations can be visualized as being a
methodical exploration of a rectangular region of the 3-dimensional vector space of switch closing
times (see Figure 9.10). For example, if switch "A" were to take 6 steps, switch "B" 5 steps, and
switch "C" 4 steps, then a total of NENERG =120=6x5x4 energizations would be internally
generated by the EMTP.
9 - 12
Switches
TCclose
NENERG
=543
Na = 5
Nc = 4
Nb = 3
Ta
TAclose
Tb
{
TBclose
Figure 9.9:
Exploration in 3-Dimensional Space
Dependency among "SYSTEMATIC" switches is permitted, and deserves an explanation. The

general concept is similar to that for "STATISTICS" switches, but with several crucial differences.
Let switch "B" be the dependent switch and switch "A" the reference switch, as before. Then the
closing time of dependent switch "B" is not independently varied, but rather is offset by a fixed
amount from the closing time of the reference switch "A":
TBclose = TAclose + TBoffset
(9.3)
Here TBoffset is a constant (it has the same value for all energizations). Rather than a beginning
(or mid) time, a step-size, and a number of steps, the input time-data for a dependent
"SYSTEMATIC" switch consists of only the constant offset time, which is to be specified in columns
15-24. A dependent "SYSTEMATIC" switch does not increase the dimensionality of the vector
9 - 13
Switches
space of switch closing times which is to be systematically explored, note. The mechanism for
indicating the dependence of a "SYSTEMATIC" switch is identical to that which is used for a
"STATISTICS" switch; the node names of the reference switch are to be punched in columns 6576
of the card for the dependent switch.
9.2.5
"MEASURING" Switch
By definition, a "MEASURING" switch is permanently closed (including any phasor steady-state

solution for initial conditions). It is used for the monitoring of current, or power and energy, in
places where these quantities are not otherwise available. This may be for EMTP output purposes
(printing or plotting), or in conjunction with use of the TYPE-91 TACS signal source (see
Section 14).
9.3 Format and Rules For Ordinary Switches

9.3.1
Time-Controlled Switch
Parameters for an ordinary switch are specified with a single card with the following format:
ITYPE
Node Name
BUS1
I2
A6
BUS2
A6
Time Criteria
TCLOSE
TOPEN,
E10.0
E10.0
Current
margin
I OUT
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234 5678901234 5678901234 567890 123456 789 0
E10.0
I1
Rule 1:
The switch type code (data card field ITYPE; Columns 1-2) is zero. Enter zero or leave
blank.
Rule 2:
Specify the two terminal nodes by names (fields BUS1 and BUS2; Columns 38 and 914, respectively). One of the nodes may be ground, indicated by a blank field for the
associated name.
Rule 3:
No switch is permitted between two voltage sources, or between one voltage source
and ground. The program will flag such a situation with an error message.
Rule 4:
If a switch connects a voltage source to a current source, then the current source is
ignored whenever the switch is closed.
9 - 14
Switches
Rule 5:
If transients start from a non-zero ac steady-state condition, make TCLOSE < 0 for
time-controlled switches which are closed in the ac steady-state condition.
Rule 6:
The switch will open for t TOPEN or when I |current margin|
Rule 7:
Output options for printing and/or plotting are specified with IOUT in column 80, as
follows:
9.3.2
IOUT
= 1 switch current output;
IOUT
= 2 switch voltage output;
IOUT
= 3 both switch current and voltage output;
IOUT
= 4 switch power and energy flow
Voltage-Controlled Flashover Switch
ITYPE
Node Name
BUS1
I2
A6
Rule 1:
Time Criteria
BUS2
TCLOSE
TDELAY
A6
E10.0
E10.0
Current
margin
E10.0
Flashover
Voltage
VFLASH
E10.0
IOUT
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234 5678901234 5678901234 567890 123456 789 0
I1
blank.
A special exception is type code ITYPE=76. This is for a "STATISTICS" switch that the
user wants to follow a uniform distribution, in a problem with one or more
"STATISTICS" switches which follow Gaussian distributions (statistics miscellaneous
data parameter "IDIST" equal to zero).
Rule 2:
associated name.
Rule 3:
9 - 15
Switches
Rule 4:
Rule 5:
Flashover-controlled switches will always be open in the ac steady-state; if TCLOSE < 0

on these switch types, the program will change it to TCLOSE=0 with the message
"TCLOSE CHANGED TO 0."
Rule 6:
After flashover, the switch will stay closed until V < VFLASH and I < |current margin|.
Rule 7:
follows:
9.3.3
IOUT
IOUT
IOUT
IOUT
Statistics Switch
ITYPE
Node Name
BUS1
I2
A6
BUS2
A6
Time Criteria
T
E10.0
E10.0
Reference
Switch Names
Current
Margin
STATISTICS
E10.0
A10
BUS5
BUS6
A6
A6
IOUT
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234 5678901234 5678901234 567890 123456 789 0
I1
Rule 1:
blank. A special exception is type code ITYPE=76. This is for a "STATISTICS" switch
that follows a uniform distribution, in a problem with one or more "STATISTICS"
switches which follow Gaussian distributions (statistics miscellaneous data parameter
"iDIST" equal to zero).
Rule 2:
associated name.
Rule 3:
9 - 16
Switches
Rule 4:
Rule 5:
For a "STATISTICS" switch, the mean closing time T (the expected value of the random closing time Tclose), as well as T (standard deviation from TCLOSE for a Normal
distribution) is specified by the user. T and the switch closing times generated from
the specified distributions should always be positive. This is because negative time in
an EMTP simulation corresponds to the sinusoidal steady-state solution, thereby making it impossible to inject switching events. To avoid such situations, (T - A) should
be positive; where A is a constant that depends on the type of probability distribution
being used:
A = SIGMAX
for Gaussian (normal) distribution. See Section 4.3

definition (user-supplied parameter which is defaulted
to 4.0 if left blank).
for uniform distribution
A fatal EMTP error termination results, if this check is not satisfied for all "STATISTICS" switches.
For a Gaussian distribution, this check ensures that any particular switch closing time turning out
negative has probability which is less than 0.005%; for a uniform distribution, this guarantees that
all switch closing times will always be positive (see Figure 9.9, Section 9.2). For dependent
switches in which Gaussian and uniform distribution offsets are mixed, the "SIGMAX" value is
applicable to the Gaussian components and the square root of three is used on all uniform
components.
Rule 6:
For a "STATISTICS" switch, the special request word "STATISTICS" must be specified in
columns 5564 (this keyword is case sensitive and must start in column 55).
Rule 7:
A specialized data-convenience option is provided by the "TARGET" feature. Any nondependent "STATISTICS" switch can be designated as a "TARGET" switch by specifying
this keyword in columns 6570 of the switch card. Only one such designation is permitted in any one data case. The effect is that all other non-dependent "STATISTICS"
switches will be internally given mean closing times which are equal to the mean closing time of the "TARGET" switch. Several details might be noted:
(A)
Only convenience of data entry is involved because the "TARGET" option does
not extend the modelling capability at all.
(B)
Nothing has been said about standard deviations T. It is only the mean
closing time T which is affected by the use of "TARGET".
(C)
Columns 1524 of non- "TARGET", non-dependent "STATISTICS" switch cards

can be left blank (since the mean closing time will come from the "TARGET"
switch).
9 - 17
Switches
Rule 8:
To indicated random opening of a switch, columns 4554 of the switch card should
have the flag "3333". Otherwise, data fields are unchanged except that "closing" is to
be read as "opening" wherever it is encountered. That is, the mean opening time T
goes in columns 15-24, the standard deviation goes in columns 2534, and the keyword "STATISTICS" must be entered in columns 5564. The use of "TARGET" or
dependency of switches is also allowed, according to the same rules as before. In fact,
the user can mix random opening and random closing within the dependency chain, if
so desired.
Rule 9:
A dependent "STATISTICS" switch can have the terminal node names of a reference
switch entered in columns 6576.
Rule 10:
follows:
9.3.4
IOUT
IOUT
IOUT
IOUT
Systematic Switch
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234 5678901234 5678901234 567890 123456 789 0
BUS1
I2
A6
Reference
Switch Names
Time Criteria
NSTEP
BUS2 TMID or TBEG

A6
E10.0
T
E10.0
E10.0
SYSTEMATIC
A10
BUS5
BUS6
IOUT
ITYPE
Node Name
A6
A6
I1
Rule 1:
blank.
Rule 2:
associated name.
Rule 3:
9 - 18
Switches
Rule 4:
Rule 5:
For either a "SYSTEMATIC" switch, the special request keyword "SYSTEMATIC" must be
specified in columns 5564 (this keyword is case sensitive and must start in column
55).
Rule 6:
For each non-dependent "SYSTEMATIC" switch card, the size of the time increment t
is specified in columns 2534. The number of steps "NSTEP" is to be entered in columns 3544.
Rule 7:
The data parameter entered in columns 1524 depends on the value of "ITEST" (miscellaneous data parameter of Section 4.3.1):
ITEST
=0
enter the mid-time TMID
=1
enter the beginning time TBEG
TBEG
TMID
Tclose
T1
T2
T3
T4
T5
T6
T7
Rule 8:
A specialized data-convenience option is provided by the "TARGET" feature. Any nondependent "SYSTEMATIC" switch can be designated as a "TARGET" switch by specifying this keyword in columns 6570 of the switch card. Only one such designation is
permitted in any one data case. The effect is that all other non-dependent "SYSTEMATIC" switches will be internally given mid-times or beginning times which are equal
to that of the "TARGET" switch. Several details might be noted:
(A)
(B)
Only convenience of the data punching is involved, for the "TARGET" option
does not extend the modeling capability at all.
Nothing has been said about the time increment t and the number of steps
It is only the mid-time or the beginning time which is affected by the
use of "TARGET".
NSTEP.
9 - 19
Switches
(C)
Columns 1524 of non-"TARGET", non-dependent "SYSTEMATIC" switch cards

can be left blank (since the beginning or mid-time will come from the
"TARGET").
Rule 9:
A dependent "SYSTEMATIC" switch is to have the terminal node names of the reference
switch entered in columns 6576.
Rule 10:
follows:
9.3.5
IOUT
IOUT
IOUT
IOUT
Measuring Switch
LK
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234 5678901234 5678901234 567890 123456 789 0
BUS1
I2
A6
MEASURING
BUS2
A6
A10
IOUT
ITYPE
Node Name
I1
Rule 1:
blank.
Rule 1:
associated name.
Rule 2:
Rule 3:
9 - 20
Switches
Rule 4:
follows:
IOUT
IOUT
IOUT
IOUT
9.4 Samples Of Ordinary Switch Specification

As an example of the preceding "ordinary" switch rules and data format specification, consider the
sample usages of the following listing:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus-->Bus--><--Tclose>---Topen-><-CURRENT><--VFLASH>----------------------->O
BUS-1 BUS-2
0.001
BUS-3 BUS-4
0.0005
0.002
CAP-1 CAP-2
0.1
1.E6
JDAYGAJDAYA
0.002
0.0004
STATISTICS
AS
ASW
0.002
0.0001
12.
SYSTEMATIC
BS
BSW
0.004
0.0005
STATISTICSTARGET
CS
CSW
0.002
0.0001
8.
SYSTEMATICTARGET
DS
DSW
0.006
0.0002
STATISTICS
ASW
ES
ESW
-0.005
SYSTEMATIC
CSW
1.
Breaker pole "BUS-1 " to "BUS-2 " closes at t = 0 and opens after t > 0.001s as
soon as current goes through 0.
2.
Breaker pole "BUS-3 " to "BUS-4 " closes at t = 0.0005s and opens after t >
0.002s as soon as absolute value of switch current is < 0.001 A or goes through
zero.
3.
Gap "GAP1 " to "GAP2 " closes whenever voltage across switch becomes >
1000000 V and opens as soon as absolute value of switch current becomes < 0.1
A or goes through zero.
4.
"STATISTICS" switch connecting node "JDAYGA" with node "JDAYA " has mean
closing time of 2.0 ms and associated standard deviation of 0.4 ms. Assuming
that statistics miscellaneous data parameter "IDIST" is zero, the probabilistic
9 - 21
Switches
distribution will be Gaussian (normal). This assumes that there is no "TARGET"
"STATISTICS" switch in the data case.
5.
Non-dependent "SYSTEMATIC" switch that connects node "AS " with node
"ASW ". The mid-closing time is 3.0 msec, assuming "ITEST" is zero and that
the data case in question has no "TARGET" switch. The closing time of the switch
is to be varied in steps of size 0.1 msec; 12 steps are to be taken.
6.
"STATISTICS" switch connecting node "BS " with node "BSW " is the "TARGET"
switch for the data case in question. All non-dependent "STATISTICS" switches of
the data case will have the same mean closing time of 4.0 msec. That is, the
information which is punched in columns 15-24 of all other-dependent
"STATISTICS" switches if overridden by the value for the present "TARGET"
switch. The standard deviation of the closing time for this "TARGET" switch is
0.5 msec.
7.
"SYSTEMATIC" switch connecting node "CS " with node "CSW " is the
"TARGET" switch for the data case. All non-dependent "SYSTEMATIC" switches
of the data case have same mid-closing time of 2.0 msec (if "ITEST" is zero), or
the same beginning closing time of 2.0 msec (if "ITEST" is unity). The
information of columns 15-24 on all other non-dependent "SYSTEMATIC" switch
cards is overridden by the value on the "TARGET" switch card (that now being
discussed). The closing time of this switch is to be varied through 8 steps; the
step-size is 0.1 msec.
8.
The switch connecting node "DS " with node "DSW " is a dependent
"STATISTICS" switch. Its reference switch is the one connecting node "AS "
with node "ASW ". Remember that the reference switch must have been
previously inputted. The closing time of the present switch is delayed form that
of its reference by a random delay time which has a mean of 6.0 msec and a
standard deviation of 0.2 msec.
9.
The switch connecting node "ES " with node "ESW " is a dependent
"SYSTEMATIC" switch. Its reference switch is the one which connects node
"CS " with node "CSW ". The closing time of the present switch is offset
(delayed) from that of its reference by a constant offset time of -5.0 msec. That
is, the present switch will always close 5.0 msec before closure of the reference
switch.
9 - 22
Switches
9.5
TACS-Controlled
Switches
Switches whose status is controlled by TACS have the following type codes:
1)
TYPE-11 TACS
controlled switch
2)
TYPE-12 TACS
controlled switch
3)
TYPE-13 TACS
controlled switch
The TYPE-11 switch is used in diode and valve applications, the TYPE-12 switch in spark gap and
triac applications, while the TYPE-13 switch is used to represent a general TACS-controlled switch.
9.5.1
TACS
Controlled Switch TYPE-11
A TYPE-11 switch, while following the simple opening/closing rules of a standard diode, may
simultaneously be controlled by any given TACS variable used as either a grid firing signal or an
overriding OPEN/CLOSE signal.
This TYPE-11 switch may represent any of the following:
1st Node
2nd Node
1st Node
2nd Node
"GRID" Firing Signal
TACS-controlled
Opening/closing Signal
Figure 9.10: Type-11 Switch applications

or a combination of those three modes, in which case either or both grid signal and overriding
signal can be active simultaneously.
9 - 23
Switches
Besides the four alphanumeric node- and variable-names (used to define 1st-NODE, 2nd-NODE and
possibly but not necessarily "GRID" signal and "OPEN/CLOSE" signal), three additional
parameters may be defined:
9.5.2
1)
the minimum ignition voltage (Vig), defaulted to 0.0
2)
the minimum holding current (IHOLD), defaulted to 0.0
3)
the de-ionization time (tDEION), defaulted to 0.0
Format and Rules for TYPE-11 Switch
BUS2
A6
A6
tDEION
IHOLD
E10.0
E10.0
(in sec.)
SAME
I2
BUS1
Vig
CLOSED
TYPE
Node Names
E10.0
A6
A4
TACS Names
GRID
A6
OPEN/
.
CLOSE
A6
ECHO
IOUT
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 5678901234 5678901234 567890 1234 567890 123456 78 9 0
I1 I1
Rule 1:
Enter 11 in columns 1-2.
Rule 2:
Specify the two terminal nodes names (fields BUS1 and BUS2 of columns 3-8 and 9-14,
respectively). One of the nodes may be ground, indicated by a blank field. The direction of the current is taken to be from "BUS1" to "BUS2". In other words, we have for
valves and diodes:
anode
always "BUS1"
cathode always "BUS2"

Rule 3:
TYPE-11 switches can be connected quite arbitrarily, but there are some restrictions that
are common to all switches. For details see Section 9 11.

Rule 4:
TACS GRID
(A)
signal and OPEN/CLOSE signal (columns 65-70 and 71-76, respectively).
If neither is specified, then the switch is a diode.

v
9 - 24
Switches
(B)
(i)
the diode starts conducting when the forward voltage across its
terminals becomes greater than the minimum ignition voltage Vig of
columns 15-24 (defaulted to zero).
(ii)
the diode ceases conducting as soon as the forward current becomes

smaller than the minimum holding current IHOLD of columns 25-34
(defaulted to zero).
if the grid signal is specified (identified by the alphanumeric TACS-variable

name in columns 65-70), then the switch is a valve:
v
(C)
(i)
which follows the opening and closing rules of the diode,
(ii)
except that the closing operation (start of conduction) will not take
place until the value of the grid signal becomes greater than zero;
(iii)
unless the valve is actually reclosing within the de-ionization time

following the previous opening, in which case the reclosing is
immediate with or without the presence of a valid grid signal. (Note
that tDEION of columns 35-44 is defaulted to zero.)
if the OPEN/CLOSE signal is specified (identified by the alphanumeric TACSvariable name in columns 71-76), then the operating rules of the diode or the
valve are overridden as soon as this signal becomes different than zero;
TACS-CONTROLLED
(i)
if positive, the switch will immediately close and stay closed as long as
the positive signal is active;
(ii)
if negative, the switch will immediately open and stay open as long as
negative signal is active;
9 - 25
Switches
(iii)
if zero, the switch returns to its regular diode or valve operating mode.
Rule 5:
The switch may be specified as closed during the steady-state initialization, if the user
enters the keyword "CLOSED" in columns 55-60.
Rule 6:
To facilitate data entry, the user may use the option "SAME" (by specifying the keyword "SAME" in columns 61-64), when data for the parametric fields (Vig, IHOLD, and
tDEION) have to be repeated on one or more cards that are placed immediately after the
card bearing the definition of these parameters.
Rule 7:
The user may obtain a printed echo of the occurrence of all opening and closing operations of the switch, by specifying a "1" in column 79. (No such diagnostic is printed if
left blank or zero.)
Rule 8:
9.5.3
IOUT
IOUT
IOUT
IOUT
Samples of TYPE-11 Switch Specification
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
11L2
R2
CLOSED
11LEFT RIGHT
FIRE2
1
11NODE1 NODE2
ORDER
11L6
R6
DELAY TRIG
1.
Diode from "L2
" to "R2 ", conducting in steady-state.
2.
Valve from "LEFT " to "RIGHT ", with "FIRE2 " as grid signal from TACS. Every
change of status will be echoed in the printout.
3.
TACS-controlled switch from "NODE1

"ORDER " has the value "0.0".
4.
Valve from "LG " to "R6 ", with grid signal "DELAY ", with its operation
overridden by the TACS-variable "TRIG " during the periods in which the value of
"TRIG " is different than zero.
9 - 26
" to "NODE2 ", operating as a diode when
Switches
9.5.4
TACS-Controlled
Switch TYPE-12
A TYPE-12 switch is used to simulate a spark gap or a triac, with firing controlled by a TACS
variable, "SPARK." An additional TACS variable "CLAMP" may be used to keep the switch open or
closed.
The operating logic of the TYPE-12 switch is as follows:
1.
"SPARK" signal from TACS:

(A)
If the field "SPARK" is left blank (see card image below) the switch is
always open.
(B)
If the "SPARK" signal < 0, the switch is kept open if it is already in an

open position. But if the switch is closed when the "SPARK" signal
arrives, the following rules apply:
(i)
if |i| > Ihold then the switch is maintained closed.
(ii) if |i| < Ihold the switch is opened.

(C)
If the "SPARK" signal > 0, the switch is kept closed if it is already in a

closed position. But if the switch is open when the "SPARK" signal
arrives, the following rules apply:
(i)
if |V| < Vig then the switch is maintained open
(ii) if |V| > Vig the switch is closed.

IHOLD was reached
V
V
"SPARK" > 0
t
"SPARK" > 0
Vsign was reached
2.
"CLAMP" signal from TACS:

9 - 27
Switches
(A)
If the field "CLAMP" is left blank (see card image below) or if the "CLAMP"
signal = 0, then the above rules for "SPARK" apply.
(B)
If the "CLAMP" signal is < 0, then the following rules apply:

(i)
the switch is kept open if it is already in an open position
(ii) the switch is opened if it is in a closed position when the "CLAMP"

signal arrives
While the switch is clamped open, all "SPARK" signals are ignored.
(C)
If the "CLAMP" signal is > 0, then the following rules apply:

(i)
the switch is kept closed if it is already in a closed position
(ii) the switch is closed if it is in an open position when the "CLAMP"

signal arrives
While the switch is clamped closed, all "SPARK" signals are ignored. The
figure below shows a switch closed by a positive "SPARK" signal and
subsequently clamped closed by a positive "CLAMP" signal.
"CLAMP" > 0
t
"SPARK" > 0
9.5.5
I2
BUS1
BUS2
A6
A6
Rule 1:
Vign
E10.0
IHOLD
E10.0
A6
Punch 12 in columns 1-2.

9 - 28
TACS Names
SPARK CLAMP .
A6
A6
ECHO
Output
Node Names
CLOSED
TYPE
1
2
3
4
5
6
7
8
12 345678 901234 5678901234 5678901234 56789012345678901234 567890 1234 567890 123456 78 9 0
I1 I1
Switches
Rule 2:
Specify the two terminal nodes by name (fields BUS1 and BUS2 of columns 3-8 and 914, respectively). One of the nodes may be ground, indicated by a blank field. The
direction of the current is taken to be from "BUS1" to "BUS2". Note that the current
opening rules observe the absolute value of the current, not the polarity.
Rule 3:
TYPE-12
Rule 4:
Minimum ignition voltage: columns 15-24, defaulted to 0.0.
switches can be connected quite arbitrarily, but there are some restrictions
common to all switches. For details see Section 9.11.
Minimum holding current: columns 25-34, defaulted to 0.0.

Rule 5:
The switch may be specified as closed during the program-calculated steady-state initialization, if the user punches the keyword "CLOSED" in columns 55-60.
Rule 6:
The user may obtain a printed echo of the occurrence of all opening and closing operations of the switch, by entering a '1' in column 79. (No such diagnostic if left blank
or zero.)
Rule 7:
9.5.6
IOUT
IOUT
IOUT
IOUT
TACS-Controlled
Switch TYPE-13
This is used to represent a switch which is controlled by any given TACS variable. The opening
and closing operations are controlled by a single OPEN/CLOSE signal defined in columns 71-76.
Open/close
Signal
TYPE-11
TYPE-12
TYPE-13
<0
OPEN
OPEN
OPEN
=0
depends on
GRID signal
depends on SPARK
signal
OPEN
>0
CLOSE
CLOSE
CLOSE
9 - 29
Switches
9.5.7
BUS1
BUS2
keyword
"CLOSED"
TACS
open/
close
signal
I2
A6
A6
A6
A6
ECHO
IOUT
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 5678901234567890123456789012345678901234 567890 1234567890 123456 78 9 0
I1 I1
Rule 1:
Set
Rule 2:
Specify the two terminal nodes by name (A6 fields BUS1 and BUS2 of columns 3-8 and
9-14, respectively). One of the nodes may be ground (blank).
Rule 3:
TYPE-13
Rule 4:
The OPEN/CLOSE signal is identified by the TACS variable, the name of which is entered
in columns 71-76.
Rule 5:
This switch is open during the steady-state initialization at t = 0.0, unless the keyword
"CLOSED" is specified in columns 55-60.
Rule 6:
The user may obtain a printed echo of the occurrence of all opening and closing operations of this switch, by entering a "1" in column 79. (No such diagnostic if column 79
is left blank or zero.)
Rule 7:
9.5.8
ITYPE
= 13 in columns 1-2.
switches can be connected quite arbitrarily, but there are some restrictions
common to all switches. For details see Section 9.11.
IOUT
IOUT
IOUT
IOUT
= 4 switch power and energy flow.
Sample of TYPE-13 Switch Specifications
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
13LEFT RIGHT
ORDER
13NODE1
CLOSED
SIG
1
9 - 30
Switches
1.
Switch from "LEFT " to "RIGHT " controlled by the TACS variable "ORDER ".
2.
Switch from "NODE1 " to ground controlled by the TACS variable "SIG " and
closed at < 0.0.
9 - 31
Section 10
Sources
10.1
General Description ............................................................................ 10-1
10.2
General Rules ...................................................................................... 10-2
10.3
Format And Rules For Conventional Sources

Of Type-11 Through Type-14 ............................................................. 10-3
10.3.1
Type-11 Step Function ................................................... 10-3
10.3.2
Type-12 Ramp Function ................................................ 10-5
10.3.3
Type-13 Ramp Function with Linear Decay .................. 10-6
10.3.4
Type-14 Sinusoidal Function ......................................... 10-8
10.4
Examples For Sources With Type-11 Through Type-14 .................... 10-11
10.5
Format And Rules For Type-15 Source .............................................. 10-12
10.6
Current Controlled dc Voltage Source (Type-16) ............................... 10-13
10.7
10.6.1
General Remarks ............................................................ 10-13
10.6.2
Format and Rules ........................................................... 10-16
10.6.3
Additional Remarks ....................................................... 10-19
10.6.4
Example ......................................................................... 10-19
Sources With User-defined Functions (Type-1 Through Type-10) .... 10-21

10.7.1
Format and Rules ........................................................... 10-21
10.8
Modulating Source (Type-17) ............................................................. 10-22
10.9
Ideal Transformer Source (Type-18) .................................................. 10-22
10.10
Slave Sources (Type-60) ..................................................................... 10-23

10.10.1
10.11
Format and Rules ........................................................... 10-24
Three-Phase Dynamic Synchronous Machine Source Component .... 10-24

10.11.1
Type-59 Dynamic Synchronous Machine

Source Component ........................................................ 10-24
10.11.2
10.12
Data Format and Rules ................................................... 10-27
U.M. Source Component ...................................................................... 10-49
10.12.1
Introduction .................................................................... 10-49
10.12.2
Note on Coil Parameters ................................................ 10-50
Section 10
10.12.3
Sources
Use of Network Representation of
Mechanical System ........................................................ 10-60
10.12.4
Steady-State Initialization .............................................. 10-62
10.12.5
Saturation Modelling ...................................................... 10-67
10.12.6
User's Guide General Remarks ................................. 10-69
10.12.7
U.M. Type-1 User's Guide .............................................. 10-71
10.12.8
10.12.9
10.12.10
10.12.11
10.12.12
10.12.13
Type-7 User's Guide ....................................................... 10-75
10.12.14
10.12.15
Type-9, 10, 11, 12 User's Guide ..................................... 10-76
10.12.16
Class 1 U.M. Data Cards ................................................. 10-76
10.12.17
Class 2 U.M. Data Cards (Machine-Table) ..................... 10-77
10.12.18
Class 3 U.M. Data Cards (Coil-Table) ............................ 10-82
10.12.19
Output Labeling ............................................................. 10-85
10.12.20
Multi-Machine Systems Sharing Common

Mechanical Network ...................................................... 10-86
10.12.21
Specific Instructions for Type-40 Induction Machine ... 10-88
10.12.22
Specific Instructions for Modelling of

Synchronous Machines of Type-59 ............................... 10-95
10.13
10.14
EMTP Load Flow ("FIX SOURCE" Usage) ........................................... 10-111
10.13.1
Background Information On EMTP Load Flow .............. 10-111
10.13.2
Format of Data Cards for EMTP Load Flow ................... 10-114
EMTP Load Flow ("FIX SOURCE1" Usage) ......................................... 10-117
10.14.1
Background Information ................................................ 10-117
10.14.2
Data Format and Rules ................................................... 10-118
Section 10
Sources
10.1
General Description
The EMTP supports a variety of source models, which include:

(A)
conventional voltage/current sources which are analytical functions of time

(Type11 through Type15). Frequently-used functions are built into the
program;
(B)
conventional voltage/current sources whose function f(t) is supplied by the

user point-by-point (Type1 through Type-10);
(C)
voltage sources that are controlled by their current output (Type16

source);
(D)
3phase dynamic synchronous machine sources (Type59 S.M. model);
(E)
Universal Machine sources for modelling many different types of electrical

machines (Type19 U.M. model);
(F)
special slave sources whose values are automatically set equal to an

appropriate TACS variable (Type60 source);
(G)
special sources for modelling ideal transformers or signal modulation (Type

17 and Type18 sources).
DC
The usage of these sources is described in Sections 10.2 through 10.12.

The EMTP has the capability to perform a steady-state solution subject to power constraints at one
or more buses ("FIX SOURCE" and "FIX SOURCE1" usage). These "EMTP Load Flow" features are
described in Sections 10.13 and 10.14. Additionally, the support program LDFLOW for more
conventional three-phase load flow capabilities (see Section 21).
Finally it should be noted that the sinusoidal (Type-14) sources are normally used to determine a
single set of initial conditions. But these sources can also be used in more than one set of initial
conditions with automatic frequency variation or "frequency scan". See Section 4.1.1 for more
details.
10 - 1
Sources
10.2
General Rules
Rule 1:
All sources are connected between node and local ground. Current sources between
two nodes can be represented very easily, as explained in Section 3.3. Voltage sources
between two nodes can also be represented as long as the internal impedance is finite,
as explained in Section 3.3. An additional option available to the user is the use of the
Ideal Transformer Source Type-18 (see Section 10.9).
Rule 2:
If sources have identical node names, then they will be added as explained in
Section 3.3. This implies series connection for voltage sources, and parallel connection for current sources. If a node has both voltage and current sources, then the current sources are ignored because they have no effect on the network (they would be
short-circuited directly through the voltage sources).
Rule 3:
All source functions must be specified in consistent units, e.g., in V and A, or in kV

and kA.
Rule 4:
The built-in source functions are zero between 0 < TSTART and for t > TSTOP (zero, not
disconnected). These two parameters make it possible to specify composite functions
which consist of segments of built-in functions, as explained in Section 10.3. The time
count for f(t) starts with TSTART.
Rule 5:
The source functions are read in or evaluated at discrete time steps t=t, 2t, ... only.
Linear interpolation between discrete points is assumed by the program.
Examples: If a 50 Hz ac source is defined as Fmax cost (solid line), then the
program would see it as a sawtooth function (dotted line) for t = 10 ms. If the source
is defined as a ramp function of Type-12 with linear decay and non-zero TSTART and
TSTOP (solid line), then the program would see it as the dotted line for t as indicated
in Figure 10.1 below.
function seen by program

defined function
f(t)
f(t)
{t}
Figure 10.1:
TSTART
Source Function Evaluation
10 - 2
TSTOP
Sources
Rule 6:
An AC steady-state solution is automatically computed if Tstart < 0 on one or more AC

source cards (Type-14). The frequency of the first ac source card with Tstart < 0 is
used as the steady-state frequency fAC. This AC source and all other AC sources with
Tstart < 0 and with a frequency which deviates by less than 0.001% from fAC will be
connected for the steady-state solution. All other sources are disconnected for the
steady-state solution.
If TMAX = 0 on the time card (see Section 4.2.1), then program execution stops after
the AC steady-state solution. If TMAX > 0, then the AC steady-state solution is used as
initial conditions for the subsequent transient simulation.
Rule 7:
AC
steady-state solutions with automatic frequency variation are invoked by a special

request "FREQUENCY SCAN" (see Section 4.5). In this case, all AC sources (Type-14)
are included in the steady-state solution with their frequency varied from fmin to fmax.
All sources other than Type = 14 are ignored.
Rule 8:
DC
steady-state conditions can be computed automatically if the DC sources are represented as AC sources f(t) = Fmax cost with a very low frequency.
A trick to place trapped charge on a disconnected transmission circuit (and associated,

isolated components), use a current source ("1" in columns 9 and 10), a low
sinusoidal frequency, and TSTART = 5432.0 in columns 61-70. With the normal EMTP
logic, this ensures that the source will not be present for the time-step loop, since a
zero current source (one never started) in effect does not exist. As for the steady-state
phasor solution, we now internally trick the EMTP into treating this source as a voltage
source (as though TSTART = -1.0 and columns 910 had been blank). The only extra
temporary complication is for degenerate problems: make sure at least one regular
phasor source is present (if not, add a dummy one with TSTART negative).
10 - 3
Sources
10.3
Format And Rules For Conventional Sources Of Type

11 Through Type14
10.3.1
Type-11 Step Function
I2
Node
Name
A6
Volt or Current
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Amplitude
TSTART
TSTOP
I2
E10.6
E10.6
E10.6
Rule 1:
Specify the source type 11 (variable TYPE in columns 1-2).
Rule 2:
f(t) = AMPLITUDE, which is an approximate step function in the case of zero initial
conditions f(0) = 0, or a DC source in the case of initial conditions f(0) = AMPLITUDE.
The step function is approximate in the sense that linear interpolation between discrete
points produces a finite rise time t (see Figure 10.2 below).
Rule 3:
Specify the node to which the source is connected by its alphanumeric name (variable
NODE NAME in columns 38). A source is always connected between node and
ground. For sources between nodes, see Section 3.3 and Section 10.9.
Rule 4:
Indicate a voltage source by specifying a non-negative integer (>0 or blank) in

columns 910. Indicate a current source by entering a negative integer (<0) in
columns 910 (current is measured into node).
Rule 5:
Specify AMPLITUDE in columns 1120.
Rule 6:
For a delayed start of a source function, use TSTART > 0. Then f(t)=0 for t < TSTART
(not disconnected) and f(t) as specified as soon as t=TSTART. Note that the time count
for the source function starts with TSTART if TSTART > 0. Negative TSTART is set to 0
by the program.
10 - 4
Sources
f(t)

Amplitude
t
t
Figure 10.2:
Step Function Representation
If initial conditions f(0) = AMPLITUDE, then this type is a DC source (see Figure 10.3
below)
f(t)

Amplitude
t
Figure 10.3: DC Source Representation
Rule 7:
or nullification of a source function after t>TSTOP use the appropriate TSTOP>0. Then
f(t) = 0 as soon as t>TSTOP (not disconnected). TSTOP=0 or blank will be interpreted
to mean TSTOP = infinity.
Rule 8:
Composite functions, consisting of pieces of available function types, can be produced

by using TSTART and TSTOP.
10 - 5
Sources
10.3.2
Type-12 Ramp Function
I2
Node
Name
A6
Volt or Current
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Amplitude
To
TSTART
TSTOP
I2
E10.6
E10.6
E10.6
E10.6
Rule 1:
Set Type=12 in columns 1-2 for a ramp function with linear rise from t = 0 to t = T0
and constant amplitude thereafter (see Figure 10.4 below, drawn for f(0) = 0).
Rule 2:
Rule 3:
Indicate a voltage source by entering a non-negative integer (>0 or blank) in

Rule 4:
Rule 5:
by the program.
f(t)

Amplitude
To
Figure 10.4:
Rule 6:
Ramp Function with Linear Rise
For nullification of a source function after t>TSTOP use the appropriate TSTOP>0.
Then f(t) = 0 as soon as t>TSTOP (not disconnected). TSTOP=0 or blank will be interpreted to mean TSTOP = infinity.
10 - 6
Sources
Rule 7:

10.3.3
Type-13 Ramp Function with Linear Decay
I2
Node
Name
Volt or Current
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Amplitude
A6
I2
E10.6
To
E10.6
A1
T1
E10.6
E10.6
TSTART
TSTOP
E10.6
E10.6
Rule 1:
Set Type=13 in columns 1-2 for a ramp function with linear decay (see Figure 10.5
below, drawn for f(0) = 0), or with linear rise if A1 > AMPLITUDE. The special case T0
= 0 is permitted.
Rule 2:
Rule 3:

Rule 4:
Rule 5:
by the program.
10 - 7
Sources
f(t)
f(t)

Amplitude

A
Amplitude
A1
t
To
T1
Special case with To=0
T1
Figure 10.5:
Complex Ramp Function
Rule 6:
Rule 7:

10.3.4
Type-14 Sinusoidal Function
I2
Node
Name
A6
Rule 1:
Volt or Current
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Amplitude
Frequency
I2
E10.6
E10.6
in degrees
E10.6
A1
TSTART
TSTOP
E10.6
E10.6
E10.6
Set Type=14 in columns 1-2 for a sinusoidal function (see Figure 10.6 below, drawn
for f(0) = 0).
10 - 8
Sources
f(t)

Figure 10.6:
Sinusoidal Source Function
Two options are available. If
Note:
A1 = 0:
f(t) = AMPLITUDE * cos [2f(t-TSTART)+o] with f =

FREQUENCY and o in degrees
A1 > 0:
f(t) = AMPLITUDE * cos[2f(t + TSTART +To)] with f =

FREQUENCY and To in seconds
For consistency, the user should specify all sinusoidal amplitudes as

peak rather than an RMS values. Unless this is done, the power flow
and loss figures of the complete steady-state phasor solution printout
(see Section 16.2) "KSSOUT", entered as "1" in column 32 of the integer
miscellaneous data card, will be only half of the correct values.
Rule 2:
Rule 3:

Rule 4:
Rule 5:
Specify frequency in Hz (variable FREQUENCY in columns 2130).
Rule 6:
for the source function starts with TSTART (see Figure 10.7 below) if TSTART > 0.
Negative TSTART is used by program (for Type-14) to indicate steady-state sources
with the time count starting at t=0.
10 - 9
Sources
f(t)
TSTART
Figure 10.7:
Delayed Cosine Source Function
Rule 7:
Rule 8:

Example 1:
f(t)
1.0

1
120
1
60
With initial condition f(0) = 1.0, the three source cards are:
1)
TYPE = 11
AMPLITUDE =
TSTART = 0
TSTOP = 1/120
TSTART = 1/120
TSTOP = BLANK
FREQUENCY = 60
A1 = 0
TSTART = 1/120
TSTOP = BLANK.
1.0
2)
TYPE = 1
AMPLITUDE =
0.5
3)
TYPE = 14
AMPLITUDE =
0.5
o = 0,
10 - 10
Sources
Example 2:
Source Type-13 with the appropriate parameters TSTART and TSTOP can be used to
simulate piecewise linear functions:
f(t)
40
20
t
5
15
30 35
37.5
45
65
95
Use 7 source cards (first two slopes are simulated with one card):
TYPE
Amplitude
T0
A1
T1
TSTART
TSTOP
1st card:
13
15
20
15
15
2nd card:
13
20
25
15
15
30
3rd card:
13
25
30
30
35
4th card:
13
30
40
2.5
35
37.5
5th card:
13
40
45
7.5
37.5
45
6th card:
13
45
50
20
45
65
7th card:
13
50
40
30
65
"blank
"
10 - 11
Sources
10.4
Examples For Sources With Type11 Through Type14
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Bus--><I<Amplitude<-FREQ---><-PHI -->---A1---->----------<---Tstart<----Tstop
11 BUS-3-1
-0.5
12 BUS-4
0.8
0.05
14 BUS-5
1.2
60.0
-0.06
-1.0
13 BUS-6
1.0
1.E-6
0.5
50.E-6
50.E-6
Explanations:
BUS3 is a node with current injection i(t) = -0.5
BUS-4 is a node with voltage excitation of the following form:
f(t)

0.8
0.05s
BUS5 is a node with voltage excitation, present in the steady state and thereafter,
e(t) = 1.2 cos (260t0.06)
BUS6 is a node with the voltage excitation of the following form:
1.0
0.5
50s
1s
10 - 12
Sources
10.5
Format And Rules For Type-15 Source
TYPE
Node
Name
VorC
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
AMPLITUDE
I2
A6
I2
E10.6
in 1--s
in 1---
TSTART
TSTOP
E10.6
E10.6
E10.6
E10.6
This source is used to represent the following surge function:

f(t) = AMPLITUDE (e t-e t)
for TSTART <t <TSTOP (zero otherwise).
The values of and in 1/s must be negative for standard impulse waveforms. Leave TSTOP
blank for TSTOP = infinity. TSTART < 0 has no meaning and is set to zero by the program.
f(t)
t
TSTART
TSTOP
Rule 1:
Specify source type (TYPE=15) in columns 12.
Rule 2:
NODE NAME) in columns 38. A source is always connected between node and
ground. For sources between nodes, see Sections 3.3 and Section 10.9.
Rule 3:
Indicate a voltage source by entering a non-negative integer (>0 or blank) in columns

9-10. Indicate a current source by entering a negative integer (<0) in columns 9-10
(current is measured into node).
Rule 4:
Specify AMPLITUDE in columns 11-20.
10 - 13
Sources
Rule 5:
Specify and in columns 2130 and 3140, respectively. These values must be in
units of 1/s and they are usually negative.
Rule 6:
Specify TSTART and TSTOP (in units of s) in columns 61-70 and 71-80, respectively.
10.6
Current Controlled DC Voltage Source (Type-16)
10.6.1
General Remarks
This source type is to be used for a DC voltage source which is controlled by its current output.
This provides a simplified model of an HVDC converter station (ripples in the DC voltages ignored).
Node Name Cathode Side
Node Name Anode Side
Figure 10.8:
DC Voltage Source Definition
Note that the DC voltage source is not defined by its positive and negative terminals but by cathode
and anode side. The positive direction of current flow is from anode to cathode. Current is not
permitted to flow in the negative direction (this is represented by a diode in the anode side).
The current regulator is assumed to be an amplifier with two input signals: one proportional to
current bias IBIAS and the other proportional to measured current i and with one output e which
determines the firing angle.
IBIAS
emax
+
G(s)
emin
10 - 14
Sources
The transfer function of the regulator is:
K ( 1 + sT2 )
G ( s ) = -----------------------------------------------( 1 + sT 2 ) ( 1 + sT 3 )
(10.1)
with limits placed on the output e in accordance with rectifier minimum firing angle and inverter
minimum extinction angle.
The dc voltage VDC is a function of einfinity, i.e.,
Vdc = k1 + k2 e
(10.2)
VDC
+150kV
Typical range
for rectifier
10
20V
Typical range
for inverter
-150kV
cos
-1
+1
As shown above (in this example, k1 = -150000 and k2 = 15000). The current regulator output e
minus a bias value (which is 10V in this example) is proportional to cos.
The inverter normally operates at minimum extinction angle at the limit emin and the rectifier
normally operates on a constant current control between the limits. The user defines steady-state
limits on VDC, which are converted by the program to limits on e with Equation (10.2). If the
converter operates at the maximum limit emax (or at the minimum limit emin) either in initial
steady-state or later during the transient simulation, it will back off the limit as soon as the
derivative de/dt becomes negative (or positive) in
10 - 15
Sources
de
d ( I BIAS i )
d e
( T1 + T 3 ) --------- = K ( I BIAS i ) + KT 2 ----------------------------- T1 T 3 ------------ e
2
dt
dt
dt
(10.3)
The value for d2e/dt2 is zero in Equation (10.3) as long as the inverter operates at the limit.
Steady-state dc initial conditions are automatically computed by the program with the value
VDC(0). If the DC initial conditions are to be zero, simply set VDC(0)=0. Since the steady-state
subroutine can only handle AC steady-state solutions, the DC voltage source is actually simulated as
VDC=VDC(0).cost with f = 0.001 Hz in the steady-state subroutine. Experience has shown that
this is sufficiently close to DC, and still makes reactances L and susceptances C large enough to
avoid problems in the AC steady-state solution. The normal steady-state operation of an HVDC
transmission link, measured somewhere at a common point (e.g., in the middle of the line) is
shown in Figure 10.9.
v vRECT -vINV
A
rectifier
inverter
rectifier
inverter
ISETTING
IORDER
A
Figure 10.9:
Steady-State Operation
For the converter operating between the limits on constant current control (which is normally the
rectifier), IBIAS is automatically computed to produce the characteristic A-A':
e ( 0 )
I BIAS = i ( 0 ) + -------------K
if emin < e (0) < emax
(10.4)
with i(0),e(0) being the dc initial conditions. For the converter operating at maximum or
minimum voltage (which is normally the inverter) the current setting ISETTING must be given,
10 - 16
Sources
which defines the point where the converter backs off the limit to constant current control. IBIAS
is automatically computed in this case from
e ( 0 )
I BIAS = I SETTING + -------------K
(10.5)
if e(0) < emax or emin
ISETTING is typically 15% lower than the current IORDER at the operating point for inverters (or
15% higher for rectifiers).
10.6.2
Format and Rules
TYPE
Node
Name
MODE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Regulator
Gain K
I2
A6
I2
E10.6
Vdc(0) or
ISETTING
E10.6
T1
T2
E10.6
E10.6
T3
E10.6
Rule 1:
Specify source type (TYPE=16) in columns 1-2.
Rule 2:
Specify NODE NAME on cathode side in columns 38.
Rule 3:
Specify MODE in columns 910.

MODE
Idc
E10.6
=1
if source operates initially on constant current control,

VDC min < VDC(0) < VDC max.
=2
if source operates initially at maximum voltage,

VDC(0) = VDC max.
=3
if source operates initially at minimum voltage,

VDC(0) = VDC min.
<1
Program stops with error message:
or >3
"ERROR IN IDENTIFICATION OF SOURCE".
Rule 4:
Specify gain K of Equation (10.1) in columns 11-20.
Rule 5:
Specify VDC(0) or ISETTING in columns 21-30.
10 - 17
Sources
If MODE=1, specify the initial steady-state DC voltage.
If MODE=2 or 3, specify the current ISETTING.
Rule 6:
Specify the time constants T1, T2 and T3 in columns 3140, 4150 and 5160, respectively. Units are s (seconds).
Rule 7:
Specify the initial steady-state DC current in columns 61-70. The initial current IDC
must be given by the user because the voltage source has an internal resistance in
series in the transients program (its value is printed in the record of source data as
"EQUIV. RESISTANCE"). This equivalent resistance results from the transformation of
the differential equations to difference equations with the trapezoidal rule of integration. The program then sets eSOURCE = e(0) + REQUIV.IDC. If the initial current is
wrong, then the initial voltage at the terminals will not be exactly e(0); this error may
or may not be negligible depending on whether |REQUIV.IDC| << |e(0)| or not. Note that
IDC is only used for the compensation just mentioned and for nothing else. If IDC < 0,
then it is set to zero internally in the program.
Node
Name
k1
k2
Vdc min
A6
E10.6
E10.6
E10.6
Vdc max
E10.6
IOUT
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890123456789 0
E10.6
I1
Rule 8:
Specify NODE NAME on the anode side in columns 3-8.
Rule 9:
Specify the coefficients K1 and K2 of Equation (10.2) in columns 11-20 and 21-30,
respectively.
Rule 10:
Specify VDC min and VDC max in columns 3140 and 41-50, respectively.
Rule 11:
Specify isolation resistance R in columns 51-60 (units are in ohms). The isolation
resistance R is used to specify a small series isolation resistor which is defined internally by the program. In the following diagram showing internal EMTP representation
of the Type16 source component, the resistor in question is that between the usersupplied anode node name "BUS1" and the internally-defined node "Type-16":
10 - 18
Sources
BUS1
Anode
TYP-16
TRASH
Requiv
BUS2
Cathode
Switch simulating diode

a) opens for current flow
right to left
b) closes for potential
drop + to
Dynamic voltage source
Figure 10.10:
Internal Representation of Type-16 Source
If field R is left blank (or set to 0.0), the EMTP will automatically supply a default
value equal to the square root of "EPSILN", where "EPSILN" is a floating-point
miscellaneous data parameter (see Section 4.2.1).
Rule 12:
Specify desired output by entering an appropriate value in the data field IOUT
(column 80).
IOUT = 1 to get branch current,
IOUT = 2 to get branch voltage,
IOUT = 3 to get both.
Request for current output will be internally converted to current-output request for
resistor R and it will show up this way in the printed column readings preceding the
transient printout.
10 - 19
Sources
10.6.3
Additional Remarks
Initial Current:
The value IBIAS in Equation (10.2) is automatically computed after return from the steady-state
subroutine. With eSOURCE as defined in Rule 3, the steady-state current i(0) is automatically
obtained. Note that this i(0) may differ from IDC if the latter was not given correctly, except if the
user specifies IDC = 0. In this case, i(0) will also be zero because the switch which simulates the
diode effect would be kept open for the steady-state solution. IBIAS is then computed from:
e ( 0 ) = K ( I BIAS i ( 0 ) )
if MODE = 1
(10.6)
or
e ( 0 ) = K ( I BIAS I SETTING )
if MODE = 2 or 3
(10.7)
Equation (10.7) is used if the steady-state voltage was at its minimum or maximum limit; in this
case the current setting ISETTING is assumed to be the value where the control would begin to back
off the limit again.
Spurious Voltage Oscillations:

Spurious voltage oscillations may occur at the terminal after current extinction unless damping
circuits are also modelled or CDA is enabled. If damping circuits do not have physically
legitimate values then CDA should be used instead. A damping circuit could consist of an RC
branch between anode and cathode (R=900 ohms, C = 0.15F).
Internally Generated Node Name:

Use of the internally generated node name "Type-16" should always be avoided by the user.
10.6.4
Example
This example describes a circuit arrangement for testing of a DC circuit breaker. One bridge of the
rectifier was connected to both poles of the DC line at the sending end, and one bridge each of the
inverter was connected to each pole at the receiving end of the DC line.
10 - 20
Sources
VR
dc line
VIP
VIN
circuit breaker
Rectifier
Inverter
The rectifier bridge operated initially at maximum voltage. It was simulated as one current
controlled voltage source with the following data:
cathode name = VR, anode name = ground,
VDC = -148000 + 14800 e for Eq. (10.2)
70 kV
VDC min = -70 kV, VDC max = 70 kV,
10
K = 1.39, ISETTING = 1000 A,
20
14.73
T1 = 4.4 s, T2 = 40 ms, T3 = 10.3 ms,
-70 kV
IDC = 900 A.
The two inverter bridges operated initially on constant current control. They were simulated as
two current controlled voltage sources with identical data (except for anode name):
cathode name = ground,
anode name = VIP on one source and VIN on
the other source,
0
-10 kV
VDC = -148000 + 14800 e for Eq. (10.2),

VDC min = -120 kV, VDC max = - 10 kV,
K = 1.39, VDC(0) = -60.42 kV,)
T1 = 4.6 s, T2 = 30 ms, T3 = 12.5 ms,
IDC = 450 A.
10 - 21
-120 kV
1.89
10
20
Sources
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
16 VR
2
1.39
1000.0
4.4
0.04
0.0103
900.0
16
-148000.0 148000.0 -70000.0
70000.0
16 VIP
1
1.39 -60420.0
4.6
0.03
0.0125
450.0
16
-148000.0 148000.0 -120000.0 -10000.0
16 VIN
1
1.39 -60420.0
4.6
0.03
0.0125
450.0
16
-148000.0 148000.0 -120000.0 -10000.0
10.7
Sources With User-defined Functions (Type-1 Through

Type-10)
Up to 10 user-defined sources can be specified point-by-point as part of the EMTP data file as
described in Section 13.2, using TACS as described in Section 13.3, by creating a user-defined
subroutine, as explained in Section 13.4.
10.7.1
Format and Rules
TYPE
Node
Name
Volt or Current
1
2
3
4
5
6
7
8
12 345678 90 1234567890123456789012345678901234567890123456789012345678901234567890
I2
A6
I2
Rule 1:
Specify source number in columns 1-2 (data field TYPE).
Rule 2:
Specify the node to which the source is connected by its alphanumeric name (data field
NODE NAME) in columns 3-8. A source is always connected between node and
Rule 3:

columns 9-10. Indicate a current source by entering a negative integer (<0) in columns
9-10 (current is measured into node).
For more details regarding the usage of user-defined sources, please refer to Section 13, namely:
1.
point-by-point as part of the EMTP data file (Section 13.2)
2.
using TACS (Section 13.30)
3.
by creating a user-defined subroutine (Section 13.4)

10 - 22
Sources
10.8
Modulating Source (Type-17)
The Type17 source is used to connect a TACS-generated modulating signal (a multiplying factor)
to the source immediately following. This is only for the time-step loop (Type-17 sources are
ignored in the phasor steady-state solution).
TYPE
1
2
3
4
5
6
7
8
12 345678 901234567890123456789012345678901234567890123456789012345678901234567890
Node
Name
I2
A6
Rule 1:
Source type is 17 (data field TYPE, columns 1-2).
Rule 2:
NODE NAME) in columns 3-8.
10.9
Ideal Transformer Source (Type-18)
The Type-18 source is used to represent a combined ideal transformer and voltage source, as
shown in Figure 10.11. The source is connected between node and ground, as usual. The
ungrounded source effect is provided by the interposing ideal transformer. To represent an
ungrounded voltage source only, set the turns ratio "n" equal to zero. For an ideal transformer
only, set the source amplitude equal to zero or a very small number, or set the source TSTART very
high. This results in shorting the source node to ground. To avoid the shorting effect, use a
current source. The source itself is defined on a preceding, conventional source card, which must
be immediately followed by the Type-18 card.
Figure 10.11:
1:n
Vs(t)
+
is
An Ideal Transformer with an Ungrounded

Voltage Source
10 - 23
Sources
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 123456 789012 345678 901234567890123456789012345678901234567890
Node
"l"
Ratio
"n"
Node
"k"
I2
A6
E10.0
A6
Node
"m"
Node
"x"
A6
A6
Rule 1:
Source type is 18 (data field TYPE, columns 1-2).
Rule 2:
Specify the terminal nodes l, k and m, by names, in columns 38, 2126 and 2732,
respectively (see Figure 10.11 above). Node j is the node name used on the immediately preceding (associated) voltage source card.
Rule 3:
Specify the transformer turns ratio n in columns 11-20.
Rule 4:
Specify, by name, an extra, fictitious node "x" which has voltage equal to the current of
the source and/or transformer. Node voltage output can be used to display this current
(printing or plotting). The A6 name is to be specified in columns 33-38.
10.10 Slave Sources (Type-60)

These are special slave sources reserved for connection to (control by) TACS variables. TACS
dynamics are fully described in Section 14. Source values are automatically set equal to the TACS
variables having the same 6-character names. Type60 sources are limited in number only by
List 4 dimensioning.
Unlike the TACS control of Type 110 EMTP sources (see Section 13.5), here the interface is
established automatically; there is no need for any specification like the "TACS EMTP SOURCES"
card of Section 4.1.5. On the other hand, the Type 110 sources are somewhat more flexible (note
in particular that the controlling TACS variables could have arbitrary names, for Type 110
sources).
10 - 24
Sources
10.10.1
Format and Rules
Node
Name
I2
A6
Volt or Cur
TYPE
1
2
3
4
5
6
7
8
12 345678 90 12345678901234567890123456789012345678901234567890 1234567890 1234567890
TSTART
TSTOP
I2
E10.6
E10.6
Rule 1:
Source type is 60 (data field TYPE, columns 12).
Rule 2:
Specify the 6-character TACS variable name to which the source output is equal in
columns 38 (data field NODE NAME).
Rule 3:

columns 910. Indicate a current by entering a negative integer (< 0) in columns 9
10 (current is measured into node).
Rule 4:
Specify TSTART and TSTOP in columns 6170 and 7180, respectively.
10.11 Three-Phase Dynamic Synchronous Machine Source

Component
Before describing what is involved in this section, let's first briefly mention what is not involved.
If the dynamics of Park's equations are not wanted, power system generators are typically
represented by sinusoidal voltage sources (Type-14 source of Section 10.3). If the rotating
machine is not a conventional, balanced, three-phase synchronous generator, the user is referred to
the universal machine (U.M.) modelling of Section 10.12.
10.11.1
Type-59 Dynamic Synchronous Machine Source Component
HP
IP
LPA
LPB
LPC
GEN
EXC
Figure 10.12:
Model of Turbine-Generator Set
10 - 25
12
E2
ixq
ira
Ji-1
Ki-1
Figure 10.13:
i+1
D1
Ji
VERTICAL
xis
q-a
VERTICAL
i-1
E1
d-axis
VERTICAL
Sources
Ji+1
Ki
Phasor Diagram (a), Turbine Part Around Mass #1 (b)
General Description of Component:

The dynamic S.M. component is internally balanced (with respect to the armature phases), and is
inherently three-phase (for more information on the mathematical details of the Type-59
synchronous Machine please see [xx], [yy], or the EMTP Theory Book).
For output purposes, most machine parameters of interest are available: velocities and angles of
the rotor masses, inter-mass shaft torques, winding currents, and certain d-q-0 variables. Such
output quantities are automatically appended to the end of the regular EMTP output vector upon the
request of the user, and hence are available for printing and plotting just like any other EMTP
variable.
Each rotor mass is allowed to have a constant mechanical power applied to it (in addition to the
torque of mechanical viscous damping and the spring connection to adjacent masses). The user
specifies proportionality factors for each mass, with the actual constant power then determined
internally by the EMTP at the time of the sinusoidal steady-state initialization (so as to produce
equilibrium). This is the basic scheme regarding mechanical input torque, in the absence of special
connections. But the user is allowed to represent prime mover (e.g., governor, boiler, etc.)
dynamics if he so chooses, by using TACS (see Section 14) to control the just-described total
mechanical power. In this case, the required TACS output variable is a normalized multiplicative
constant (equal to unity if it is to produce no effect) for scaling the otherwise-constant mechanical
power. Should the user desire a constant-torque representation of the prime mover, this can be
readily provided by the TACS connection (TACS would divide the steady-state speed by the
instantaneous speed to produce the required control signal).
Even without any detailed explanation of where they come from, it might be desirable to document
the set of equations which is used to model the S.M. We have:
10 - 26
Sources
m
d-axis
ia
a
if
f
vf
n
q
kq
axis
g
b
ib
kd
n
c
Figure 10.14:
ic
Schematic Diagram of a Synchronous Machine
Counter-Clockwise Steady-State Rotation:

d m
2
----------- = -------- 2 F REQ
dt
NP
(10.8)
Common neutral connection point for the three Wye-connected

armature windings. This point may be grounded through a
series R-L branch.
Mechanical angle of the rotor (the angle between the direct axis
of the rotor and the axis of armature phase "a"). Sketch is for a
2pole machine, for simplicity.
ia, ib, ic
Armature currents. Note that these are positive when flowing

out of the machine (generator convention).
if
Field current (current in winding number 1 of the direct axis of

the rotor). This is positive into the coil, in accord with the
direction of steady-state flow.
f, kd
Rotor windings number 1 and 2 of the direct axis (d-axis),

respectively.
g, kq
Rotor windings number 1 and 2 of the quadrature axis (q-axis),

respectively.
10 - 27
Sources
Notes about mechanical equation (see Figure 10.13b):
10.11.2
1.
Coefficient names are as defined in the section about Class 4 S.M. data cards.
2.
This equation is Newton's law in rotational form for mass number "k", assuming
the most general case.
3.
If case number "k" is not the generator rotor, omit the electromechanical torque
Tem.
4.
If mass number "k" is at one end of the shaft system, either mass number k-1 or
mass number k+1 (or both, in the case of a single mass system) will not exist. In
this case, the associated terms of mutual coupling ("DSM" and "HSP") are defined
to be zero.
5.
The synchronous mechanical frequency "f" enters only through the speeddeviation self-damping term.
6.
As used here, k is the absolute angle of mass number "k" of the shaft system.
7.
Tkmech is the externally-applied mechanical torque on the mass number "k", in

the direction of +.
The specification of a dynamic synchronous machine source component within an EMTP data case
requires a number of data cards. These shall now be described in order of data input, in groups,
according to the following classification:
Class 1 S.M. Data Cards:

First come three cards which specify the component type code, the voltage magnitude and angle in
the steady-state, and names for the network nodes to which the armature windings are to be
connected. The first of these three cards is for phase "a", as per the following format:
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
BUS
VOLT
FREQ
ANGLE
I2
A6
E10.6
E10.6
E10.6
10 - 28
Sources
TYPE
(1-2)
BUS
(3-8)
VOLT
(11-20)
FREQ
(21-30)
ANGLE
(31-40)
Type code for the dynamic synchronous machine

component. Enter TYPE=59 in columns 1-2.
Six-character EMTP network node name, entered in
columns 38. The armature winding of phase "a" is to be
connected to this node of the electric network.
The steady-state voltage magnitude at the terminals of the
machine. This is peak voltage (1.414 times RMS voltage),
line to neutral (.577 times line-to-line). Use units of volts.
The electrical frequency of the generator in Hz, for steadystate operation. For North American systems, this is 60.
The steady-state voltage phasor angle at the terminals of
the machine, for phase "a". Units are degrees. This
machine component, all other ones, and all sinusoidal
voltage or current sources of Type-14 are assumed to all
have the same reference for phase angles.
Conventional positive sequence is assumed, as this term is
used in North America. That is, phase "b" voltage lags
phase "a" voltage by 120 degrees (and "c" leads "a" by 120
degrees. See sketch below.
Vc
Va
v(t) = Re { Vej t }
Vb
The second and third cards in Class 1 use the same format described above:
TYPE
Blank.
(1-2)
BUS
(3-8)
Six-character EMTP network node name. The appropriate

armature winding (phase b or c) is to be connected to this
node.
10 - 29
Sources
(10-11)
The steady-state voltage magnitude at the above-named

machine terminal (network node). Use units of volts.
FREQ
Blank.
VOLT
(21-30)
The steady-state voltage phase angle at the above-named
machine terminal. Units are degrees.
ANGLE
(31-40)
If the values of VOLT and ANGLE on cards number two and three are not specified (left blank), the
program assumes the presence of a 3-phase balanced source at the machine terminals. Otherwise,
the presence of an unbalanced 3phase source is assumed. This allows the user the specification of
unbalanced steady-state conditions.

Following Class 1 data cards come optional special-request cards (if any) which are associated
with the machine. There are three such possible cards, with format and meaning described below.
Ordering of such cards (if two or more) within the Class 2 grouping is arbitrary. Remember that
only non-blank (non-zero) fields serve to redefine the built-in default tables.
(A)
"TOLERANCES" special request card
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
TOLERANCES EPSUBA
A10
E10.0
EPSUBA
(11-20)
EPOMEG
EPOGEL
NIOMAX
E10.0
E10.0
I 10
Number specifying the ratio between the built-in damping

resistors and the resistive models of the inductive elements
paralleled by these resistors, i.e., the following
relationship holds true for each inductive element:
Rp
EPSUBA = -----------------------(2 L ) t
(10.9)
If columns 11-20 are left blank, a default value of

EPSUBA= 100.0 would be assigned. For the rare cares of
numerical instability, the recommended value is 20.0 to
50.0.
10 - 30
Sources
EPOMEG
(21-30)
Tolerance associated with the iterative calculation of S.M.

rotor speed at each time-step of the simulation.
The calculation is assumed to have failed when the
relative speed correction is greater than this tolerance
after NIOMAX iterations. The program execution is then
terminated with an appropriate error message.
EPOGEL
(31-40)
NIOMAX
(51-60)
Tolerance associated with the iterative calculation of S.M.

rotor speed at each time-step. The calculation is assumed
to have converged when the relative speed correction is
less than this tolerance. If the relative speed correction is
larger than this tolerance but smaller than EPOMEG, the
solution is assumed to have converged marginally and an
appropriate warning is printed.
Maximum number of iterations which are allowed for the
calculation of the S.M. rotor speed, at any time-step
(default NIOMAX = 10).
It should be stressed that these are scalar variables only; they apply to the entire date case (all
machines) rather than to the specific machine with which they were defined.
(B)
"PARAMETER FITTING" special-request card
The PARAMETER FITTING card is used to distinguish between different types of data. The
presence of this card signals manufacturer's data, while its absence indicates the use of per unit
inductances and resistances. Depending upon the value of FM, the EMTP logic will select either
the simplified procedure or attempt to precisely calculate the machine parameters.
1
2
3
4
5
6
7
8
123456789012345678901234 56789012 345678901234567890123456789012345678901234567890
PARAMETER FITTING
A24
FM
E8.0
Documentation of the mathematics used for the exact calculation of machine parameters is
contained in Reference 8.
If the original machine parameters are unrealistic, the exact calculation may not be possible, and a
warning message to this effect will be printed, (see message number 39 of Section 17.1). Nothing
is lost in this case, however, since the EMTP will recover, and will simply use the results of
simplified data conversion procedure.
If standard manufacturer data is being used, then there are several different cases of interest:
10 - 31
Sources
Case 0:
To disable the exact calculation, set FM = 2.0.
Case 1:
If the user has a valid full set of S.M. data parameters (i.e., full transient and subtransient data which is not somehow degenerate), then usage of parameter optimization is
optional; if used, "FM" should always be given a value of unity (FM = 1.0).
Case 2:
If the user has data in which X'q = Xq and T'qo T"qo such parameters can be shown to
be inconsistent. There are two options available to the EMTP user who enters data that
satisfies such conditions.
Case 3:
(A)
If use is made of parameter "FM" somewhere in the range .95<FM<1.0, then the
EMTP will use all q-axis coils. Internally, however, the constraint Xq' = FM.Xq
is observed (despite the user-provided equality of these two parameters).
(B)
If use is made of parameter "FM" equal to unity (FM=1.0), then the q-axis
damper winding is dropped from the model; this leaves just one q-axis coil.
It is not uncommon for the user to have no data values for X'q and T'qo. In this case, he
can still run the simulation if he does three things to the data:
1.
2.
3.
Use FM = 1.0;
Specify X'q equal to the known Xq value;
Set T'qo = 0.0
In this case, the EMTP will drop the q-axis damper winding from the model, leaving
just one q-axis coil.
Case 4:
Just like Case 2, only with "q" (referring to the quadrature axis) replaced by "d" (referring to the direct axis).
Case 5:
Just like Case 3, only with "q" replaced by "d".
Case 6:
If the user wants to model a machine without any dampers at all on the q-axis, he
should provide a common value for Xq, X'q, and X"q.
(C)
"DELTA CONNECTION" special request card
If the armature windings of the dynamic S.M. are delta-connected (rather than Wye-connected),
then this status must be communicated to the EMTP by a special-request card which bears the text
"DELTA CONNECTION" in columns 116:
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
DELTA CONNECTION
10 - 32
Sources
In the absence of such a card, the machine is assumed to be Wye-connected (which is by far the
most common situation for large power system generators). Except for the possible presence of
this one special-request card, no other portion of the S.M. data specification explicitly makes
reference to how the armature windings are connected.
For a machine with delta-connected armature windings, the labelling of variables is not altered
from that used for a Wye-connection; so the user must be very careful in interpreting the printout.
There are four situations, depending upon whether one considers initial conditions or time-step
loop printout, and whether coil variables or Park's variables are considered. Of these four, two are
incorrectly labelled.
1.
For the time-step loop output, "ID", "IQ", and "IO" are indeed armature-winding
variables (albeit in Park's coordinates). But "IA", "IB", and "IC" are erroneously
labeled (see sketch), for they are armature coil variable:. e.g., what is labeled
"IA" in the S.M. initial condition printout is actually iab, the current from "a" to
"b" in the winding.
IA
IB
IC
IA
IB
IC
c
Figure 10.15: Erroneous InitialCondition Labeling, Delta
Connection
2.
c
Figure 10.16:
Conventional Wye
Labeling
For the steady-state initial-conditions output, the situation is identical to that in

Point 1).
Restriction: The option of "DELTA CONNECTION", when used for multiple machines, applies
to all machines connected to that bus. Consequently it is not possible to have a combination of
Wye-connected and Delta-connected machines at the same bus. If such a rare configuration
should ever arise, the user is required to isolate the machines by a small reactance or resistance.
10 - 33
Sources

Next come either four or five cards, which specify the electrical parameters of the synchronous
machine. The first of this group has the following format:
KEXC
KMAC
NUMAS
1
2
3
4
5
6
7
8
12 34 56 7890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
NP
SMOUTP
SMOUTQ
RMVA
RKV
AGLINE
S1
S2
I2 I2 I2
I4
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
NUMAS
(1-2)
KMAC
(3-4)
KEXC
(5-6)
NP
(7-10)
SMOUTP
(11-20)
SMOUTQ
(21-30)
RMVA
(31-40)
The number of connected masses on the shaft system of

this generator.
The mass number which corresponds to the generator (or
motor) rotor, within the interconnected mass-spring shaft
system. Masses are to be numbered by the user for
identification, beginning with number one on either end,
and continuing sequentially (2, 3, ...) to the other end of
the shaft. See Figure 10.12.
The mass number which corresponds to the exciter on the
shaft system. If no exciter exists, leave this field blank.
The number of poles (not pole pairs) which characterize
the machine rotor. The electrical frequency of the
machine is equal to the mechanical frequency times NP/2.
Proportionality factor which is used only to split the real
power among the generators constituting a multiple
machine during the machine initialization. If a single
machine, enter 1.0. If a dual machine, suppose that the
user specifies value PA for this half of the dual, and value
PB for the other half; then the fraction PA/(PA + PB) of the
total steady-state real power output will be assigned to
this half of the dual during machine initialization.
The same as "SMOUTP", only for reactive rather than real
power.
The total 3-phase volt-ampere rating of the machine, in
units of MVA (million volt-amperes).
10 - 34
Sources
RKV
(41-50)
The rated line-to-line voltage of the machine, in units of

RMS KV. Taken as a pair, "RKV" and "RMVA" define the
base values upon which per unit machine parameters are
assumed to apply, it may be noted. Should the machine in
question have a delta-connected armature, specify RKV as
3 times the rated winding voltage in KV RMS.
AGLINE
(51-60)
Value of the field current in amperes which will produce

rated armature voltage (1.0 per unit) on the air gap line.
This is an indirect specification of the mutual inductance
between the field and the armature of the machine. See
Figure 10.17. If the S.M. is saturable, append an extra
minus sign (making "AGLINE" negative) as a flag.
air gap line

(characteristic if the
machine iron is not
subject to saturation)
|vt| p.u.
AD2
1.0
AD1
S1
S2
curve of terminal
voltage phasor magnitude
as a function of field
current, at rated speed
and no load.
Field current, if
AGLINE
Figure 10.17:
S1
(61-70)
Definition of Saturation

AD1 rated armature voltage (AD1 per unit) on the no-load
saturation curve. See Figure 10.17. This field, as well as
"S2", can be left blank if the S.M. is not saturable.
10 - 35
Sources
S2
(71-80)
Like "S1", only for AD2 of rated voltage (AD2 per unit).
The parameters "AGLINE", "S1" and "S2" just inputted
actually apply to the d-axis of the machine.
A second card follows call it Card 1A (as an extension of Card 1) to provide for the
following parameters:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 12345678901234567890123456789012345678901234567890
AD1
AD2
E10.6
E10.6
If there is to be no saturation modelling at all, this card can be left blank.

AD1
(11-20)
AD2
(21-30)
Per unit voltage at which current S1 was measured.

Should this field be left blank, the default value of 1.0
will be automatically assigned.
Like AD1, except for S2. Should this field be left blank,
The default value of 1.2 will be automatically assigned.
The remaining cards of Class 3 S.M. data depend upon whether standard manufacturer-supplied
data are being used. If so ("PARAMETER FITTING" card was used), two additional cards having the
following format complete the Class 2 data:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Ra
Xl
Xd
Xq
Xd
Xq
X"d
X"q
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Tdo
Tqo
T"do
T"qo
Xo
Rn
Xn
Xc
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
Ra
Armature resistance, in per unit. This must be nonnegative.
10 - 36
Sources
Xl
Armature leakage reactance, in per unit. If unknown, use

0.95 times the smallest of X"d, X"q, and Xo.
Xd
Direct-axis (d-axis) synchronous reactance, in per unit.
Xq
Quadrature-axis (q-axis) synchronous reactance, in per

unit.
X'd
Direct-axis (d-axis) transient reactance, in per unit.
X'q
Quadrature-axis (q-axis) transient reactance, in per unit.
X"d
Direct-axis (d-axis) subtransient reactance, in per unit.
X"q
Quadrature-axis (q-axis) subtransient reactance, in per

unit.
T'do
Direct-axis (d-axis) open-circuit transient time constant,

in seconds.
T'qo
Quadrature-axis (q-axis) open-circuit transient time

constant, in seconds.
T"do
Direct-axis (d-axis) open-circuit subtransient time

T"qo
Quadrature-axis (q-axis) open-circuit subtransient time

Xo
Zero-sequence reactance, in per unit. If this is unknown,

it is suggested that the value of Xl/0.95 be used. See note
below.
Rn
The neutral grounding resistance, in ohms. See note

below.
Xn
The neutral grounding reactance, in ohms. See note

below.
Xc
Canay reactance, in per unit. If unknown, leave blank.

The program will, in such case, automatically assign the
default value of Xc, i.e., Xc = Xl.
10 - 37
Sources
ia
ib
ic
Rn
Xn
Figure 10.18:
Note:
Standard Armature Connection
The machine is assumed to have a Wye-connected armature, with the neutral

connected to ground through the impedance Rn + jXn. If the machine is in fact
ungrounded (but still Wye-connected), simply use a large grounding impedance. Use
100 per unit or more, either Rn or Xn.
On the other hand, should the user have chosen to describe the machine by means of per unit
inductance and resistance matrices (no "PARAMETER FITTING" card used), then the Class 2 S.M.
data is completed with three cards of the following format:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Xa f
Xf
E10.6
E10.6
Xf k d
E10.6
Xd
Xa k d
E10.6
E10.6
Xk d
E10.6
Xl
E10.6
Numbering of windings on the direct axis (d-axis):

#1 field winding of machine, #2 d-axis damper winding
Xf
The self reactance of the field winding (circuit number 1

on the direct axis (d-axis) of the rotor, in per unit.
10 - 38
Sources
Note:
Xaf
The mutual reactance coefficient between the armature

and the field winding (circuit number 1 on the direct axis
(d-axis) of the rotor), in per unit.
Xfkd
The mutual reactance coefficient between the field

winding (circuit number 1 on the direct axis (d-axis) of
the rotor) and the direct axis damper winding (circuit
number 2 on the direct axis (d-axis) of the rotor), in per
unit.
Xd
The self reactance coefficient for the direct axis (d-axis)

of the armature, in per unit.
Xakd

and the direct axis damper winding (circuit number 2 on
the direct axis (daxis) of the rotor), in per unit.
Xkd
The self reactance of the direct axis damper winding

(circuit number 2 on the direct axis (d-axis of the rotor),
in per unit.
Direct-axis parameters, Xaf, Xfkd, and Xakd are generally assumed to all be equal. But
this is not a constraint of the EMTP code. Unequal values can be used, if available.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
Xg
Xa g
Xg k q
Xq
E10.6
E10.6
E10.6
E10.6
Xa k q
E10.6
Xk q
E10.6
Numbering of windings on the quadrature axis (q-axis):

#1 q-axis damper winding, #2 eddy current winding
Xg
The self reactance of the q-axis damper winding (circuit

number 1 on the quadrature axis (q-axis) of the rotor), in
per unit.
Xag

and the qaxis damper winding (circuit number 1 on the
quadrature axis (qaxis) of the rotor), in per unit.
Xgkq
The mutual reactance coefficient between the two

circuits on the quadrature axis (q-axis) of the rotor, in per
unit.
10 - 39
Sources
Note:
Xq
The self reactance coefficient for the quadrature axis (qaxis) of the armature, in per unit.
Xakq

and the eddy-current winding (circuit number 2 on the
quadrature axis (q-axis) of the rotor), in per unit.
Xkq
The self reactance of the eddy-circuit winding (circuit

per unit.
Quadrature-axis parameters Xag, Xgkq, and Xakq are generally assumed to all be equal.
But this is not a constraint of the EMTP code. Unequal values can be used, if available.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Xo
Ra
Rf
Rk d
Rg
E10.6
E10.6
E10.6
E10.6
E10.6
Rk q
E10.6
Rn
Xn
E10.6
E10.6
Xo

it is suggested that the value of Xl/0.95 be used. See note
about Wye-connection few pages back.
Ra
Armature resistance, in per unit.
Rf
Resistance of the field winding (winding number 1 on

the direct axis), in per unit.
Rkd
Resistance of the damper winding on the direct-axis (daxis) of the rotor, in per unit. This is the resistance of
winding number 2 on the d-axis of the rotor.
Rg
Resistance of the q-axis damper winding (winding

per unit.
Rkq
Resistance of the eddy-current winding (winding number

2 on the quadrature axis (q-axis) of the rotor), in per unit.
Rn
The neutral grounding resistance, in ohms. See note

about Wye-connected armature.
Xn
The neutral grounding reactance, in ohms. See note

about Wye-connected armature.
Xl
Leakage reactance, in per unit.
10 - 40
Sources

Next in order of data input come the mass cards which contain mechanical parameters for the shaft
system. There is to be one such card for each mass, entered according to the format immediately
below. In number, there are "NUMAS" such cards (see first card of Class 3 data, columns 1-2);
ordering is actually immaterial (the mass cards can be shuffled), though it is generally clearer to
stack such cards in naturally-increasing order of the mass number "ML".
1
2
3
4
5
6
7
8
12 34567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
ML
EXTRS
HICO
DSR
DSM
HSP
DSD
I2
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
ML
(1-2)
EXTRS
(11-20)
The mass number. Recall that masses should be

numbered beginning with number one on either end of
the shaft system, and continuing sequentially (2, 3, ...) to
the other end. See Figure 10.12.
Entered in columns 1120 is the fraction of the total
external mechanical torque (power) which is associated
with this mass:
Generator: For a generator, enter the fraction of the
total external mechanical torque of this shaft system
which is applied to this particular mass. If a pump, enter
a negative value.
Motor: For a motor, enter the fraction of the total
external mechanical torque of the shaft system which is
developed by this particular mass. If a pump, enter a
positive value.
HICO
(21-30)
DSR
(31-40)
The moment of inertia (WR2) of mass number "ML".

6
This is to be in units of ( 10 lb ft ) ( rad s ) .

The speed-deviation self-damping coefficient for mass
number "ML". By definition,
T i = DSR ( i s )
10 - 41
(10.10)
Sources
where Ti is the particular damping torque under

consideration for mass number i = ML, i is the angular
velocity of mass i, and s is synchronous mechanical
velocity for this shaft system. This is to be in units of
( lb ft ) ( rad s ) .
DSM
(41-50)
The mutual-damping coefficient. This pertains to the

present mass (assumed to be number i = ML) and the mass
with the next higher number (i+1). The damping in
question is a function of the velocity difference between
the two masses:
T = DSR ( i i + i )
(10.11)
The mutual-damping coefficient is to be entered in units

of ( lb ft ) ( rad s ) .
HSP
(51-60)
The spring constant. This pertains to the elastic

connection between the present mass (assumed to be
number i = ML) and the mass with the next higher
6
number (i+1). Units are ( 10 lb ft ) rad .

DSD
(61-70)
The absolute-speed self-damping coefficient for this

mass. By definition,
T = DSR i
(10.12)
where Ti is the associated damping torque on mass

number i, and i is the angular velocity (absolute) of this
mass. Units are ( lb ft ) ( rad s ) .
Note that fields "DSM" and "HSP" contain data which really do not belong exclusively to mass
number i = ML. It pertains equally to the mass i+1. For mass card of the highest-numbered mass
(number "NUMAS"), then, these two fields have no meaning, and should be left blank.
Terminate the Mass Cards (Class 4 Data Cards) with a blank card.
10 - 42
Sources
Class 5 S.M. Data Cards (Output Requests):

1)
General Remarks
The Class 5 (output request) cards follow the blank card terminating the Class 4 (mass) cards.
The requests for the output of S.M. variables are divided into five separate groups:
1.
Electrical variables;
2.
Mechanical angles of shaft masses;
3.
Mechanical speed deviations of shaft masses;
4.
Mechanical torques on shaft sections;
5.
Machine parameters and initial conditions.
To request output of any S.M. variable, the user has to specify the output group to which the
variable belongs and a unique number identifying that variable within that output group. The rules
for assigning those numbers are explained in the following sections.
There is one common data card format for all different output groups. The user must specify the
desired S.M. output in the following format:
Group
ALL
1
2
3
4
5
6
7
8
12 3 4 5678 901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
N1
N2
N3
N4
N5
I1 I1
I6
I6
I6
I6
I6
Flag identifying output group to which the variables

specified on this card belong.
GROUP
(3)
Flag identifying a request for the output of all possible

variables in this output group:
ALL
(4)
N1-N12
=0 or
blank
selective specification of variables in columns 980;
=1
all possible variables in this class are to be outputted,

columns 980 will not be scanned;
Identification of variables to be outputted.
10 - 43
Sources
The output request cards can be stacked in any order, i.e., a request for output group i does not
have to precede the request for output group i+1. Similarly, the variable numbers (columns 980)
can be specified in any order. It is also possible to stack any number of cards specifying the
requests for different variables within any output group.
A blank card terminates the Class 5 S.M. data cards (output requests).
2)
Specification of Variables in Output Group 1
Presently there are 15 possible variables in this group. The actual number will vary depending on
the complexity of the model used.
The following Table summarizes the possible outputs in Group 1:
EMTP
Request
Number
Units
Output
Name
ID
"ID "
current in the d-axis armature winding;
IQ
"IQ "
current in the q-axis armature winding;
IO
"IO "
current in the zero-axis armature winding;
IF
"IF "
field winding current (winding #1 on the

direct axis);
ID2
"IKD "
current in the d-axis damper winding

(winding #2 on the direct axis)
IQ1
"IG "
current in the q-axis damper winding

(winding #1 on the quadrature axis);
IQ2
"IKQ "
current in the q-axis eddy-circuit winding

IA
"IA "
current in the phase 'a' armature winding;
IB
"IB
"
current in the phase 'b' armature winding;
IC
10
"IC
"
current in the phase 'c' armature winding;
VF
11
"EFD "
Variable
voltage applied to the field winding; this

will be a constant (dc) voltage, unless the
user explicitly specifies a connection to
TACS exciter dynamics as part of the
Class 6 S.M. data cards;
Table 1: Summary of Variables in Output Group 1
10 - 44
Sources
MFORCE
12
"MFORCE
"
the total mmf in the air-gap of the machine;
MANGLE
13
RAD
"MANGL
E"
angle between the q- and the d-axis

components of the total mmf (MFORCE);
TEG
14
Nm106
"TQ GEN"
electrodynamic torque of the machine;
TEXC
15
Nm106
"TQ EXC"
electromechanical torque of the exciter.
Table 1: Summary of Variables in Output Group 1
3)
This output group contains the mechanical angles of the different masses in the lumped mass
model of the turbine-generator set. The number of possible variables in this group depends on the
number of masses used in the representation of the S.M. (see parameter NUMAS in Class 3 s.m.
data cards).
Every mass of the turbine-generator set has been assigned a number during the specification of its
parameters (Class 4 s.m. data cards). The same number is to be used when requesting the output
of the mechanical angle of that mass. The output is in units of degrees.
4)
This output group contains the deviations of the mechanical speed of the different masses of the
machine shaft from the synchronous speed. Similarly to Output Group 2, specify the appropriate
mass number to obtain the desired output in units of rad/s.
5)
This output group contains the mechanical torques between the different shaft sections of the
turbine-generator set. The shaft torque number i is the torque on the shaft section connecting
masses number i and i+1 (there is, therefore, no output possible for a single-mass model). The
units of the output are million Newton-meters (Nm106).
6)
This output group can be printed only immediately following the steady-state solution. At
present, there are two possible requests in this group obtained by specifying 1 and/or 2 in any of
the variable number fields (columns 980) of Group 5 card:
1)
to request the printout of all machine parameters (reactances, resistances, shaft

data, etc.) following the steady-state solution;
10 - 45
Sources
2)
to request the complete printout of the initial conditions of a machine.
Note:
Do not forget to terminate the output requests with a blank card.
For EMTP output and plotting purposes, and just-delineated output variables are actually identified
by a pair of 6-character names. The second name (lower name of printed pair) identifies the
variable type mnemonically, as documented above. The first name (upper name of printed pair)
identifies the generator in question, in order of data input. For example, "MACH 3" would be for
the third machine. A specific example of such column headings of printed EMTP time-step-loop.
44
first
next
next
32 output variables pertain to dynamic synchronous machines, with names generated internally
STEP
TIME
JEFA
JEFB
JEFC
ROCA
ROCB
ROCC
MACA1
MACB1
MACC1
RJ0A
TERRA
ROCA
COMROC
ROCB
COMROC
MACH 1
ID
MACH 1
IQ
MACH 1
I0
MACH 1
IF
MACH 1
IKD
MACH 1
IG
MACH 1
IKQ
MACH 1
IA
MACH 1
IB
MACH 1
IC
MACH 1
EFD
MACH 1
MFORCE
MACH 1
MANG
MACH 1
TQ GEN
MACH 1
ANG 1
MACH 1
VEL 1
MACH 2
ID
MACH 2
IQ
MACH 2
I0
MACH 2
IF
MACH 2
IKD
MACH 2
IG
MACH 2
IKQ
MACH 2
IA
MACH 2
MACH 2
MACH 2
MACH 2
MACH 2
MACH 2
IB
IC
EFD
MFORCE
MANG
TQ GEN
0 0.000000 0.597706E+06-0.174582E+06-0.428338E+06 0.573396E+06 0.308519E+05-0.601402E+06
0.000000E+00 0.200639E+03-0.346962E+03-0.531940E+05 0.119511E+05-0.672124E-10
0.562678E+02 0.324215E+05 0.102060E+05-0.426276E+05-0.664511E+03 0.748410E+02
-0.568434E-13-0.531940E+05 0.119511E+05-0.672124E-10 0.770638E+04-0.509259E+02
0.102060E+05-0.426276E+05-0.664511E+03 0.748410E+02 0.873732E+00 0.362662E+01
1 0.000200 0.584982E+06-0.128961E+06-0.461446E+06 0.542021E+06 0.821850E+05-0.621363E+06
MACH 2
MACH 2
ANG 1
VEL 1
0.823733E+04 0.143815E+05-0.226188E+05
0.770638E+04-0.509259E+02 0.910237E+01
0.873732E+00 0.362662E+01 0.210592E+03
0.910237E+01 0.562678E+02 0.324215E+05
0.210592E+03-0.568434E-13
0.648591E+04 0.157221E+05-0.222080E+05
Class 6 S.M. Data Cards

Next in order of data input come cards which describe any interface connections between the
machine under consideration, and TACS. The general format for the one or more such cards of
Class 6 is as follows:
1
2
3
4
5
6
7
8
12 345678 901234 567 890123456789012345678901234567890123456789012345678901234567890
KK BUS
K1
I2
I3
Rule 1:
A6
If the field voltage of the machine is to be controlled by exciter dynamics which have
been modelled using TACS, then the following is required:
KK
(1-2)
BUS
(3-8)
Enter "71" in columns 1-2, as a special request for the

exciter connection.
The 6-character name of a TACS variable is to be entered
in columns 3-8. The numerical value of this TACS
variable will then be used by the EMTP logic for the field
voltage vf of this machine.
10 - 46
Sources
If no such card is used, the EMTP logic will simply hold the field voltage for this
machine constant, at whatever value was dictated by the initial conditions.
Rule 2:
If the mechanical power applied to any mass on the shaft system is to be controlled by
dynamics which are modelled using TACS, then the following card is required:
(1-2)
Enter "72" in columns 1-2, as a special request for the

TACS control of mechanical power.
BUS
The 6-character name of a TACS variable is to be
KK
(3-8)
entered in columns 3-8. The numerical value of this
TACS variable will then be used by the EMTP logic as a
multiplicative factor for scaling the otherwise-constant
(steady-state) power values.
KI
The mass number whose externally applied mechanical

power is to be controlled by TACS variable BUS.
Should the user want to code the logic of this TACS connections, but temporarily
bypass any associated dynamics, he can use the "UNITY". Recall that this is the name
of the built-in TACS source which has output identically equal to 1.0.
If no such card is used for mass #N, the EMTP logic will simply hold the external
mechanical power applied to that mass constant (at whatever value was dictated by the
initial condition calculation). Consequently, the user can selectively control the
mechanical power on different shaft sections.
Rule 3:
If the internal electrical machine variables are to be passed into TACS from the
machine, then the following is required. There can be any number (including zero) of
such cards:
KK
(1-2)
Enter "73" in columns 1-2 as a special request for the

internal electrical machine variables.
(3-8)
The 6-character name of the TACS source whose value is

to be equal to the desired electrical machine variable at
each time-step.
KI
Enter variable number in columns 15-17.
BUS
10 - 47
Sources
The following variables can be passed to TACS:
TACS
Variable
Request
Number
Units
ID
current in the d-axis armature winding;
IQ
current in the q-axis armature winding;
IO
current in the zero-axis armature winding;
IF
field winding current (winding #1 on the

direct axis);
ID2

(winding #2 on the direct axis);
IQ1

IQ2
current in the q-axis eddy-current winding

VD
voltage applied to the d-axis;
VQ
voltage applied to the q-axis;
VO
10
voltage applied to the zero-axis;
VF
11
voltage applied to the field winding; this will

be a constant (dc) voltage, unless the user
explicitly specifies a connection to TACS
exciter dynamics;
12
the total mmf in the air-gap of the machine;
13
RAD
angle between the q- and the d-axis

components of the total mmf (MFORCE);
TEG
14
NM106
electrodynamic torque of the machine;
TEXC
15
NM106
electromechanical torque of the exciter;
PSID
16
Weber-turn
d-axis flux linkage;
PSIQ
17
Weber-turn
q-axis flux linkage;
MFORCE
MANGLE
Table 2: Summary of Variables Accessible to TACS

A separate data card is needed for each variable.
10 - 48
Sources
If one of more mechanical angles i, angular velocities i, or shaft torques Ti of the
shaft system of the machine are to be passed into TACS, then one data card must be
entered for each such variable. Angles are in units of absolute radians (e.g., increasing
by 2f every second for 60 Hz steady-state operation), angular velocities are in radians/second, and shaft torques are in million Newton-meters. The data card format is:
Rule 4:
KK
(1-2)
BUS
(3-8)
KI
Enter "74" in columns 1-2 as a special request for the

access to mechanical variables of the S.M.
The EMTP reads from columns 3-8 a 6character
alphanumeric name. This field contains the name of the
TACS source whose value is to be controlled by the
mechanical machine variable in question.
The EMTP reads from columns 15-17 the code number
which identified the mechanical variable in question:
for i, enter i
for i, enter NUMAS + i
for Ti, enter 2 NUMAS + i
where i is the mass number specified in Class 4 S.M. data cards (mass cards); and
NUMAS number of connected masses on the shaft system of this S.M. (specified in
Class 3 s.m. data cards);
Rule 5:
Ordering of the different possible cards which have just been described is immaterial;
they can be shuffled, without altering the result.
The end of all such cards as have just been described is to be indicated by a special
terminator card. Field "KK" is to be left blank, and field "BUS" should contain the
"FINISH". Should this machine have no connections at all to TACS, then this "FINISH"
record alone makes up the Class 6 S.M. data.
1
2
3
4
5
6
7
8
12 345678 901234 567 890123456789012345678901234567890123456789012345678901234567890
FINISH
A6
10 - 49
Sources
Class 7 S.M. Data Card:

If the machine in question is not paralleled by one or more additional dynamic S.M. source
components, then there is no Class 7 S.M. data. In this case, the "FINISH" card of the Class 6 data
represents the final data card for this dynamic synchronous machine.
But suppose that a second dynamic S.M. is to be connected in parallel (armature windings
connected to the same buses) with the one just inputted. Then the just-inputted "FINISH" card of
Rule 6 must be modified to read "FINISH PART", indicating to the EMTP that S.M. data for this 3phase generator bus (to which armatures are connected) has only been partially completed. Class 7
data then consists of S.M. data cards for the second machine, beginning with Class 2 and ending
with Class 6.
This procedure can be generalized to apply to as many machines as the user wants to parallel on
the same generator bus. For M machines, the first M-1 has data cards which end with a Rule 6 data
card reading "FINISH PART"; the final one ends with just "FINISH", indicating to the EMTP that no
other machines follow on that generator bus. Only the first of the M machines has any Class 1 data
cards (since this applies to the common generator bus); all M-1 following machines skip this data.
How to Modify the Limits on Masses, Generators, Output Variables:

The Type-59 S.M. code is variably dimensioned. List 17 (see Section 2.5.1) defines the maximum
number of machines while List 16 defines the maximum number of masses to be modelled on all
machines. The size of the integer output vector (ISMOUT) is controlled by List 11.
10.12 U.M. Source Component

10.12.1
Introduction
The U.M. module of the EMTP currently can be used to represent 12 major types of electric
machines, as will be described in Section 10.12.6. Despite the universality of the program set-up,
a serious attempt is made not to burden the user with this universality if only the use of one or a
few of the available types is desired. Therefore, a user's guide is given for each U.M. type. Then
taking up these instructions for the desired U.M. type, reference will be found to the rules for
setting up the three classes of U.M. data cards as described in the section regarding the format of
the U.M. data cards. It is also possible to run the U.M. with the Type-59 S.M. data input. In this
case, no understanding about the rules pertaining to the U.M. is required at all. Instead, only the
rules as described in the section "U.M. with S.M. Type-59 Data Input" apply. The following points
can be remarked:
10 - 50
Sources
Number of machines and coils, compensation:

The number of U.M. machines which can be used is arbitrary. For each machine the following
rules apply. The maximum number of coils on the power side is restricted to three. This justified
by today's usage of electric machines. These three coils are three-phase compensated, i.e., they
can be externally connected to each other in one electrical network. As far as compensation of the
excitation coils is concerned, the general rule is that only the first three excitation coils are threephase compensated, i.e., only the first three excitation coils are permitted to be externally
connected to each other in one single electric network. If additional excitation coils are used, then
each of these coils must be totally disconnected from other coils used. The above outlined
restrictions can be overcome by the approach of using stub lines or inserting a TACS element
between U.M. machines or between coils which are not supposed to be connected to each other
(see Section 8.6.2 for comments about the stub line usage).
10.12.2
Note on Coil Parameters
The purpose of this note is to supply conversion formulas for users who have data of coilparameters specified in terms of self and mutual inductances, which due to the U.M. data input,
have to be expressed in terms of leakage and main inductances.
In this next discussion the following notation will be adopted;
Llki
Lmi
Ri
A)
=
=
=
leakage inductance of coil i [H]

main inductance of coil i [H]
resistance of coil i [Ohm]
Two-Coil System:
The idea of how to obtain the leakage and main inductances from the self and mutual inductances
will be illustrated for a transformer. Although a transformer is not an electrical machine, the basic
mechanism of deriving the desired conversion formulas is essentially the same. This is because an
inductance is a parameter that describes the relation between a coil current and the magnetic flux
enclosed by the coil contour. By considering a 2-coil transformer, the central data can be most
easily understood.
10 - 51
Sources
R1
R2
i1
V1
i2
L11
L22
V2
L12
Figure 10.19: Two-Coil Transformer
For the transformer as shown schematically in the figure above, the voltage equations can be
written as:
d 1
v 1 = R 1 i 1 --------dt
(10.13)
d 2
v 2 = R 2 i 2 --------dt
The current-flux relation in terms of self-inductances, L11 and L22, and of mutual inductance L12,
can be expressed as:
L 11 L 12 I 1
L 12 L 22 i 2
(10.13)
with:
N1 N2
L 12 = L 21 = -------------Rm
and
(10.14)
Ni
L ii = L lki + ------R m = L lki + L mi
where N = number of turns and Rm = magnetic reluctance.
To express the equations in terms of leakage and mutual inductances, one of the "sides" of the
transformer has to be reduced (referred) to the other. Let us reduce the secondary (index 2) to the
10 - 52
Sources
primary side (index 1). Then introduce the following reduction factor, which for transformers
specifically, is called turns ratio:
N
a 2 = ------1N2
(10.15)
Now rewrite the voltage relations (Equation 10.13) as:

d 1
v 1 = R 1 i 1 --------dt
a2 v2 =
i2
2
-
( a 2 R 2 ) ---a
2

d ( a2 2 )
------------------dt
(10.16)
and also rewrite the current-flux relations (Equation 10.14) as:
a2 2
L 11 a 2 L 12
I1
2
a 2 L 12 a 2 L 22 I 2 a 2
(10.17)
By introducing the so-called reduced secondary variables:
v' 2 = a 2 v 2 ,
2
R' 2 = a 2 R 2
i2
i' 2 = ----a2
and
(10.18)
2
L' 22 = a 2 L 22
and realizing from Equation (10.15) that

L m1 = a 2 L12
(10.19)
the transformer can be described by the equations:

d
v 1 = R 1 i 1 --------1dt
(10.20)
d' 2
v' 2 = R 2 i 2 ---------dt
10 - 53
Sources
' 2
L 11 L m1 i 1
L m1 L' 22 L 2
(10.21)
At this point L11 and L22 can be partitioned as:

L 11 = L lk + L m1
(10.22)
L'22 = a 2 L 22 = a 2 L + Lm1 = L' lk2 + Lm1

lk2
so that from Equation (10.21) we have the current-flux relation:
1 = L 11 i 1 + L m1 ( i 1 + i 2 )
(10.23)
' 2 = L'12 i 2 + L m1 ( i 1 + i 2 )
The leakage inductances are then identified as Llk1 and Llk2) and both coils have now the same
main inductance Lm1.
On the basis of the formulation of Equation (10.20) and Equation (10.22) the transformer can be
interpreted as having the following connected network:
i1
R1
Llk1
Llk2
v1
R2
i2
+
v2
Lm1
Several conclusions can now be made:

1.
Physically this formulation corresponds with the concept of 2 coils linked by a

main flux in addition to their individual leakage fluxes. The usefulness of this
formulation to numerical implementation is the fact that practically only the main
inductance Lm1 is susceptive to changes if the magnetic material becomes
saturated.
10 - 54
Sources
2.
If the reduction factor a2 is given, then the conversion formulas needed to obtain
the main and leakage inductances from given self and mutual inductances, are:
L m1 = a 2 L 12
L lk1 = L 11 Lm1
(10.24)
L' lk2 = a 2 L 22 L m1
3.
If the reduction factor is not given, then the following approximate conversion
formulas are suggested:
L m1 = 0,9L 11
L lk1 = 0,1L 11
L m1
a 2 = --------L 12
(10.25)
2
L' lk2 = a 2 L 22 L m1
Sometimes the rated currents I1B and I2B are given. In this case the reduction factor
can be approximated by:
I 2B
a 2 = ------I 1B
or
v 1B
--------v 2B
(10.26)
The conversion formulas of Equation (10.25) can be used to obtain the desired main
and leakage inductances.
4.
All parameters and variables of the secondary side, including the network to
which it is connected, have to be referred to the primary side (see
Equation 10.19), i.e., their voltages are to be multiplied by a2, their currents are
to be divided by a2, their resistances and inductances are to be multiplied by a22
B)
N-Coil System:
The concept of the 2-coil system linked by a main flux in addition to their individual leakage flux,
can be easily extended to the general ncoil system. In this case, there are to be defined n-1
reduction factors (see Equation 10.16).
N
a j = ------1Nj
j = 2, , n
(10.27)
10 - 55
Sources
These reduction factors are then implemented in the system configurations in the same way as for
the 2-coil system (Equation (10.27) and Equation (10.18)).
As a result of these operations:
1.
We are able to obtain the main flux concept as represented by one main
inductance Lm1 in addition to the individual leakage fluxes of each coil as
represented by
Llk1, Llk2 , , Llkn
2.
The formulation in terms of main and leakage inductances requires the reduction
of all coils to one coil, which in above examples is coil 1. This means that all
coils (plus networks connected to these coils) other than coil 1 should have: their
voltages multiplied by the corresponding reduction factor, their currents divided
by the corresponding reduction factor, their resistances and inductances
multiplied by the corresponding square of the reduction factor.
3.
If all reduction factors are given, then the conversion formulas to obtain the main
and leakage inductances from a given self and mutual inductances are:
L m1 = a 2 L 12
Llk1 = L 12 L m1
(10.28)
L' lki = a i L ii L m1
4.
for i = 2, 3, , n
If the reduction factors are not given, then the following approximate conversion
is suggested:
L m1 = 0,9L 11
L lk1 = 0,1L 11
Lm1
a i = --------L 1i
(10.29)
L'lki = a i L ii L m1
for i = 2, 3, ,
n
5.
Sometimes the rated currents IiB are given. In this case the reduction factors can
be approximated by:
I iB
a i = ------I 1B
or
v 1B
--------v iB
(10.30)
10 - 56
Sources
Having determined these reduction factors, the formulas in Equation 10.29 can be
used to obtain the main and leakage inductances.
6.
C)
If only I1B and I2B are given, then

(A)
find a2 from Equation (10.31)
(B)
find Lm1 = a2 L12
(C)
find the remaining ai's from: ai = Lm1 / L1i
Extension to Electric Machines:
The concept of coils on the same magnetic shaft as linked by a main flux in addition to the
individual leakage fluxes can be carried over to the coil-arrangement of electric machines. In the
Park domain these magnetic shafts are the direct and quadrature axis. Then for each of these axis,
the conversion formulas as given in Equation (10.29) through (10.31) can be applied. A rigorous
account of the reduction process to the power side d and qcoils can be found, for instance, in
Reference 40.
Basically the main and leakage inductances are introduced in the same way as explained earlier,
except that proper account has to be taken of the winding distribution factors rather than by
considering the number of turns only.
Let us now consider the case that all excitation coils are coils which have magnetic shafts either in
the direct daxis or in the quadrature qaxis, or in both axes. See Special Notes on Direct Current
Machines and UM Type-4. It is convenient to formulate the main leakage inductances on the basis
of reduction of all excitation coils to the power coils which are positioned on the same magnetic
shaft.
If the power side is a three-phase system, then their equivalent coil-parameters in the Park domain
have to be determined first. This is a well-known procedure, and let us call these equivalent
power coils the 0-coil, the d-coil, and the q-coil.
Now let us consider the following example. Suppose there are four excitation coils, two coils D
and F on the d-axis and two coils Q and G on the q-axis. The system is schematically shown in
Figure 10.20:
10 - 57
Sources
d-axis
q-axis
d- and q-coils, power coils; F-, D-, G-, Q-coils: excitation coils
Figure 10.19:
System with Four Excitation Coils
Now suppose that all self and mutual inductances are available. Then in order to find the main and
leakage inductances as required by the UM, the same approach can be taken as outlined for the ncoil system.
It is reemphasized that all parameters and variables of the excitation coils, including the network
to which they are connected, have to be reduced according to rules as explained earlier:
1.
their voltages multiplied by the reduction factor.
2.
their currents divided by the reduction factor.
3.
their resistances and inductances multiplied by the square of the reduction factor.
The power coil variables and parameters remain unchanged in the reduction process.
10 - 58
Sources
D)
Special Note on U.M. Type-4:
Because a Type4 machine has a three-phase excitation system, it does not directly fit the general
discussion as held earlier, where all excitation coils have magnet shafts in line with the direct and
quadrature axis. This is solved internally in the EMTP by an extended Park transformation. As a
result, the general comments made earlier hold and can be applied.
The induction machine is balanced in its design with respect to its three a, b, c phases. The very
simple conversion formulas are described below with the following notation for the (given) self
and mutual inductances of the machine.
LSS
stator self-inductance
MSS
mutual inductance between 2 stator coils
LRR
rotor self-inductance
MRR
mutual inductance between 2 rotor coils
MSR
maximum value of mutual inductance between stator and a rotor coil of

the same phase.
The power coils are the stator coils, and the excitation coils are the rotor coils.
The main and leakage inductances required by the UM can be obtained as follows:
(A)
Realize first that the real a, b, c coils have to be transformed to equivalent d, q

coils for both the stator and rotor side. The self and mutual inductances of
these equivalent coils are:
(i)
self inductance of stator 0-coil:
(ii) self inductance of rotor 0-coil:
SS
Lo = L
RR
Lo
(iii) self inductance of stator d and q-coil:

SS
SS
L d = Lq = L
SS
+M
SS
(iv) self inductance of rotor d- and q-coil:

RR
LL d
R1
= LLq
= LL
10 - 59
RR
+M
RR
= L
SS
RR
2M
SS
2M
RR
Sources
(v) mutual inductance between stator and rotor coil on the same
RS
RS
3 SR
magnetic shaft (either d- or q-axis): L dd = L qq = --- M
2
If the reduction factor a is given, then the conversion formulas to obtain
the main and leakage inductances can be obtained by applying the general
theory discussed earlier, to give:
3 SR
L md = Lmq = a --- M
2
S
L lkd = L lkq = ( L
R
SS
+M
L lkd' = L lkq' = a ( L
RR
SS
) L md
+M
RR
(10.31)
) Lmd
The variables supplied by a prime are reduced variables.

(B)
If the reduction factor is not given, the conversion formulas can be

approximated by:
L md = L mq = 0,9 ( L
S
Ld1 = L q1 = 0,1 ( L
a = (L
R
SS
+M
R
SS
SS
+M
+M
SS
SS
3 SR
) --- M
2
L d1' = L q1' = a ( L
(C)
SS
RR
+M
RR
(10.32)
) Lmd
All excitation variables have to be reduced in the same sense as described in

previous discussion. It is to be noted that the zero excitation coil has to be
reduced also. Thus:
R
2
RR
L o = a Lo
E)
Special Note on DC Machines:
For some of the types of dc machines the situation arises that a power coil and an excitation coil
might not be on the same magnetic axis. In this case the partitioning into main and leakage flux
can be made arbitrarily without any numerical consequences.
10 - 60
Sources
10.12.3
Use of Network Representation of Mechanical System
Users of the U.M. model have the option to employ the network representation of the mechanical
system. This network representation can then be handled by the Electrical Network section of the
EMTP. This approach provides the means to use the efficient sparsity oriented solution method of
the EMTP, in addition to its variable dimensioning feature. Moreover, no time-step delay is
required in interfacing the U.M. model with the mechanical system.
For each rotor mass, there is a capacitor to ground; for each spring interconnecting rotor masses,
there is an inductor between the capacitor nodes. Applied (load) torque becomes injected current
into such a capacitor node. Etc. as per the following correspondence:
Mechanical quantity
Electrical quantity
(torque on mass)
(current into node)
wm
(angular speed)
(node voltage)
(angle)
(capacitor charge)
(moment of inertia)
(capacitance to ground)
(spring constant)
1/L
(reciprocal of inductance)
(viscous damping)
1/R
(conductance)
10 - 61
Sources
v
+
i
2
dv
i = C -----dt
d 2
d
T = J -------------- = J ------dt
dt
2
k +
1
i = --- ( v k v m ) d
L
T = K ( 1 2 )
= K ( 1 2 ) d
Vk
T = D
Figure 10.20:
R
ik
1
i k = --R V k
Network Representation of Mechanical System
In summary: For each mass of the shaft system, define an electric network node with a capacitor
of value J (the moment of inertia) to ground. If there is absolute-speed damping on this mass, a
resistor having conductance D (the viscous damping constant) is put in parallel with the capacitor.
If there is an externally applied mechanical torque such as a mechanical load on this mass, a
current source is put in parallel with the capacitor such that a mechanical load becomes a negative
injection. If there are two or more masses, an inductor is added to connect the associated adjacent
capacitors; value of the inductance is 1/K (reciprocal of the spring constant connecting the two
masses). If there is speed-deviation damping associated with this coupling, the inductance is
parallelled by a conductance of value DMUT (the viscous damping constant of this effect). There is
no element representing the U.M. driving torque, since the EMTP provides this
connection automatically.
In terms of EMTP mathematics, the U.M. driving torque looks like a hidden nonlinear element
which is connected from ground to the node of the capacitor that represents the rotor mass of the
U.M. Hence do not expect the currents in branches that are connected to this node to sum to zero,
10 - 62
Sources
for one branch is hidden. As for output of such mechanical variables, they are handled by column80 requests on the branch cards representing the R-L-C elements that are used. Remember that
shaft torques are inductor currents, node voltages are mechanical speeds, etc. If mechanical
angles are wanted, one can pass the node voltages into TACS and integrate.
10.12.4
A)
Steady-State Initialization
Approach for Data Initialization:
The twelve U.M. types can be classified into three major classes of electric machinery:
synchronous machines, induction machines, and direct current machines. The doubly-fed
machines can be represented by either a synchronous or an induction machine type where both the
stator and rotor are connected to external electric networks containing sources of the desired
frequencies. All cases are currently subject to the common restriction that a completely correct
solution of the steady-state initialization feature will result only if the electric network to which
the power coils are connected is linear and balanced. The reason is that the current coding of
steady-state initialization is limited to processing the positive sequence component only. If the
network is nonlinear or unbalanced, then no program stop will occur. In this case the program
simply runs from an incorrect initial condition and needs to run a number of integration steps
before eventually settling down to the correct steady-state equilibrium. For successive runs, the
START AGAIN feature could be used to take advantage of once established correct steady-state
equilibrium.
B)
Approach for Synchronous and Direct Current Machines:
In initializing the different U.M. types of synchronous and direct current machines, the basic
approach is similar to that taken by Type-59 S.M. model. The crucial data input that the user is
requested to specify are the desired amplitude and angle of the machine terminal voltage.
The following steps are executed by the automatic initialization procedure:
1.
The electric network to which the power coils are connected is provided with
voltage sources reflecting the above-mentioned user-supplied data input.
Solving the steady-state equations of this network establishes the current of all
power coils.
2.
The current of the excitation (field) coil is adjusted such that the power coils
currents and voltages match those as given from step (1).
3.
The electromagnetic torque is found from the coil currents that are all known at
this stage.
10 - 63
Sources
4.
C)
Sources in the network connected to the excitation coil and the sources in the
mechanical network are adjusted to accommodate the excitation current and
electromagnetic torque as found in steps (2) and (3).
Approach for Induction Machines:
The approach of the previous section is not applicable to the steady-state initialization of induction
machines. The basic reason is that by the user's specification of the machine terminal voltage and
by the execution of step (1), each machine power coil is impressed with a predetermined voltage
and current as well as power factor. However, as opposed to synchronous and direct current
machines, the required adjustment of the currents in the excitation (rotor) coils of step (2) is not
possible due to the lack of external sources in the excitation (rotor) circuits. Neither is it possible
to accommodate these required currents by a particular value of rotor speed. This can be
understood from inspection of the circle diagrams of induction machines, reflecting the relation of
stator and rotor phasor currents for all rotor speeds for a given stator terminal voltage.
The approach followed for the U.M. induction machine types is to require the user to specify the
desired rotor speed instead of the terminal voltage (which is required for synchronous and direct
current machines). The desired rotor speed is specified in terms of the slip (in %), defined as:
o np m
slip = ---------------------------- 100
o
[%]
(10.33)
with
o
angular frequency of the electric network to which the power

(stator) coils are connected (=2 STATFR)
np
number of pole-pairs
angular speed of the rotor mass (=2 .RPM/60)
The slip is positive for motor operation and negative for generator operation.
The following steps are executed by the automatic initialization procedure for induction machines:
1.
The program acquires the Thevenin voltages and impedances of the electric
network to which the U.M. power coils are connected. These Thevenin
parameters are transformed into the frequency domain of the internal U.M.
equations (this frequency is different from the frequency of the electric network).
2.
With the given slip, the frequency of the excitation coils and the connecting
networks is known. The program acquires the Thevenin voltages and
impedances of the networks.
10 - 64
Sources
3.
Solution of all U.M. equations in the frequency domain establishes the currents
of all coils, and thus all currents that have to flow into the various networks
connected to the U.M.
4.
With the currents found in step (3), the electromagnetic torque and the terminal
voltages of all coils are easily obtained.
5.
Sources in the mechanical network are adjusted to accommodate the

electromagnetic torque determined in step (4).
The above procedure does not require any iteration to find the machine variables.
D)
Approach for Doubly-Fed Machines:
It was mentioned earlier that any synchronous or induction type of the U.M. can be used to
represent a doubly-fed machine by connecting external sources of any desired frequency to both
the stator and rotor side of the machine. For automatic initialization, however, the user is obliged
to employ the induction types of the U.M. This will not result in any loss of generality.
As a consequence, the same approach as outlined in the previous section is taken for the automatic
initialization of doubly-fed machines. The user is requested to specify the rotor speed in terms of
the slip. This determines the frequency of the rotor circuit and all machine variables based on the
described approach. The amplitude of any source in the electric network connected to the U.M.
excitation coils (stator) and in the electric network connected to the U.M. excitation coils (rotor)
can be freely set by the user with the EMTP Type-14 sources. As far as frequencies are concerned,
only the frequency of the sources in the network connected to the stator coils (corresponding to the
value set by the user in the frequency argument of the Type-14 sources) will be accepted. The
frequency argument of all Type-14 sources in the electric network that is connected to the rotor
coils can be set to any arbitrary positive value. The program simply ignores this specification and
instead will set this frequency equal to the correct frequency of the rotor circuit as calculated from
the steady-state condition with given stator frequency and the slip that the user was requested to
specify. This steady-state condition for the frequency is the well-known relation:
np m + r = o
(10.34)
angular stator frequency
angular frequency of rotor electric circuit
angular speed of rotor mass
np
number of pole-pairs.
10 - 65
Sources
E)
Rules for U.M. Automatic Initialization:
It will be outlined in the next sections how to make use of the U.M. automatic steady-state
initialization option. The following rules or restrictions need to be taken into consideration in
using this option:
1.
If a U.M. coil is connected to some network of the EMTP that contains a number
of EMTP sources, then only the EMTP Type14 sources are allowed to be
activated with TSTART < 0, and only these sources will be regarded in the
initialization process. All other types of EMTP or TACS sources can only be
included if they are activated at TSTART > 0. The frequencies of all EMTP Type14 sources in one connected subnetwork need to be of the same value. The
initialization of dc networks containing dc sources only happens if these dc
sources are simulated by EMTP Type14 sources with TSTART < 0 and all
frequencies set to an adequately low value.
Note:
2.
The automatic steady-state initialization option is only honoured if the

mechanical network representation feature is used. In this electric network
analog of the mechanical system, as many EMTP Type14 sources to represent
applied torques (currents) or speeds (voltages) are allowed to be activated at
TSTART < 0. It is to be noted that in the steady-state this network is a DC network
and the needed DC sources are to be simulated by the Type14 source with the
frequencies set to a proper and equal low value. The user is requested to specify
at least one of these Type14 sources to be adjustable in its amplitude. The value
of this amplitude will be calculated by the program to accommodate the
electromagnetic torque that the machine is supposed to produce. The user will be
requested to specify the name of the node to which this adjustable source is
connected. If more than one source is declared to be adjustable, the ratios of the
source amplitudes with respect to the amplitude of the first specified adjustable
source can also be freely set. Thus, the demanded electromagnetic torque can be
freely distributed over an arbitrary number of sources. Due to the rule described
in point 1), TACS or other types of EMTP sources may be included to the network,
provided that they are activated at TSTART > 0, and thus do not take part in the
initialization process.
Note:
3.
Any U.M. node is not allowed to be connected directly to an external

EMTP source. Insert a negligible small resistance as an approximation.
The node representing the mass of the electric machine is not to be

connected directly to an external EMTP source. A negligible small
resistance can be inserted if so desired.
Rule exclusively for synchronous and direct current machine types of the U.M.:
Of all EMTP subnetworks connected to the different coils on the excitation side of
the machine, only the subnetwork that is connected to the field coil is allowed to
have a non-zero Type14 source with TSTART < 0. This field coil is to be
10 - 66
Sources
identified with the first coil of the d-axis coils that is to be used. Since this
Type14 source is to simulate a DC source, the frequency argument is to be set to
an adequately low value. The amplitude can be set to any positive value, since
the correct value will be calculated by the program to accommodate the field
current that is needed to generate the desired terminal voltage of the machine.
The user will be requested to specify the name of the node to which this source
is connected. Other EMTP or TACS sources are allowed to be included, provided
that they are all activated at TSTART > 0.
F)
U.M. Interface with EMTP Load Flow (FIX SOURCE):
The EMTP Load Flow (see Section 10.13) is restricted to three-phase systems. Consequently, only
U.M. types with three-phase armatures (power sides) can be initialized by the EMTP load-flow.
These U.M. types are:
1.
U.M. Type-1 synchronous machine;
2.
U.M. Type-3 induction machine;
3.
U.M. Type4 induction machine or doubly-fed machine;
4.
U.M. with S.M. Type59 data input.
The following restrictions apply only to the induction and doubly-fed machinery:
1.
For both U.M. Type 3 and 4 the user has the freedom to specify the initial speed
of the machine through the usual specification of the initial slip on the assigned
data card as described in the general U.M. data input rules. This specified speed
will not be modified by the load-flow calculations.
2.
The induction machine operating at a given speed behaves exactly like an

impedance. Therefore, as is the case with impedances, no direct adjustment of
the induction machine power output nor input can be made by the load-flow
process. The power flow as well as voltage conditions are in the load-flow
process only controllable by available sources in the network such as Type14
sources and, synchronous machines.
3.
U.M. Type3 as well as Type-4 can be used to simulate doubly-fed machinery,
where both the power and excitation coils are connected in some way to external
networks containing certain sources. In this case these U.M. types could be
utilized as controllable sources in the load-flow process. However, only U.M.
Type-4 is permitted to be used as such. The U.M. interfacing logic creates, prior
to the load-flow calculations, the equivalent circuit of U.M. Type-4. The logic
also provides the proper connection of this circuit with the electric network on
the power side of the machine, and with the electric network on the excitation
10 - 67
Sources
side of the machine. Consequently, if the user included a balanced three-phase
set of Type-14 voltage sources to the electric network on the excitation side, then
these sources may be used in the load-flow calculation process for the purpose of
accommodating a desired power flow or voltage condition of the entire electric
network (see rules for FIX-SOURCE in Section 10.13 and Section 10.14). The
frequency of these Type-14 sources will be automatically adjusted by the U.M.
initialization logic to match the initial slip value as specified by the user with
U.M. data input in the usual way. It is remarked that obviously the power flow
through the excitation coils of the machine is equal to the well-known slip power,
i.e., the armature power (power side) times the value of the absolute slip. The
use of U.M. Type-3 in the doubly-fed machine mode is in the load-flow process
considered to be only equivalent to a fixed impedance and a fixed source.
The theoretical background for the U.M.-load-flow interface can be briefly explained as follows.
The interfacing for the synchronous machine mode of the U.M. is trivial because the insertion of a
single step prior to the existing U.M. steady-state initialization procedure is all that needs to be
implemented. In this step the load-flow calculations are introduced with start values taken from
the machine terminal voltages which are specified by the user with the machine data input in the
usual way. The machine is represented with a balanced set of three-phase Type14 sources and
completion of the mentioned step provides directly the adjusted source arguments corresponding
with the desired power flow or voltage condition. The subsequent series of calculations follow the
earlier explained logic.
The above-outlined procedure for synchronous machines is not applicable to induction machines.
This was explained earlier with regard to the steady-state initialization procedure of induction
machines. The reason is that induction machines cannot be imposed with an arbitrary state of the
armature voltages and currents, even if the machine speed (or slip) is assumed to be adjustable
from minus infinite to plus infinite. This phenomenon leads to the need of determining the
equivalent impedance or network of the induction machine prior to the initiation of the load-flow
calculation process. Extension to induction machines which, on the excitation side are connected
to an external network containing certain sources (this doubly-fed machinery), requires the
determination of the equivalent sources as well, before the load-flow calculations can be initiated.
The determination of these equivalent sources may be omitted if the equivalent circuit of the
machine is created such that a topological connection between the electric networks on the
armature (power side) and on the excitation side can be realized. This approach was taken for U.M.
Type4, but is not possible for U.M. Type-3 due to the fact that armature and excitation windings
for this machine do not have an equal number of phases.
To achieve this goal logic has been set up to subject the same equations to a multi-pass procedure,
handling the pre- and well and post-load flow calculation process.
10.12.5
Saturation Modelling
To simulate the saturation effects, the U.M. approximates the saturation curve as two piecewise
linear segments. The user can either include the saturation for the daxis, or qaxis, or for both.
10 - 68
Sources
The saturation parameters for a daxis situation curve are FLXSD, LMUD, and LMSD as indicated in
Figure 10.22 below. The qaxis saturation parameters are similarly defined.
md
[Weber/m2]
FLXSD
LMSD (Lmd saturated)

LMUD (Lmd unsaturated)
imd [A]
Figure 10.21:
Approximated d-axis saturation curve
These daxis saturation parameters are to be specified on card (2) of the Class 2 U.M. data cards,
and the qaxis saturation parameters on card (3).
The decoupled approach of direct and quadrature axis saturation to the simulation of the saturation
phenomenon works reasonably well for example with synchronous or direct current machines
with a definite field coil in one axis. However, when the electromagnetic circuit structures of both
the stator and the rotor are symmetric, as with most induction machines, this decoupled approach
leads to unacceptable results. Therefore, the total saturation option is also available.
A)
Residual Flux:
The U.M. module includes the residual flux option. The two-segment saturation curves for the d
axis, qaxis and total saturation are extended with one additional segment. Figure 10.23 below
shows the saturation curve for the d-axis.
In addition to the usual saturation parameters FLXSD, LMUD and LMSD, the parameter FLXRD,
which is the daxis residual flux, is required to be specified. For the qaxis the corresponding
saturation and residual flux parameters are FLXSQ, LMUQ, LMSQ and FLXRQ.
10 - 69
Sources
md
FLXSD
FLXRD
Figure 10.22:
10.12.6
LMSD
LMUD
imd
Approximated d-axis saturation curve with

residual flux FLXRD
User's Guide General Remarks
This section describes the procedure for setting up a data case for any of the 12 available U.M.
types:
Type-1:
synchronous machine, 3-phase power side
Type-2:
synchronous machine, 2-phase power side
Type-3:
induction machine, 3-phase power side
Type-4:
induction machine, 3-phase power side and 3phase

excitation side. There is also a special option for the
input of induction machine data using nameplate data.
This is the Type-40 machine.
Type-5:
induction machine, 2phase power side
Type-6:
single-phase AC machine, induction or synchronous, 1

phase excitation
Type-7:
same as type 6 but 2phase excitation
Type-8:
DC
machine, separately excited
Type-9:
DC
machine, series compound (long shunt) field
Type-10:
DC
machine, series field
Type-11:
DC
machine, parallel compound (short shunt) field
Type-12:
DC
machine, parallel field
10 - 70
Sources
Note:
Except for Types 4, 6, 7 and 40, all machine types can

have an arbitrary number of d and qaxis excitation
coils. The 3phase excitation axes of Type-4 and Type
40 are 120 degrees shifted.
There are three classes of U.M. data cards: classes 1, 2 and 3. Each class consists of one or more
data cards. Class 1 of the U.M. data cards contain the general common specifications for all U.M.
machines. Class 2 and Class 3 U.M. data cards contain the information for each U.M. machine to
be used.
Class 2 contains parameters of the machine-table, such as the speed or the rotor moment of inertia
of the machine. Class 3 contains the required parameters of a coil, for example, the coil resistance
value. The class 3 data are stored in the coil table.
The U.M. data cards are to be placed behind all other used types of EMTP sources.
The general format of these U.M. data cards is as follows:
1.
Other EMTP cards, including the request card for changing the default U.M.
dimension (see Section 4.16).
2.
Class 1 U.M. Data Cards (general specifications for all U.M. machines)
consisting of two cards.
3.
BLANK CARD ENDING CLASS 1 U.M. DATA.
4.
Cards of U.M. #1, an arbitrary U.M. machine assigned to be machine number 1.

Class 2 U.M. Data Cards - 3 or more cards constituting the machine-table.
Class 3 U.M. Data Cards - as many cards as the number of coils used.
5.
Cards of other U.M. machines to be included, with each machine specified in

terms of Class 2 and Class 3 U.M. Data Cards (see (4)).
6.
BLANK CARD ENDING ALL U.M. DATA.
7.
Other EMTP cards.
10 - 71
Sources
10.12.7
U.M. Type-1 User's Guide
U.M. Type-1
= Synchronous machine, 3phase power side, arbitrary number of excitation

coils in d and qaxis.
Step 1:
Set up the general specification data cards according to the instructions of

Class 1 U.M. Data Cards.
Step 2:
Set up the machine-table data cards according to the instructions of

Step 3:
Set up the coil-table data cards according to the instructions of Class 3

U.M. Data Cards.
Exception to these instructions is with regard to the specification of the
resistances and leakage inductances in the coil cards representing the
power coils:
(A)
Resistances of the power coils A, B and C are required to be equal to each

other. In the case of non-equal resistances, the following procedure has to
be followed. Insert to the U.M. equal resistances with values equal to the
smallest resistance. Then account for the remaining resistances of the
coils by connecting in the Electrical Network resistance branches in series
with the U.M. coil. This is also a valid approach for studies of frequency
dependency of resistances.
(B)
Specify the leakage inductances LLEAK on the coil cards according to the
rules;
on coil card A:LLEAK
total inductances of zero

sequence component
on coil card B:LLEAK
daxis leakage inductance
on coil card C:LLEAK
qaxis leakage inductance
Note:
If no zero sequence component current flow is desired, then leave blank

the resistance and inductance of coil A.
Remark: These inductances can be found from the self- and mutual-inductances
by applying a similarity transformation with the Park's transformation
matrix. See, for example, Reference 1.
10 - 72
Sources
10.12.8
U.M. Type-2 =
Synchronous machine, 2-phase power side, arbitrary number of excitation

Step 1:

Step 2:

Step 3:

U.M. Data Cards.
Exception to these instructions is with regard to the coil cards for the
power coils:
(A)
The 2-phase power coils have to be chosen as power coil B and C. Since
power coil A is not used, the coil card representing power coil A is to be
completely left blank.
(B)
Resistances and leakage inductances which have to be specified in the

coil cards representing the power coils B and C, must follow the same
instructions as outlined in Step 3 of u.m. Type 1 (except of course that
reference to coil card of power coil A has to be disregarded).
10.12.9
U.M. Type-3
= Induction machine, 3-phase power side, arbitrary number of excitation

Step 1:

Step 2:

Step 3:

U.M. Data Cards.
Exception to these instructions is with regard to the specification of the
resistances RESIS and the leakage inductances in the coil cards
representing the power coils. The instruction to be followed is outlined
in Step 3 of u.m. Type-1.
10 - 73
Sources
10.12.10
U.M. Type-4
= Induction machine, 3phase power side and 3phase excitation side.
Step 1:

Step 2:

Exception to these instructions is with regard to the specification of
NCLD and NCLQ for card (1), both of which have to be left blank.
Step 3:

U.M. Data Cards, where rules (3.B) and (3.C) have to be disregarded.
Then additionally the following rules apply:
(A)
Instead of the mentioned rules 3.B and 3.C, the rule here is that the power
coil cards have to be followed by excitation coil cards in the sequence of
coil B, coil C and coil A. Note that the power coil cards are supposed to
be arranged in the sequence of coil A, coil B and coil C according to rule
3.A of Class 3 U.M. data cards.
(B)
Specify the resistances RESIS and the leakage inductances LLEAK on the
coil cards representing the power coils A, B, C according to the same rules
as outlined for Step 3 of U.M. Type-1.
(C)
Same as B) for the coil cards representing the excitation coils A, B, and C.
10.12.11
U.M. Type-5
= Induction machine, 2phase power side, arbitrary number of excitation

coils.
Set up the data case for this type according to the same instructions is outlined for U.M. Type-2.
Note: Positive sequence for 2-phase systems is obtained by:

vb = v sin t
vc = v sin (t + 30)
(10.35)
This corresponds to positive sequence in 3-phase system:
10 - 74
Sources
va = v sin t
vb = v sin (t 120)
vc = v sin (t + 120)
(10.36)
10.12.12
U.M. Type-6
= Single-phase AC machine, induction or synchronous, 1phase excitation.
Step 1:

Step 2:

Exception to these instructions is with regard card (1) variables NCLD
and NCLQ. Due to the internal U.M. coil arrangement, these variables
have to be specified as:
NCLD = 1, NCLQ = leave blank or 1 if an auxiliary starting coil is desired.
Step 3:

U.M. data cards, but completely disregard rule number 3 about the
sequence of coil cards belonging to one machine. Instead the following
rules have to be followed:
(A)
First insert 2 blank cards.
(B)
Then comes the coil card representing the excitation coil. The current
through this coil will be indicated in the output as IPC.
(C)
Now comes the coil card representing the power coil. If NCLQ has been
set equal to one, then this power coil card has to be followed with another
coil card representing the auxiliary starting coil. The currents through the
power coil and auxiliary coil will be indicated in the output as
respectively IE1 and IE2.
10 - 75
Sources
10.12.13
Type-7 User's Guide
U.M. Type-7
= Single-phase AC machine, induction or synchronous, 2phase excitation.
The instructions to set up the data case are the same as for U.M. Type-6, except for rules A) and B)
of Step 3, which now become:
(A)
First insert 1 blank card.
(B)
Then come 2 coil cards representing the two excitation coil, which are
spatially 90 degrees shifted from each other. The current through these
excitation coils will be indicated in the output as respectively IPB and IPC.
10.12.14
U.M. Type-8
machine, separately excited, arbitrary number of excitation coils in d

and qaxis.
DC
Step 1:

Step 2:
Set up the machine-table data cards according to the instructions of Class

2 U.M. Data Cards. In specifying NCLD and NCLQ on card (1) it is
important to be aware of the fact that the power coil (armature coil) is on
the q-axis. The usual way is to arrange the excitation coils (field coils) on
the daxis.
Step 3:

U.M. Data Cards.
Exception to these instructions is with regard to rule (3A), which instead
must be replaced by the following rule: First insert 2 blank cards, and
then follow with the coil card representing the power coil (armature coil),
which is located on the qaxis.
10 - 76
Sources
10.12.15
Type-9, 10, 11, 12 User's Guide

U.M. Type
DC
machine, series compound (long shunt)
10
DC
machine, series
11
DC
12
DC
machine, parallel compound (short

shunt)
machine, parallel
Step 1:

Class 1 U.M. Data cards.
Step 2:

Exception to these instructions is with regard to card (1) variables
NCLD and NCLQ, which are to be specified as:
NCLD = 2, NCLQ = leave blank.
Remark: This means that there are 2 excitation (field) coils in the daxis.
Step 3:
(A)
(B)

U.M. Data Cards, but completely disregard rule number (3) about the
sequence of coil cards belonging to one machine. Instead the following
rules have to be followed:
First insert 2 blank cards.
Then comes the coil card representing the power (armature) coil. The
U.M. considers this coil to be located on the qaxis.
(C)
Finally come 2 coil cards representing the 2 excitation (field) coils which
is considered by the U.M. to be located on the daxis. The first of these
two coil cards specifies a shunt coil and the second a series coil. If no
shunt coil or no series coil is used, then the corresponding coil card is to
be left blank completely.
10 - 77
Sources
10.12.16
Class 1 U.M. Data Cards
This class contains two cards, containing information pertaining to all U.M. machines to be used.
This class has to be ended by a blank termination card.
Card (1):
indicates the request for the use of a U.M. model.
1
2
3
4
5
6
7
8
12 345678901234567890123456789012345678901234567890123456789012345678901234567890
19
I2
Card (2):
general specification
INPU
INITUM
ICOMP
1
2
3
4
5
6
7
8
1 2 345678901234 5 67890123456789012345678901234567890123456789012345678901234567890
I1 I1
I1
INPU
(1)
INITU
M
the SI units are assumed (V, A, W, s, Nm, m, etc.). If 1 is

specified, then the per unit system is to be used as described in
Reference 1, pp. 92-98, pp. 414418. If 2 is specified, then SI
units are used and option for residual magnetism is present.
Enter 1 to request automatic steady-state initialization.
Otherwise leave blank.
(2)
ICOMP
=0 or blank
compensation is used for interfacing U.M. with

the external electric network.
=1
Prediction instead of compensation is used for

the interface between U.M. and network. By
using the prediction option, an arbitrary number
of machines can be connected on the power side
to the same network.
(15)
10 - 78
Sources
10.12.17
Class 2 U.M. Data Cards (Machine-Table)
Class 2 of U.M. data cards specify general information about each U.M. model to be used (this is
why the cards of this class are called machine cards as opposed to coil cards which form the Class
3 U.M. data cards). This class of data cards consist of three cards. Four cards are required if
automatic steady-state initialization is requested.
Card (1) of Class 2 data cards:
MECH
NODE
NAME
TACS
NAME
A6
A6
I2 I2 I2 I1 I1 I1
JTYPE
NPPAIIR
TQOUT
OMOUT
THOUT
NCLQ
NCLD
JTYPE
1
2
3
4
5
6
7
8
12 34 56 7 8 9 012345 678901 23 45678901234567 89012345678901 23456789012345 678901234567890
RJ
DCOEF
EPSOM
FREQ
I2
E14.5
E14.5
E14.5
E14.5
The U.M. type number desired to be used.
(1-2)
NCLD
Number of daxis excitation coils.
(3-4)
NCLQ
Number of qaxis excitation coils.
(5-6)
TQOUT
=0
no output request
=1
torque output is requested
=2
torque and daxis main flux output are requested
=3
torque, daxis main flux and daxis magnetizing

current are requested
(7)
Request for output of rotor speed [rad/sec]
OMOUT
(8)
=0
no output request
=1
speed is requested
=2
speed and q-axis main flux are requested
10 - 79
Sources
=3
Request for output of angle indicating rotor

position with respect to stator [rad]. For JTYPE = 1
and 2, this angle becomes the torque angle.
THOUT
(9)
MECH.
NODE
NAME
(10-15)
=0
no output
=1
output requested
Name of the node to which the U.M. mass (= capacitor) is

connected if mechanical network option is used, i.e.,
mechanical shaft system is represented by its electrical
network analog. If this network option is not used, then
leave this space blank.
Note:
TACS
(1621)
The U.M. sign convention corresponds with the

generator convention. Thus for generator operation,
a positive mechanical input is required.
Name of the TACS variable which forms the mechanical

input torque to the U.M., when the mechanical system is not
simulated by an electrical network representation. Note that
the U.M. sign convention is based on the generator
convention (see above).
Note:
NPPAIR
speed, q-axis main flux and q-axis magnetizing

current are requested
Leave this space blank if the mechanical network

option is used.
Number of pole pairs.
(2223)
RJ
(2437)
DCOEF
(38-51)
EPSOM
(5265)
FREQ
(6679)
Rotor moment of inertia in [Nms2/rad] or in per unit.

Leave blank, if mechanical network option is used.
Damping coefficient in [Nms/rad] or in per unit. Leave
blank if mechanical network option is used.
Convergence margin for the rotor speed iteration process. If
left blank, then by default EPSOM = 0.01% of synchronous
speed or rated speed for dc machines.
Specify the steady-state frequency of the network to which
the power coils of the machines are connected. If left blank,
a default value = STATFR (defined in the module "SYSDEP")
will be used.
10 - 80
Sources
OMEGM
LMUD
E14.5
E14.5
OMEGM
(1-14)
LMUD
(1528)
JSATD
(29)
JSATD
1
2
3
4
5
6
7
8
12345678901234 56789012345678 9 01234567890123 45678901234567 89012345678901 234567890
LMSD
FLXSD
FLXRD
I1
E14.5
E14.5
E14.5
Initial condition of the mechanical speed in [rad/sec] or

in per unit.
Unsaturated d-axis main inductances in [H] or in per
unit.
Request for implementation of daxis saturation if JSATD
is set to 1. Leave blank if no daxis saturation is desired.
Set to 5 if total saturation is desired.
Remark: The quantities following JSATD on the data

card can be all left blank if no daxis
saturation is desired. Otherwise these
quantities pertain to specifications of the d
axis saturation curve as discussed in the
Section 10.12.5.
LMSD
(3043)
FLXSD
(4457)
FLXRD
(5871)
Saturated daxis main inductance in [H] or in per unit.

Leave blank if no daxis saturation is desired. Otherwise
see REMARK pertaining to JSATD and Figure 10.22
(Section 10.12.5).
Leave blank if no daxis saturation is desired.
Otherwise, see Remark pertaining to JSATD and
Figure 10.22 (Section 10.12.5).
Leave blank if option for residual flux was not requested.
Otherwise, enter the desired value of the daxis residual
flux, or the residual flux for total saturation.
10 - 81
Sources
THETAM
LMUQ
E14.5
E14.5
JSATQ
1
2
3
4
5
6
7
8
12345678901234 56789012345678 9 01234567890123 45678901234567 89012345678901 234567890
LMSQ
FLXSQ
FLXRQ
I1
E14.5
E14.5
E14.5
(114)
Initial rotor position with respect to stator in

mechanical [rad]. For synchronous machines, i.e., JTYPE
= 1 and 2, this angle is the torque angle in electrical [rad].
LMUQ
Unsaturated qaxis main inductance in [H] or in per unit.
THETAM
(1528)
JSATQ,
LMSQ,
FLXSQ,
(2957)
FLXRQ
(5871)
All these quantities can be left blank, if total saturation or

no qaxis saturation effect is desired. Otherwise these
qaxis saturation quantities are completely analogous
with those of the daxis, i.e., JSATD, LMSD, FLXSD.
Please consult card (2) where these quantities are
documented.
Enter the desired value of the qaxis residual flux if
residual flux option was requested, leave blank
otherwise. Leave also blank if total saturation option
was requested.

This card is to be omitted if no automatic steady-state initialization was requested for (INITUM of
Class 1 data cards).
1
2
3
4
5
6
7
8
12345678901234 56789012345678 901234 567890 1234567890123456789012345678901234567890
AMPLUM
ANGLUM
BUSF
BUSM
E14.5
E14.5
A6
A6
AMPLUM
(114)
If synchronous or direct current machine, specify the

amplitude the terminal voltage of the machine on the
power side.
10 - 82
Sources
If induction or doubly-fed machine, specify the slip
(in %).
ANGLUM
(1528)
If a synchronous machine, specify the phase angle (in

degrees) of the terminal voltage of the machine on the
power side. For 3-phase machines, this angle is the
voltage angle for power coil A. Angles of power coils B
and C is taken respectively as lagging and leading with
120 degrees. For 2phase machines, this angle is the
voltage angle for power coil B, angle of power coil C is
taken as leading with 90 degrees.
If not a synchronous machine, leave this space blank.
BUSF
(2934)
If a synchronous or direct current machine, specify the

node name of the EMTP Type14 source that is desired to
be adjusted by the program in order to accommodate the
steady-state field current. This field coil is to be
identified with the first daxis excitation coil.
If an induction or doubly-fed machine, leave this space
blank.
BUSM
(3540)
Specify the node name of the EMTP Type14 source in

the mechanical network that is desired to be adjusted by
the program to accommodate the steady-state
electromagnetic torque. If more adjustable sources are
desired, then see card (5).
Card (5) and higher of Class 2 data cards:

These cards are to be omitted if card (4) is not used. Otherwise for each additional EMTP Type14
source of the mechanical network to be adjusted by the program to accommodate the total
electromagnetic torque, a card of the following format is required.
MORE
This is to be typed in columns 1-4.
(14)
BUSM
Same meaning as BUSM on card (4).
(3540)
DISTRF
(4154)
Enter the ratio of the value of this adjustable EMTP

Type14 source and the value of the adjustable source
BUSM as specified on card (4); also supply a negative
sign if of opposite polarity.
10 - 83
Sources
10.12.18
Class 3 U.M. Data Cards (Coil-Table)
The Class 2 U.M. data cards have to be followed immediately with the U.M. Class 3 data cards.
Each card of this class contains information on a single coil. Hence it is referred to as a "coil card".
The data format of a coil card is as follows:
RESIS
LLEAK
BUS1
BUS2
XTACS
CUROUT
1
2
3
4
5
6
7
8
12345678901234 56789012345678 901234 567890 123456 7 89012345678901 2345678901234567890
CUR
E14.5
E14.5
A6
A6
A6
I1
E14.5
(114)
Resistance in ohms or in per unit. Cannot be equal to

zero.
LLEAK
Leakage inductance in [H] or in per unit.
RESIS
(15-28)
BUS1
(2934)
BUS2
(3540)
XTACS
(4146)
CUROUT
(47)
CUR
(4861)
Node name of the Electrical Network to which one

terminal of the coil is connected. Leave blank if local
ground connection or if no connection with the Electrical
Network is desired.
Node name of the Electrical Network to which the other
terminal of the coil is connected. Leave blank if local
ground connection or if no connection with the Electrical
Network is desired.
Name of the TACS variable which is used as a voltage
source in series with the coil.
Request for output of the coil current if set to 1; leave
blank if no output is desired. If set to 2, then the power
coils are requested to provide their current output in the
Park domain, i.e., io, id, iq.
Initial coil current in [A] or in per unit. Leave blank if
automatic steady-state initialization is requested (see
INITUM of Class 1 data cards).
General Rules with respect to the coil cards:

1.
The following sign-convention has to be realized in assigning BUS1, BUS2,

XTACS and CUR.
10 - 84
Sources
icoil
U.M. coil
BUS1
Vcoil
XTACS
BUS2
Positive coil current means that it goes from BUS2 and BUS1.
The terminal coil voltage is:
vcoil = vBUS1 - BUS2 + vXTACS
2.
The coil cards specifying the coils of one machine follow immediately the
Class 2 U.M. data cards of the machine.
3.
The following rules apply to the sequence of coil cards belonging to one
machine:
(A)
First come three mandatory coil cards representing the power coils in the
sequence of coil A, coil B and coil C. If less than three power coils are
needed, insert blank cards for the unused coils.
(B)
Then come NCLD coil cards representing the daxis excitation coils, with
the value of NCLD as specified in the machine-table (see Class 2 u.m. data
cards).
Remark: Neither coil nor blank cards are to be inserted if NCLD = 0.

(C)
Finally come NCLQ coil cards representing the qaxis excitation coils,
with the value of NCLQ as specified in the machine-table.
Remark: Neither coil nor blank cards are to be inserted if NCLQ = 0.

4.
If the option for full compensation is requested, then all U.M. machines have to
be electrically separated from each other. The "stub line" strategy can be
applied. This restriction is not present if the prediction option is requested. (See
Class 1 u.m. data cards).
(A)
The power coils can all be connected to each other in one electrical
network, but cannot be connected to the excitation coils. Again, use stub
line or TACS element to avoid this restriction.
10 - 85
Sources
(B)
5.
10.12.19
Only the first three excitation coils can be connected to each other in one
electric network; if more than three excitation coils are used, then
excitation coil number 4 and higher must be completely disconnected
from each other. This restriction can be avoided by using a stub line or a
TACS element.
Provide a blank termination card behind the last coil card of the last machine,
indicating the termination of all U.M. data cards.
Output Labeling
All U.M. output are labelled by two names:

1.
Upper label indicates the machine number;
2.
Lower label indicates the corresponding variable.
Example:
UM-4
UM-4
TQGEN
IPA
The U.M. variables which can be requested for output are:
TQGEN
Electromechanical torque based on generator convention in per

unit or Nm. If motor, TQGEN is negative; if generator, TQGEN is
positive.
OMEGM
Mechanical rotor speed in rad/s.
THETAM
Mechanical rotor angle in rad; for Type 1 and 2 (synchronous

machine) this angle is the torque angle in rad.
IPA, IPB, IPC
Currents of power coils A, B and C in per unit or A.
Note: For all types of direct current machines, the power

(armature) coil current is indicated by IPC.
For U.M. Type-6 and 7 (= single phase ac machines; see their
User's Guide), these labels indicate excitation currents.
IE1,IE2, etc.
Currents of the excitation coils in per unit or A.
Note: For U.M. Type-4 (induction machine, 3-phase stator and

rotor), the 3-phase rotor currents in coils B, C, and A are
respectively indicated by IE1, IE2, IE3.
10 - 86
Sources
For U.M. Type-6 and 7 (single phase ac machines), these labels
indicate the power coil currents.
IPO, IPD,
IPQ
Currents of the power coils in the Park domain.
FLUXMD,
FLUXMQ
Magnetic fluxes in the d and qaxes, respectively.
IMD, IMQ
Magnetizing currents in the d and qaxes, respectively.
10.12.20
Multi-Machine Systems Sharing Common Mechanical Network
Up to three U.M. machines of possibly different types connected to a single shaft system. It can
be applied to a general network of mechanical components, arbitrary in configuration as well as in
size. Such an example is a wind turbine for which the blades and the hub are critical mechanical
components besides the different parts of the turbine shaft.
The option is only honoured if the automatic U.M. steady-state initialization feature is requested
with the U.M. data input. This feature is internally called for if the U.M. is supplied with the Type
59 S.M. data input, for which this mechanical network sharing option may also be requested.
The theory behind the chosen approach is a rather straightforward extension of the method of
handling the mechanical system of a single U.M. (see reference 14). The method is extended to a
three-phase compensation approach, rather than a single-phase one, in interfacing the U.M.
electromagnetic coupling field equations with the mechanical network representation. In addition,
linear prediction of the electromagnetic torque production by each U.M. is included for the purpose
of determining the "Thevenin speed" as affected by the other U.M.'s which share the mechanical
network. This linear prediction can be expected to be reliable considering its intended use for
calculating the speed. Relatively smooth behaviour of machines speeds is a reasonable
assumption due to the relatively higher time-constants of the mechanical network components as
compared to those of the electrical network. Extensive experimentation with this new option has
confirmed the desired stability of the solutions. In fact, a one time-step delay rather than linear
prediction was observed to be sufficient in most practical cases.
Usage Rules:
1.
One card, referred to as "share"-card with a format to be described later in rule 5,

is needed to be inserted with the data input of each U.M. which belongs to a set of
up to three U.M.'s sharing a common mechanical network. An arbitrary number
of these sets can be included in a single data case. The U.M.'s which are not
provided with a share-card will be considered to have as usual only one U.M.
connected to the specified mechanical system. Any U.M. specified in the data
10 - 87
Sources
case, whether or not it belongs to a set of U.M.'s sharing a common mechanical
network, may be connected electrically to each other.
2.
Insert the share-card as follows: for U.M.'s with pure U.M. data input the sharecard is to precede card (4) of Class 2 U.M. data cards; for U.M.'s with Type-59
S.M. data input the share-card is to precede the Class 4 Type59 S.M. data cards
(mass cards).
3.
In order to distinguish one U.M. from the other, the EMTP logic assigns a number
to each U.M. This number is in accordance with the sequence as chosen by the
user in specifying the data input of the U.M.'s included in the data case.
4.
The U.M. numbering system as explained in the previous rule is important with
respect to the following rules regarding the mandatory structure of each set of
U.M.'s sharing a common mechanical network:
(A)
It is not required to specify the data input of the U.M.'s belonging to a set
sharing a common mechanical network in subsequent order.
For instance, if 10 U.M.'s are included to a certain data case and three
out of these 10 U.M.'s form a set which shares a common mechanical
network, then the order of specifying those three U.M.'s can be chosen
to have them become U.M. number 2, 4, and 8. However, if one of the
U.M.'s in this set is a U.M. provided with Type59 S.M. data input, then
the data input of this U.M. has to be supplied in the data case such that
it becomes the lowest numbered U.M. of the U.M.'s which belong to the
set. Thus in terms of the example, this U.M. with the Type59 S.M. data
input has to become U.M. number 2.
(B)
The relevant information about the mechanical network shared by a set of

U.M.'s is required to be supplied to the lowest numbered U.M. in this set.
For the case of pure U.M. data input, this information contains the bus
names of nodes to which the external torques are applied, i.e., card (4) and
card (5) of Class 2 U.M. data cards. For the U.M. with S.M. Type59 data
input, this information is with respect to the output request and the
structure of the entire mechanical network, i.e., Class 4 S.M. data cards
(mass cards).
(C)
The information referred to in the previous rule (4B) is to be omitted in

the input of the other U.M.'s in the set which are not chosen to be the
lowest numbered U.M. This rule is realized by taking the following steps
with respect to the data input of these U.M.'s (U.M. number 4 and 8 of the
example in rule 4A):
(i)
For U.M.'s with pure U.M. data input: remove all specifications
regarding the external torques applied to the mechanical system by
leaving blank the bus name "BUSM" on card (4) and omitting
10 - 88
Sources
completely all cards (5) of the Class 2 U.M. data cards.
(ii) For U.M.'s with Type59 S.M. data input: remove all Class 4 Type
59 S.M. data cards (mass cards).
Note:
5.
An important consequence of these rules is that if one of the U.M.'s in

the set of U.M.'s sharing a common mechanical network is a U.M.
provided with Type59 S.M. data input, then the structure of this
mechanical network will be restricted to correspond with the structure
of the Type59 S.M. mechanical system.
Format of the share-card:
1
2
3
4
5
6
7
8
12345 6 789012 345678 90123456789012345678901234567890123456789012345678901234567890
SHARE
NUM2
NUM3
A5
I6
I6
SHARE
(15)
NUM2
(712)
NUM3
(13-18)
10.12.21
Enter "SHARE" to indicate that the current U.M. will

share the data input of its mechanical network with that
of the universal machines indicated in fields NUM2 and
NUM3.
This is the number of another U.M. sharing the
mechanical network.
Number of a third U.M. sharing the mechanical network.
Leave blank if only two U.M.s share the mechanical
network.
Specific Instructions for Type-40 Induction Machine
This subsection contains specific user instructions for modelling of induction machine of Type-40.
A)
Data Applicable to the Problem as a Whole (Class I data cards):
The U.M. tables are stored within the total allocation of List 25 of EMTP variable dimensioning.
Make sure that the EMTP has been dimensioned with non-unity storage for this working area.
10 - 89
Sources
The first data card belonging to the U.M. is a request for the allocation of List 25 storage among
U.M.'s four different tables. This card must precede the miscellaneous data cards, and should
follow the BEGIN NEW DATA CASE card that begins the data case. Within those limits, ordering is
immaterial (all such special request cards at the front can come in any order). The card with free
format is (see Section 4.1.6 for more details):
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
ABSOLUTE U.M. DIMENSIONS, I, J, K, L
where "I", "J", "K", and "L" are symbols for explicitly-entered integers with the following
meaning:
I
maximum total number of coils (for all machines)
maximum number of machines
maximum number of U.M. outputs (total for all machines)
maximum number of U.M. 6character names (total for all machines)
All machine data is grouped together, and should follow the last conventional (non-dynamic)
EMTP source card, just as is the case with the Type59 S.M. modelling. The beginning of U.M. data
is signalled by two data cards:
Card 1:
Enter "19" in columns 1-2, as a special request for U.M.

modelling. The rest of the card is left blank.
Card 2:
Column 15 with a flag which indicates whether or not a

compensation based interface is to be used:
=1
prediction based interface
=0 or blank
compensation based interface
There is nothing else on this second card.

Card 3:
B)
Blank card to indicate the end of the general specifications

for all U.M.'s to be used.
Data Cards for General Machine Specification:
Next (after Card 3 of Section A) come data cards specifying the machine type, stator connections,
etc. There are three such cards for each Type40 induction machine.
10 - 90
Sources
Card #1:
JTYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 12345678901234567890123456789012345678901234567890
BUS
SLIP
FREQ
I2
A6
E10.6
E10.6
JTYPE
(1-2)
BUS
(38)
Type code. Enter "40" for 3phase induction machine

with 3phase rotor using nameplate data.
Six character EMTP network node name to which phase
"a" of the machine stator is connected.
(1120)
Initial operating slip of the induction machine in percent

(e.g., 100.0 for standstill; 0.0 for synchronous speed).
FREQ
Frequency of the network connected to the power side.
SLIP
(2130)
Cards #2 and 3:
These two identical cards identify the connection of phases b and c of the stator.
BUS
(38)
C)

"b" or "c" of the machine stator is connected.
Optional Data Cards for Design Ratio:
This optional data card specifies induction machine design ratio.
1
2
3
4
5
6
7
8
12345678901234567890 12345678 9012345678901234567890123456789012345678901234567890
DESIGN RATIO
A20
FMIN
(21-28)
FMIN
E8.0
Design ratio of the induction machine. This value is

around 1.0 for double-cage induction motors and 0.5-0.6
for deep-bar motors. The default value assigned by the
EMTP is 1.0.
This optional card must be used with the keyword "DESIGN RATIO" in columns 1-12.
10 - 91
Sources
D)
Data Cards with Nameplate Ratings:
This grouping of data cards contains the information on the nameplate rating of the machine and
comprises of three cards.
Card #1: Full Load Rating:

1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 123456789012345678901234567890
PFLD
VFLD
PFFLD
EFFLD
SFLD
E10.0
E10.0
E10.0
E10.0
E10.0
PFLD
(4-10)
Full load Volt-Ampere rating of the induction motor; in

VA (alternatively machine rating may be specified in
Horse Power (HP) in which case PFLD must be entered as
a negative number).
(1120)
Full load line-line voltage rating of the machine; in volts

rms.
PFFLD
Full load power factor of the machine.
VFLD
(2130)
EFFLD
Full load efficiency of the machine.
(3140)
Full load fractional slip of the machine.
SFLD
(4150)
Card #2: Starting Information:

1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
CSTA
TRAT
VRED
CRED
E10.0
E10.0
E10.0
E10.0
CSTA
(1-10)
TRAT
(11-20)
Starting current at rated voltage; in per unit of the full

load rated current.
Starting torque at rated voltage; in per unit of the full load
rated torque.
10 - 92
Sources
VRED
(21-30)
CRED
(31-40)
Reduced voltage at which another starting test has been

performed in per unit of the rated voltage. Default value
is 0.8 per unit.
Starting current at the reduced voltage VRED; in per unit
of the full load rated current. Default value is 0.78 CSTA.
Card #3: Leakage Saturation Data:

1
2
3
4
5
6
7
8
1234567890 1234567890 123456789012345678901234567890123456789012345678901234567890
CSAT
TQMAX
E10.0
E10.0
CSAT
(1-10)
TQMAX
(11-20)
Saturation threshold current for the leakage reactance;

in per unit of the full load rated current. Typically CSAT
is around 2 per unit.
Induction motor maximum (breakdown) torque; in per
unit of full load torque.
A negative value for CSAT or TQMAX signals that the
unsaturated values of leakage reactances will be used for
the entire EMTP solution. This makes the solution more
efficient and will be useful for cases where induction
machine does not enter saturation in the simulation
period.
These two variables are not independent from one
another and hence, only one of them may be specified. If
the value of CSAT is specified, the value of TQMAX will
be totally ignored. Also note that once machine
parameters are calculated in such a way to satisfy starting
and full load conditions the value of maximum torque
becomes more or less fixed and can only vary between
tight limits. Hence, if TQMAX is being specified instead
of CSAT, the user should confirm that the value is
reasonable.
E)
Optional Saturation Data Card:
This optional data card contains information on magnetic saturation of the machine.
10 - 93
Sources
1
2
3
4
5
6
7
8
12345678 90 1234567890 1234567890 1234567890 1234567890 123456789012345678901234567890
MAGNETIC
VLIN
VFLAT
CLIN
CFLAT
A8
E10.0
E10.0
E10.0
E10.0
VLIN
(11-20)
Stator line-line voltage on the linear part of the stator no

load volt-ampere characteristic; in Vrms (usually
VLIN < 0.6 VFLD).
(21-30)
Stator line-line voltage on the flat (saturated) part of the

stator no load volt-ampere characteristic; in Volts rms
(usually: 1.1 VFLD < VLFAT < 1.3 VFLD).
CLIN
Stator current corresponding to VLIN; Amps rms.
VFLAT
(3140)
CFLAT
Stator current corresponding to VLFAT; Amps rms.
(4150)
This optional data card must be used with the keyword "MAGNETIC" columns 1-8.
F)
Data Cards with Mechanical Parameters:
This grouping is dedicated to the specification of the mechanical parameters of the machine and
contains one card:
1
2
3
4
5
6
7
8
12 34567890 1234567890 123456 789012345678901234567890123456789012345678901234567890
NP
HINERT
BUS M
I2
E10.0
A6
NP
(12)
HINERT
(11-20)
Number of poles in the induction machine (not pole

pairs).
Inertia of the machine rotor; in kg-m2 (HINERT can be
specified in lb-ft2 if entered a negative number).
10 - 94
Sources
BUSM
(21=26)
G)
Six-character name assigned to the rotor mass. Note that

as the mechanical system of the machine is modelled by
its electrical analogue, the rotor mass name "BUSM"
becomes directly available to EMTP for connecting a
current source (load torque) of any waveform or other
electrical elements that are duals of mechanical
components (e.g., load torque of TLOAD = kw will
correspond to a resistance of R = 1/K connected between
the BUSM" and the ground).
Data Cards with Output Requests:
This grouping consists of one or two data cards.
Card #1:
1
2
3
4
5
6
7
8
12 34 56 78 90 12 34 567890123456789012345678901234567890123456789012345678901234567890
IP IS IR IE IN IA IQ
I2 I2 I2 I2 I2 I2 I2
IP
(1-2)
IS
Flag for printing of machine parameters in per unit (on

machine rating)
=1
print parameters
= 0 or blank
no parameters printout
Flag for printing machine stator currents in Amperes
(34)
IR
=1
print stator currents
= 0 or blank
no current printout
Flag for printing machine rotor currents in Amperes
(56)
IE
(78)
=1
print rotor currents
= 0 or blank
no current printout
Flag for printing machine electromagnetic torque in

Newton-meters
10 - 95
Sources
IV
(910)
=1
print torque
= 0 or blank
no torque printout
Flag for printing machine rotational velocity in radians/

second
=1
print velocity
= 0 or blank
no velocity printout
Flag for printing machine rotor angle in radians
IA
(1112)
IQ
(13-14)
=1
print rotor angle
= 0 or blank
no angle printout
Flag for printer plotting of the torque-slip curve of the

machine for comparison with the manufacturer's curve
=1
plot the torque-slip curve
= 0 or blank
no curve plotting
Card #2 (optional):
This data card must be present only when IQ=1 was specified on the previous card. This card
specifies the voltages (in per unit) for which the torque-slip curve is plotted.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
VLTQ1
VLTQ2
VLTQ3
VLTQ4
VLTQ5
VLTQ6
VLTQ7
VLTQ8
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
VLTQi
Voltage at which the torque-slip curve is to be plotted; in per unit of full

load voltage rating. Up to 8 voltages can be specified.
The data cards for additional UM models must immediately follow the last grouping of data cards
of the induction machine. A blank card signals the end of U.M. data cards. Then comes the blank
card ending EMTP sources, and the rest of the data case.
10.12.22
Specific Instructions for Modelling of Synchronous

Machines of Type59
This subsection contains specific user instruction for the modelling of synchronous machines.
10 - 96
Sources
If the user has existing Type59 S.M. data cards (Section 10.11) these can be converted easily so
as to be solved by the U.M. code. The user needs only to study sections marked with "**" in the
beginning of a section.
A)
Data Applicable to all U.M. Machines:
All machine data is grouped together, and should follow the last conventional (non-dynamic)
EMTP source cards, just as is the case with the Type-59 S.M. modelling. The beginning of U.M.
data is signalled by two data cards:
Card 1:
Enter "19" in columns 1-2, as a special request for U.M. modelling. The
rest of the card is left blank.
Card 2:
Enter column 15 with a flag which indicates whether or not a

compensation based interface is to be used:
=1
prediction based interface
=0 or blank
compensation based interface
There is nothing on this second card.

Card 3:
Blank card to indicate the end of the general specifications for all U.M.'s
to be used.
The above 3 cards are used only once for any particular data case.
B)
Data Cards for General Machine Specification:
Next (after Card 3 of Section A) come data cards specifying the machine type, stator connections,
etc. There are three such cards for each synchronous machine.
Card #1:
JTYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
BUS
VOLT
FREQ
ANGLE
I2
A6
E10.6
E10.6
E10.6
JTYPE
Type code. Enter "59" for 3-phase synchronous machine.
(12)
10 - 97
Sources
BUS
(3-8)
VOLT
(11-20)

"a" of the machine stator is connected.
The steady-state voltage magnitude at the terminals of the
machine. This is peak voltage ( 2 times RMS), line to
neutral. (1/ 3 times line-to-line). Use units of [volts].
FREQ
(21-30)
ANGLE
(31-40)
The electrical frequency of the generator in Hz, for

steady-state operation.
The steady-state voltage phasor angle at the terminals of
the machine, for phase "a". Units are degrees. This
machine component, all other ones, and all sinusoidal
voltage or current sources of Type-14 are assumed to all
have the same reference for phase angles, of course.
Conventional positive sequence is assumed, as this term
is used in North America. That is, phase "b" voltage lags
phase "a" voltage by 120 degrees (and "c" leads "a" by
120 degrees).
Cards #2 and #3: In Class 1 are used to specify the connection of phases "b" and "c" of the
machine. The format for those records is identical to that of Card #1 except that only the data field
"BUS" (columns 3-8) is used:
BUS
(3-8)
Note:

"b" ("c") of the machine stator is connected.
The program assumes the presence of a 3-phase balanced
source at the synchronous machine terminal.
Class 2:
The class of data cards is used for special-request cards (if any). All the special request cards used
with Type59 S.M. can be used but only one is partly honoured.
The "PARAMETER FITTING" keyword (columns 1-17) is only used to distinguish between different
options of specifying machine electrical data. If this record is used, U.M. expects standard
manufacturer's data and uses the conventional data conversion methods (used in most transient
stability programs) to derive the electrical parameters. In the absence of this special request card,
the user must directly supply machine electrical parameters (self and mutual reactances, winding
resistances).
10 - 98
Sources
Class 3A:
Class 3A consists of 2 data cards which specify the general parameters of the synchronous
machine:
Card #1:
Ratings
1
2
3
4
5
6
7
8
12 34 56 7890 12345678901234567890 1234567890 1234567890 1234567890 1234567890 1234567890
NM KM KE NP
RMVA
RKJ
AGLINE
S1
S2
I2 I2 I2 I4
E10.6
E10.6
E10.6
E10.6
E10.6
NM
(1-2)
KM
(3-4)
KE
(5-6)
NP
(7-10)
RMVA
(31-40)
RKV
(41-50)
AGLINE
(51-60)
S1
(61-70)
The number of connected masses on the shaft system of

this generator.
The mass number which corresponds to the generator (or
motor) rotor, within the interconnected mass-spring shaft
system. Masses are to be numbered by the user for
identification, beginning with number one on either end,
and continuing sequentially (2,3,...) to the other end of
the shaft. See Figure 10.18 (Section 10.11).
The mass number which corresponds to the exciter on the
shaft system. If no exciter exists, leave this field blank.
The number of poles (not pole pairs) which characterize
the machine rotor.
The total 3-phase volt-ampere rating of the machine, in
units of MVA.
The rated line-to-line voltage of the machine, in units of
RMS kV.
rated armature voltage (1.0 per unit) on the air gap line.
This is an indirect specification of the mutual inductance
between the field and the armature of the machine. If the
S.M. is saturable, append an extra minus sign (making
"AGLINE" negative) as a flag.
rated armature voltage on the no-load saturation curve.
This field, as well as "S2", can be left blank if the S.M. is
not saturable.
10 - 99
Sources
S2
(7180)
Like "S1", only for 1.2 per unit of rated voltage.
The parameters "AGLINE", "S1", and "S2" just inputted actually apply to the daxis of the machine.
A second card follows (as an extension of) to provide for the following parameters:
Card #2:
Q-axis saturation data
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890 1234567890 1234567890 1234567890
AGLQ
(51-60)
Class 3B:
AGLQ
S1Q
S2Q
E10.6
E10.6
E10.6
Identical to AGLINE, except for the q-axis. For cases where

the user does not have the q-axis saturation data the following
special values for AGLQ can be used:
= 1.0
q-axis is saturable, with saturation

characteristics internally generated internally by
the U.M. similar to that of the d-axis. The
saturation of both axis will be based on the
machine overall MMF.
= 0.0
(or record 2 missing); q-axis is not saturable and

d-axis saturates (if saturable) based on the total
MMF in the daxis alone.
= 5.0
q-axis is not saturable and the d-axis saturates

(if saturable) based on the machine overall
MMF.
= 5.0
q-axis is also saturable (U.M. generates

internally the saturation characteristics),
however, each axis saturates based on the total
MMF in that axis alone (decoupled saturation of
the d- and qaxis).
S1Q, S2Q:
Identical to S1 and S2 of Card #1, except for the

q-axis.
Machine Electric Parameters
The form of the data cards of Class 3B data depends upon the presence of the keyword
"PARAMETER FITTING" (Class-2 data).
10 - 100
Sources
Case A:
Standard manufacturer-supplied data (PARAMETER FITTING keyword

used):
Card #1:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Ra
Xl
Xd
Xq
Xd
Xq
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
X"d
E10.6
Ra
Armature resistance, in per unit. This must be positive.
Xl
Armature leakage reactance, in per unit. If unknown, use

0.95 times the smallest of X"d, X"q, and Xo.
Xd
Direct-axis (d-axis) synchronous reactance, in per unit.
Xq
Quadrature-axis (q-axis) synchronous reactance, in per

unit.
X'd
Direct-axis (d-axis) transient reactance, in per unit.
X'q
Quadrature-axis (q-axis) transient reactance, in per unit.
X"d
Direct-axis (d-axis) subtransient reactance, in per unit.
X"q
Quadrature-axis (q-axis) subtransient reactance, in per

unit.
X"q
E10.6
Card #2:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Tdo
Tqo
E10.6
E10.6
T"do
T"qo
Xo
Rn
Xn
E10.6
E10.6
E10.6
E10.6
E10.6
T'do
Direct-axis (d-axis) open-circuit transient time

constant, in seconds
T'qo
Quadrature-axis (q-axis) open-circuit transient time

T"do
Direct-axis (d-axis) open-circuit subtransient time

10 - 101
Sources
T"qo
Quadrature-axis (q-axis) open-circuit subtransient time

Xo

it is suggested that the value of Xl/0.95 be used.
Rn
The neutral grounding resistance, in per unit.
Xn
The neutral grounding reactance, in per unit.
Case B:
Machine electrical parameters (PARAMETER FITTING request not used):
Card #1:
d- axis reactances.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
Xf
E10.6
Xa f
Xf k d
Xd
Xa k d
Xk d
E10.6
E10.6
E10.6
E10.6
E10.6
Xf
(1-10)
The self reactance of the field winding, in per unit.
Xaf
(11-20)

and the field winding, in per unit.
Xfkd
(21-30)
The mutual reactance coefficient between the field

winding and the direct axis damper winding, in per unit.
Xd
(31-40)
The self reactance coefficient for the direct axis (d-axis)

of the armature, in per unit.
Xakd
(41-50)

and the direct axis damper winding, in per unit.
Xkd
(51-60)
The self reactance of the direct axis damper winding,

in per unit.
Card #2:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
Xq
Xa g
Xg k q
Xq
Xa kq
Xkq
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
10 - 102
Sources
Xq
(1-10)
The self reactance of q-axis damper winding, in per unit.
Xag
(11-20)

and the qaxis damper winding, in per unit.
Xgkq
(21-30)
The mutual reactance coefficient between the two

circuits on the quadrature axis (q-axis) of the rotor, in per
unit.
Xq
(31-40)
The self reactance coefficient for the quadrature axis (qaxis) of the armature, in per unit.
Xakq
(41-50)

and the eddy-current winding, in per unit.
Xkq
(51-60)
The self reactance of the eddy-circuit winding, in per

unit.
Card #3:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
Xo
Ra
Rf
Rkd
Rg
Rkq
Rn
Xn
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
Xo
(1-10)
Zero-sequence reactance, in per unit. If this unknown, it

is suggested that the value of Xl/0.95 be used.
Ra
(11-20)
Armature resistance, in per unit.
Rf
(21-30)
Resistance of the field winding, in per unit.
Rkd
(31-40)
Resistance of the damper winding on the direct-axis (daxis) of the rotor, in per unit.
Rg
(41-50)
Resistance of the q-axis damper winding, in per unit.
Rkq
(51-60)
Resistance of the eddy-current winding, in per unit.
Rn
(61-70)
The neutral grounding resistance, in per unit.
10 - 103
Sources
Xn
(71-80)
The neutral grounding reactance, in per unit.
Class 4:
Consists of NM (NM: number of rotor masses on machine shaft from Card #1 of Class 3A) records
specifying the mechanical parameters of the shaft system. There is to be one record for each mass
in the following format:
1
2
3
4
5
6
7
8
12 34567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 123456 7890
ML
EXTRS
HICO
DSR
DSM
HSP
DSD
BUS
I2
E10.6
E10.6
E10.6
E10.6
E10.6
E10.6
A6
ML
(1-2)
EXTRS
(11-20)
The mass number. Masses are to be numbered by the

user beginning with number one on either end of the shaft
system, and continuing sequentially (2, 3, ...) to the other
end. See Figure 10.12 (Section 10.11).
Specified in columns 11-20 is to be the fraction of the
total external mechanical torque (power) which is
associated with this mass:
Generator: For a generator, enter the fraction of the

total external mechanical torque of this shaft system
which is applied to this particular mass. If a pump, enter
a negative value.
Motor: For a motor, enter the fraction of the total

external mechanical torque of the shaft system which is
developed by this particular mass. If a pump, enter a
positive value.
HICO
(21-30)
DSR
(31-40)
The moment of inertia (WR2) of mass number "ML".

This is to be in units of 106 lb-ft/(rad/s2)[(million poundfeet)/(rad/ sec**2)].
The speed-deviation self-damping coefficient for mass
number "ML". By definition, Ti = DSR (i - s) where Ti
is the particular damping torque under consideration for
mass number i=ML, i is the angular velocity of mass i,
and s is the synchronous mechanical velocity for this
shaft system. This is to be in units of lb-ft/(rad/s)
[(pound-feet) / (radians/seconds)].
10 - 104
Sources
DSM
(41-50)
The mutual-damping coefficient. This pertains to the

present mass (assumed to be number i=ML) and the mass
with the next higher number (i+1). The damping in
question is a function of the velocity difference between
the two masses:
T = DSM (i - i+1)
The mutual-damping coefficient is to be entered in units
of lb-ft/(rad/s) [(pound-feet) / (radians/second)].
HSP
(51-60)
DSD
(61-70)
The spring constant. This pertains to the elastic

connection between the present mass (assumed to be
number i=ML) and the mass with the next higher number
(i+1). Units are 106 lb-ft/rad [(million pound-feet) /
(radians)].
The absolute-speed self-damping coefficient for this
mass. By definition,
Ti = DSD i
where Ti is the associated damping torque on mass
number i, and i is the angular velocity (absolute) of this
mass. Units for this data are lb-ft/(rad/s) [(pound-feet) /
(radians/second)].
**BUS:
Six character EMTP node name assigned to mass ML. In

U.M., mechanical systems are actually modelled by their
electrical analogues which are, in turn, solved as part of
the electrical network in the EMTP. Hence, this name will
be assigned to the electrical node corresponding to mass
ML. The voltage of this electrical node will be the
mechanical speed of mass ML and this voltage must be
passed to TACS (Type-90 TACS source) if the mechanical
speed of mass ML is required for processing in TACS.
The current between the two nodes representing two
adjacent masses corresponds to the shaft torque between
those masses. This current should be used for postprocessing of the shaft torque in such cases as fatigue
analysis: This record must be repeated "NM" times. Also
note that:
1. Fields "DSM" and "HSP" contain data which really do
not belong exclusively to mass number i = ML. It
pertains equally to the mass i+1. For mass card of the
highest-numbered mass (number "NM"), then, these two
fields have no meaning, and are to be left blank.
10 - 105
Sources
2. EXTRS must be specified for turbine masses in multimass representation of the rotor.
3. Specification of node name BUS is highly
recommended.
Class 5:
Class 5 of the machine data cards specifies the output requests. Two different options for
requesting the output of machine variables are permitted:
**Option 1:
To get access to this old-style request, do not terminate the mass (Class 4 U.M.) data
cards with a blank. Following the last mass card, insert a card which is to be entered
according to the format described immediately below.
Only those synchronous machine variables which are explicitly requested for output
using this card will be a part of the EMTP output vector, which is used for both printing
and plotting. The request scheme, unless otherwise noted below, is:
"0"leave blank, or set to zero, if no such output is desired;
"1"enter unity for output in SI units.
JIDQO
JFI
JD2
JQ1
JQ2
JETM
JETE
JIABC
JSAT
1
2
3
4
5
6
7
8
12 3 4 5 6 7 8 9 0 1 2 34567890123456789012345678901234567890123456789012345678901234567890
I1 I1 I1 I1 I1
I1 I1 I1 I1
**JIDQO
(3)
Use of this column-3 to output armature (stator)

currents in Park's coordinates will prevent simultaneous
request for such output in phase coordinates (by means
of a column-11 entry for JIABC).
JFI
Controls printout of the current in the field winding of

the machine. The EMTP name which is used for output
identification is "IE1".
(4)
JD2
(5)
JQ1
(6)
Controls printout of the current in the damper winding of

the direct axis (d-axis) of the machine. The EMTP name
which is used for output identification is "IE2".
Controls printout of the current in the q-axis damper
winding. The EMTP name which is used for output
10 - 106
Sources
JQ2
(7)
JETM
(9)
**JETE
(10)
Controls printout of the current in the eddy-current

winding. The EMTP name which is used for output
Controls printout of the electromagnetic torque on the
machine rotor "TQGEN".
Input value of "1" to request the exciter-torque printout.
The Column10 entry is converted from exciter-torque
specification to a request for output of generator torque
angle in electrical degrees if a "3" is inputted. This angle
is equal to the relative generator mass angle times the
number of pole pairs. The output names used is
"THETAM". This torque angle is a relative angle with
respect to the synchronously rotating reference frame; for
no-load, it has a value of +90 degrees (as with the S.M.
code).
**JIABC
(11)
JSAT
(12)
This column11 request to print armature (stator)

currents in phase coordinates (Ia, Ib, Ic) is ignored if
armature currents in Park's coordinates (Id, Iq, Io) have
already been requested by nonzero column3 entry of
JIDQO. But if such output appears, it will be identified
using an extra "P" to signify the power side of the
machine. For example, assuming machine number 7:
UM-7
UM-7
IPA
IPB
UM-7
IPC
This column12 entry controls the saturation printout,

using direct and quadrature-axis components. Further it
is honoured only if the column9 specification for JETM
has been left blank (or zero). Again using machine
number 7 as an example, variable identification will be
as follows:
UM-7
UM-7
FLUXMD
FLUXMQ
= 1 output main fluxes in d- and q-axis.

= 2 output stator fluxes (main plus leakage) in d- and
q-axis (these values are printed with Type59
model)
= 3 output overall main flux (geometric sum of the dand qaxis main fluxes) and overall stator.
10 - 107
Sources
Flux (geometric sum of the d- and qaxis stator fluxes).
The variable names in all these cases remain the same
(FLXMD, FLXMQ).
For all the above cases, the total magnetization currents in both the d- and qaxes will also be
printed out (IMD, IMG)
**THETI
(13...)
Controls printout of the rotor speeds and shaft torques:

=1
absolute angular speed of mass "i" will be printed

rads/sec.
=2
shaft torque between masses "i" and "i+1" will be

printed in Newton-meters.
=3
Both angular speeds and shaft torques will be

printed.
Note that:
1.
Speeds will be printed as the voltage of the node name assigned to mass "i"
(variable "bus" of the Class 4 data records).
2.
Shaft torques will be printed as the current between the node names of the masses
across the shaft section.
It is not possible to directly output the angle of the mass, note. If angle is desired, the user must
first pass angular velocity into TACS, integrate this, and then output the result. Since the user
named the masses, this is simple: use a Type90 TACS source with the same A6 name (since node
voltage is the electrical analog of mechanical speed). Alternatively, the Class6 U.M. interface
capability below.
Note that U.M. speeds are in absolute radians/second, rather than relative radians/second as with
Type59 S.M.
Option 2:
In this case Class 5 consists of a minimum of one and a maximum of three data cards that must
follow Class 4 data records. A "BLANK" record must, however, separate the records of Class 4 and
Class 5 data.
All data cards in this data class are to be inputted according to the following format:
10 - 108
Sources
Group
All
1
2
3
4
5
6
7
8
12 3 4 5678 901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
N1
N2
N3
N4
N5
I1 I1
I6
I6
I6
I6
I6
GROUP
(3)
Flag identifying output group to which the variables specified

on this card belong.
ALL
Flag identifying a request for the output of all possible

variables in this output group:
(4)
N1-N12
= 0 or blank
selective specification of variables in

columns 9-80;
=1
all possible variables in this class are to be

outputted, columns 9-80 will not be scanned;
Identification of variables to be outputted.
The available grouping of the machine variables are presented in the following table:
Variable Ni
Group
No. Name
1. Electrical
Variable
IE
IE
IE
IE
IP
B
IPC
IP
A
8 9 10
11
12
13
14
15
TQGEN
OMEG
M
IMD
IMQ
FLUXM
D
FLUXM
Q
TQGEN
OMEG
M
IMD
IMQ
FLUXM
D
FLUXM
Q
TQGEN
OMEG
M
IMD
IMQ
FLUXM
D
FLUXM
Q
TQGEN
OMEG
M
IMD
IMQ
FLUXM
D
FLUXM
Q
TQGEN
OMEG
M
THETA
M
2. Mechanical
Speeds of
Mass
8 9 10
11
12
13
14
15
3. Torque
Across
Shaft
8 9 10
11
12
13
14
15
Table 3: Grouping of Output Variables
10 - 109
Sources
A "BLANK" record must terminate the output request records of Option 2. No such record will be
necessary if output request in Option 1 is used.
Notes on Output Request:

In general, the direct output requests for the output of the following S.M. parameters and variables
are not honoured by the U.M.:
1.
Machine electrical and mechanical parameters. These values may be recovered

from the diagnostic printouts with IPRSUP = 10.
2.
Initial conditions of the machine internal variables. These could be partly

recovered from the time-step printouts of t=0 sec.
3.
Machine rotor angles, which can be recovered by passing rotor speed to TACS
and integrating it.
4.
Machine field voltage, which can be recovered by passing this variable to TACS
(Class 6 data records) and having it printed as a TACS variable.
5.
Armature currents in both abc and Park's coordinates (only one set is allowed).
6.
Electrical Output requests 11, 12, 13 differ in the contents of FLUXMD and
FLUXMQ.
11 provides Main Flux in D and Q components.
12 provides Stator Flux in D and Q components.
13 provides the magnitude of Main and Stator Flux in FLUXMD and FLUXMQ
respectively.
7.
Electrical Output requests 11, 12, 13, and 14 interact, or are mutually exclusive
in that only the last request (farthest to the right) on the card or series of cards
will be honoured.
8.
The Torque Angle request, THETAM of Electrical Variable 14 will be honoured

only in the absence of any request for exciter torque of variable 15.
Class 6:
This class of machine data cards specifies U.M. interface with TACS. As many records as required
can be used in the following format:
1
2
3
4
5
6
7
8
12 345678 901234567890123456789012345678901234567890123456789012345678901234567890
KK BUS
I2
A6
10 - 110
Sources
The interpretation of variables "KK" (integer) and "BUS" (character) is as in the following table:
KK
71
INTERFACING
VARIABLE
BUS
TACS bus name whose
numerical value will become

the field voltage (not the field
voltage multiplier) of the
machine
72
TACS bus name whose
numerical value becomes the

multiplier for the machine
overall mechanical power
73
TACS bus name whose
numerical value becomes equal

to the machine field voltage
74
TACS bus name whose

to the machine field current
i
TACS bus name whose

to the absolute angle of the
mass "i"
NM+i
TACS bus name whose

speed of mass "i"
2NM+i
TACS bus name whose

shaft torque between masses
"i" and "i+1"
Passing exciter field voltage

from TACS to U.M.
Passing governors
mechanical power from
TACS to U.M.
Passing machine field
voltage from U.M. to TACS
Passing machine field
current from U.M. to TACS
Passing mass angles from
U.M. to TACS
Passing mass speeds from

U.M. to TACS
Passing shaft torques from
U.M. to TACS
Table 4: Summary of TACS U.M. Interfaces
Notes On Interface with TACS:

Transfer of mechanical variables from U.M. to TACS may be achieved directly without the use of
the Class 6 data records. The following alternative methods for establishing the interface with
TACS can be easily understood once it is realized that the mechanical system of the machine is
modelled by its electrical equivalent.
10 - 111
Sources
1.
Mechanical speed of each mass can be transferred to TACS as the voltage of the
node name assigned to that mass, according to Class 4 data records (Type-90;
TACS source).
2.
Shaft torque between two masses can be transferred to TACS as the switch current
between the node names assigned to those masses (Type-91 TACS source). This
switch is automatically created by the U.M. when the output of shaft torques are
requested.
3.
Absolute angle of each mass can be calculated in TACS from the integration of
the speed of that mass already transferred to TACS.
The end of the TACS interface cards must be signalled by the following data card:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
FINISH
C.
Ending:
The data cards U.M. model must immediately follow Class 6 data cards of the present machine. A
"BLANK" card signals the end of all U.M. data cards. Then comes the blank card ending EMTP
sources, and the rest of the data case.
10.13 EMTP Load Flow ("FIX SOURCE" Usage)

The steady-state phasor solution for initial conditions of the electric network can be accomplished
so as to observe power constraints at one or more buses. By analogy with the dominant steadystate program of system planning departments, this has been named the "load flow" feature of the
EMTP. In effect, it is multi-phase load flow capability (albeit in somewhat restricted form).
10.13.1
Background Information On EMTP Load Flow
Each single-phase node of a network involves four real variables of interest: real power injection
PK, reactive power injection QK, phasor voltage magnitude VK, and phasor voltage angle THETAK.
Since there is one real-power and one reactive power constraint equation for the node, this implies
that two of these variables can be specified arbitrarily, and the other two can then be solved for.
The user can choose among the following constraints at an EMTP source node:
(A)
Specify PK and QK; the EMTP solves for VK and THETAK. This is ordinary load
modelling of conventional, single-phase load flow usage.
10 - 112
Sources
(B)
Specify PK and VK; the EMTP solves for QK and the THETAK. This is ordinary
generator modelling of convention.
(C)
Specify THETAK and QK; the EMTP solves for PK and VK.
There is no limit on the number of buses which are so constrained, although at least one source
should be unconstrained. The unconstrained source is commonly taken as the reference for other
phasors (THETAK is fixed at zero). In the terminology of single-phase load flows, this is the "slack
bus".
For those who intend to terminate execution once the load flow is complete (i.e., for TMAX nonpositive), life is particularly simple. No EMTP source cards are required, and power constraints
can be applied to any node of the network. The only program limitation is that the source table
(List 4) must be sized to equal or exceed the total number of power-constrained nodes of the
network. The reason for this constraint is that one source is automatically defined internally for
each power-constraint card.
When a transient simulation is to follow the EMTP load flow (i.e., if TMAX is positive), the user
should understand that his power constraints will be forgotten once the time-step loop is entered.
With power constraints applied only to nodes with Type14 voltage sources, the angle and/or
magnitude of this source will be automatically adjusted prior to entry into the time-step loop -adjusted to correspond to the load flow solution. This may or may not result in approximate
continued observance of the steady-state power constraints during the transient simulation. The
result is a function of nonlinearities, and of unpredictable transients, which are quite beyond any
general analysis. The user should simply keep in mind that load flow constraints only provide for
the setting of constants (e.g., generator angles) at time zero, and have no other direct effect on any
subsequent simulation.
As presently implemented, power constraints are only allowed at nodes of known voltage, or at
the armature nodes of rotating AC machinery. Such constraints cannot be applied to current
sources. For cases without any associated transient simulation, this is not a restriction, since the
required voltage sources are defined internally, to whatever nodes the user wants to apply power
constraints. But for cases involving transient simulation, current sources are simply not allowed.
A power constraint of the EMTP load flow may only be applied at a network bus. It is not possible
to constrain a line flow, a group of line flows (the familiar area interchange control), directly.
Neither is there any automatic adjustment of transformer or phase shifter taps at the present time.
But certain limits on voltage magnitude and angle at a power constrained node can be observed.
First, at a load bus which normally would have PK and QK fixed, the reactive-power constraint will
only be maintained within user-specified limits on voltage VK, and the real-power constraint will
only be held within user-specified limits on the angle THETAK. Second, at a generator bus which
normally would have PK and VK fixed, the real-power constraint will only be held within userspecified limits on the angle THETAK. Third and finally, at a bus which normally would have
fixed QK and THETAK, the reactive-power constraint will only be held within user-specified limits
on voltage VK.
10 - 113
Sources
Thus far, there has been a mixture of talk about network buses and individual nodes. These may or
may not be synonymous, for purposes of EMTP load flow usage. The possible difference is due to
multi-phase constraints. The user can gang three nodes together, and control as a unit, if he likes.
In this case, all three voltages are assumed to be balanced, positive-sequence phasors. Any
specified real or reactive power is a total 3phase injection, which will be split equally among the
three phases only if the rest of the problem is balanced. But if either the network or some
excitation is unbalanced, note that the 3phase injections will not be balanced, either. In addition
to this conventional 3phase usage, it is possible to drop the third phase (phase "c", which lags
phase "a" by 240 degrees), so that only phases "a" and "b" are ganged together. This could be of
use to those studying outages (the loss of one phase).
For each network bus which is to have a power constraint, there is one EMTP data card which
specifies the associated local control parameters. Following the last such data card, there is an
extra card of overall control parameters, which shall be called the miscellaneous data card of the
load flow. This data structure is summarized as follows:
Card for first power constraint of load flow
Card for second power constraint of load flow
< < Etc. > >
Card for the last power constraint of load flow
Miscellaneous data card for load flow
One 3phase bus requires only one source card, if the user accepts the 3phase logic which this
implies. Alternatively, each phase could be controlled separately, thereby requiring three EMTP
data cards. In any case, taken together, all such cards constitute data peculiar to the EMTP load
flow -- data which affects no other aspect of the simulation. This data follows the blank card
ending sources (Section 3.1), and it precedes the initial condition cards for the electric network (if
any, see Section 11) and the node voltage output requests (Section 12).
EMTP load flow capability is flagged by the special-request card reading "FIX SOURCE". (see
Section 4.1.7). This request must precede all miscellaneous data cards.
Iterative convergence of the EMTP load flow is monitored on LUNIT6, where the largest correction
of each iteration is displayed, 20 numbers per line. There also is an option (see miscellaneous data
parameter NNOUT) for the parallel printing of the numbers of the power constraints which
produced these greatest corrections. Numbering of this identification corresponds one-for-one
with the user-inputted data cards. As an example of such output, consider the first eight columns
of the first two rows of twenty, which might appear as follows (taken from BENCHMARK DC-26):
vchang(k)
0.016
0.015
0.014
0.013
0.012
0.012
0.012
fix source
10 - 114
Sources
Note that convergence is slow and that the generator having the worst correction is always the
same (DC-26 involves just a single load flow source). If program output goes to the disk, then the
user will not see such output until later, so timing is irrelevant. At most 20 iterations will occur
between such outputs, thereby assuring the user that the computation is progressing. The user has
control over the frequency of this output (variable NFLOUT of the miscellaneous data card
described below).
It is important to check for convergence, since as presently implemented, the EMTP simulation will
continue whether the iteration has converged or not! Should the user-supplied control tolerances
not all be met, a warning message will be printed. But execution will not be stopped.
The solution at the power-constrained buses can be obtained with variable NPRINT of the
miscellaneous data card (the final data card of the load flow). Such output precedes the phasor
branch flows, with the following being a representative sample:
Row
Node
Name
A1
10
Voltage
Degrees
Real Power
Reactive
Power
0.113908E+02
-30.000
0.399846E+02
-0.451654E+01
B1
0.113908E+02
-150.000
0.399846E+02
-0.451654E+01
C1
0.113908E+02
90.000
0.399846E+02
-0.451654E+01
Yet there is really nothing special to see, since such phasor values have always been available via
the regular steady-state printout (e.g., by the display of injections which follow phasor branch
flows).
Examples of EMTP load flow usage can be found in standard test cases BENCHMARK DC25 and
DC26, which involve synchronous machines. Since the Type59 S.M. is represented by nothing
other than three Type14 sinusoidal sources in the steady-state, it is trivial to apply power
constraints to these using EMTP load flow capability. The U.M. is not quite so simple, due to its
generality; but it, too, is compatible with the EMTP load flow for induction and synchronous
modes
10.13.2
Format of Data Cards for EMTP Load Flow
First, there must be one data card for each power constraint. Ordering of these cards is arbitrary.
Each such data card is to be specified according to the following rules:
10 - 115
Sources
NEK
1
2
3
4
5
6
7
8
12 345678 901234 567890 1234567890123456 7890123456789012 34567890 12345678 901234 567890
BUS1
BUS2
BUS3
Pk or k
I2
A6
A6
A6
E16.0
NEK
(12)
BUS1
(3-8)
BUS2
(9-14)
BUS3
(1520)
PK OR
THETAK
(53-60)
VMAX
(61-68)
Vmax
min
max
E8.0
E8.0
E6.0
E6.0
Enter the one, two, or three names of network nodes which

are involved in this power constraint. For single-phase
usage. Leave BUS2 and BUS3 blank. For 3phase usage (the
other common case), name the three nodes in natural,
positive-sequence order (first "a", then "b", and finally "c").
A)
In case injected real power PK is to be constrained (for

NEK=0 or 1), enter the desired value in units of
power (units of voltage times units of current). For a
multi-phase constraint, this is the total of the two or
three individual injections at the bus of interest.
B)
For the remaining, less-common case having NEK=2,

enter the desired fixed angle THETAK in degrees).
A)
In case injected reactive power QK is to be constrained

(for NEK=0 or 2), enter the desired value in units of
power (units of voltage times units of current). For a
multi-phase constraint, this is the total of the two or
three individual injections at the bus of interest.
B)
In case voltage VK is to be constrained (for NEK=1),

enter the desired value in units of peak voltage.
(37-52)
VMIN
E16.0
Vmin
Type code of power constraint. Choose among:

0 to constrain PK and QK (unknown VK, THETAK);
1 to constrain PK and VK (unknown QK, THETAK);
2 to constrain QK and THETAK (unknown PK, VK).
(21-36)
QK OR
VK
Q k or V k
Enter the minimum voltage VK, for those cases where it is

not fixed. This is used for NEK=0 or 2, and the value is in
units of peak voltage. A blank or zero value means that no
minimum will be imposed.
Enter maximum voltage VK, etc. (see VMIN). A blank or zero
means that no maximum will be imposed (internally, VMAX
is set to infinity).
10 - 116
Sources
QMIN
(69-74)
QMAX
(75-80)
Enter the minimum angle, THETAK, for those cases where it

is not fixed. This is used for NEK=0 or 1, and the value is in
degrees. A blank or zero value means that no minimum will
be imposed (internally, QMIN is set to minus infinity).
Enter the maximum angle, THETAK, ... (see QMIN). A blank
or zero means that no maximum will be imposed (internally,
QMAX is set to infinity).
After the last such power constraint card, add the following miscellaneous data card which is
recognized by the leading blank field (columns 1-8):
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 5678901234567890
NNOUT
NITERA
NFLOUT
I8
I8
I8
NNOUT
(9-16)
NITERA
(17-24)
NFLOUT
(25-32)
NPRINT
(33-40)
NPRINT
I8
RALCHK
CFITEV
CFITEA
E8.0
E8.0
E8.0
This parameter is a flag for control of additional

interactive output during the iteration. A value of unity
will add variable identification in parallel with the
always-present interactive printout of the largest variable
correction at each iteration. The preceding explanation
of convergence monitoring illustrated such usage. But if
no such identification of the worst corrections is desired,
enter zero.
The limiting number of iterations for the load flow
solution. A non-positive entry will be given a default
value of 500.
The buffer size for accumulating the output which is
associated with the interactive convergence monitoring
(the worst correction at each iteration). Any non-positive
value, or any value in excess of 20, will be set equal to
the default value of 20. .
A flag which allows for the selection or suppression of
special tabular printout for the nodes with power
constraints. Enter unity to obtain such values (VK,
THETAK, PK, QK); zero or blank will suppress such
solution output.
10 - 117
Sources
RALCHK
(41-48)
CFITEV
(49-56)
CFITEA
(57-64)
A relative convergence tolerance which controls the

accuracy of the solution. For termination of the
iteration, all power mismatches must be less than
RALCHK times the largest scheduled power (either P or
Q) of the network. A non-positive value will be taken as
a request for the default value 0.01.
"Acceleration factor" associated with the conversion of a
change in reactive power to a change in voltage
magnitude. In the absence of knowledge about this
parameter, leave the field blank (which represents a
request for the default value of 0.2).
"Acceleration factor" associated with the conversion of a
change in real power to a change in voltage angle. In the
absence of knowledge about this parameter, leave the
field blank (which represents a request for the default
value of 2.5).
10.14 EMTP Load Flow ("FIX SOURCE1" Usage)

The steady-state phasor solution for initial conditions of the electric network can be accomplished
so as to observe power constraints at one or more buses. By analogy with the dominant steadystate program of system planning departments, this has been named the "load flow" feature of the
EMTP. In effect, is multi-phase load flow capability (albeit in somewhat restricted form).
10.14.1
Background Information
The primary intention of the implemented feature is to initialize three-phase electric networks
under balanced or unbalanced steady-state conditions for given power constraints at system buses.
The main differences between FIX SOURCE1 and FIX SOURCE are:
FIX SOURCE 1 converges more quickly, than FIX SOURCE, and
FIX SOURCE1 requires zero and positive sequence impedances as input.
Otherwise, both options are functionally similar. in FIX SOURCE 1, the user can choose from
among the following constraints at an EMTP load-flow bus:
(A)
To specify the bus three-phase real and reactive power injection (PK and QK,
respectively); the EMTP in this case will solve for the voltage magnitude (VK)
and the phase angle (THETAK) for each of the three phases. The bus is referred
to as a PQ bus.
10 - 118
Sources
(B)
To specify the bus three-phase real power injection (PK) and the amplitude of
the positive-sequence line-to-neutral voltage VK the EMTP will then solve for
QK and THETAK. This bus is referred to as a PV bus.
Upon completion of the load flow solution, for perfectly balanced networks, the calculated VK and
THETAK for the three phases will constitute balanced positive-sequence phasor sets. The specified
three-phase PK and QK will be equally divided among the three phases.
For unbalanced networks, the calculated voltage phasors will be unbalanced and the specified
three-phase PK and QK will be divided accordingly among the three phases. The magnitude of the
positive-sequence voltage component at the PV bus will, however, be close to its specified value.
Power constraints are allowed at buses connected to the Type-59 synchronous machine model and
any three-phase bus in the network is defined as the Type-80 bus. In addition to the power
constraints, users have to provide the negative- and zero-sequence impedances behind the PQ and
PV buses. It is assumed that the positive and negative sequence impedances are equal in value.
When a transient simulation is to follow the load flow, users should know that the power
constraints imposed on the Type-80 buses will not be respected once the time-step loop is entered.
The Type-80 buses will be represented by balanced Type-14 current sources in parallel with the
supplied sequence impedances.
The convergence is reasonably fast for well-behaved networks. However, there are cases when a
solution cannot be reached for given constraints. It is important that the load flow user checks the
printout, since as presently implemented, the EMTP simulation will continue whether the iteration
has converged or not.
10.14.2
Special Request Card:

To activate the three-phase load flow feature, the user must provide a special request card labelled
"FIX SOURCE1" before the miscellaneous data cards (see Section 4.1.7).
Data input sequence:

The data cards to use FIX SOURCE1 are ordered as follows:
1.
2.
3.
4.
all other source cards

Type-80 cards
cards containing power constraints on the Type-59 synchronous machine model
blank card to end the source cards
10 - 119
Sources
Type-80 card:
Two data cards are required for each Type-80 bus. A description of the required input for each
card is shown as follows:
Card 1:
ITYPE
1
2
3
4
5
6
7
8
12 345678 90 123456 78 901234 567890123456 789012345678 901234567890 123456789012 34567890
BUSA
I2
A6
BUSB
A6
BUSC
PK
QK
VK
THETAK
A6
E12.3
E12.3
E12.3
E12.3
enter "80" as an indication of the special source type.
ITYPE
(1-2)
Identification of the three-phase bus involved in the
power constraint. The three bus names should be in a
natural positive-sequence order (columns 3-8, 11-16, 1824, respectively).
BUSA
(3-8)
BUSB
(11-16)
BUSC
(19-24)
Three-phase real power injection (units of PK depend on
the units of voltage times the units of current).
PK
(25-36)
Three-phase reactive power injection (units of QK

depend on the units of voltage times the units of current).
QK
(37-48)
For PQ bus, the initial guess for the positive-sequence

line-to-neutral voltage amplitude (V or kV).
VK
(49-60)
For PV bus, the specified positive-sequence line-toneutral voltage amplitude (V or kV).

THETAK
(61-72)
The initial guess for the phase angle for phase A

(degrees).
Card 2:
IPV
1
2
3
4
5
6
7
8
12 3456789012 3456789012 3456789012 345678901 234567890123456789012345678901234567890
RO
XO
R2
X2
I2
F10.3
F10.3
F10.3
F10.3
10 - 120
Sources
Flag indicating the type of constraint on the Type-80 bus.
0 for PQ bus
1 for PV bus
IPV
(1-2)
R0,X0
(8-12)
(13-22)
R2,X2
(23-32)
(33-41)
Zero-sequence impedance of the network behind the

Type-80 bus (ohms or mH depending on XOPT).
Negative-sequence impedance of the network behind the
Type-80 bus (ohms or mH depending on XOPT).
Note: For an unknown zero-sequence impedance, set its

value equal to that of the negative-sequence impedance.
Bus Connected to a Type-59 Synchronous Machine:

Six data cards, with two cards for each phase, are required for the three-phase bus connected to a
Type-59 synchronous machine. The bus names should be in a natural positive-sequence order. A
description of the data input for each of these cards is shown below:
Card 1:
ITYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
BUS1
VK
FREQ
THETAK
I2
A6
E10.6
E10.6
E10.6
ITYPE
enter "59" as an identification of the special source type.
(1-2)
BUS1
Node name for phase A.
(3-8)
(11-20)
For PQ bus. The initial guess for the positive-sequence

line-to-neutral voltage amplitude (V or kV).
FREQ
Frequency of the system (Hz).
VK
(21-30)
THETAK
Initial guess of the phase angle for phase A (degrees).
(31-40)
10 - 121
Sources
Card 2:
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
BUS1
A6
PK
QK
E10.0
E10.0
BUS1
R
E10.0
PVBUS
E10.0
E10.0
Node name for phase A.
(3-8)
PK
Enter one third of the total real power injection.
(11-20)
QK
(21-30)
For a PQ bus, enter one third of the total reactive power

injection.
For a PV bus, enter the positive-sequence voltage
constraint.
R,X
Zero-sequence impedance behind the PV or PQ bus.
(31-40)
(41-50)
PVBUS
(51-60)
Flag indicating the type of the constraint on the bus:

0 for PQ bus
1 for PV bus
Card 3:
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
BUS1
A6
BUS1
(3-8)
Node name for phase B.
10 - 122
Sources
Card 4:
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
BUS1
PK
A6
E10.0
BUS1
QK
E10.0
R
E10.0
PVBUS
E10.0
E10.0
Node name for phase B.
(3-8)
PK
(11-20)
QK
(21-30)
Enter one third of the total three-phase real power

injection.
For a PQ bus, enter one third of the total three-phase
reactive power injection.
constraint.
R,X
Negative-sequence impedance behind the PV or PQ bus.
(31-40)
(41-50)
PVBUS
(51-60)
Flag indicating the type of the constraint on the bus.

0 for PQ bus
1 for PV bus
Card 5:
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
BUS1
A6
BUS1
Node name for phase C.
(3-8)
Card 6:
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
BUS1
PK
A6
E10.0
QK
PVBUS
E10.0
E10.0
10 - 123
Sources
BUS1
Node name for phase C.
(3-8)
PK
Enter one third of the total three-phase power injection.
(21-20)
QK
(21-30)
For a PQ bus, enter one third of the total three-phase

reactive power injection.
constraint.
PVBUS
(51-60)
Code indicating the type of the constraint on the bus:

0 for PQ bus
1 for PV bus
10 - 124
Section 11
11.1
General Remarks ................................................................................. 11-1
11.2
Warning On Incomplete Initial-Conditions Specification

Capability ............................................................................................ 11-3
11.3
General Rules ...................................................................................... 11-4
11.4
Format and Rules ................................................................................ 11-5

11.4.1
Cards for Node Voltages and Initial Condition

Steady-State Frequency .................................................. 11-5
11.4.2
Cards for Currents in Linear Branches .......................... 11-6
11.4.3
Cards for Currents in Nonlinear Inductance

Branches ......................................................................... 11-8
11.4.4
Examples ........................................................................ 11-9
Section 11
11.1
General Remarks
Initial conditions are automatically set to zero by the program if the case does not start from an AC
steady-state. They are automatically set to the correct initial values if the network is linear and
starts from an AC steady-state (source Type-14 with TSTART <0) or if the network contains
current-controlled DC voltage sources (see Section 10.6). See Section 8.1 and Section 8.6 for the
treatment of nonlinearities in the latter two cases.
Initial conditions which are neither all zero nor all linear AC steady-state (at one frequency) must
be supplied by the user. An exception to this rule is the "INITIALIZATION WITH HARMONICS" of
Section 4.1.1. Any initial conditions supplied by the user will override the respective zero or AC
steady-state initial conditions computed by the program. Typical applications of user-supplied
initial conditions are:
1.
Continuation of a transient simulation beyond TMAX of the first run.

In this case, have the program enter all voltages and currents at the very last
time step of the first run (IPUN > 0 on the second miscellaneous data card), and
use this data deck in the new runs as user-supplied initial conditions (the
program enters the values with the same format as described in this section).
This approach saves computer time if a number of studies are identical up to a
certain instant of time, but differ thereafter. The new runs do not require the
same t as the first run. Continuation works only if all of the following
conditions are met:
2.
(A)
the passive network has not been changed between the previous run and
the new run, including the order in which the branch cards are read in;
(B)
there are no branches in parallel with identical pairs of node names;
(C)
there are no branches with distributed parameters. Recall that the

program will not enter branch currents for any distributed parameter
branch. The user can, however, manually specify the steady-state (at one
frequency) current flow in such branches (see Rule 1 of Section 11.3).
New Cases.
If the user knows the initial conditions for a new case, he can supply them as a
part of the input. This could be used to model:
11 - 1

(A)
Linear systems with AC sources of different frequencies. A typical

situation where this arises is AC sources which have a fundamental
frequency component as well as harmonic components. In this case, use
the program repeatedly to get AC steady-state solutions for the various
frequencies, (e.g., first for the fundamental, then for the third harmonic,
etc.), and add up the values by hand to get the initial conditions (AC
steady-state output is in RMS values in the form VRMS ej therefore,
v ( O ) = V RMS 2 cos . This approach works only if:
(i)
the network does not contain lines with distributed parameters, and
(ii) if nonlinear and piecewise linear inductances do not saturate after

superposition of the individual solutions (otherwise, initial
conditions will be more or less inaccurate because superposition is
only valid in a linear network).
(B)
Trapped charge. In typical switching surge studies, the line to be

energized may have trapped charge on it, while the feeding network
behind the circuit breaker will normally be in AC steady-state. In this case
let the program automatically calculate the AC steady-state initial
conditions, which will give correct initial conditions in the feeding
network part and incorrect zero initial conditions on the switched line.
Override the latter incorrect values with user-supplied initial conditions.
switched line
feeding network
Figure 11.1: Example of Case with a Trapped Charge

For example, if the trapped charge leaves DC voltages of Va = 80kV, Vb = 40kV,
Vc = -30kV on the switched line, then supply initial voltage values of
V1a(0) = 80000
V2a(0) = 80000
11 - 2
User-Supplied Inital Conditions

V1b(0) = -40000
V2b(0) = -40000
V1c(0) = -30000
V2c(0) = -30000
and initial current values of

i1a(0) = 0.0
i2a(0) = 0.0
i1b(0) = 0.0
i2b(0) = 0.0
i1c(0) = 0.0
i2c(0) = 0.0
(for an explanation why zero values for currents must be supplied see Section 11.4.2).
It is advisable to let the program determine initial conditions whenever possible, that is, for AC
steady-state as well as for DC steady-state. In the latter case, use cosine functions with very low
frequency to simulate DC sources (e.g., f = 0.001 Hz works very well in most cases).
11.2
Warning On Incomplete Initial-Conditions

Specification Capability
Unfortunately, the manual specification of initial conditions (as per Section 11.3) is not possible at
the present time for many EMTP components. Specifically excluded are the following
components:
(A)
Saturable "TRANSFORMER" component of Section 6.4
(B)
Frequency-dependent branch modelling of Section 6
(C)
JMARTI Frequency-dependent line model of Section 7.4
(D)
Type-99 pseudo-nonlinear resistance of Section 8.3
(E)
Type-98 pseudo-nonlinear reactor of Section 8.4
(F)
Type-97 staircase resistance of Section 8.2
(G)
Type-16 controlled DC voltage source of Section 10.6
(H)
Dynamic synchronous machine (S.M. and U.M.) components of Section 10.11,

and Section 10.12
(I)
TACS modelling (see Section 14 for TACS-specific steady-state initialization
11 - 3

rules)
For typical modern production studies, the manual specification (override) of initial conditions is
not possible.
11.3
General Rules
Rule 1:
If the network contains lines with distributed parameters, then the initial conditions
must be those of a DC steady-state (e.g., trapped charge) or of a linear AC steady-state.
In the latter case, all values for voltages and currents must be specified as crest (not
RMS) phasor values x(0) + jy(0), where x(0) is the instantaneous value at t = 0 and
2
2
x( 0 ) + y( 0 )
cos(t + tan-1 (y(0) / x(0))
is the instantaneous value at t 0.0. The frequency finitial for the initial AC steady-state
must be entered on the very first node voltage card. For DC steady-state, set finitial = 0
(or blank) and supply instantaneous values x(0) (y(0) ignored if finitial = 0).
Rule 2:
In networks without distributed parameters, any initial conditions are acceptable. In

this case, all initial voltages and currents are specified as instantaneous values x(0)
(ignore format fields provided for the imaginary part y(0) of phasor values and the
steady-state frequency as described in Rule 1).
Rule 3:
Cards for initial conditions must be stacked in the following order:

1.
Cards for node voltage ("2" in column 2);
2.
Cards for currents in linear elements ("3" in column 2);
3.
Cards for currents in time-varying and nonlinear elements ("4" in column 2).
The above order for stacking cards is important because the program computes the
initial conditions in a branch as soon as the initial currents are read in, from these
currents as well as from the node voltages at both ends. Because of this, branch cards
must also be provided for branches with currents zero if they contain shunt or series
capacitances because initial conditions may still be nonzero if the voltages are
nonzero.
If the network contains parallel branches with identical pairs of node names, then all
initial currents for linear elements with that pair of node names will go into the first
linear branch found in the list. All initial currents for nonlinear branches with identical
pairs of node names will likewise go into the first nonlinear branch only. This leads to
wrong initial conditions. If there are only two branches in parallel, then this problem
can be avoided by identifying the second branch in reverse order (e.g., first branch
11 - 4

from A1 to B1, second branch from B1 to A1).
currents in nonlinear and

time varying branches
3
currents in
linear branches
node voltages
Figure 11.1:
Order of Initial Conditions Specification
11.4
Format and Rules
11.4.1
Cards for Node Voltages and Initial Condition Steady-State

Frequency
1
2
3
4
5
6
7
8
12 345678 901234567890123 456789012345678 901234567890123 456789012345678901234567890
ID
NODE
NAME
Re {Vcrest} or v(o)
Im {Vcrest}
finitial (Hz)
I2
A6
E15.8
E15.8
E15.8
Rule 1:
Specify ID = 2 (data field ID in columns 1-2).
Rule 2:
Specify node by its alphanumeric name (data field NODE NAME in columns 3-8).
Rule 3:
Use columns 9-23 to specify real part of voltage crest phasor value Vcrest if initial conditions are linear AC steady-state (instantaneous value defined as v(t) = Re(Vcrest
ejt)), or instantaneous value v(0) for arbitrary initial conditions recognized by
finitial = 0.
11 - 5

Rule 4:
Use columns 24-38 to specify imaginary part of crest phasor value Vcrest (only needed
if initial conditions are AC steady-state and if the network contains distributed parameter lines).
Rule 5:
Use columns 39-53 to specify frequency of the initial AC steady-state in units of Hz.
To be entered on the very first card for node voltages. Set finitial = 0 or blank if initial
conditions are DC steady-state. This parameter is ignored unless the network contains
distributed parameter lines.
11.4.2
Cards for Currents in Linear Branches
1
2
3
4
5
6
7
8
12 345678 901234 567890123456789 012345678901234 5678901234567890 12345678901234 567890
ID NODEk NODEm
I2
A6
A6
E15.8
E15.8
E15.8
E15.8
Parameters:
ID
Specify ID = 3 (data field ID in columns 1-2)
NODE
NAMES
Specify names in the same sequence as on branch card (if

node names on branch card are "A to B", then use "A to
B" here also, and not "B to A" which would be regarded
as a different branch). Use columns 3-8 and 9-14,
respectively.
A,B,C,D
Specify parameters, A,B,C,D as follows:

(a) For lumped series R,L,C
ikm
ikm(0) in A
vCAP(0) in B
vCAP = vl - vm
No initial current card is needed if the branch is purely

resistive (L=0, 1/C = 0) or if current through branch
and capacitor voltage are both zero.
11 - 6

(b) For symmetric pi-circuits, and for transformers and
other inductively coupled branches:
k
iko
imo
ikm
ikm(0) in A
iko(0) in B
imo(0) in C
A card for currents is needed for pi-circuits if either the

initial node voltages in k and m, or the initial currents, or
both, are nonzero. If all currents are zero the card is still
needed then the node voltages are nonzero; otherwise,
the initial conditions in the shunt capacitances are not set
properly. For multiphase pi-circuits or inductively
coupled branches, specify one card for each phase in the
same sequence as branch cards were read in. In the
multiphase case, iko and imo are the sum of the shunt
currents which go from the respective phase to all other
phases as well as to ground.
(c) For distributed parameter lines:
im
ik
Re(Ik crest) in A
Im(Ik crest) in B
Re(Im crest) in C
Im(Im crest) in D
The definition of crest phasor values is the same as in

Section 11.4.1. A card for currents is needed if either the
initial node voltages, or the initial currents, or both, are
nonzero. If all currents are zero the card is still needed if
the node voltages are nonzero; otherwise the initial
conditions in the shunt capacitances are not set properly.
For multiphase lines, specify one card for each phase in
the same sequence as branch cards were read in.
11 - 7
11.4.3
Cards for Currents in Nonlinear Inductance Branches
1
2
3
4
5
6
7
8
12 345678 901234 567890123456789 012345678901234 5678901234567890 12345678901234 567890
ID NODEk NODEm
I2
A6
A6
E15.8
E15.8
Parameters:
ID
Specify ID = 4 (data field ID in columns 1-2)
NODE
NAMES
Specify pair of node names in same the sequence as on

branch card (see Section 11.4.2)
Initial current ikm(0) in nonlinear inductance (no cards

are needed for initial currents in time-varying and
nonlinear resistances).
Leave this field blank if initial flux km is uniquely

defined by ikm(0) and the nonlinear characteristic (case 1
below). Specify km(0) in this field if characteristic has
infinite slope at km(0), in which case km(0) would not
be known unless supplied (case 2 at right).
km (0)
ikm (0)
ikm (0)
Case 1
Case 2
Figure 11.2: Initial Conditions for a Nonlinear Inductance
11 - 8
11.4.4
Examples
First Example:
A1
A2
A3
A4
B2 B3 B4
B1
C2 C3 C4
C1

TCLOSE > 0
Figure 11.3: Test System for a Trapped Charge Case

Assume that AC steady-state initial conditions are automatically computed by the program, and
that the line from 3 to 4 has trapped charge on it. Then initial DC conditions must only be read in
for the line from 3 to 4 to override what is automatically calculated (which would be zero in this
case if the switches are open in the initial steady-state). Assume that the trapped charge leads to
DC voltages of 3000 (units of voltage) in phase A, 2800 in phase B, and 1000 in phase C.
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C INITIAL NODE VOLTAGES
C <NAME-<---- Vreal----<-----Vimag----<---finitial--2A3
3000.0
2A4
3000.0
2B3
-2800.0
2B4
-2800.0
2C3
-1000.0
2C3
-1000.0
C INITIAL BRANCH CURRENTS
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <BUSK-<BUSM-<------ A ----<------ B ------<------ C ------<------ D -----3A3
A4
3B3
B4
3C3
C4
11 - 9
Second Example:
L1
R1
iL
R2
ic
C1
Lm
RFe
Figure 11.4: Test System for Multi-Frequency Initial Conditions

Assume that an unloaded transformer (represented by resistance R2, nonlinear magnetizing
inductance Lm and RFe for iron losses) is energized from a system (represented by an equivalent
circuit with L1=795.8 mH, R1 = 40 ohms, and C1=0.6F)
Assume further that the source contains a third harmonic, i.e.,
vsource = 311101 cos (t-89.55o) + 6222.03 (cos 3t -89.55o)
There is one solution with the source of fundamental frequency:
vsource(0) =
2461.58 V
v1(0)
0.002357 V
iL(0)
61.54 A
and another solution with the source of third harmonic with the results
vsource(0) =
49.23 V
v1(0)
338.36 V
iL(0)
6.10 A
The initial conditions for node voltages are therefore

vsource(0) =
2510.81 V
V1(0)
338.36 V
11 - 10

and for the branch current
iL(0)
67.64 A
If the nonlinear inductance is represented with an infinite slope at i=0 as shown, then a residual
flux of 900 Vs could also be read in as initial conditions.
1225 Vs
983 Vs
i
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C INITIAL NODE VOLTAGES
C <NAME-<---- Vreal----<-----Vimag----<---finitial--2SOURCE
2510.81
21
-338.36
C INITIAL BRANCH CURRENTS
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <BUSK-<BUSM-<------ A ----<------ B -----3SOURCE1
67.64
31
67.64
-338.36
C NONLINEAR BRANCH INITIAL CONDITIONS
42
900.00
11 - 11
Section 12 Output-Variable Specification for the TimeStep Loop

12.1
Introduction ......................................................................................... 12-1
12.2
Output-Variable Specification For Conventional Data Cases ............ 12-1
12.3
12.2.1
Node Voltage Output ..................................................... 12-1
12.2.2
Branch Output ................................................................ 12-2
12.2.3
Synchronous Machine and TACS Outputs ...................... 12-5
Output-Variables Specification for "STATISTICS"

or "SYSTEMATIC" Data Case ............................................................... 12-6
12.4
12.3.1
General Introduction ...................................................... 12-6
12.3.2
Optional Miscellaneous Statistics Data Card ................. 12-7
12.3.3
Output Scaling Specification Cards ............................... 12-8
Explanation of "STATISTICS" Printed Output ..................................... 12-10
Section 12
Output-Variable Specification for the Time-Step Loop
12.1
Introduction
The results of an EMTP simulation can be obtained in the form of tabulated quantities in the
standard output file, as printer plots (also in the output file), as graphs generated with the builtin Postscript batch-mode plotting option, or as graphs produced by a pl4 file post-processing
program such as EMTPOUT.
The EMTP generates output only for those variables for which specific output requests have been
made. Output requests depend on the type of simulation: single deterministic simulation or
statistical.
12.2
Output-Variable Specification For Conventional Data

Cases
Values of voltage, current, power, and/or energy variables are computed by the EMTP at discrete
time instants t=0, t, 2t, etc. Such values are then either printed (with frequency controlled by
miscellaneous data parameter "IOUT"; see Section 4.2.2), or they are written to disk into the pl4
binary plot file with frequency "IPLOT" (miscellaneous data parameter; see Sections 4.2.2), for
purposes of later plotting after the simulation is finished. Three basic types of output requests can
be identified:
12.2.1
Node Voltage Output
The user must use one or more cards specifying all node names for which the node voltages
(voltage to ground) will be printed or plotted.
(A)
To request all node voltages use a single output specification card, with just a
"1" in column 2; do not terminate with a blank card.
(B)
To request selective node-voltage output, list the node names on one or more
cards which have columns 1-2 blank, with at least the first field (variable
BUS1, read from columns 3-8) non-blank. Terminate such cards by a blank
card.
12 - 1
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
BUS1
BUS2
BUS3
BUS4
BUS5
BUS6
A6
A6
A6
A6
A6
A6
Note that of the 13 potential fields per card, only the first must always be non-blank
(except for the terminating blank card); otherwise fields can randomly be left blank,
and will be passed over. For example, the following three cards request node-voltage
output for the six nodes named ALPHA, BETA, GAMMA, DELTA, PSI and ZETA:
C
1
2
3
4
5
6 ...
C 3456789 123456789 123456789 123456789 123456789 123456789 1 ...
C <name<-name<-name<-name<-name<-name<-name<-name<-name<-name ...
ALPHA
BETA
GAMMA
DELTA
PSI
ZETA
BLANK CARD TERMINATING NODE VOLTAGE SPECIFICATION
(C)
12.2.2
No node voltages desired: Supply just a single blank card (which is taken as
the blank card ending the non-existent selective node-voltage specification
cards).
Branch Output
Branch and switch variables can be printed or plotted by entering "1" in column 80 on the
respective branch cards. No further specification is required at this point. The printed heading for
branch-variable output consists of pairs of node names, one above the other. This ordering (upper
to lower) indicates the polarity of the output. Suppose that node k represents the upper node of the
pair, and node m the lower one; then
(A)
Branch voltage output is vkm(t) = vk(t) - vm(t)
(B)
Branch current output is ikm(t), the flow from node k to node m.
For non-switch elements, output power is the instantaneous power consumed by the branch.
Output energy is the integral of this power since t=0. For switches, the power and energy are those
that flow through the switch:
(C)
Non-switch power output is Pkm(t) = (Vk(t) Vm(t)) ikm(t)
(D)
Switch power output is Pkm(t) = vk(t) ikm(t)

12 - 2

t
(E)
Energy output is
Pkm(u) du
0
In addition to these column-80 requests, it is possible to request branch and switch output along
with node-voltage outputs. The format for this added alternate branch and switch output
capability is as follows:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
NAM1
NAM2
NAM3
NAM4
NAM5
NAM6
I2
A6
A6
A6
A6
A6
A6
(A)
ITYPE
= 1 Only current outputs for branches or switches

= 2 Only voltage outputs for branches or switches
= 3 Current and voltage outputs for branches or switches
= 4 Power and energy outputs for branches or switches
(B)
List the 6-character names of branches or switches on one or more cards

beginning in column 3. Of the 13 potential names per card, any blank ones
will be ignored by the program.
(C)
The user can mix any branch or switch output request cards with any node
voltage output request cards.
(D)
If selective node voltage output requests exist, then one blank card is needed to
terminate the specification of EMTP output variables.
(E)
On the other hand, if the user has requested the output of all EMTP node
voltages by means of "1" in column 2, then:
(i)
If this "1" card is the last output request card, then do not follow by a
blank card.
(ii)
If this "1" card is not the last output request card, columns 3 through 8
of this card must be non-blank. Also, a terminating blank card must
follow the last EMTP output request card.
12 - 3

Following are two examples of the output request cards:
C
1
2
3
4
5
6 ...
C 3456789 123456789 123456789 123456789 123456789 123456789 1 ...
1
-2BRCH10
BRCH30
-1SWCH1 SWCH2 SWCH3
-3
BRCH20
-4
BRCH40
BLANK CARD TERMINATING OUTPUT REQUESTS
C
1
2
3
4
5
6 ...
C 3456789 123456789 123456789 123456789 123456789 123456789 1 ...
-1BRCH1 BRCH2
-3BRCH20
-2BRCH1 BRCH3
BLANK CARD TERMINATING OUTPUT
Plotting creates no special problem, unless output for parallel branches should be requested. Then
as per Rule 3 of Section 5.1.2, only the first current (energy) output of the two or more in parallel is
accessible.
As a practical restriction, the above limitation is not serious. One can always insert a near-zeroresistance or reactance branch, thereby eliminating any parallel connection. For example, should
power P(t) and energy E(t) be desired for a nonlinear resistance, in addition to the branch current
and voltage, the configuration shown in Figure 12.1 below, could be used. Resistance R2 is chosen
to be "small," in accord with the rule concerning system-dependent minimum values (see Section
5.1.6). Current output (1 in column 80) is requested on the branch card for resistance R2.
Resistance R1 can be chosen to be very large, and branch voltage output is to be requested for it
(2 in column 80). The nonlinear element has 4 in column 80, to give its power and energy
consumption.
R2 0
R1 = 1018
Figure 12.1:
Elimination of a Parallel Connection

12 - 4
12.2.3
Synchronous Machine and TACS Outputs
In all, there are five general groupings of output variables for EMTP printed output of the time-step
loop. The just-described node voltages come first, then branch voltages come second, and branch
currents come third. Then come two additional classes of EMTP output variables: synchronous or
universal machine output variables (fourth grouping), and TACS output quantities (fifth and final
grouping).
Output requests for synchronous machine Type-59 S.M. variables are part of the synchronous
machine input data (see Section 10.11, Class 5 data). Variables are identified by a pair of 6character alphanumeric identifying names for each quantity; these are generated internally. For
plotting purposes, synchronous machine variables are treated as though they were branch currents
(e.g., using a "9" in column 3 of batch-mode plot cards).
Handling of output requests for the universal machine (Type-19 U.M.) of Section 10.12 is similar
to that of the Type-59 Synchronous Machine.
Output requests for TACS variables are part of the TACS input data which immediately follows the
EMTP miscellaneous data cards and their extensions. See Section 14 for details. One 6-character
alphanumeric name identifies the desired TACS variable. But a name-pair is used for the variable
identification, with the TACS name being the second name of this pair. The first name "TACS" is
automatically supplied internally by the EMTP. For example, "TIMEX" is the name of the built-in
variable which carries the current simulation time; for output purposes, the pair of names "TACS"
and "TIMEX" would be used. For plotting purposes, TACS variables are treated as though they
were branch currents (e.g., using a "9" in column 3 of batch-mode plot cards).
The above description presumes a hybrid problem with both TACS and some electric network
representation. For a TACS-only data case, TACS variables are treated like node voltages (Type-1
or 4 for plotting purposes), and only a single name (that of the TACS variable) is used for
identification.
The following sample of EMTP line printer output for the beginning of the time-step loop
illustrates the five classes of output variables just described. Since there was no TACS
representation, the fifth class or grouping is empty, note:
44
first
next
next
32 output variables pertain to dynamic synchronous machines, with names generated internally
STEP
TIME
JEFA
JEFB
JEFC
ROCA
ROCB
ROCC
MACA1
MACB1
MACC1
RJ0A
TERRA
ROCA
COMROC
ROCB
COMROC
MACH 1
ID
MACH 1
IQ
MACH 1
I0
MACH 1
IF
MACH 1
IKD
MACH 1
IG
MACH 1
IKQ
MACH 1
IA
MACH 1
IB
MACH 1
IC
MACH 1
EFD
MACH 1
MFORCE
MACH 1
MANG
MACH 1
TQ GEN
MACH 1
ANG 1
MACH 1
VEL 1
MACH 2
ID
MACH 2
IQ
MACH 2
I0
MACH 2
IF
MACH 2
IKD
MACH 2
IG
MACH 2
IKQ
MACH 2
IA
12 - 5

MACH 2
IB
MACH 2
IC
MACH 2
EFD
MACH 2
MFORCE
MACH 2
MANG
MACH 2
TQ GEN
MACH 2
ANG 1
MACH 2
VEL 1
12.3
Output-Variables Specification for "STATISTICS" or

"SYSTEMATIC" Data Case
12.3.1
General Introduction
A "STATISTICS" or "SYSTEMATIC" overvoltage study is distinct from all other EMTP data cases in
that multiple simulations are automatically generated internally by the EMTP. Integer
miscellaneous data parameter "NENERG" (see Section 4.2.2) defines the number of energizations
which are to be simulated. Each energization gives a different solution, due to differences in the
closing times for "STATISTICS" or "SYSTEMATIC" switches (see Section 9.2.3 and Section 9.2.4).
For each energization, only the vector of variable maxima is available for printing, and a statistical
tabulation of all |NENERG| such vectors is automatically provided at the conclusion of all such
internally-generated simulations. This is the statistical overvoltage output line printer output
which occurs for either "STATISTICS" or "SYSTEMATIC" data cases. There is no associated plotting
(unlike conventional cases, where printing and plotting go together). It is the purpose of the
present section to explain how the user requests such overvoltage output, and what the printout
looks like. But before describing the output specification cards for the statistical overvoltage
tabulation, it should be mentioned that there is a base-case solution which precedes the |NENERG|
energizations. For a "STATISTICS" data case, this is a zero-th energization with all variances
temporarily set to zero. For a "SYSTEMATIC" data case, either all beginning or all mid-closing
times of the switches are used whichever the user has chosen to input (see Section 9.3.4).
Before appending the requests for statistical overvoltage tabulations, the user must complete the
data deck for the base case, pretending that it is a separate conventional simulation. The overall
(total) data structure has the following components and ordering:
1.
EMTP source cards, terminated by a blank card.
2.
Node-voltage output specification card or cards, applicable to the base case and
statistical tabulation (see Section 12.2). If node voltages are selectively
requested, this data will be terminated by a blank card. These cards are required
even when the "OMIT BASE CASE" option of running statistics case is used.
3.
Batch-mode plot-specification cards, applicable to the base case solution only.

These are terminated by a blank card. The plot cards must be omitted for "OMIT
BASE CASE" option, and a blank card is all it needs.
4.
Optional miscellaneous statistics data card (see Section 12.3.2).
5.
Output scaling specification cards for purposes of statistical tabulation,

terminated by a blank card (see Section 12.3.3).
12 - 6
blank card
statistics output
scaling cards
blank card
optional miscellaneous
statistics data card
batch mode plotting
blank card
node voltage
output specification
EMTP data deck including
source cards
12.3.2
Optional Miscellaneous Statistics Data Card
The optional miscellaneous statistics data card has the keyword "MISC. STATISTICS DATA"
entered in columns 41 to 61.
1
2
3
4
5
6
7
8
1234567890123456789012345 6789012 34567890 123456789012345678901 2345678901234567890
AINCR
F8.0
XMAXMX
MISC. STATISTICS DATA
F8.0
A21
The purpose of this optional card is twofold. First, for overriding values of AINCR and/or
XMAXMX on the original statistics miscellaneous data card (see Section 4.3.1). Second, AINCR
can be entered as a negative integer N: this means that the statistical variables entered on the
following output scaling cards are to be tabulated within N compartments. Any number, if any, of
these cards can be specified and values of AINCR and XMAXMX are effective until overridden by
another such card.
12 - 7
12.3.3
Output Scaling Specification Cards
IBROPT
1
2
3
4
5
6
7
8
12 345678901234 567890 123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
BASE
BUS1
BUS2
BUS3
BUS4
BUS5
BUS6
BUS7
BUS8
I2
E12.0
A6
A6
A6
A6
A6
A6
A6
A6
IBROPT
(1-2)
BASE
(3-14)
BUS1BUS11
(15-80)
BUS9 BUS10 BUS11

A6
Type code for this output-specification card, telling the

type of variables in question:
= 0
This card is for node-voltage specifications.
= 1
This card is for branch voltage specifications.
= 2
This card is for branch current specifications.
= 3
This card is for power specifications.
= 4
This card is for energy specifications.
Base value which is to be used for the per unit statistical

tabulation of all variables specified on this data card.
The units of base value for a given type of variable
should be the same as that of the variables on the card. If
AINCR is positive, a blank or zero then BASE is given the
following default values:
A)
If IBROPT is 0 or 1, the default value is equal to

the voltage which the EMTP read from the user's
first source card, assuming it was a Type-14
sinusoidal voltage source. See Section 10.3, field
"AMPLITUDE" of columns 1120 of the source
card.
B)
The default value is 1 if IBROPT is 2, 3 or 4.

However, if AINCR is a negative value, the field
of BASE is ignored. The minimum value of the
solutions for a given variable is then chosen to
be the base value for the tabulation of that
variable.
Node names which specify the desired output variable

for the statistical study. There are two distinct cases,
depending on the value of 'IBROPT':
12 - 8
A6
A6

A)
If IBROPT = 0, the variables in question are node

names which are to have node-voltage (to
ground) output. Blank fields are ignored.
B)
If IBROPT does not equal 0, the variables in

question are the node names, taken in pairs,
which specify the branch voltages, branch
current, power or energy consumed in the
branch. The final field "BUS11" is ignored.
Any pair of two blank fields is ignored.
If there are more than 11 node voltage output variables or more than five branch output variables
for a given base value, continuation to five more cards is allowed. To do this, the user should
specify "CONT." in columns 76-80 on the output scaling specification card which follows. The
continuation cards use the same format except for the fields "IBROPT" and "BASE" which can be
left blank.
There are no ordering restrictions on such "STATISTICS" output scaling specification cards, among
themselves. Their end is signalled by a blank card. Statistical tabulations for output variables will
occur in the order that the specifications are encountered on these input cards.
Illustrative example of data input:
C
1
2
3
4
5
6
7
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <---BASE----<-bus-<-bus-<-bus-<-bus-<-bus-<-bus-<-bus-<-bus-<-bus-<-bus-<-bus-1
100.0 GENA
A1
-1
GENC
C1 ENDA
A10
ASW10 BSW10 CSW10
-20.0
MISC. STATISTICS DATA
-3
B1 BSW1
-2
0.0 GENB
B1 ENDA
A10
-4
0.0
B1 BSW1
1.
Branch voltage between nodes " GENA" and "

requested.
2.
Two branch voltages are requested: between nodes " GENC" and " C1",
" ENDA" and " A10". The base value for the output tabulation for these two
branch voltages is blank, which means the "AMPLITUDE" specified on the first
type-14 sinusoidal voltage source is to be used as the base value.
3.
Node voltages output have been requested for nodes " ASW10", " BSW10" and
" CSW10" with default base value same as the one described in (2).
12 - 9
A1" at base value of 100 volts is
12.4
4.
Here is an optional miscellaneous statistics data card on which AINCR= 20 and

XMAXMX=blank. This means XMAXMX takes the same value as inputted on the
first "STATISTICS" miscellaneous data card at the top of the data case. Negative
20 for AINCR means that 20 compartments will be used for tabulating following
output variables. These two new values for AINCR and XMAXMX are effective
until another such card appears.
5.
Power flow through the branch connecting nodes " B1" and BSW1" is
requested. Since AINCR= 20, the field of base value is ignored, and 20
compartments will be used for tabulating the results for this power flow.
Minimum value among the power flow solutions for this branch is chosen to be
the base value.
6.
Two branch currents are requested: between nodes of " GENB" and " B1",
" ENDA" and " A10". Since AINCR is 20, the rule of deciding the base value
and compartments for statistical tabulation of these two branch currents is the
same as described in (5) above.
7.
Energy consumed in the branch connecting nodes " B1" and " BSW1" is
requested. Again, since AINCR is a negative integer, the rule for tabulation is the
same as in (5).
Explanation of "STATISTICS" Printed Output
The printed outputs available to the architect of a statistical overvoltage study include the
following:
Output a:Switch Closing Times for Each Energization.
Printout of the switch closing times Tclose is automatically provided if the user enters
the statistics miscellaneous data variable "ISW" equal to unity. See Section 4.3.1 and
Section 17.2.
Output b:Maximum Voltages of Each Energization
Printout of the vector of maximum variable values for each energization is
automatically provided. This is comparable to what the user gets by setting
miscellaneous data parameter "MAXOUT" (see Section 4.2.2) equal to unity for a
conventional deterministic simulation. The "reference angle" (see Section 9.2.3) is
also automatically provided, as part of this printout.
An illustrative sample of "Output a" and "Output b" for a small statistical overvoltage study
appears as follows:
12 - 10

THE DATA CASE NOW READY TO BE SOLVED IS A STATISTICAL OVERVOLTAGE STUDY WHICH INVOLVES
6
ENERGIZATIONS
(PARAMETER NENERG OF COLUMNS 65-72 OF THE 2ND MISC. DATA CARD).
SWITCH CLOSING OR OPENING TIMES FOR THE
SPECIALLY-FLAGGED SWITCHES (WITH STATISTICS PUNCHED IN COLUMNS 55-64) ARE VARIED RANDOMLY ACCORDING TO A NORMAL DISTRIBUTION.
THE USER CAN SELECT EITHER UNIFORM OR NORMAL DISTRIBUTIONS, BASED ON THE VALUE OF PARAMETER IDIST OF COLUMNS
17-24 OF THE SPECIAL STATISTICS MISC. DATA CARD.
THE FOLLOWING IS A LISTING OF SWITCHES WHOSE CLOSING TIMES ARE
TO BE STATISTICALLY VARIED, ALONG WITH THE ASSOCIATED MEAN AND STANDARD DEVIATION FOR THE DISTRIBUTION, AS REQUESTED BY THE USER.
THE USER PUNCHED STATISTICS MISCELLANEOUS DATA PARAMETER XMAXMX = -0.200E+01 ,
WITH THE NEGATIVE VALUE REPRESENTING
A REQUEST FOR USE OF THE 100 BUILT-IN RANDOM NUMBERS.
AS SUCH, THE ANSWER IS REALLY DETERMINISTIC, THEN (A SECOND SOLUTION
ON ANY COMPUTER SHOULD GIVE THE SAME ANSWER).
ENTRY
SWITCH
FROM
TO
COLUMNS 15-24
COLUMNS 25-34
REFERENCE SWITCH NO.
NUMBER
NUMBER
BUS
BUS
(IN SECONDS)
(IN SECONDS)
(0 IMPLIES INDEPENDENCE)
1
1
A1
ASW1
0.002000
0.000100
0
2
2
B1
BSW1
0.004000
0.000500
0
3
3
C1
CSW1
0.006000
0.001000
0
NOW IN ADDITION TO SWITCH-TIME VARIATION CAUSED BY EACH SWITCHS OWN DISTRIBUTION, THERE IS THE ADDED RANDOM
DELAY WHICH IS THE SAME FOR ALL SWITCHES, REFERED TO BY THE TERM REFERENCE ANGLE.
DISTRIBUTION FOR THIS
ANGLE IS UNIFORM OVER THE TIME INTERVAL FROM
0.00000E+00 TO
0.36000E+03 DEGREES BASED ON
0.600E+02 HZ
FREQUENCY.
THIS WAS ALL SPECIFIED BY THE USER USING FIELDS DEGMIN , DEGMAX , AND STATFR (COLS. 41-64)
OF THE SPECIAL STATISTICS MISC. DATA CARD.
HOWEVER, IF THE PARAMETER ITEST (FOUND IN COL. 9-16 OF THE STATISTICS MISC. DATA CARD) IS GREATER THAN ZERO,
THE REFERENCE ANGLE IS SET TO ZERO AND, CONSEQUENTLY, THERE IS NO ADDED RANDOM DELAY.
IN THE FOLLOWING PRINTOUT, SWITCH TIMES WHICH ARE USED FOR EACH ENERGIZATION ARE TABULATED, FOLLOWED BY PEAK
OUTPUT-VARIABLE VALUES WHICH OCCUR FOR THE ASSOCIATED SIMULATION.
FORMAT FOR PEAK VALUES IS IDENTICAL TO THE
PRINTOUT FOR MAXIMUM VARIABLE VALUES OF A CONVENTIONAL STUDY
EXCEPT THAT PRINTOUT OF THE REFERENCE ANGLE IN DEGREES HAS BEEN ADDED AT THE EXTREME LEFT.

10
first
next
next
Branch power consumption (power flow, if a switch) is treated like a branch voltage for this grouping
Branch energy consumption (energy flow, if a switch) is treated like a branch current for this grouping
Columns of POWER---ENERGY data:
9
10
ASW10
REF. ANGLE
BSW10
CSW10
GENA
A1
GENC
C1
ENDA
A10
B1
BSW1
B1
BSW1
ENDA
A10
RANDOM SWITCHING TIMES FOR ENERGIZATION NUMBER
1
1
0.211320E-02
2
0.421433E-02
3
0.656836E-02
0.0000
-0.535147E+03 0.538317E+03 0.557000E+03 0.125970E+03 0.100276E+03 0.830743E+02 0.442313E+03 0.103401E+01 0.142884E+01
-0.194250E+01
TIMES OF MAXIMA :
0.770000E-02 0.520000E-02 0.114000E-01 0.230000E-02 0.670000E-02 0.810000E-02 0.140000E-01 0.200000E-01 0.470000E-02
0.172000E-01
RANDOM SWITCHING TIMES FOR ENERGIZATION NUMBER
2
1
0.202111E-02
2
0.412600E-02
3
0.578814E-02
0.0000
-0.530097E+03 0.538891E+03 0.537257E+03 0.131475E+03-0.719084E+02-0.527593E+02 0.423384E+03 0.924623E+00 0.144094E+01
-0.186786E+01
TIMES OF MAXIMA :
0.750000E-02 0.510000E-02 0.114000E-01 0.220000E-02 0.127000E-01 0.860000E-02 0.460000E-02 0.200000E-01 0.460000E-02
0.170000E-01
Output c: Statistical Distribution of Voltages, Currents, Powers, and Energies.

Statistical distributions of maximum output-variables values are automatically
provided. These are both one at a time, and also as part of the distribution for the
union of all such individual outputs which have the same base value.
12 - 11
GENB
B1

An illustrative sample of "Output c" for a small statistical overvoltage study of NENERG=6
energizations appears as follows. Of the several tabulations, only that for node voltage "ASW10" is
shown:
STATISTICAL DISTRIBUTION OF PEAK VOLTAGE AT NODE ASW10 .

BASE VOLTAGE FOR PER-UNIT PRINTOUT
=
0.30300E+03
INTERVAL
VOLTAGE
VOLTAGE IN
FREQUENCY
CUMULATIVE
PER CENT
NUMBER
IN PER UNIT
PHYSICAL UNITS
(DENSITY)
FREQUENCY
.GE. CURRENT
VALUE
24
1.15000
0.348450E+03
0
0
100.000
25
1.20000
0.363600E+03
1
1
83.333
26
1.25000
0.378750E+03
0
1
83.333
27
1.30000
0.393900E+03
1
2
66.667
28
1.35000
0.409050E+03
0
2
66.667
29
1.40000
0.424200E+03
0
2
66.667
30
1.45000
0.439350E+03
0
2
66.667
31
1.50000
0.454500E+03
0
2
66.667
32
1.55000
0.469650E+03
0
2
66.667
33
1.60000
0.484800E+03
0
2
66.667
34
1.65000
0.499950E+03
2
4
33.333
35
1.70000
0.515100E+03
1
5
16.667
36
1.75000
0.530250E+03
1
6
0.000
DISTRIBUTION PARAMETERS FOR THE ABOVE DATA.
MEAN =
VARIANCE =
STD DEVIATION =
Output d:
GROUPED DATA
1.5416667
0.0534167
0.2311205
UNGROUPED DATA
1.5661356
0.0510193
0.2258746
Mini Line Printer Plots of Switch Closing Times.
For "STATISTICS" data cases only, the end of the statistical tabulation of overvoltage
output is signaled by line printer plots of the switch closing times. Each "STATISTICS"
switch has such a mini plot, the size of which is always the same: 30 columns wide by
25 lines (of the printer) long. Up to three such plots can be spaced horizontally on the
page if the following one or two switches have identically the same variance as the first
one (which goes on the left). This will normally be the situation for the poles of a
3-phase breaker, and the user will generally want all three graphs to be spaced
horizontally across the page. If the user wants otherwise, he can slightly perturb one or
more variances (a switch with a variance which differs from the preceding one will be
placed below on the left, automatically).
12 - 12

Two illustrations of the printer plots of "Output d" are shown immediately below:
SWITCH B1
TO
BSW1
SWITCH C1
0.4414E+00 SWITCH
CLOSINGS/COL
TO
CSW1
COLUMNS
SWITCH CLOSINGS/COL
0.4414E+00
5
20
25
10
15
COLUMNS
30
+----+----+----+-20
--+----+----+
AD
02
--+----+----+
*
*
*
*
1D A
AD
AD
1 DA
1 D A
1 AD
1
A
1
AD
0.400E1
AD
1
02
AD
AD
1 A D
1A D
1
10
15
30
+----+----+----+--
*
*
*
DA
AD
1D A
AD
1 DA
1 DA
1 AD
1
25
0.600E-
D A
D A
AD
1 A D
1A D
1 D A
1D A
1D A
AD
*
*
*
*
TIME
0.3333E-03
SEC/LINE
A
1D A
AD
AD
*
*
*
*
TIME
0.1667E-03
SEC/LINE
12 - 13
Section 13
User-Supplied Source Functions

(Type 1-10 Sources)
13.1
Introduction ......................................................................................... 13-1
13.2
Empirical Definition of f(t) At Every Time Step

(Procedure 1) ....................................................................................... 13-2
13.3
Definition of Type 1-10 EMTP Sources Using TACS

(Procedure 2) ....................................................................................... 13-3
13.4
Analytical Definition of f(t)

(Procedure 3) ....................................................................................... 13-3
13.5
Sample Usage of Analytic Type 1-10 Source Definition .................... 13-4
Section 13
User-Specified Source Functions (Type 1-10 Sources)
13.1
Introduction
In Section 10, it was mentioned that source types 1 through 10 are reserved for user-defined
functions. The user can use any of the following three procedures for this purpose:
Procedure 1:
Define the source functions f(t) empirically at every time step at which
it is nonzero, and enter such points as part of the EMTP data case.
Details are given in Section 13.2.
Procedure 2:
Equate one or more source functions f(t) to a TACS variable value.

See Section 4.1.5.
Procedure 3:
Define one or more source functions f(t) analytically, and develop the
FORTRAN code which returns the desired value for any given input
time. This becomes the user's own special subroutine "ANALYT", for
which he is responsible. The user must replace the original EMTP
module of this same name by his own module, before executing the
data case in question. A special request card reading "ANALYTIC
SOURCES" is also required (see Section 4.1.1).
The three procedures above apply to Type 1-10 sources in the order they are numbered. That is,
Procedure 1 first, etc.:
1.
First a data card is read, if it exists (if the bounding flag "9999" has not yet been
encountered); all ten source values are defined from there.
2.
Next, any "TACS EMTP SOURCES" definitions will override the Procedure 1
assignments for those sources.
3.
Finally, module "ANALYT" (if user-supplied and active) will modify the result
of the preceding step in a way which is known only to the user. Note that further
analytical processing of TACS dynamics is in this way possible, before the
values actually are applied to the Type 1-10 EMTP sources.
13 - 1
13.2
Empirical Definition Of f(t) At Every Time Step

(Procedure 1)
To enter the values f(t) on cards for every time step the following procedure must be followed. See
Section 3 and Section 10.7 for correct position of these data cards within the data case in question.
Rules:
Rule 1:
The values of f(t) at t = t, 2t, 3t, ... must be supplied in the correct time order with
one card for each time step. Note that the first card is for t = t (not t=0, which belongs
to the initial conditions).
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 789012345678901234567890
Format for empirically specifying source types 1 -10
Type1
Type2
Type3
Type4
Type5
Type6
Type7
E8.0
E8.0
E8.0
E8.0
E8.0
E8.0
E8.0
Rule 2:
If all 10 source functions become zero prior to t=TMAX, then enter "9999" in columns
5-8 on that card where they are all zero for the first time. Then the functions in that
and all subsequent time steps will be regarded as zero. No further source cards are
allowed after the 9999-card.
Rule 3:
If at least one of the source functions in the 10 fields is non-zero all the way to TMAX,
then a stack of exactly n cards with no blank card for termination must be supplied,
where:
TMAX
n = -----------------t
Examples:
rounded to nearest integer.
TMAX = 0.0105, t = 0.005 n = 2

TMAX = 0.0130, t = 0.005 n = 3
TMAX = 0.0098, t = 0.005 n = 2
13 - 2
13.3
Definition Of Type 1-10 EMTP Sources Using TACS

(Procedure 2)
Procedure 2 (the use of TACS) for the definition of Type 1-10 EMTP sources is straightforward.
The user must define the desired TACS dynamics, as explained in the TACS data instructions of
Section 14. The user must also indicate which source is to be controlled by which TACS variable,
using a "TACS EMTP SOURCES" card of Section 4.1.5.
13.4
Analytical Definition of f(t) (Procedure 3)
Either as an alternative, or as a supplement, to the representation of Section 13.2, the user is

permitted to define one or more of the ten possible source functions f(t) in FORTRAN, within
subroutine "ANALYT". The following rules and restrictions apply:
Rule 1:
Begin the data case with an "ANALYTIC SOURCES USAGE" data card (see Section
4.1.1). To appear before the miscellaneous data cards, this record tells the EMTP to
employ subroutine "ANALYT" of Rule 5.
Rule 2:
Modify subroutine "ANALYT" to define the one or more needed functions for any time
value t. As default, this subroutine is used to perform "RAMPING" of different variables.
Rule 3:
Any function definitions f(t) within subroutine "ANALYT" are applied by the EMTP
after the empirical definitions of Section 13.2, thereby overriding such empirical definitions. It is this sequencing which shows how and why it is possible to use a combination of the two definition procedures:
(A)
When using the "ANALYTIC SOURCES USAGE" option, data cards of

Section 13.2 must always be supplied. If no functions f(t) are to be empirically
defined, then it is just the single 9999-card which is required.
(B)
Upon the call to subroutine "ANALYT", vector VOLTBC(10) will contain values
as they were read from the ten source fields of the Section 13.2 empirical
source-definition data cards. Or, if the terminating 9999-card has already been
read, VOLTBC(10) will be identical zero. These input values for vector
VOLTBC(10) can be used in the logic of subroutine "ANALYT", if desired.
13 - 3
13.5
Sample Usage Of Analytic Type 1-10 Source Definition
Suppose that the following two sources are desired, as part of some particular EMTP data case:
Source 1: A voltage source v(t) which is defined empirically (point by point),

except that when such values are found to be non-positive, the following
alternative assignment applies:
v(t) = max { 3.6, 5.4 cos2 (120t + 60o) }
Source 2: A sinusoidal voltage source, where the phase angle varies periodically,
cyclically. Such might not be an unreasonable first approximation for
a generator which is "swinging" in the transient stability sense, due to
a system disturbance. Suppose we want:
v(t) = 303000 cos(120t + (t))
(t) = 30o + 10o cos(5t + 90o)
The user can supply any number (from zero up) of data cards defining source number 1
empirically, terminated by a 9999-card if such definitions do not extend through the final time step.
The following logic, when integrated into the subroutine "ANALYT", will then complete the job of
defining the two desired sources:
IF(VOLTBC(1) .LE. 0.0) THEN
D1 = TWOPI * (60.0 * T +1.0/6.0)
D1 = -5.4 * COS(D1)**2
IF(-3.6.GT.D1) D1 = -3.6
END IF
D1 = COS(5.0 * T +TWOPI / 4.0)
D1 = D1 * 10.0 + 30.0
D1 = TWOPI * ( 60.0 * T = d1 / 360.0
VOLTBC(2) = 30300.0 * COS(D1)
13 - 4
Section 14
14.1
"TACS" - Transient Analysis of Control

Systems

14.1.1
Modelling Assumptions ................................................. 14-2
14.1.2
Time step TACS solution ................................................ 14-3
14.1.3
Ordering Techniques ...................................................... 14-4
14.1.4
Initial conditions ............................................................ 14-6
14.1.5
TACS input ...................................................................... 14-7
14.2
Signal Sources ..................................................................................... 14-8
14.3
Transfer Functions .............................................................................. 14-19

14.3.1
Specification of S-blocks and Z-blocks ......................... 14-21
14.4
Supplemental Devices ......................................................................... 14-23
14.5
Supplemental Variables And Pseudo-FORTRAN ................................. 14-58

14.5.1
Subset of fortran available ............................................. 14-58
14.5.2
IF-THEN-ELSE Structures ................................................ 14-59
14.6
FORTRAN Interface .............................................................................. 14-61
14.7
Initial Conditions ................................................................................. 14-66
14.8
Output Specifications .......................................................................... 14-67
Section 14
TACS-Transient Analysis of Control Systems
This Section contains rules associated with the usage of the latest revision of TACS. TACS is an
acronym for Transient Analysis of Control Systems, an EMTP modelling capability for analog
control devices, and algebraic and logical processors.
14.1 General Description

TACS is used to solve the differential and algebraic equations associated with an arbitrary
interconnection of the following elements:
Transfer functions in s, with up to 5 input signals and one output signal (S block)
Gains, with up to 5 input signal and one output signal (Z block)
Static or dynamic limiters associated with the transfer functions and the gains
Nonlinear devices (supplemental devices)
Pseudo-FORTRAN capabilities (supplemental variables)
FORTRAN interface
Input signal sources
Input and output interface with EMTP
The basic rules that must be followed in preparing the data for a simulation are:
obtain a block-diagram of the system to input into TACS
each block output and signal must be assigned a unique name
identify TACS building blocks that permit construction of the block-diagram
input the TACS statements in the natural order of their computation
14 - 1

Any TACS signal can be passed to the electrical network as a voltage or current source, or as a
signal controlling the operation of a switch, valve or diode.
Interaction from EMTP to TACS occurs through EMTP defined TACS sources: network node voltages
(type 90), switch currents (type 91), machine variables (type 92), and switch status (type 93).
Interfacing with user-supplied relay models is achieved with the type 94 source.
14.1.1
Modelling Assumptions
The information used by EMTP to advance to time t is based on the TACS solution Tout for time
t-T, where T is the step size. Based on the network solution Ein at time t, TACS computes its
output signals. At each time step of the simulation, the evaluation of the control variables in TACS
follows the evaluation of the electric variables in EMTP. This sequence implies a delay of one time
step in the application of control signals to the electric system. This is shown in Figure 14.1. TACS
output to the EMTP may be in fact one or more time steps late because there are internal time delays
in the TACS solution. The same value of the step size is used in TACS and EMTP.
The concept of separate EMTP and TACS simulation is the result of preserving solution efficiency,
and as much as possible modularity. Basically, the TACS solution is simultaneous for linear blocks
(as much as possible), and sequential for nonlinear blocks or functions. The TACS equations are
sparse but asymmetric.
voltages and currents
EMTP
Ein
voltage & current

sources
switch trigger
TACS
Tout
delay T
Figure 14.1:
Solution Sequence
Typical control systems are a mixture of linear and nonlinear elements. However, the proportion
of nonlinear elements in TACS is typically much higher than in EMTP. In order to preserve solution
speed, the TACS solution is non-iterative but introduces internal TACS time delays to take into
account nonlinear components of the control circuit. This is a pseudo-nonlinear representation.
Not all nonlinear components require individual time delays. For example, series nonlinearities in
a loop do not have cumulative internal time delays (i.e., only one time delay), while some
topologies can be solved without time delays. There are no iterations during the solution. Instead,
the equations are ordered in such a way as to minimize any time delays that need to be introduced
14 - 2
TACS- Transient Analysis of Control Systems

in loops involving nonlinear blocks. When there is freedom, the equations are then ordered to
minimize fill-ins during the factorizations.
The internal TACS time delays can often be minimized or compensated with proper control
modelling. Time delays may affect the accuracy or stability of the TACS solution, but have less
impact on the stability of the EMTP-TACS interface.
14.1.2
Time step TACS solution
matrices are asymmetric, therefore, triangularization yields two distinct upper and lower
matrices. Because there are typically several nonlinearities in a control system, the TACS solution
is not as simple as triangularization separate of the time step loop, and a forward-backward
substitution in the time step loop. As in the EMTP, it is necessary to distinguish linear components
(explicitly declared transfer functions) from the nonlinearities. Nonlinearities in TACS context
include supplemental FORTRAN and pseudo-FORTRAN expressions, which may express a linear
relationship between variables.
TACS
Before entering the time step loop, all TACS transfer functions in the s-domain are converted into
algebraic difference equations in the time domain through the trapezoidal rule of integration.
These equations can be written as A x = b, where A is n x n and non symmetric, this is shown in
Figure 14.2 (history terms merged on the RHS). Basically, after a simultaneous solution of the
linear components is obtained, all supplemental devices are sequentially taken into account in a
manner that reduces the internal time delays, without iterations.
A11
A22
X1
A21
A22
X2
U11
X1
B2
B2
X2
Figure 14.2: TACS simultaneous solution of linear components

Triangularization of the matrix A is performed once (Row Gaussian Elimination is used to solve
the equations in the time step loop), then at each time step the unknown variables are found by a
forward substitution on b, followed by back-substitution to obtain the state variables x. The backsubstitution process is interrupted, as needed, to update the supplemental variables whenever
possible. That is, when TACS does the back-substitution to solve for the output of linear
components, then TACS checks the link list for output signals driving supplemental variables or
devices. If any, then TACS updates the relevant supplemental variables or devices.
14 - 3
14.1.3
Ordering Techniques
Ordering techniques completely differ from those of EMTP. Loops containing only function blocks
are not sensitive to ordering. In fact, if there is at most one limiter present, the internal TACS
solution is simultaneous. In most cases however, the correct ordering of all TACS variables, except
the sources, is crucial for reducing the internal TACS time delays and for getting as accurate results
as possible without iterations. It is not simply a matter of reducing the fill-ins. Ordering is thus
used to minimize the number of internal TACS time delays (prioritized over fill-ins in the
triangularization), to keep the number of operations in the time step loop as low as possible and,
together with sparsity storage, to reduce the size of the memory for storing the triangularized
matrices.
For the time step control system solution, TACS forms a matrix equation representing the linear
transfer functions and tries to solve the entire system simultaneously by forward-backward
substitution. To speed-up the computations no iteration of the nonlinearities is implemented in
TACS, neither is the re-factorization of the matrix equation. To resolve the nonlinearities without
iterations time delays are introduced wherever required. Nonlinear function blocks are not directly
included in the simultaneous solution. Their respective output signals are evaluated sequentially as
soon as their input signals become available at some stage during the back substitution. The
(single) limiter of a transfer function loop is ordered such that it is the first variable found in the
back-substitution. Dynamic limits of linear transfer functions changing status would require refactorization, but the effect is included in the right-hand side of the equations to resolve (based on
compensation).
The following ordering rules were in effect in v2.0 of TACS (EMTP):
1.
Sources, sequentially as read-in.
2.
Supplemental variables/devices whose input signals have been evaluated.
3.
Transfer functions (S,Z blocks) and supplemental devices whose input signals
have been evaluated. These are selected recursively, until no one remains.
4.
Supplemental devices and transfer functions with limits. This begins with the
selection of a function block with limits, followed by supplemental variables/
devices and other function blocks for which the input signals are known. The
process is repeated until no function block with limits remains. (Transfer
function loops with more than one limiter are broken up with time-delays).
5.
Remaining blocks, starting with the function block with the most input signals.
Next, supplemental variables/devices and any other function blocks for which
the input signals are now known. This process is repeated until all function
blocks have been selected.
In the time step solution of TACS, after all transfer functions were solved, history terms were
updated.
14 - 4

In version 3.0 of the EMTP the non-iterative TACS solution scheme is kept, but control system time
delays are introduced at different locations than with the original ordering algorithm because, in
some cases, the ordering of v2.0 failed to provide correct control system solution. The concept of
'super-block' is introduced: loops or nested loops composed of transfer functions and at most one
limiter.
Whether a system element is a function block or a supplemental variable/device does not impact
the ordering sequence, which only depends on the relationship, or the connection, between all
control system elements. The following ordering rules are in effect in v3.0:
1.
Get rid of the influence from the sources (ordered at top) and output variables
(ordered at bottom), so that the remaining ordering is independent from those
variables which may be set by users arbitrarily.
2.
Pick a block with the maximum number of input signals. Next, any other block
for which the input signals are now known (forward ordering). This process is
repeated until all blocks have been selected or a loop has been identified.
3.
The blocks left unordered after step (2) do not constitute a loop with the
elements just ordered. The program locates input signals from the ordered
blocks and puts it ahead of the element it has output to (backward ordering).
4.
From the ordered list the program identifies the correct order for sequentially
calculating the control system, finds where TACS implicit time-delays occur and
identifies loops.
5.
If a loop consists of function blocks only and has at most one limiter, it is called
a super-block and flagged as such so that it will later be solved simultaneously.
The super-block is re-ordered and its limiter is ordered first. The loop
identification is continued until all super-blocks are identified.
6.
If a loop consists of function blocks but has more than one limiter, the loop is
explicitly broken up by introducing explicit approximations to a one time step
delay so that history terms are made coherent in the time step solution.
When doing back substitution, starting with v3.0, TACS calculates each transfer function block and
updates its history terms sequentially following the ordering sequence. The supplemental
variables/devices linked to a function block are updated as soon as a function block has been
evaluated. For a super-block, TACS updates history terms for all the function blocks inside the
super-block only when the super-block has been completely solved.
In v2.0 (and older) supplemental devices and variables had to be broken down into different
groups (input, output and inside) which were solved sequentially within each group, in the order
the data cards were read in. It is not necessary in version 3.0 to distinguish these groups. The code
88 can be used for all supplemental variables.
14 - 5
14.1.4
Initial Conditions
In order to get a satisfactory simulation of the control system, there must be a steady-state solution
before the simulation starts. The initial conditions include initial values for all TACS variables and
history terms for transfer functions, as well as history terms for some supplemental devices (Type
53, 58, 59, 62, 64 and 65). However, the program does not need (and will not calculate) any DC
steady-state output values unless there is at least one Type 11, 92-DC and 93-DC source active at
t = 0 (TSTART < 0.0). Similarly, no AC steady-state output is calculated unless initially active Type
14 or 92-sources exist.
The steady-state initialization of the electrical network in the EMTP is calculated before the TACS
steady-state initialization. It is useful to note that in the electrical network, the user must initialize
any TACS variables (which are not yet calculated for time t = 0) by means of user-defined initial
conditions in the EMTP, or other sort of stratagem. In the TACS steady-state solution, the interface
signal of sources type 90, 91, 92, 93 are defined and accessible at time t = 0, since they have
already been calculated by the EMTP. The respective EMTP and TACS steady-state solutions cannot
be thought of as an interacting process, but as a sequential operation.
In any event, TACS always recognizes user-defined initial conditions. TACS always takes care of
initializing the DC and AC history terms of the transfer functions and transport delays.
In previous releases of TACS, the initialization algorithm was based on the steady-state solution of
transfer functions, except integrators, and the user had to provide DC initial conditions for any
other variables if needed. The AC contribution of all supplemental variables and devices was
totally ignored. For either the DC or AC steady-state TACS used two separate simultaneous solutions
of the transfer functions. The initial values and history terms for DC and AC steady-state were
superimposed to obtain final initial conditions before entering into the time step loop. Initialization
was one of the most troublesome issues because TACS could not guarantee fair initial values, in
some cases widely erroneous values were computed, causing additional transients. The algorithm
could not handle initial conditions of FORTRAN-like expressions at all. For the devices, some initial
conditions could be calculated, but not all of them. An integrator may not have a steady-state DC
solution, but it may have an AC solution, but this was excluded from the algorithm. In fact the
output of an integrator could be in error, even no transients took place in the time step loop, due to
incorrect initial conditions when the simulation started, i.e., initial condition errors were not always
self-correcting.
Starting from version 3.0, the concept of simultaneous solutions for the DC and AC steady-state
solution is abandoned since in most cases the simulation cannot be done without supplemental
variables/devices. A simultaneous solution is only used for super-blocks. Otherwise, the ordering
sequence is used to sequentially update the DC and AC initial values of the variables, including DC
and AC values for most supplemental variables/devices. For AC initial values, the value of each
frequency is calculated and stored separately. Then all DC and all AC history terms are computed
and superposition is applied to get the final initial values. When the algorithm detects that
superposition is no longer valid, because not all control systems can be solved exactly, a warning
message is issued and the simulation starts with 'approximate' initial conditions. A loop has been
implemented to minimize the impact of ordering in evaluating steady-state conditions of nonlinear
loops.
14 - 6

The algorithm uses Fourier analysis to handle TACS sources type 23 and 24. For supplemental
variables, i.e., pseudo-FORTRAN, initial values are computed for expressions of the form:
(AC or DC) + (AC or DC)
(AC or DC) (AC or DC)
(AC or DC) (AC or DC)
(AC) (AC)
(AC or DC) / DC
(AC) (AC) results in two AC values with different frequencies, or one AC and one DC component.
For other supplemental variables not in these categories, the initial conditions are set to zero.
If initial values for integrators are not provided, the algorithm assumes a DC value of zero while it
solves for its AC steady-state.
Note: Even though the steady-state initialization is not complete, version 3.0 is an improvement
over previous releases. Nevertheless, the user must remain aware of potential initialization
transients.
14.1.5
TACS input
The various classes or types of TACS data are described separately in the next sections. TACS data
can come in arbitrary order though.
Some special requests are specific to TACS and they are described in Section 4. To summarize
they are:
TACS EMTP SOURCES
TACS INIT
TACS ORDER
TACS STAND ALONE
TACS HYBRID
The end of TACS data must be specified by a blank line.
Note: Starting from version 3.0, the names SUBXXX, IF-xxx, DELxxx, ELSE and THEN ( where
xxx is a number ) are regarded as reserved and should not be used as input, output, named limits or
named control signals of TACS elements.
14 - 7
14.2 Signal Sources

signal sources are signal generators that can be used as input to other blocks and elements of
the system.
TACS
Some sources have reserved pre-defined names and do not require any source specification by the
user: it is sufficient to use them in the data file. TACS sources defined in EMTP (Request: TACS
EMTP SOURCES) can also be used as the pre-defined sources.
The pre-defined sources are, with all names left-justified:
TIMEX
simulation time in seconds ( = 0 at initialization )
ISTEP
number of the present time step of the simulation
DELTAT
integration step size in seconds
FREQHZ
frequency of first TACS sinusoidal source read-in, Hz
OMEGAR
2 FREQHZ, rad/sec
ZERO
0.0
MINUS1
1.0
PLUS1
+ 1.0
UNITY
+ 1.0
INFNTY
+ (very large number defined at installation)
PI
Every other TACS built-in source is defined on a single line of the data file, as a separate system
component having an alphanumeric name identifying usage of the source. The user specifies the
type and the name of the source according to the rules indicated below. The TACS built-in sources
are:
Type
Description
11
level signal
14
cosine signal
23
pulse signal
24
ramp signal
14 - 8

These sources can be started once and stopped once. Other sources are available from the EMTPTACS interfaces either from the special request for TACS EMTP SOURCEs or:
Type
Description
90
EMTP node voltage
91
EMTP switch current
92
selected variable of special EMTP components
93
EMTP switch status
94
selected variable of user-supplied digital relay models
By connecting an EMTP source to a node having a 1.0 shunt resistance, this source can be used
in TACS through the above type-90 specification.
Specification of sources:
All sources use the same card layout. The fields and format used vary according to the source
type.
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
E10.0
E10.0
The following default values are used for T-start and T-stop, which are rounded-up at the highest
multiple of the step-size:
T-start
T-stop
blank or 0
source is activated at t 0
value > 0
source is activated at t value
value < 0
source is activated at t 0
blank or 0
source is never de-activated
value
source is de-activated (=0) at t value
14 - 9
Type 11:
level signal
signal
signal
signal
T-start < 0
T-start
T-start T-stop
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
TYPE
Source type = 11
(1-2)
OUT
Output name
(3-8)
A
(11-20)
Amplitude A of signal
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
14 - 10
cosine signal = A cos ( 2f t + 2 /360 )
Type 14:
signal
A
T-start
T-stop
-A
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
E10.0
E10.0
TYPE
Source type = 14
(1-2)
OUT
Output name
(3-8)
A
(11-20)
B
(21-30)
Frequency, Hz
C
(31-40)
Phase shift Y, degrees
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
14 - 11
Type 23:
pulse signal
width
signal
T-start
T-stop
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
E10.0
E10.0
TYPE
Source type = 23
(1-2)
OUT
Output name
(3-8)
A
(11-20)
B
(21-30)
Period T, seconds
C
(31-40)
Width, seconds ( default value = step-size )
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
The initialization uses a Fourier decomposition and multi-frequency initialization.
14 - 12
Type 24:
ramp signal
signal
T
A
T-start
T-stop
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
E10.0
TYPE
Source type = 24
(1-2)
OUT
Output name
(3-8)
A
(11-20)
B
(21-30)
Period T, seconds
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
The initialization uses a Fourier decomposition and multi-frequency initialization.
14 - 13
Type 90:
EMTP node voltage
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
E10.0
TYPE
Source type = 90
(1-2)
(3-8)
Output name (node name already defined

in EMTP)
A
(11-20)
0. or blank for AC voltages

1.0 for DC voltage
B
(21-30)
frequency of AC voltage
0. or blank for DC voltage
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
OUT
The output value of the source is made equal to the value of the voltage (with respect to TERRA), as
measured at the node which bears the same name in the EMTP network. The information in the field
21-30 is used at initialization.
14 - 14
Type 91:
EMTP switch current
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
E10.0
E10.0
TYPE
Source type = 91
(1-2)
(3-8)
Output name (node name to which an

EMTP switch is connected)
A
(11-20)
0. or blank for AC current

1.0 for DC current
B
(21-30)
frequency of AC current
0. or blank for DC current
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
OUT
The output value of the source is made equal to the current flowing from the LEFT-node to the
RIGHT-node, as defined on the EMTP switch card. In cases where more than one switch is attached
to the said node, reference is assumed to be to the first of these switches, in the order of the data
input. The information in the field 21-30 is used at initialization.
14 - 15
Type 92:
Selected variable of special EMTP components
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
TYPE
Source type= 92
(1-2)
OUT
(3-8)
Output name (as defined in the data of the

special component)
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
This source type is reserved for passing to TACS the values of selected variables of special EMTP
components, for example, rotor angles and angular velocities of the Type-59 synchronous
machine.
14 - 16
Type 93:
EMTP switch status
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
TYPE
Source type = 93
(1-2)
OUT
Output name (node name to which an
(3-8)
EMTP switch is connected)
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
The output value of the source is made equal to the conduction status of the switch between the
LEFT-node to the RIGHT-node, as defined on the EMTP switch card. The output is 0 when the switch
is open, and 1 when the switch is closed. In cases where more than one switch are attached to the
said node, reference is assumed to be to the first of these switches, in the order of the data input.
14 - 17
Type 94:
Selected variable of user-supplied digital relay

models
TYPE
1
2
3
4
5
6
7
8
12 345678 90 1234567890 1234567890 1234567890 12345678901234567890 1234567890 1234567890
OUT
I2
A6
E10.0
E10.0
TYPE
Source type = 94
(1-2)
OUT
(3-8)
Output name (name used in the

FORTRAN code of the digital relay)
D
(61-70)
T-start, seconds
E
(71-80)
T-stop, seconds
The output value of the source is made equal to the variable in the user-supplied FORTRAN routine
COMPREL which serves to model digital relays.
14 - 18
14.3 Transfer Functions

This element is used to describe a transfer function of arbitrary order in the Laplace operator 's',
i.e. in the frequency domain. This applies to continuous time systems. The nth order transfer
function is defined as in the figure below. There can be up to 5 input signals u, but there is only
one output signal Y which identifies the block. See Section 14.3.1 for input format description.
The order N of the transfer function is max(n,m), which must be n for the transfer function to be
realizable.
high
U1
U2
G(s)
low
U5
m
N0 + N 1 s + + N m s
G ( s ) = gain -------------------------------------------------------------n
D 0 + D1 s + + D n s
There can be limits imposed on the transfer functions. These are of two types:
Static limiter: (soft, windup) - The output value of an order-zero or Z-block is always
calculated without considering any existing limit, and then, if the output value violates
either limit, the output is clamped at the violated limit.
Dynamic limiter: (hard, non-windup) - The output value of an S-block is calculated
depending on the limit violation status and dynamic limits should be used only when
the numerator of G(s) is of order zero. If a dynamic limit is violated, the derivatives of
the output are set to zero.
Above first order, or if there are zeroes in G(s), a warning message advises the user that the limits
or the zeroes are neglected. It is recommended to break down the transfer function into its
elementary components in order to be realistic in the handling of limits. In fact it could be argued
that limits should only be imposed on gains or on integrators.
Low and/or high limits of fixed or calculated value (named limit) can be imposed on order-zero
and first order blocks. The correct model of limiters is important to adequately model how the
transfer function comes in or out of saturation when either the input or the named limits take
different values.
14 - 19

If named limits become invalid (e.g. low limit > high limit), TACS issues a warning that the limits
have criss-crossed. The simulation is not stopped however because this may be due to initialization
errors. The user should certainly inspect the TACS data file if such a message is issued.
The user can specify different combinations of fixed and named limits. TACS interprets the
information as shown below. Cases 1-2 have no limits. A single limit is used for cases 3-8, while
two limits (lower and upper) are used in cases 9-12. The last 5 cases are degenerate cases.
A COMBINATION OF
Fixed Low
Fixed High
Named Low Named High Low Limit
0.
0.
0.
nameL(t)
nameH(t)
nameL(t)
0.
0.
nameH(t)
nameH
nameL(t)
nameH(t)
nameH
nameH(t)
nameL(t)
0.
0.
nameH(tx)
nameH(tx)
0.
nameL
6
7
nameH
1.
nameL
8
9
1.
L
11
nameL
L
12
13
14
L (> 0).
nameL
H (< L)
15
16
nameH
10
H (< 0.)
L
nameH (1)
17
nameL (2)
High Limit
1
2
IS INTERPRETED AS
Note:
1.
2.
for NameH(t) < L, at t = tx

for NameL(t) > H, at t = tx
There is a maximum limit of number of independent S-Z-block loops in TACS. If this number is
exceeded, a descriptive message will be issued.
14 - 20
14.3.1
Specification of S-blocks and Z-blocks
Line 1:
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
N
OUT
IN1
IN2
IN3
IN4
IN5
I2
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
E6.0
A6
A6
N
(1-2)
Order N of block (1)
OUT
Output name Y
(3-8)
(11)
+ or , according to the sign of u1
IN1
(12-17)
Name of 1st input signal u1
(19)
IN2
(20-25)
Name of 2nd input signal u2
(27)
IN3
(28-33)
Name of 3rd input signal u3
(35)
IN4
(36-41)
Name of 4th input signal u4
(43)
IN5
(44-49)
14 - 21

A
(51-56)
Gain The default value for the gain is one

(+1.0)
B
(57-62)
Fixed low limit (2)
C
(63-68)
Fixed high limit (2)
D
(69-74)
Named low limit (2)
E
(75-80)
Named high limit (2)
Note:
1.
The order N on the transfer function is max(n,m), which must be n for the
transfer function to be realizable. The special case where N = 0 is the order-zero
block or Z-block, otherwise TACS refers to the block as an S-block. The value of
N is right-justified.
2.
Limits on a S-block are dynamic, while limits of Z-blocks are static.
For a Z-block only the first line of data is needed. For an S-block it is necessary to specify the N
and D parameters, values not specified are defaulted to zero.
Lines 2 and 3:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
N0
N1
N2
N3
N4
N5
N6
N7
D0
D1
D2
D3
D4
D5
D6
D7
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
Line 2
Line 3
(1-10)
N0
D0
(11-20)
N1
D1
(21-30)
N2
D2
(31-40)
N3
D3
(41-50)
N4
D4
14 - 22
E10.0
E10.0

(51-60)
N5
D5
(61-70)
N6
D6
(71-80)
N7
D7
14.4 Supplemental Devices

These elements are called supplemental simply because they differ from the linear S-blocks and
Z-blocks. Each device is identified by the alphanumeric name of its output and its total input is the
sum of any or all signed input signals (except for devices type 60, 61 and 63 where each input
signal is interpreted individually). The sign character ( + or ) must be specified for all non-blank
input.
The different devices available are:
Type
Description
50
frequency meter
51
relay-operated switch
52
level-triggered switch
53
transport delay
54
pulse delay
55
digitizer
56
point-by-point user-defined non-linearity
57
multi-operation time-sequenced switch
58
controlled integrator
59
derivator
60
input-IF selector
61
signal selector
62
sample and track
63
instantaneous min/max
64
min/max tracking
65
accumulator and counter
66
RMS meter
14 - 23

Supplemental devices are identified by the 88 code in columns 1-2, and by the device type in
columns 9-10. Columns 1-2 may also contain 98 or 99, to be compatible with previous releases,
but there is no discrimination made. Miscellaneous improvements to the devices have been made
for version 3.0.
Specification of supplemental devices:

All devices use the same card layout. The fields and formats used vary according to the device
type.
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
E6.0
A6
A6
Type 50:
frequency meter
U1
U2
frequency meter
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
TYPE
88
(1-2)
OUT
Output name Y
(3-8)
CODE
50
(9-10)
14 - 24
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Initial frequency, Hz
The output will take this value until two
zero-crossings have occurred.
B
(57-62)
Maximum acceptable change in %
The time between successive zero-crossings of the net input signal U is measured and converted to
a frequency value in Hz. The device will automatically reject higher-frequency non-characteristic
oscillations in U by simple averaging with the previous value, but the user may choose to process
the input signal through a low-pass filter.
Linear interpolation between values of the net input changing sign at two successive time steps is
used to obtain the times of zero crossings.
In previous releases of TACS the frequency meter was malfunctioning if U had a DC offset, but this
has been corrected in v3.0.
14 - 25
Type 51:
Relay-operated switch
U1
U2
abs
U5
Sx = named variable + bias

For normally open relay:
For normally closed relay:
If |S(t)| Sx(t) then

Y(t) = K U
else
Y(t) = 0
endif
If |S(t)| < Sx(t) then

Y(t) = 0
else
Y(t) = K U
endif
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
I6
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
51
(11)
IN1
(12-17)
14 - 26
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Value of gain K ( default = 1.0 )
B
(57-62)
bias
C
(63-68)
Desired switch operation code (-2,-1,0,1,2)

0 or 2 normally open switch
1 normally closed switch
-2 normally open switch but closed at time 0-1 normally closed switch and closed at
time 0-
D
(69-74)
Name of NAMED level (default = ZERO)
E
(74-80)
Name of driving signal S
14 - 27
Type 52:
Level-triggered switch
U1
U2
U5
Sx = named variable + bias

For normally open relay:
For normally closed relay:
If S(t) Sx(t) then

Y(t) = K U
else
Y(t) = 0
endif
If S(t) < Sx(t) then

Y(t) = 0
else
Y(t) = K U
endif
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
I6
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
52
(11)
IN1
(12-17)
(19)
14 - 28

IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
+ or -, according to the sign of u5
IN5
(44-49)
A
(51-56)
Value of gain K ( default = 1.0 )
B
(57-62)
bias
C
(63-68)
Desired switch operation code (-2,-1,0,1,2)

0 or 2 normally open switch
1 normally closed switch
-2 normally open switch but closed at time 0-1 normally closed switch and closed at
time 0-
D
(69-74)
Name of NAMED level (default = ZERO)
E
(74-80)
Name of driving signal S
14 - 29
Type 53:
Transport delay
U1
U2
delay of T sec
Y(t) = U(t - T(t))
T(t) = Tf + NAMED(t)
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 56789 0
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
E6.0
A6
A5
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
53
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
14 - 30

IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
DC history value
B
(57-62)
Value of fixed delay Tf, sec
C
(63-68)
Maximum value of delay, sec
D
(69-74)
Name of NAMED delay (default = ZERO)
E
(75-79)
"VALUE"
This keyword must be supplied so that the
DC history in field A is used, otherwise
TACS calculates the DC and AC history
terms.
If a NAMED delay is used, the user must specify the maximum value of the delay so that enough
memory can be allocated before starting the simulation. This device can be storage hungry.
14 - 31
Type 54:
Pulse delay
U1
U2
level
detector
delay of
T seconds
Y(t) = P(t-T(tx))
T(tx) = Tf + NAMED(tx)
T(tx)
U5
1.0
tx
Control of initial response:
1.0
1.0
t
T-on
T-off
t
T-on
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
E6.0
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
54
14 - 32
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Value of T-on, sec ( default = large negative

number )
B
(57-62)
Value of fixed delay Tf, sec
C
(63-68)
Value of T-off, sec ( default = large

negative number )
D
(69-74)
Name of NAMED delay (default = ZERO)
This device is much more memory efficient than device 53. The frequency of the input signal U
must be such that the period of U > ( delay + pulse_width).
The values of T-on and T-off are used to make it possible for the device to respond to an input
signal which partly or totally occurred before t=0 . The program does not make any distinction
between a blank field or value 0.0. Therefore, for T-on or T-off equal to 0.0, a value that is
negative and very close to it ( < t ) must be used instead.
14 - 33
Type 55:
Digitizer
U1
U2
n levels
If { X(t) level1 } then

Y(t) = level1
elseif { levelk < X(t) levelk+1 for any k<n } then
Y(t) = levelk
else
Y(t) = leveln
endif
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012345678901234567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
Level1
Level2
Leveln
9999.
First line:
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
(11)
88
Output name Y
55
14 - 34

IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Gain K ( default = 1.0 )
Following lines: ( one line of input per level of discretization, in increasing order )
Line
Columns
Description
Format
(1-16)
Value of level1
E16.0
(1-16)
Value of level2
E16.0
...
...
n+1
(1-16)
Value of leveln
n+2
(13-16)
9999. ( flag indicating end of device )
...
E16.0
This device discretizes the continuous input signal U according to the levels, and the input gain
specified.
14 - 35
Type 56:
Point-by-point user-defined non-linearity
U1
U2
n pairs of (X,Y) points
The points do not have to

be uniformly spaced.
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012345678901234567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
X1
Y1
X2
Y2
Yn
Xn
9999.
First line:
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
56
(11)
IN1
(12-17)
(19)
14 - 36

IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Following lines: ( one line of input per (X,Y) pair, in increasing order of X )
Line
2
...
n+1
n+2
Columns
Description
Format
(1-16)
Value of X1
E16.0
(17-32)
Value of Y1
E16.0
(1-16)
Value of X2
E16.0
(17-32)
Value of Y2
E16.0
...
...
(1-16)
Value of Xn
E16.0
(17-32)
Value of Yn
E16.0
(13-16)
Use as many pairs of (X,Y) points as needed to cover the expected variation. Linear interpolation
is used between two pairs of points.
14 - 37
Type 57:
Multi-operation time-sequenced switch
U1
U2
T-close-1
T-open-1
T-close-2
T-open-2
T-last
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012345678901234567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
T-close-1
T-open-1
T-close-2
T-last
9999.
First line:
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
57
(11)
IN1
(12-17)
(19)
14 - 38

IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Following lines ( one line of input per event, in increasing order of time ) :
Line
Columns
Description
Format
(1-16)
Value of T-close-1, sec
E16.0
(1-16)
Value of T-open-1, sec
E16.0
(1-16)
Value of T-close-2, sec
E16.0
(1-16)
Value of T-open-2, sec
E16.0
...
...
...
n+1
(1-16)
Value of last operation ( nth operation), sec
n+2
(13-16)
E16.0
This switch is open at initialization, but it is possible to assign any initial condition to its output.
14 - 39
Type 58:
Controlled integrator
U1
U2
D0 + D1 s
If { control(t) 0.0 or no control signal } then
1
Y = -------------------------X
D 0 + D1 s
else
U5
Y = reset(t)
endif
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
E6.0
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
58
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
14 - 40

IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
B
(57-62)
D0 ( default = 0.0 )
C
(63-68)
D1 ( default = 0.0 )
D
(69-74)
Name of the control signal ( if any )
E
(75-80)
Name of the reset signal ( default = ZERO )
This device can also be used as a counter, but the user should then consider device-65 instead. To
use device-58 as a counter, use COUNTR as the control signal name and leave fields A, B, C and E
blank. Then we obtain:
Y(t) = Y( t - T ) + X(t)
where T is the step-size.
14 - 41
Type 59:
Derivator
U1
U2
X ( t ) X ( t T )
Y ( t ) = ----------------------------------------T
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012345678901234567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
(11)
88
Output name Y
59
IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
14 - 42
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
The step-size is T. This device will be properly initialized by TACS provided the initial conditions
for all its input data are initially available. In any event, the user may choose to supply the initial
conditions.
14 - 43
Type 60:
Input-IF selector
X1
U1
X2
U2
X3
U3
If { S1(t) < (S2(t) +bias) } then

Y(t) = X1
elseif { S1(t) = (S2(t) + bias) } then
Y(t) = X2
else
Y(t) = X3
endif
S1
S2 + bias
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
A
1
A6
A
1
A6
A
1
A6
E6.0
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
60
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
14 - 44

IN3
(28-33)
A
(51-56)
bias ( default = 0.0 )
D
(69-74)
Name of signal S1 (default = ZERO )
E
(75-80)
Name of signal S2 ( default = ZERO )
The signal Sl may be the output of a complex expression. If Sl is blank, then the switch is
controlled as:
If { (S2(t) + bias) > 0 } then
Y(t) = X1
elseif { (S2(t) + bias) = 0 } then
Y(t) = X2
else
Y(t) = X3
endif
14 - 45
Type 61:
Signal selector
LOW
U1
U2
S(t)
Y(t)
Default
X1
< 0.5
LOW
X2
1.0 0.5
X1
0.0
2.0 0.5
X2
0.0
3.0 0.5
X3
0.0
4.0 0.5
X4
0.0
5.0 0.5
X5
0.0
6.0 0.5
X6
0.0
HIGH
X3
U3
X4
U4
X5
U5
X6
U6
HIGH
6.5
S
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
61
(11)
IN1
(12-17)
(19)
14 - 46

IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
B
(57-62)
LOW
C
(63-68)
HIGH
D
(69-74)
Name of 6th input signal u6 ( default =

ZERO )
E
(75-80)
Name of selector variable S
14 - 47
Type 62:
Sample and track
U1
U2
HOLD
SAMPLE
TRACK
1.0
TRACK SAMPLE
U5
i.c.
i.c.
OUTPUT
OUTPUT
INPUT
INPUT
>0
TRACK
0
>0
SAMPLE
0
Sampling of a new value is performed on the rising front of the SAMPLE signal (negative to
positive), which must then be reset to < 0 before the next sampling occurs. Tracking starts on the
rising front of the TRACK signal (negative to positive) and continues until TRACK signal supersedes
the SAMPLE signal, should both signals be > 0 simultaneously.
In the absence of both TRACK and SAMPLE signals; the output is in the HOLD mode.
Tracking becomes equivalent to sampling if the TRACK signal becomes positive for only one time
step.
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A6
A6
TYPE
(1-2)
88
14 - 48

OUT
(3-8)
CODE
(9-10)
Output name Y
62
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
D
(69-74)
Name of TRACK signal ( default = ZERO )
E
(75-80)
Name of SAMPLE signal ( default = ZERO )
Initial conditions must be user-supplied, otherwise 0. is assumed.
14 - 49
Type 63:
Instantaneous min/max
X1
U1
MIN
X2
U2
X3
U3
or
X4
U4
MAX
X5
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
63
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
14 - 50

IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
D
(57-62)
1.0, for minimum selection

+1.0, for maximum selection
14 - 51
Type 64:
Min/max tracking
U1
U2
1.0
MIN
or
MAX
RESET
U5
HOLD
reset_value
If { RESET(t) > 0 } then

Y(t) = reset_value
elseif { HOLD(t)> 0 } then
Y(t) = Y(t - T)
else
If { MIN operation } then
Y(t) = min ( Y(t - T), X(t) )
else
Y(t) = max ( Y(t - T), X(t) )
endif
endif
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
E6.0
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
64
(11)
IN1
(12-17)
14 - 52
(19)
IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
B
(57-62)
1.0, for minimum selection

+1.0, for maximum selection
C
(63-68)
reset_value
D
(69-74)
Name of the RESET signal ( default = ZERO

)
E
(75-80)
Name of the HOLD signal ( default = ZERO )
The step-size is T. Initial conditions can be user-supplied; Otherwise 0.0 is assumed, unless all
input signals are initially known.
14 - 53
Type 65:
Accumulator and counter
U1
U2
1.0
Accumulator
or
Counter
RESET
U5
HOLD
reset_value
If { RESET(t) > 0. } then

Y(t) = reset_value
elseif { HOLD(t)> 0. } then
Y(t) = Y(t - T)
else
Y(t) = Y(t - T) + X(t)
endif
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012 345678 901234 567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
A6
A6
TYPE
(1-2)
OUT
(3-8)
CODE
(9-10)
88
Output name Y
65
(11)
IN1
(12-17)
(19)
14 - 54

IN2
(20-25)
(27)
IN3
(28-33)
(35)
IN4
(36-41)
(43)
IN5
(44-49)
C
(63-68)
reset_value
D
(69-74)
Name of the RESET signal ( default = ZERO

)
E
(75-80)
Name of the HOLD signal ( default = ZERO )
If the input to this device is made PLUS1, then we obtain a step counter, with or without reset.
The step-size is T. Initial conditions can be user-supplied; otherwise 0. is assumed, unless all
input data are initially known.
14 - 55
Type 66:
RMS meter
U1
U2
RMS
meter
U5
TYPE
OUT
CODE
1
2
3
4
5
6
7
8
12 345678 90 1 234567 8 9 012345 6 7 890123 4 5 678901 2 3 456789 0 123456 789012345678901234567890
I2
A6
I2
IN1
IN2
IN3
IN4
IN5
A
1
A6
A
1
A6
A
1
A6
A
1
A6
A
1
A6
E6.0
TYPE
88
(1-2)
OUT
OUTPUT NAME Y
(3-8)
CODE
66
(9-10)
(11)
IN1
(12-17)
(19)
IN2
(20-25)
(27)
IN3
(28-33)
14 - 56
(35)
IN4
(36-41)
(43)
IN5
(44-49)
A
(51-56)
Base frequency, Hz
The storage needed for this device depends on the frequency and the step size. The dimension of
the array used for storage of input points is:
1
points = 1 + ------------------------------------frequency t
where the step-size is T.
14 - 57
14.5 Supplemental Variables And Pseudo-FORTRAN

These supplemental variables are the output of pseudo-FORTRAN statements within TACS. The freeformat input may take several lines. Input can also be with the older fixed-format, which is
somewhat faster at execution but has limited capabilities. TACS can still accept old fixed format
input, but the format is not documented in this manual and its use is not recommended.
1
2
3
4
5
6
7
8
12 345678 90 1 234567890123456789012345678901234567890123456789012345678901234567890
88
OUT
I2
A6
A
1
free-format FORTRAN expression
First line of input:

Supplemental variables are identified by the 88 code in columns 1-2, and by the "=" sign in column
11. Columns 1-2 may also contain 98 or 99, to be compatible with previous releases, but there is no
discrimination made. The pseudo-FORTRAN expression starts in column 12.
Columns
Description
(1-2)
88
(3-8)
Output name Y
(11)
(12-80)
pseudo-FORTRAN expression
Continuation lines:
The pseudo-FORTRAN expression can be continued on as many lines as necessary, provided the
continuation character "$" is specified on each line to be continued. Columns 1-80 are available for
input on these cards.
14.5.1
Subset of FORTRAN available
All the rules for constructing standard FORTRAN expressions apply but only a limited number of
FORTRAN statements are available (e.g., DO loops, GOTO, SUBROUTINES are not allowed), while
additional functions are provided. All names are at most 6 characters long and cannot contain the
symbols "+", "-", "*", "/", ")", "(" or "$". All blank spaces in names or numbers are eliminated. For
14 - 58

example "1.0 E6" will be interpreted as 2 fields: 1.0 followed by E6, this is meaningless and an
error message will be generated.
All operators or functions must be in upper case. All variables are real variables even those
normally associated with logical relational operators. For example, x = ( .NOT. variable ) is 1 (for
.TRUE.) if variable > 0, else x = 0. Arguments specified as integers are handled as real variables.
Valid arguments are either a TACS name, a numerical constant, or another function.
The number of nested parentheses in each expression cannot exceed a rather large number, fixed
by TACS, and which is checked when constructing the expression. If a message regarding the
number of nested parenthesis is issued, the user should break the expression into smaller ones.
Some valid names used in function blocks, sources or devices cannot be recognized in free format
pseudo-FORTRAN expressions, e.g.:
Names with embedded operators, such as "+" or ".NE."
Names starting with a ".", or ending by a "." (in fact names with "." should be avoided)
Names with embedded "$"
14.5.2
IF-THEN-ELSE Structures
Starting from version 3.0, "IF-THEN-ELSE" structures have become available. They allow the
program, during the time step solution, to execute different paths depending on the logical values
of the IF argument (this can be a variable or a valid expression).
Inside an "IF-THEN-ELSE", there can only be pseudo-FORTRAN statements and they can have 88, 98
or 99 in columns 1-2, or they can have S and a blank or 2 blanks in these columns.
Either the IF or ELSE sections can be left blank.
''IF-THEN-ELSE" structures cannot be nested.
If the total number of "IF-THEN-ELSE" structures is exceeded, a message will be issued and the
actions to be taken will be indicated. Each IF has a logical expression associated with it. It is called
IF-xxx, where xxx is a sequential number starting from 001. The variable IF-xxx is available for
reporting.
When using "IF-THEN-ELSE" the user must be careful to create valid and meaningful structures.
Consider for example:
...
IF (X1 .GE. 1.0) THEN
SF1 = 0.5*X
14 - 59

ELSE
SF2 = - 0.5*X
ENDIF
...
This structure is valid provided that X and X1 are defined elsewhere in TACS. Note that it is
interpreted as:
IF X1(t) > 1.0 then
F1(t) = 0.5 * X(t)
F2(t) = F2(t-DT)
ELSE
F1(t) = F1(t-DT)
F2(t) = - 0.5 * X(t)
ENDIF
The user should be aware that it is not possible to redefine the output of pseudo-FORTRAN
expressions that are inside an IF-THEN-ELSE structure anywhere in the data deck, not even inside
another IF-THEN-ELSE structure.
Pseudo-FORTRAN operators and functions:

Algebraic operators
+, -, *, /, **
Relational operators
.EQ., .NE., .LT., .LE., .GE., .GT., .OR., .AND., .NOT.
FORTRAN
functions
SIN, COS, TAN, COTAN

ASIN, ACOS, ATAN
(argument in rad)
(answer in rad)
sinh, cosh, tanh

exp, log, log10, sqrt, abs
TACS
functions
TRUNC(x)
integer part of x
MLNUS(x)
-x
INVRS(x)
1.0/x
RAD(x)
conversion from degrees to radians
DEG(x)
conversion from radians to degrees
SIGN(x)
-1.0 if x < 0, else = +1.0
NOT(x)
0.0 if x > 0, else = +1.0
SEQ6(x)
modulo_6( INT(x) ), that is

for x =
SEQ6(x)
14 - 60
... -1 0 1 2 3 4 5 6 7 8 ...
= ... 5 6 1 2 3 4 5 6 1 2 ...

RAN(x)
random number where x is an arbitrary

number or an existing TACS name
if-then-else
IF(
pseudo-FORTRAN control expression ) THEN

VAR1
= (pseudo-FORTRAN expression)
VAR2
else
VAR3
VAR4
endif
14.6 FORTRAN Interface

Although pseudo-FORTRAN is useful and input-flexible, the implementation suffers from various
characteristics on top of the limitations already imposed by the TACS syntax (e.g. each block has a
single output signal)
it is slow.
it has a limited number of FORTRAN statements (SUBROUTINE, GOTO and DO loops

are missing).
large pseudo-FORTRAN data files require considerable amount of storage.
except for statements inside IF-THEN-ELSE, every TACS statement is executed at each
time step, which may not be the case in the real control system.
Version 3.0 provides a TACS-FORTRAN interface which offers the following control flexibility (at
the expense of more complex input rules, i.e. need to compile and link):
users can model anything inside FORTRAN subroutines.
the entire subroutines do not have to be executed at each time step.
there are multi-input and multi-output signals associated with the block which is
used to identify the subroutine.
FORTRAN
routines are ordered in the control system as per their I/O variables.
14 - 61

The TACS routine INFACE allows the interface between TACS and the user-supplied routines to be
made. This is an interface routine for transferring input and output variables between the TACS
program and user written FORTRAN subroutines. Each time TACS needs the user FORTRAN routine
to do calculation during the time step solution, TACS will call this routine, pass the routine name as
a character string and pass the input variables to user FORTRAN routine. Then, the output variables
will be passed back to TACS.
The user FORTRAN routine name is passed as a routine parameter. The input variables, output
variables and other three special variables - ISTEP, DELTAT and LUN006 - are passed through named
COMMON blocks. The parameter order in the array ENARG-OTARG must correspond to the order of
the input-output parameters of the TACS FORTRAN routine call. Internally, each user-supplied
subroutine usage is assigned the name SUBxxx, where xxx is a sequential number starting at 001.
Although TACS could handle FORTRAN routine names as big as 36 characters, it is strongly
suggested to first check the maximum size your FORTRAN compiler can handle. The standard
FORTRAN 77 can handle only 6 character names. If your FORTRAN compiler can handle more, you
may use more, but not more than 36 characters.
The maximum number of input and output parameters in each user subroutine is limited at 18
(Parameter MAXINT) of the routine INFACE.
After creating the interface routine, compile it together with your own routines. Then, link them
with the EMTP object libraries. Be sure to type the user routine names exactly the same as in your
TACS input data cases and your own FORTRAN routines.
Usage of FORTRAN interface:

Interface routine code :
C------------------------------------------------------------------------------C
1
2
3
4
5
6
7
8
C2345678901234567890123456789012345678901234567890123456789012345678901234567890
C-------------------------------------------------------------------------------
SUBROUTINE INFACE( RTNAM )

C
INCLUDE 'implidp.f77'
INCLUDE 'fortint.f77'
C
DOUBLE PRECISION ENARG, OTARG
C
C
C
C
C
C
C
C
Next COMMON block is for input parameters. The parameter order

in the array ENARG is following the order in the corresponding
data card for input parameters of user FORTRAN routine call.
COMMON /INPPAR/ ENARG(MAXINT)
Next COMMON block is for output parameters. The parameter order
in the array OTARG is following the order in the corresponding
data card for input parameters of user FORTRAN routine call.
COMMON /OUTPAR/ OTARG(MAXINT)
14 - 62

C
C
C
C
C
C
Next COMMON block is for some special variables that user routine
may need in users' subroutine.
DELTAT - time step in seconds
ISTEP - number of time step
LUN006 - number of logical unit to be used for output of messages
in EMTP printout file
COMMON /TIMVAR/ DELTAT, ISTEP, LUN006
C
CHARACTER*(*) RTNAM
C
C
C
C
C
C
C
C
C
C
When you need call your own FORTRAN routine from TACS, please
replace USER-SUBROUTINE_x with your own routine name.
==================================================================
NEXT IS AN EXAMPLE OF USER FORTRAN ROUTINE CALLS
IF ( RTNAM .EQ. 'USER_ROUTINE_1 ' ) THEN
CALL USER_ROUTINE_1
==================================================================
NEXT IS ANOTHER EXAMPLE OF USER FORTRAN ROUTINE CALLS
ELSE IF ( RTNAM .EQ. 'USER_ROUTINE_2' ) THEN
CALL USER_ROUTINE_2
==================================================================
ELSE
( Routine not found - output error message )
WRITE (LUN006,1000) RTNAM
CALL STOPTP
{ STOPTP is an EMTP routine }
END IF
C
RETURN
C
C
( Error message format )

1000 FORMAT ( 5X, 'THE EMTP HAS DIFFICULTY TO MATCH AN USER ',
1 'FORTRAN CALL. THE ROUTINE NAME SUPPOSED TO BE:'// 10X,A36//
2 5X, 'PLEASE CHECK THE USER FORTRAN ROUTINE NAMES IN YOUR ',
3 'INPUT DATA FILE AND IN THE INTERFACE ROUTINE "INFACE" AND '/
4 5X, 'YOUR OWN FORTRAN ROUTINES. MAKE SURE NOT HAVE ANY ',
5 'TYPING MISTAKES.')
C
END
To make a call to a user-defined FORTRAN subroutine the user has to add to the TACS data deck at
least 3 cards for each routine, according to the input format below:
14 - 63
X1
U1
U2
U3
SUBROUTINE
U4
X2
X3
X4
1
2
3
4
5
6
7
8
1 2 345678 9 0 123456 7 8 901234 5 6 789012 3 4 567890 1 2 345678 9 0 123456 7 8 901234 5 6 789012 3456789 0
IN1
OUT1 +
IN2
OUT2 +
IN3
OUT3 +
IN4
OUT4 +
IN5
OUT5 +
IN6
OUT6 +
IN7
OUT7 +
IN8
OUT8 +
IN9
OUT9
First card:
Columns
Description
(1-10)
SUBROUTINE
(13-51)
Format
( keyword )
A10
A36
NAME_OF_SUBROUTINE
Input description card(s):

Columns
Description
Format
(2)
A1
(3-8)
A6
(10)
A1
(11-16)
A6
(18)
A1
(19-24)
A6
(26)
A1
14 - 64
CONT CONT
NAME_OF_SUBROUTINE
SUBROUTINE

(27-32)
A6
(34)
A1
(35-40)
A6
(42)
A1
(43-48)
A6
(50)
A1
(51-56)
A6
(58)
A1
(59-64)
A6
(66)
A1
(67-72)
A6
(80)
Continuation flag
A1
If there are more than 9 input variables use "$" at this

column and follow this card with another with the same
format.
Output description card(s):
Columns
Description
Format
(3-8)
Name of 1st output Y1
A6
(11-16)
Name of 2nd output Y2
A6
(19-24)
Name of 3rd output Y3
A6
(27-32)
Name of 4th output Y4
A6
(35-40)
A6
(43-48)
A6
(51-56)
A6
14 - 65

(59-64)
A6
(67-72)
A6
(80)
Continuation flag
A1
If there are more than 9 output variables use "$" at this

column and follow this card with another with the same
format.
14.7 Initial Conditions

The initial value of variables at t = 0.0 not covered by the program initialization calculations, can
be provided by the user using the format shown below.
Typically, the following elements will require user-supplied initial conditions for t=0.0. The
default value of unspecified initial conditions is 0.0.
DC values of integrators.
output of some devices.
DC values for Type 23 and 24 sources, if the internal Fourier analysis of these sources
is not sufficient.
Syntax:
1
2
3
4
5
6
7
8
12 345678 90 1234567890 123456789012345678901234567890123456789012345678901234567890
77 NAME
VALUE
I2
E10.0
A6
(1-2)
77 ( flag for initial conditions )
NAME
(3-8)
Name of variable for which DC initial

conditions are provided
VALUE
Initial condition
(11-20)
The initial conditions may be specified anywhere in the TACS data file.
14 - 66
14.8 Output Specifications

TACS variables are available for output at each simulation time step. The printing and plotting
intervals are controlled by the parameters IOUT and IPLOT, as with EMTP variables (miscellaneous
input section).
Diagnostic printout of TACS tables is also controllable to obtain information relevant to the
internal data structure (variable IPRSUP or keyword DIAGNOSTIC).
To specify which TACS variables are to be available as output, use as many output specification
lines as needed:
Syntax ( selected output ) :

1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
33
NAME1
NAME2
NAME3
NAME4
NAME5
NAME6
NAME7
NAME8
I2
A6
A6
A6
A6
A6
A6
A6
A6
(1-2)
33 ( flag for output )
NAME1
Name of output variable
NAME9 NAME10 NAME11 NAME12 NAME13

A6
A6
A6
A6
A6
(3-8)
NAME2
(9-14)
...
NAME13
...
(75-80)
Any number of output requests can be made on the Type-33 line (up to the maximum allowed by
program dimensioning) .
To output all TACS variables, use a single output request card instead :
Syntax (global output) :

Columns
2
Field Name
-------------
Description
1 (flag requesting all variables)
14 - 67
Format
I1

In the "plot" file of TACS HYBRID simulations, TACS variables are classified as branch currents or
Type-9 variables. The identification of variable NAME is TACS, NAME
In the "plot" file of TACS STAND ALONE simulations, TACS variables are classified as node
voltages (Type-1 variables), and only the name of the TACS variable is used as identifier.
14 - 68
Section 15
15.1
15.2
15.3
Batch Mode Plotting ........................................................................... 15-1

15.1.1
General Considerations .................................................. 15-1
15.1.2
Postscript Basics ............................................................ 15-3
15.1.3
Input Data Card Format For Batch-Mode Plotting ........ 15-4
Fourier Analysis .................................................................................. 15-25

15.2.1
Card For Fourier Series Computation ............................ 15-25
15.2.2
Background for EMTP Fourier Series Computation ....... 15-28
Fatigue Analysis .................................................................................. 15-30

15.3.1
15.4
Output Processing
Cards for Fatigue Analysis ............................................. 15-30
Output Processing with MATLAB ........................................................ 15-37

15.4.1
Installation ...................................................................... 15-37
15.4.2
Usage Rules .................................................................... 15-38
15.4.3
Command panel ............................................................. 15-39
Section 15
Output Processing
The EMTP has built-in output processing capabilities such as batch-mode plotting, Fourier and
fatigue analysis. Post-processing operates on the plot or "pl4" file, and can take place at the end of
a simulation or by reading a previously-generated plot file (see Section 4.1.4).
15.1 Batch Mode Plotting

15.1.1
General Considerations
Rule 1:
Any number of plots can be made from any one solved data case. Up to four curves
per plot are permitted. Only curves of the same type (among the five different classes:
node voltage, branch voltage, branch current, branch power, branch energy) can generally be plotted on the same graph.
Rule 2:
Three batch plotting modes are available:

(A)
A high quality graph printable on any Postscript-compatible printer (i.e., an

ACSII file with Postscript commands). This will be referred to as the
"POSTSCRIPT PLOT" mode of operation.
(B)
A relatively simple graph generated on the output file. This will be referred to
as the PRINTER PLOT.
(C)
Both Postscript and PRINTER plots. This will be referred to as the

"POSTSCRIPT PRINTER" mode of operation.
It is possible to switch back and forth between these three modes during the course of
these plot requests.
Rule 3:
Branch current, power or energy plots for two or more elements which are connected
in parallel are not possible at the present time. Since plot branches are identified only
by terminal node-pair names, the program will always pick the first of the two or more
qualifying candidates which it finds in the output vector. A commonly used trick to
output from two parallel branches connected between nodes A and B, is to specify one
branch between nodes A and B while the other is specified between nodes B and A.
15 - 1
Output Processing
Rule 4:
There is no limit on the number of data points which go into the production of each
plot.
Rule 5:
There are a number of plot request cards which control the appearance of the plotted
output. These cards are described in Section 15.1.3. A BLANK card indicates the end
of all plot (as well as FOURIER and FATIGUE) specification cards.
Rule 6:
When a Postscript plot is generated, the program first discards those data points which
will not visually affect the plot, according to the following "smoothing" algorithm.
lin
e
d
on
st
d te
line
d6
d5
5
tte
d3
d4
s ec
4=n
pl o
al
ac
tu
e
lin
gh
ou
r
th
ts
in
po
d
an
d3<
d4>
The first point of raw plot data is always retained, and an imaginary line is drawn from
it through the second point of plot data. Starting with the third point, then, the distance
"d" of each succeeding point from the line is calculated, and compared to a smoothing
tolerance "":
(A)
As long as d , the distance calculation is simply repeated for the following

point.
(B)
Eventually, for some n, it is found that the n-th point is further removed than
from the line (d > ). In this case, the (n-1)th point is saved, and all points
between the (n-1)th point and the last previously-saved point are discarded.
This is based on the principle that throwing away such points will not
significantly affect the resulting graph visually, since they lie on, or almost on,
the line about which the distances were calculated.
Such distance-testing is then continued recursively, following the sample just outlined.
It is begun by drawing a new line from the (n-1)-st point through the n-th point, and
then checking distances from this new line for points numbered n+1 onward.
The one exception to the distance comparison based on "" as just described is that
pertaining to relative minima or relative maxima; extrema points are always saved,
regardless of their distance from the line. In effect, = 0 is used for the calculation at
extrema.
15 - 2
Output Processing
The value of smoothing tolerance "" is under user control, by means of a special
"SMOOTH" request card.
Note that PRINTER plotting is not affected by smoothing.
15.1.2
Postscript Basics
The implementation of Postscript batch mode plotting in version 3.0 of the EMTP is a direct
translation of the old Calcomp style plots from the days when the EMTP only ran on mainframe
computers. Although there is no "standard" for hardcopy devices, the widespread availability and
relative low cost of Postscript (and Postscript compatible) laser printers have provided the closest
approximation to a hardware/software standard that is possible nowadays.
Postscript is a page description language used to produce plain ASCII files with commands that
Postscript printers can interpret. It is called a page definition language because it assumes a
drawing surface (or page) which typically corresponds to an 8.5" x 11" sheet of paper. The unit of
measure in Postscript is the point which corresponds to 72 points per inch. An 8.5" x 11" drawing
surface is then 792 by 612 points (landscape orientation), and the origin is in the lower left corner
of the page.
The Postscript files created by the EMTP can be imported by most word processors and desktop
publishing programs. They should be imported as Encapsulated Postscript (EPS) files.
Producing Output:
Postscript output is a plain ASCII file that contains the page description of one or more images
(plots). Normally, the Postscript interpreter inside the printer checks the first line of the file to
decide if it has received a legitimate Postscript file. Some laser printers always assume that the
received file is a Postscript file.
By default, the EMTP will produce a Postscript file with the same name as the pl4 file but with the
extension "ps". For example, if the pl4 file is called "abcd.pl4", the resulting Postscript file will be
called "abcd.ps". If the pl4 file is called "abcdef" (period is absent), the Postscript file will be
called "abcdef.ps". Additionally, the user can specify an arbitrary file name using the control
parameter file "emtpps.par" (see Postscript Filename card in Section 15.1.3).
Once the Postscript file has been created (e.g., "xxx.ps"), simply send it to the printer. The exact
syntax varies from system to system. Here are a few examples.
copy xxx.ps lpt1 (DOS)
print xxx.ps (DOS, VAX-VMS)
pr -Pprintername xxx.ps(Unix)
15 - 3
Output Processing
As indicated above, the Postscript files created in this implementation of batch mode plotting are
actually simple EPS files, and as such, they can be imported directly by most word processors and
desktop publishing programs. To import a plot into a word processing program there should only
be one plot per file.
Font List:
It is possible to use any valid Postscript font in different parts of a plot (see Headings font and
Annotations font cards). A small but nearly universal list of Postscript fonts is given below.
AvantGarde-Book
AvantGarde-BookOblique
Avantgarde -Demi
AvantGarde-DemiOblique
Bookman-Demi
Bookman-DemiItalic
Bookman-Light
Bookman-LightItalic
Courier
Courier-Bold
Courier-BoldOblique
Courier-Oblique
Helvetica
Helvetica-Bold
Helvetica-BoldOblique
Helvetica-Narrow
Helvetica-Narrow Bold
Helvetica-Narrow-BoldOblique
Helvetica-Narrow-Oblique
Helvetica-Oblique
NewCenturySchlbk-Bold
NewCenturySchlbk-BoldItalic
NewCenturySchlbk-Italic
NewCenturySchlbk-Roman
Palatino-Bold
Palatino-BoldItalic
Palatino-Italic
Palatino-Roman
Times-Bold
Times-BoldItalic
Times-Italic
Times-Roman
ZapfChancery-MediumItalic
15.1.3
Input Data Card Format For Batch-Mode Plotting
The following cards permit control of the plotted images. These cards are optional except for the
plot specification card and the BLANK card to indicate the end of the plot specification cards. The
FOURIER and FATIGUE request cards can also be specified within this card group. The cards in this
group are:
Case-Title Card
Subheading Text Card
Plot Specification Card
15 - 4
Output Processing
Card for input of 3 or 4 Branch-Variable Names
Card for Extended Precision of Floating-Point Fields
Card for Graph-Size Adjustment
Card for Redefinition of Smoothing Tolerance ""
Card for Changing the Graphic Mode Selection
Card for Changing Pens and Grid Status
Card for Changing Line Limit on Sparse Printer Plots
Card to Superimpose Plots and/or Suppress Vertical-Axis Labelling
Card for Turning On or Off the Plot Identifier Symbol
Card for Plotting one EMTP Variable versus Another
Card to Begin Each Printer Plot on a New Page
Headings font specification
Annotations font specification
Postscript Parameters
Postscript filename
Card to Mark the Termination of all Plot Cards
Case-Title Card:
This card has a "2" in column number 2, followed by up to 78 arbitrary characters of case-title
text. This text will then appear at the top of all plots which are requested after this definition
until redefined at some later stage of the plot-data input. Blank text is automatically provided on
all plots which precede the user's first definition. The format for data input is:
1
2
3
4
5
6
7
8
12 345678901234567890123456789012345678901234567890123456789012345678901234567890
2
Arbitrary 78 character case-title text
15 - 5
Output Processing
Subheading Text Card:

Zero to four such data cards for defining graph subheading text are permitted. This text will appear
as a subheading on only the immediately-following graph. Each line is 78 characters in length.
Each subheading data card should have columns 1-2 blank, with text entered in columns 3-80, as
per the following format:
1
2
3
4
5
6
7
8
12 345678901234567890123456789012345678901234567890123456789012345678901234567890
Arbitrary 78 character case-title text
Note that misspelled plot commands may appear as sub-headings in a plot (e.g., "POSCRIPT PLOT"
would appear as a sub-heading).
Plot Specification Card:

This is the only plot card type actually required to produce a graph. One graph will be produced on
the Postscript and/or line printer for each such plot card. The card format is as follows:
GTYPE
UNITS
X-scale
FLAG
1
2
3
4
5
6
8
12 3 4 567 8901 2345 67890 1234 567890 123456 789012 345678 9012345678901234 5678901234567890
Xmin Xmax Ymin
I2 I1 I1 E3.0 E4.0 E4.0
FLAG
Ymax BUS1
E5.0 E4.0
A6
BUS2
BUS3
BUS4
Xlabel
Ylabel
A6
A6
A6
A16
A16
Enter "1" in column 2.
(1-2)
GTYPE
Type of EMTP variable being requested:
(3)
GTYPE=4
For node-voltage plots. Set GTYPE=4 for

TACS STAND ALONE variables (see Section
14.8).
GTYPE=8
For voltage difference or power of branches

or switches. Set GTYPE=8 for TACS hybrid
simulations (see Section 14.8).
15 - 6
Output Processing
GTYPE=9
UNITS
(4)
For currents or energies of branches or

switches. Remember that dynamic
synchronous and universal machine variables
are all plotted as though they were branch
currents, so they use a GTYPE=9 as well.
This code indicates which units are desired for the time
axis. Floating-point time specifications to follow are then
assumed to be in these units:
"1"
for degrees.
"2"
for cycles of fundamental frequency STATFR.
"3"
for seconds.
"4"
for milliseconds.
"5"
for microseconds.
"6"
for frequency in Hz (frequency scan plots

only).
"7"
for the base-ten logarithm of frequency in Hz

(frequency scan plots only).
Note:
UNITS=1 and UNITS=2 are based on the

synchronous power frequency "STATFR" (see
Section 4.1.8 for redefinition).

X-scale
(5-7)
The desired horizontal scale in the time units (whatever

they may be, from the column 4 request) per inch.
For example, x-scale = 3.0, and UNITS=4, then time axis
scaling gives 3.0 msec/inch.
A blank or zero field has special significance, as a request
for extended floating-point precision for plot-card input
fields (see Card for Extended Precision of Floating-Point
Fields).
Xmin
(8-11)
The starting value where the x-axis will start, in the units
selected in the UNITS field.
Xmax
(12-15)
The final value where the x-axis will end, in the units
selected in the UNITS field.
The curves will be plotted from the left graph limit x-min
to the right graph limit x-max only; these limits
themselves will generally lie somewhere in the study
time-range of zero to "TMAX".
15 - 7
Output Processing
Ymin
(16-20)
Value at bottom of vertical axis. This is the minimum

ordinate on the scale. Leave blank if automatic scaling is
desired.
Ymax
(21-24)
Value at top of vertical axis. This is the maximum

ordinate on the scale. Leave blank if automatic scaling is
desired.
Note that Parameters Ymin and Ymax only have meaning
for Postscript plots.
In cases where the minimum and maximum values of the
one or more curves for a given Postscript plot are not
known in advance, it is usually convenient to let the EMTP
decide upon the appropriate vertical scaling. This is
requested by leaving fields Ymin and Ymax blank.
In order for the "tic" marks on the vertical axis of a
Postscript plot to correspond to round numbers, the
variable span of that axis (Ymin - Ymax) should be
divisible evenly by the axis length in inches. The axis
length is under user control (see Graph size adjustment
card. Default value is 8.0 inches.
BUS1
EMTP network node names. These identify the curves to
(25-30)
be plotted, as follows:
BUS2
(31-36)
BUS3
(37-42)
BUS4
(43-48)
1)
For node-voltage plots, enter the name or

names of the nodes in question. Any of the
four name-fields can be left blank, if not
needed.
2)
For branch or switch variables (element

voltage difference, current, power, or
energy), enter either one (e.g., NODEA and
NODEB) or two node-pair names which
identify the branches in question.
If more than two branch or switch variables
are desired on the same graph, the user
should specify the special request word
"BRANCH" in columns 25-30; then the EMTP
will read the node-pair names for up to four
elements from a following card (see Card
input for 3 or 4 branch-variable names).
Xlabel
(49-64)
Text used to label the y-axis of the plot.
15 - 8
Output Processing
Ylabel
(65-80)
Text used to label the x-axis of the plot.
The curves that can appear on the same graph are identified by the following symbols and node
name. In the upper right hand corner of every plot, there will be a legend which shows the
symbol, the curve number, and a sample line drawn for each curve. The output frequency of the
curve-indication symbols on the Postscript plots corresponds to every one inch for the unsmoothed
(raw) data.
Curve
Number
Printer Postscript
Plot
Plot
Name for
node voltage
Node-pair names for

branch or switch
output
BUS1
BUS1 BUS2
BUS2
BUS3 BUS4
BUS3
BUS5 BUS6 (using

"BRANCH" request)
BUS4
BUS7 BUS8 (using

"BRANCH" request)
Intersection
of 2 or more
curves
Example:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <<<xs<xmn<xmx<-ymn<ymx<-bus1<-bus2<-bus3<-bus4<--Xlabel-------<-----Ylabel--1442. 8. 30. -50. 50. THREE FOUR FIVE
TWO-CYCLE TEST VOLTAGE PER UNIT
1942. 8. 30.
A1
A2
B1
B2
CURRENT
Four voltages are plotted on the first plot and two currents on the second plot. The scale on the
second plot is chosen by the EMTP.
15 - 9
Output Processing
Card for Input of 3 or 4 Branch-Variable Names:

If the special request word "BRANCH' is entered in columns 25-30 of the plot specification card,
then the variable identification is not made using BUS1 to BUS4 (columns 25-48); instead, the
EMTP will read BUS1 to BUS8 from an additional inserted immediately after the corresponding plot
specification card:
1
2
3
4
5
6
7
8
1 23456789012345678901234 567890 123456 789012 345678 901234 567890 123456 789012 34567890
2nd Branch
3rd Branch
4th Branch
NCRV
1st Branch
BUS1
BUS2
BUS3
BUS4
BUS5
BUS6
BUS7
BUS8
I1
A6
A6
A6
A6
A6
A6
A6
A6
NCRV
(1)
Leave blank if all curves are to be plotted on the same

graph.
A nonzero integer between 1 and 4 will be taken to
mean the number of curves on the same graph. Such
usage is rare.
BUS1 BUS2
(25-36)
BUS3 BUS4
(37-48)
BUS5 BUS6
(49-60)
BUS7 BUS8
(61-72)
Node-pair names for the first branch or switch which is

to have its variable plotted on the graph (columns 25-30
and 31-36).
Node-pair names for the second branch or switch which
is to have its variable plotted on the graph (columns 3742 and 43-48).
Node-pair names for the third branch or switch which is
to have its variable plotted on the graph (columns 49-54
and 55-60).
Node-pair names for the fourth branch or switch which
is to have its variable plotted on the graph (columns 6166 and 67-72).
Note: If any one of the four node-pair identifications contains two blank words (12 blank
characters), the entire request is ignored.
15 - 10
Output Processing
Card for Extended Precision of Floating-Point Fields:

The five floating-point fields of the plot specification card "xscale", "Xmin", "Xmax", "Ymin",
and "Ymax", in the plot specification card may not have sufficiently wide fields for all practical
applications. For full 16-column wide specification of all five floating point plot-specification
card parameters, leave field "xscale" in the plot specification card blank (column 5-7), and specify
"xscale", "Xmin", "Xmax", "Ymin", and "Ymax", with the format described below:
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678 9012345678901234 5678901234567890
Xscale
XMIN
E16.0
E16.0
XMAX
E16.0
YMIN
YMAX
E16.0
E16.0
Exception:
If the "BRANCH" option to specify more than two branch output, and the extended-precision
requests are desired for the same plot, then the following procedure should be followed:
1.
Plot specification card with field "xscale" blank (columns 5-7) and "BRANCH"
keyword specified in columns 25-30.
2.
Card for input of three or four branch-variable names.
3.
Followed by the extended-precision card described above.
Card for Graph-Size Adjustment:

The bottom margin and vertical-axis height of a Postscript plot can be changed using a special
request card having the following format:
I2
SPAN
BEGIN
ITYPE
1
2
3
4
5
6
7
8
12 34567 8901 2345 678901234 567890 12345678901234567890123456789012345678901234567890
E4.0 E4.0
ITYPE
HEIGHT
A6
Set ITYPE=1.
(1-2)
15 - 11
Output Processing
New lower margin in inches.
BEGIN
(8-11)
New graph height in inches.
SPAN
(12-15)
Special request keyword "HEIGHT".
HEIGHT
(25-30)
All plots requested after input of this card will have this specified bottom margin and vertical-axis
height, unless changed by another such card at a later stage of the plot-data input cards.
If the total distance represented by bottom margin "BEGIN" plus vertical axis height "SPAN" is
greater than the width of the assumed plotting paper width "SZPLT" (see Section 4.1.4), a warning
message is produced and the request is ignored. Line printer plotting is unaffected by the presence
or absence of this card.
Note: The Postscript parameter specification card can be used to change the origin and size of
the plot more precisely. The graph size adjustment card described here is used primarily for
backward compatibility with older versions of the program.
Card for Redefinition of Smoothing Tolerance "":

The numerical value for smoothing tolerance "" (which was described in Rule 6 of Section 15.1.1)
is given a default value of 0.01 inches by the EMTP at the beginning of each new data case.
However, its value is under user control, and can be changed from one plot to another by use of a
special request card of the following format:
ITYPE
EPS
SMOOTH
1
2
3
4
5
6
7
8
12 3456789012345 67890 1234 567890 12345678901234567890123456789012345678901234567890
I2
E5.0
A6
ITYPE
Set ITYPE = 1.
(1-2)
(16-20)
"EPS", the desired new smoothing tolerance in

inches.
SMOOTH
The special request keyword "SMOOTH".
EPS
(25-30)
15 - 12
Output Processing
This tolerance remains in effect until it is re-defined with another request card. Values of "EPS"
larger than 0.1 inches trigger a warning message. If the tolerance field "EPS" is left blank or set to
less than 0.0001 inches, the EMTP sets the tolerance "" to zero, and no smoothing occurs: all raw
data points are plotted.
Line printer plotting is unaffected by the presence or absence of this card.
Card for Changing the Graphic Mode Selection:

At any stage of the plotting data-card input, the user can request that plotting be done just on the
Postscript, just on the line printer, or on both of these devices (see Rule 2, Section 15.1.1). A
request for one of the three alternative plotting modes requires a card having one of the following
formats:
1
2
3
4
5
6
7
8
12 345678901234567 890123456789012345678901234567890123456789012345678901234567890
KEYWORD
Keyword
(3-17)
Special request keyword for the desired output

mode:
"PRINTER PLOT" for printed plot output (default).
"POSTSCRIPT PLOT" for Postscript output only
"POSTSCRIPT PRINTER" for both printer plot and
Postscript output.
The EMTP is initialized at the beginning of each new data case to produce just printer plots. This
or a later requested mode will remain in effect for the data case in question until changed by
another such request card as above.
Card for Changing Pens and Grid Status:

Different curves of any plot can be drawn with different pens (or traces). The option of a grid for
plots is also available. Grid lines are drawn every inch, both horizontally and vertically, as
extension to the "tic" marks of the axes.
Dashed line patterns are implemented in the form of pen changes. Different pen numbers are
interpreted as different dashed patterns. The user is allowed control of Postscript-specific features
like text fonts and global scaling and translation by using the Postscript parameter specification
card described below.
15 - 13
Output Processing
User control of the two preceding options is provided by a single card, with the following format:
1
2
3
4
5
6
7
8
12 3456789012 3456 78901234 56789012 34567890 12345678 90123456 789012345678901234567890
PEN CHOICE
A10
KPGRID
KPEN1
KPEN2
KPEN3
KPEN4
I8
I8
I8
I8
I8
PEN CHOICE
Special request keyword "PEN CHOICE".
(3-12)
KPGRID
KGRID > 0 sets the pattern used for drawing the
(17-24)
grid. A zero or blank field will suppress the grid;

i.e., none will be drawn.
KPEN1 (25-32)
KPEN2 (33-40)
KPEN3 (41-48)
KPEN4 (49-56)
KPEN(I), the integer pen number which is to be
used to draw curve number "I" of the plot. A zero

or blank field is taken to mean no change in the
pen number (from the preceding assignment) for
the curve associated with that data field.
The requested pen choices and grid specification will remain in effect throughout the data case in
question, until altered by another such request card. By default there is no grid, and all curves are
drawn with pen number 1, unless an explicit request to the contrary follows. Pen choices 10 to 99
produce unevenly spaced line/space patterns. For example, 92 produces a line of length 9 and a
space of length 2. A pen choice of zero or blank produces a solid black trace.
Card for Changing Line Limit on Sparse Printer Plots:

For curves where the number of raw plot data points exceeds the number of printed lines of a line
printer plot, there is no problem with line limits on printer plots; printer plots can be of any length
in this case. But if the user requests more printed lines of time axis than there are raw plot-data
points for that interval, then a special limit "LINLIM" comes into play. Such a "sparse" printer plot
will be terminated when this line limit is exceeded where "sparse" is defined to mean that there
are fewer raw data points than printed lines. They are characterized visually by lines of plot output
which have no curve symbols.
Sparse plot line-limit "LINLIM" has been given a default value of 100. This limit can be redefined
by the user, however, using a special-request card of the following format:
1
2
3
4
5
6
7
8
12 3456789012345678901234 56789012 345678901234567890123456789012345678901234567890
PLOT LINE LIMIT
A22
LINLIM
I8
15 - 14
Output Processing
Such a user-requested "LINLIM" will remain in effect throughout the data case in question, until
altered by another such request card.
Card to Superimpose Plots and/or Suppress Vertical-Axis Labeling:

The special request card "SUPERIMPOSE" can be used for the following purposes:
1.
the plotting of more than four curves on the same graph, or
2.
the plotting of different variable types (e.g., node voltages and branch currents)
on the same graph, or
3.
the addition of a vertical offset for one or more curves of a graph, or suppression
of the vertical-axis labelling (leaving just the vertical line with tic marks).
The format of the "SUPERIMPOSE" special request card is as follows:

1
2
3
4
5
6
7
8
12 3456789012345678901234 56789012 34567890 12345678 90123456789012345678901234567890
SUPERIMPOSE
A22
SUPERIMPOSE
MPL1
MPL4
MPL5
I8
I8
I8
Special request keyword "SUPERIMPOSE".
(3-13)
MPL1
(25-32)
The number of plot specification cards which are

to be superimposed. Any non-positive integer
(e.g., zero or blank) will be defaulted to unity.
MPL4
Vertical offset between curves (see Usage 3).
(33-40)
MPL5
(41-48)
A flag for suppression of the vertical axis labeling.

If no such labeling is desired, enter a "1" in column
48; otherwise, leave the field blank, or zero.
This "SUPERIMPOSE" feature is very powerful, but not very easy to use. To use this option
effectively, the following operational principles, should be kept in mind:
Principle 1:
The superimposition feature is based on suppression of the originchanging operation which normally occurs upon the completion of a
plot. If the change of origin is omitted, the following plot will then be
drawn on top, i.e., superimposed, of the preceding one. Thus the
15 - 15
Output Processing
superimposed plots are really separate, independent EMTP plots; they
are produced separately, just as would be the case if there were no
"SUPERIMPOSE" card.
Principle 2:
All non-curve drawing for the second and later of a group of

superimposed plots is automatically suppressed. Thus the grid (if
requested), the axes, the date and time, etc., and any title will only be
drawn once -- for the first of the superimposed plots. The scaling of the
second and later superimposed plots is not documented on the plot.
The SUPERIMPOSE option can be used as follows:
Usage 1:
To plot more than four EMTP variables on the same graph: place the plot
requests in as many plot specification cards as needed, and preceded
them by the appropriate "SUPERIMPOSE" card.
As an example of this usage, consider the required data cards for seven
node voltages on the same graph:
Example:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
SUPERIMPOSE
3
C <<<xs<xmn<xmx<-ymn<ymx<-bus1<-bus2<-bus3<-bus4<--Xlabel-------<-----Ylabel---144 8. 0.0 80. -20. 30. NODE1 NODE5 NODE9
144 8. 0.0 80. -20. 30. NODE3 NODE3
144 8. 0.0 80. -20. 30. BLACK
BLUE
Usage 2:
To mix different plot-variable classes (e.g., node voltages and branch

currents) on the same graphs: Use separate plot specification cards for
variables within any one class, and precede them by the appropriate
"SUPERIMPOSE" card.
As an example of this usage, consider the data cards which will plot two
node voltages and one branch current on the same graph:
Example:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
SUPERIMPOSE
2
C <<<xs<xmn<xmx<-ymn<ymx<-bus1<-bus2<-bus3<-bus4<--Xlabel-------<-----Ylabel---144 8. 0.0 80. -20. 30. NODE3 NODE4
194 8. 0.0 80. -4. 4. BUSK BUSM
15 - 16
Output Processing
Usage 3:
Offset two or more curves which are similar and would be difficult to
distinguish one from the other: Use separate plot specification cards
and offset them by adding a constant to Ymin and Ymax, and precede
them by the appropriate "SUPERIMPOSE" card.
As an example of this usage, consider the case of three node voltages.
The second is to be offset one inch above the first, and the third is to
offset one inch above the second. Note that in this example the vertical
axis labeling is suppressed.
Example:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
SUPERIMPOSE
3
1
1
C <<<xs<xmn<xmx<-ymn<ymx<-bus1<-bus2<-bus3<-bus4<--Xlabel-------<-----Ylabel---144 8. 0.0 80. -20. 20. NODE1
144 8. 0.0 80. -25. 15. NODE1
144 8. 0.0 80. -30. 10. NODE1
Usage 4:
To suppress all axis labeling on a conventional plot: Use a plot

specification cards preceded by a "SUPERIMPOSE" card having a "1"
in columns 40 and 48.
As an example of this usage, consider the case of two node voltages, for which no vertical-axis
labeling is desired:
Example:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
SUPERIMPOSE
1
1
C <<<xs<xmn<xmx<-ymn<ymx<-bus1<-bus2<-bus3<-bus4<--Xlabel-------<-----Ylabel---144 8. 0.0 80. -20. 20. NODE1
Vertical-axis scaling is unambiguous if the scaling is specified explicitly. However, if automatic

scaling is specified there is no way to interpret the scaling of the second to last variable since the
corresponding axis are suppressed. As a general rule, it is advisable to provide vertical-axis
scaling explicitly on plots which involve the "SUPERIMPOSE" usage.
15 - 17
Output Processing
Card for Turning On or Off the Plot Identifier Symbols:

The plot identifier symbols are printed out at intervals of equal number of plot points. The interval
is determined automatically by the program. As a result of the SMOOTH procedure, plot points are
not uniformly distributed with time, but tend to concentrate in areas of higher curve dynamics.
This results in concentration of the identifiers in the same locations, and can have a distracting
effect. By use of the special request words, "IDENTIFY ON" and "IDENTIFY OFF", the use of the
identifier symbols can be toggled, at any point in the plot requests. The default condition is
IDENTIFY ON. The format is as follows:
1
2
3
4
5
6
7
8
12 345678901234567890123456789012345678901234567890123456789012345678901234567890
IDENTIFY ON
IDENTIFY OFF
Card for Plotting one EMTP Variable versus Another:

The special request card "X-Y PLOT" permits plotting any EMTP variable against another. The
rules for the "X-Y PLOT" feature are as follows:
Rule 1:
Only the following special request cards are accepted:
Case-Title Card
Subheading Text Card
Card for Changing Pens and Grid Status
Card for input of 3 or 4 Branch-Variable Names
Headings font specification
Annotations font specification
Postscript Parameters
Three cards signal the request for "X-Y PLOT" usage:
Rule 2:
Card 1:
1
2
3
4
5
6
7
8
12 34567890 1234 567890123456789012345678 901234567890123456789012345678901234567890
X-Y PLOT
Title
A8
A24
15 - 18
Output Processing
X-Y Plot
(3-10)
Special request keyword "X-Y PLOT".
Title
(15-38)
Title for the horizontal axis label.
Card 2:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012345678901234567890123456789012345678901234567890
LENX
XMIN
XMAX
E8.0
E8.0
E8.0
(1-8)
Length of the horizontal "X" axis, in inches. Zero or

blank is defaulted to eight (8.0) inches.
XMIN
Minimum "X" value.
LENX
(9-16)
XMAX
Maximum "X" value.
(17-24)
Note: If both "XMIN" and "XMAX" blank, the EMTP will

automatically choose these values to cover the curve
being plotted.
Card 3:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 1234567890123456789012345678901234567890
LENY
YMIN
YMAX
DIV
SYM
E8.0
E8.0
E8.0
E8.0
E8.0
(1-8)
Length of the horizontal "Y" axis, in inches. Zero or

blank is defaulted to eight (8.0) inches.
YMIN
Minimum "Y" value.
LENY
(9-16)
15 - 19
Output Processing
YMAX
Maximum "Y" value.
(17-24)
DIV
Number of divisions per inch of the graph paper
(25-32)
which is being drawn on. A zero or blank field is
defaulted to ten divisions/inch.
SYM
One less than the number of symbols which is used
(33-40)
to identify the curves. Zero or blank is interpreted to
mean that no symbols shall be placed on the curves for
purposes of identification.
Rule 3:
The above cards are to be followed by a modified plot-specification card:
GTYPE
UNITS
FLAG
1
2
3
4
5
6
8
12 3 4 567 8901 2345 678901234 567890 123456 789012 345678 9012345678901234 5678901234567890
I2 I1 I1
Tmin Tmax
BUS1
BUS2
BUS3
BUS4
Xlabel
Ylabel
E4.0 E4.0
A6
A6
A6
A6
A16
A16
FLAG
Enter "1" in column 2.
(1-2)
GTYPE
Type of EMTP variable being requested:
(3)
UNITS
(4)
GTYPE=4
For node-voltage plots. Set GTYPE=4 for

TACS STAND ALONE variables (see Section
14.8).
GTYPE=8
For voltage difference or power of branches

or switches. Set GTYPE=8 for TACS hybrid
simulations (see Section 14.8).
GTYPE=9
For currents or energies of branches or

switches. Remember that dynamic
synchronous and universal machine variables
are all plotted as though they were branch
currents, so they use a GTYPE=9 as well.
This code indicates which units are desired for the time
axis. Floating-point time specifications to follow are then
assumed to be in these units:
"1"
for degrees.
15 - 20
Output Processing
"2"
"3"
for seconds.
"4"
for milliseconds.
"5"
for microseconds.
"6"
for frequency in Hz (frequency scan plots

only).
"7"
for the base-ten logarithm of frequency in Hz

(frequency scan plots only).
Note:
UNITS=1 and UNITS=2 are based on the

synchronous power frequency "STATFR" (see
Section 4.1.8 for redefinition).

For example, x-scale = 3.0, and UNITS=4, then time axis
scaling gives 3.0 msec/inch.
Tmin
(8-11)
Minimum or beginning time of the points to be plotted, in

units of the UNITS specification.
Tmax
(12-15)
Maximum or ending time of the points to be plotted, in

units of the UNITS specification.
BUS1
Six-character names of EMTP variables which are to be

plotted. The first two variables are for one curve, as an
ordered (X,Y) pair of variables.
(25-30)
BUS2
(31-36)
BUS3
(37-42)
BUS4
(43-48)
For node voltages, BUS1 and BUS2 identify the first curve
(with the X-name appearing first); BUS3 and BUS4 identify
the second curve (if a second exists). For branch voltages
or branch currents, a pair of names is required for variable
identification, therefore, only one with branch-variables
can be specified with this card. For more than one curve of
branch variables, the "BRANCH" option can be used.
Xlabel
(49-64)
Text used to label the y-axis of the plot.
Ylabel
(65-80)
Text used to label the x-axis of the plot.
15 - 21
Output Processing
Example:
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
2 Test of type 96 pseudo-nonlinear hysteretic reactance element
Flux vs Current
X-Y PLOT
Type-96 current in amps
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C -LENX-<--XMIN-<--XMAX5.0
-4.0
4.0
C -LENY-<--YMIN-<--YMAX-<--DIV--<-SYM--6.8
-1.7
1.3
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <<<xs<Tmn<Tmx
<-bus1<-bus2<-bus3<-bus4<--Xlabel-------<-----Ylabel---194 8. 0.0 25.
TACS CURR TACS FLUX Hysteresis in volt-sec
C Next 3 cards are used to flag return to standard plotting
X-Y PLOT
C -LENX-<--XMIN-<--XMAX9999.
-4.0
4.0
C -LENY-<--YMIN-<--YMAX-<--DIV--<-SYM--6.8
-1.7
1.3
Note: To switch back to conventional plotting after X-Y plot usage, repeat the three cards of the
request which switches to X-Y PLOT usage in the first place, only with LENX=9999. This flags a
return to conventional (versus time) plotting is desired.
In order for the Y-axis labelling (numbers) of a printer plot to be correct, LENY=13 is required. For
any other Y-axis length, Y-axis numbers of a printer plot of the X-Y PLOT type should be
disregarded. The curve itself should be validly constructed according to the user's requested axis
scaling, however.
Card to Begin Each PRINTER PLOT on a New Page:

A card with the "PAGE" keyword in columns 3-6 will produce a page skip on the printer output just
prior to the output of each new PRINTER PLOT.
1
2
3
4
5
6
7
8
12 3456 78901234567890123456789012345678901234567890123456789012345678901234567890
PAGE
15 - 22
Output Processing
Headings Font:
1
2
3
4
5
6
7
8
12 345678901234567890 123456 789012345678901234567890123456789012345678901234567890
H-FONT
hfac
A18
H-FONT
E6.0
FONT NAME
(3-20)
Keyword to specify the font used in headings and subheadings.
hfac
(21-26)
Scaling factor for text used in headings and sub-headings.

Default = 1.0.
FONTNAME
Valid Postscript font name. Default = /Helvetica. (see

Section 15.12 for valid font names). Note that the "/"
character must precede a valid font name.
Annotations Font:
1
2
3
4
5
6
7
8
12 345678901234567890 123456 789012345678901234567890123456789012345678901234567890
A-FONT
afac
A18
E6.0
FONT NAME
(3-20)
Keyword to specify the font used in annotations, such as

variable names, axis labels, etc.
afac
(21-26)
Scaling factor for text used in annotations, such as

variable names, axis labels, etc. Default =1.0.
FONTNAME
Valid Postscript font name. Default = /Helvetica. (see

Section 15.12 for valid font names). Note that the "/"
character must precede a valid font name.
A-FONT
15 - 23
Output Processing
Postscript Control Parameters:

A graph can be scaled, rotated and translated arbitrarily within the plotting surface, but the
program will not detect absurd image manipulation (e.g., outside the plotting page).
1
2
3
4
5
6
7
8
12 345678901234567890 123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
PS-PARAMETER
A18
tics
htrl
vtrl
hzfac
vfac
grot
xlab
ylab
line
E6.0
E6.0
E6.0
E6.0
E6.0
E6.0
E6.0
E6.0
E6.0
PS-PARAMETER
(3-20)
Keyword to signal that the default values for

Postscript-specific parameters will be modified.
tics
(21-26)
Point size of axis tic marks, default = 10.
htrl
(27-32)
Global horizontal translation in points, default=0.
vtrl
(33-38)
Global vertical translation in points, default = 0.
hzfac
(39-44)
Global horizontal scaling factor, default = 1.0.
vfac
(45-50)
Global vertical scaling factor, default = 1.0.
grot
(51-56)
Global rotation in degrees, default = 0.0.
xlab
(57-62)
Flag to control the position of x-axis label.

xlab = 0 for centre x-axis label.
xlab = 1 for flush right.
ylab
(63-68)
Flag to control the position of the y-axis label.

ylab = 0 for centre y-axis label.
ylab = 1 for flush top.
line
(69-74)
line width in points, default = 1.0.
15 - 24
Output Processing
Postscript Filename:
1
3
4
5
6
7
8
12 345678901234567890 123456789012345678901234567890123456789012345678901234567890
PS-FILENAME
FILENAME
A18
PS-FILENAME
(3-20)
FILENAME
Keyword to specify a non-default filename for Postscript

plotted output. The default filename is created with the
same filename root as the EMTP input data with the
extension ".ps".
Non-default filename used for the Postscript output file.
(21-80)
Card to Mark the Termination of all Plot Cards:

A blank card is used to signal the end of all data cards that are associated with plotting. This also
completes the data for the data case being set up (see Section 3.1).
15.2 Fourier Analysis

15.2.1
Card For Fourier Series Computation
The Fourier series computation of the EMTP can only be accessed via this EMTP plot data
classification, using the special request card (see below) which switches from plotting to Fourier
series:
1
2
3
4
5
6
7
8
12 3456789012 345678901234 56789012 345678901234567890123456789012345678901234567890
FOURIER ON
NFOUR
A10
I8
15 - 25
Output Processing
FOURIER ON
Special request keyword "FOURIER ON".
(3-12)
Number of harmonics to be printed. If left blank, 30 will
be printed (default value). For KPL equidistant points in
the fundamental period, only KPL/2 harmonics are
computed. If the user specifies a larger NFOUR, it will
automatically be reduced to this maximum.
NFOUR
(25-31)
Next come the cards that identify the EMTP variables and a time window over which the Fourier
analysis is to be performed. Format is shown below:
HMAX
HMIN
FLAG
TYPE
UNITS
HPI
1
2
3
4
5
6
7
8
12 3 4 567 8901 2345 678901234 567890 123456 78901234567890123456789012345678901234567890
I2 I1 I1 E3.0 E4.0 E4.0
FLAG
BUS1
BUS2
A6
A6
Set FLAG = 1.
(1-2)
TYPE
(3)
Set TYPE to "4", "8", or "9", depending upon whether one is to

Fourier analyze a node voltage, a branch voltage, or a branch
current, respectively.
UNITS
Flag indicating which units are desired for time parameters HPI,
(4)
HMIN, and HMAX below:
"1"
for degrees.
"2"
"3"
for seconds.
"4"
for milliseconds.
"5"
for microseconds.
"6"
for frequency in Hz (frequency scan plots only).
"7"
for the base-ten logarithm of frequency in Hz (frequency

scan plots only).
Note:
UNITS=1 and UNITS=2 are based on the synchronous

power frequency "STATFR" (see Section 4.1.8 for
redefinition).
15 - 26
Output Processing
HPI
(5-7)
HMIN
(8-11)
Horizontal scale HPI. This data is not used for the Fourier
calculation, but a credible value is required to successfully pass
through plot card data checking logic.
The beginning time "HMIN" of the one-cycle window that is to be
used for the Fourier analysis, in units specified in the UNITS field.
(12-15)
The ending time "HMAX" of the one-cycle window, that is to be

used for the Fourier analysis, in units specified in the UNITS field.
BUS1
First EMTP variable name.
(25-30)
BUS2
(32-36)
Second EMTP variable name, if any. For the Fourier analysis of a

node voltage, only BUS1 is used; but for branch voltage or current,
BUS2 is also required to identify the variable.
One specification card is used for each Fourier analysis. Any number of Fourier analysis can be
requested. On output, a table is produced with the harmonic number (first column), cosine
coefficient (second column), sine coefficient (third column), complex amplitude (fourth column),
and relative size (fifth column).
Example:
REQUEST FOR LINE PRINTER PLOT.
1 PRINTER PLOT
FOURIER SERIES STARTED. NFOUR =
10 1 FOURIER ON
10
Comment line.
1C
NOTE HPI=7.6 IS SPECIAL REQUEST FOR REDUCTION Comment line.
1C
TO HMAX UNTIL MATCH WITH HMIN POINT IS FOUND.
PLOT CARD. 0.760E+01 0.150E+01 0.120E+02
1.5 12.
VALUE
1 1437.6
BACK UP FINAL TIME FROM POINT NUMBER 11 TO POINT NUMBER 8,SO THAT SIGNAL WILL BE NEARLY PERIODIC.
THE FOLLOWING UNUSED POINT, WHICH HAS VALUE CLOSEST TO POINT NUMBER 1, EQUALS
0.850000E+00
BEGIN FOURIER SERIES CALCULATION USING
8
EQUIDISTANT POINTS.
BEGINNING TWO POINTS =
0.8485282063E+00
0.0000000000E+00
ENDING TWO POINTS =
-0.1565690041E+01
0.3410000086E+01
COEFFICIENTS OF RESULTANT FOURIER SERIES, WITH "COMPLEX AMPLITUDE" BEING THE SQUARE ROOT
OF THE SUM OF THE SQUARES OF THE TWO PRECEDING ENTRIES.
THE FINAL COLUMN APPLIES TO THIS AMPLITUDE.
HARMONIC
COSINE
SINE
COMPLEX
FRACTION OF
NUMBER
COEFFICIENT
COEFFICIENT
AMPLITUDE
FUNDAMENTAL
0
0.1012495253E+00
0.0000000000E+00
0.1012495253E+00
0.08666142
1
0.1133138623E+01
-0.2846091995E+00
0.1168334514E+01
1.00000000
2
0.5000010580E+00
-0.9025000222E+00
0.1031749654E+01
0.88309439
3
-0.2846105064E+00
-0.8502972136E+00
0.8966652061E+00
0.76747301
4
-0.6012504939E+00
0.0000000000E+00
0.6012504939E+00
0.51462187
It is important to define exactly one cycle of the periodic wave, with no overlap. Because of the
floating-point counting of the EMTP (T = T + DELTAT), this may not be straightforward. The first
two points and the last two points are printed out to verify that the period selected was the desired
one. All plot points for T < HMIN or T> HMAX are rejected. Occasionally, it may be necessary to
add or subtract half a time step to the range specified with HMIN and HMAX.
Since there is no interpolation on the EMTP plot file points, it is important that time-step size
DELTAT be an exact multiple of the period.
15 - 27
Output Processing
To return to conventional EMTP plotting after the use of Fourier transformation, use the "FOURIER
OFF" request card shown below.
1
2
3
4
5
6
7
8
12 34567890123 4567890123456789012345678901234567890123456789012345678901234567890
FOURIER OFF
15.2.2
Background for EMTP Fourier Series Computation
Given a periodic function f(x):

f(x)
1 Period
This periodic function can be expressed as a series of trigonometric functions:

f(x) =
ai cos ( ix ) + bi sin ( ix )
i=0
i=0
(15.1)
with ao = dc offset,
2
a1 + b1
a +b
= amplitude
of fundamental frequency and
= amplitude of i -th harmonic
(15.2)
(15.3)
To perform Fourier analysis on an EMTP output variable, the curve must be defined within one
period by n points at equal distances along x (n>1):
15 - 28
Output Processing
1 period
This last value

equals point one.
Do not specify.
4 5 6 7

8 9 10
11
17
16
12
15
14
13
The EMTP will compute the values ao, am (cosine-coefficients) and bo, , bm (sinecoefficients) of Equation (15.1) as well as the magnitudes Co, , Cm, with
ci =
(15.4)
ai + ai
n
If n is even, then m = --2
(15.5)
(n 1)
If n is odd, then m = ---------------2
(15.6)
The resulting finite series

m
F( x ) =
ai cos ( ix ) + bi sin ( ix )
i=0
(15.7)
i=0
passes through the n given points ("exactly", except for round-off errors) and provides a smooth
interpolation between points with the least possible number of harmonics.
15 - 29
Output Processing
15.3 Fatigue Analysis

15.3.1
Cards for Fatigue Analysis
The Fatigue damage computation of the EMTP can only be accessed via this EMTP plot data
classification, using the special request card (see below) which switches from plotting to fatigue
analysis.
1
2
3
4
5
6
7
8
12 3456789012 34567890123456789012345678901234567890123456789012345678901234567890
FATIGUE ON
FATIGUE ON
(3-12)
Special request keyword "FATIGUE ON". This card can be

placed anywhere among the cards used for batch-mode
plotting.
Next comes Card #1 (general excitation data card):
Card #1:
TYPE
UNITS
FLAG
1
2
3
4
5
6
7
8
12 3 4 567 8901 2345 678901234 567890 123456 78901234567890123456789012345678901234567890
I 2 I1 I1
HMIN HMAX
NAME1 NAME2
E4.0 E4.0
FLAG
A6
A6
Set FLAG = 1.
(1-2)
TYPE
(3)
UNITS
(4)
Set TYPE = 9 to indicate a synchronous (universal)

machine variable.
Indicate which units of time are to be used in the
study.
"2"
for cycle of fundamental frequency STATFR.
"3"
for seconds.
"4"
for milliseconds.
"5"
for microseconds.
15 - 30
Output Processing
HMIN
(8-11)
HMAX
(12-15)
NODE1
(25-30)
NODE2
(31-36)
The beginning time HMIN where the torque signal

will start, in units specified in field UNITS.
The final time HMAX where the torque signal will
end, in units specified in field UNITS.
The name of the torque signal for which the fatigue
analysis is to be performed. The name of the torque
signal depends on the machine model (S.M. Type59
or U.M.) used to generate the torque data. For
example, to analyze the shaft section connecting
masses #2 and #3 (see Section 10.11 and Section
10.12), then use the names "MACH 1" and "TOR 2 "
for data generated by the Type59 S.M. model, and
the names "MASS-2" and "MASS-3" for data
generated by the U.M. model.
The data field NAME1 can also be used to flag the presence of the user defined data for the
excitation signal. This option is described later in this subsection.
Next come cards to specify the machine shaft data:
Card #2:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
RADS
GEL
E10.5
E10.5
RADS
(1120)
GEL
(21-30)
STRBRK
(31-40)
STRBRK
E10.5
Radius of the shaft nominal section adjacent to the

notch (meters in SI units; inches in British units).
Stress-strain history is initially calculated for this
nominal section.
Elastic shear modules of the nominal section
adjacent to the notch (N/m2 in SI units; PSI in
British units).
Stress value at the knee point of the cyclic stressstrain curve for the nominal section
(N/m2 in SI units; PSI in British units).
15 - 31
Output Processing
The value for STRBRK can be obtained by plotting the cyclic stress-strain curve for the nominal
section. The EMTP will choose the highest stress value of this curve for STRBRK if no value is
provided by the user.
In case all these values have already been provided for the previous shaft section, shaft data can be
repeated by using the following keyword:
1
2
3
4
5
6
7
8
12 3456 78901234567890123456789012345678901234567890123456789012345678901234567890
SAME
Card #3:
Specifies the number of points in the cyclic stress-strain curve for the shaft material.
1
2
3
4
5
6
7
8
123456 7890 1234567890123456789012345678901234567890123456789012345678901234567890
NYSN
I4
NYSN
(7-10)
The number of points on the cyclic stress-strain

curve to be supplied by the user.
Alternatively, if this curve has already been specified for the previous shaft section, use the
keyword "SAME" as in Card #2 above.
15 - 32
Output Processing
Card #4:
Point-by-point representation of the cyclic stress-strain curve for the shaft material is an elastoplastic deformation in an increasing order (see Figure 15.1 below). If the keyword "SAME" was
used for Card #3, skip Card #4 completely.
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345678901234567890123456789012345678901234567890
STRSS
STRNN
E15.5
E15.5
Stress value on cyclic curve (N/m2 in SI units; PSI in

British units).
STRSS
(1-15)
Strain value corresponding to the STRSS (m/m in SI units;

Inch/Inch in British units).
STRNN
(16-30)
This card must be repeated "NYSN" times.
Shear stress amplitude (psi)
50,000
STRBRK
25,000
0
0.0
0.015
Shear strain amplitude in/in
Figure 15.1:
Torsional Cyclic Stress-Strain Curve
15 - 33
0.030
Output Processing
Card #5:
Specifies the number of points on the strain-life diagram (shear strain amplitude versus cycles to
crack initiation curve for the shaft material at the notch).
1
2
3
4
5
6
7
8
123456 7890 1234567890123456789012345678901234567890123456789012345678901234567890
NYSL
I4
NYSL
The number of points on the strain-life diagram.
(7-10)
Again, if the same information has already been provided for the previous shaft section, the
keyword "SAME" may be used as in Card #2.
Card #6:
Point-by-point representation of the strain-life diagram for the shaft material at the notch (see
Figure 15.4 below). No data is necessary if the keyword "SAME" was used in Card #5.
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345678901234567890123456789012345678901234567890
CYCLE
STRLF
E15.5
E15.5
CYCLE
Cycles to failure.
(1-15)
STRLF
(16-30)
Strain value for CYCLE (m/m in SI units, inch/inch in

British units).
This card must be repeated "NYSL" times.
15 - 34
Output Processing
Shear strain amplitude in/in
10-3
10-2
0
102
103
104
105
106
107
108
Reversals to failure
Figure 15.2:
Torsional Strain-Life Curve
Card #7:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345678901234567890123456789012345678901234567890
FATCON
TRQCON
E15.5
E15.5
FATCON
(1-15)
TRQCON
(16-30)
Theoretical stress concentration factor relating stressstrain history of the notch to that of the nominal shaft
section.
Torque multiplier for the unit conversion of torque data to
that of shaft material. Type59 S.M. model calculates the
torque in millions of N-m (Newton-meters) and U.M.
model calculates it directly in N-m, therefore:
Material Data In SI
Type59 S.M
TRQCON
106
Material Data In British
U.M.
Type59 S.M
U.M.
1.0
8.8508 x 106
8.8508
15 - 35
Output Processing
This card must be present for each shaft section even if an identical shaft section precedes the
section in question.
Card #1 through Card #7 may be repeated as many times as desired so that all shaft sections are
analyzed for loss of life due to fatigue.
User Supplied Data:

As mentioned earlier, the user can supply excitation data or the stress-strain history of the shaft
section. This is achieved be specifying a special request keyword in the field NAME1
(Columns 25-30, Card #1). The following two keywords are accepted by the EMTP:
TORQUE:
This keyword is used if the information on the applied

torque is to be supplied directly by the user (after
specifying data on shaft characteristics).
STRESS:
This keyword signals that the user will directly supply the
stress-strain history of the shaft section (after specifying
data on shaft characteristics).
When any of the above two keywords is used, the fields UNITS (column 4), HMIN (columns 8-11)
and HMAX (columns 12-15) in Card #1can be left blank. Cards #2 through #7 have to be specified
as explained above. However, there are additional cards to follow Card #7.
Card #8:
This card is used only if a keyword was used for variable "NAME1" on Card #1 indicating direct
input of the torque or stress-strain data.
1
2
3
4
5
6
7
8
123456 7890 1234567890123456789012345678901234567890123456789012345678901234567890
KPL
I4
KPL
The number of data points for:
(710)
CASE A:
Torque data if NAME1 = "TORQUE"
CASE B:
Stress-strain data if NAME1 = "STRESS"
15 - 36
Output Processing
Card #9:
Depending on the keyword "name1" it may contain:
(A)
The applied torque or stress-strain data on the nominal shaft section;
(B)
The stress-strain history of the shaft section.
Case A - Torque data:
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
TORQUE
E15.5
TORQUE
Point-by-point value of the applied torque.
(115)
Case B - Stress-strain data:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345678901234567890123456789012345678901234567890
STESS
STAIN
E15.5
E15.5
STESS
The applied stress (point-by-point).
(115)
STAIN
The resulting strain (point-by-point).
(1630)
Here, the same units as those used for the shaft data must be used in which case TRQCON = 1.0
(see Card #7).
Cards #8 and #9 have to be repeated as many times as Cards #1 through 7.
15 - 37
Output Processing
To return to conventional EMTP plotting after the use of fatigue analysis, use the "FATIGUE OFF"
request which follows.
1
2
3
4
5
6
7
8
12 34567890123 4567890123456789012345678901234567890123456789012345678901234567890
FATIGUE OFF
15.4 Output Processing with MATLAB

MATLAB is a commercially-available scientific software package that permits the manipulation
and analysis of data. EMTP simulation results can be imported into MATLAB using MCAT. MCAT is
a utility program written in MATLAB m-files and works on all platforms supported by MATLAB.
15.4.1
Installation
All the MATLAB m-files necessary to run MCAT are usually located in a single directory. The
location of this directory is specified in file "setpath.m". This file can be modified with a text
editor to specify the location of MCAT m-files in any given installation.
Run "setpath" in MATLAB to set the MATLAB path or insert its contents in a MATLAB startup file
for automatic installation at each MATLAB session.
15.4.2
Usage Rules
To run MCAT, the EMTP must first generate two support files: a file containing EMTP simulation
output and a file containing the rules needed to interpret the contents of this output file. To
generate these support files the following rules apply:
1.
Set ICAT=3 in the EMTP miscellaneous parameters (see Section 4.2.2). With
ICAT=3 it is now IOUT and not IPLOT that controls the frequency of the plotted
output. IPUN=-1 in miscellaneous data cards can now be used to vary the plotted
output frequency.
2.
The two support files are automatically generated by EMTP. The naming
convention of these files is based on the name of the input data file for the emtp
simulation (e.g., "datafile.dat"):
(A)
File "datafilM.m contains MATLAB commands for waveform

identification.
15 - 38
Output Processing
(B)
15.4.3
File "datafilX.mdt" is a binary file with columns of numbers that contain

the results of the EMTP simulation.
3.
After the emtp simulation is finished, run "emtpcat" from within MATLAB and
select "datafilM", the rest is automatic. Note that the command help emtpcat
displays emtpcat command line arguments.
4.
MCAT can be used for any EMTP case type: transient analysis and frequency
scan. Note that the keyword TACS_STAND_ALONE is not supported, but any
other cases with TACS_HYBRID are supported by MCAT.
5.
All plots generated with MCAT are based on an x-y plot. By default x is time,
but it can be set to become any other EMTP simulation variable. Both x and y are
matrices that can have any number of columns and any number of lines.
Entering the command "global x y" in MATLAB after pushing the PLOT button,
will make the x and y matrices available to MATLABs workspace.
6.
Function mcat2cop can be used to prepare the created graphic figure for copying
and pasting into a word processor document as a bitmap.
Command panel
A brief description of the command panel of mcat follows (from top to down and left to right):
Menu Output type: selects the type of EMTP output variable:

vn:
vb:
ib:
p:
e:
tacs:
um:
sm:
vnmag:
vnang:
cvn:
node voltages
branch voltages
branch currents
power
energy
TACS data
universal machine output
synchronous machine output
voltage phasor magnitude (frequency scan)
voltage phasor angle (frequency scan)
node voltages on a corona line
Static text data type identification: identifies the currently selected output variable
type. Pop-up of variable names: identifies all available variables of the above type.
Push button get file: to retrieve another EMTP file, data from several files can be
made available in a single session. Pop-up of file names: pop-up of currently available file names.
15 - 39
Output Processing
Check box legend in: if checked (default) the legend is inside the plot border, if
unchecked the legend is placed outside the plot border.
Pop-up same plot: selects the type of plot presentation.
Push button Select y: selects one or more y data from the pop-up of variable names.
Push button Clear: to clear all selections.
Push button Select x: selects one or more x data from the pop-up of variable names.
No selection means that x is the simulation time. The number of x selections must be
the same as the number of y selections or only the first x selection will be used for all
y selections.
Push button PLOT: to plot the selected waveforms.
Static text Selected y: shows current y selections, scrollable.
Static text Selected x: shows current x selections, scrollable.
Push button QUIT: to quit this session.
15 - 40
Section 16
16.1
Input Data Listing ............................................................................... 16-1

16.1.1
Comment Card ............................................................... 16-1
16.1.2
Cards to Begin a New Data Case ................................... 16-1
16.1.3
Specially-Requested Extensions to
Miscellaneous Data Cards .............................................. 16-6
16.2
16.1.4
TACS Data ...................................................................... 16-7
16.1.5
Lumped Series R-L-C Branches ..................................... 16-8
16.1.6
Pi-Equivalents ................................................................ 16-8
16.1.7
Coupled R-L Branches .................................................... 16-10
16.1.8
"CASCADED PI" Steady-State Feature ............................ 16-10
16.1.9
Saturable Transformer Component ................................ 16-12
16.1.10
Distributed-Parameter Transmission Line ..................... 16-13
16.1.11
Multi-Phase ZnO Surge Arresters .................................. 16-13
16.1.12
Branch Element Types 91-93 and 96-99 ........................ 16-14
16.1.13
Branches Using the Reference-Branch Idea .................. 16-15
16.1.14
Switches and Switched Elements ................................... 16-16
16.1.15
Sources ........................................................................... 16-17
16.1.16
Overriding Initial Conditions ......................................... 16-21
16.1.17
Output Specifications Cards ........................................... 16-22
16.1.18
Cards for Batch-Mode Plotting ...................................... 16-22
16.1.19
Blank Termination Cards ............................................... 16-24
Sinusoidal Steady-State Solution Output ............................................ 16-25

16.2.1
Definition of Power and Power Losses in AC

Steady-State Solution ..................................................... 16-26
16.2.2
Example of a Complete Steady-State

Solution Printout ............................................................ 16-27
16.2.3
Optional Type-59 S.M. Steady-State

Solution Printout ............................................................ 16-30
16.2.4
Abbreviated "FREQUENCY SCAN" Printout ................... 16-31
Section 16
16.3
Record Of Transient Results .............................................................. 16-33

16.3.1
Printed Transients Solution Variables ........................... 16-34
16.3.2
Sample Illustrative Output ............................................. 16-36
Section 16
This section is an aid for the interpretation of the contents of the EMTP output file.
16.1 Input Data Listing

A listing of all input data cards is always provided as part of the printed output:
1.
All 80 columns are printed, character by character, in columns 52 through 131.
2.
The character "1" is printed in column 51 of the line, to mark the beginning of
the card image.
3.
Columns 1 through 50 are reserved for interpretation of the input card in

question. Sections 16.1.1 to 16.1.19 summarize what is printed out for different
types of data cards. Note that the input interpretation is shown in a different
font, and that the numerical values in the output file are shown italicized within
square brackets.
4.
There is only one line of printing for each data card read. The contents of the
output file are meant to be printed with a printer that recognizes the character
"+" in column 1 as the overprint directive.
See Section 16.3.2 for an example of a typical output file.
16.1.1
Comment Card
Comments within and EMTP input data file have the characters "C " in columns 1-2 and are
echoed in the output file:
COMMENT CARD
16.1.2
There are two different cards which are associated with separating different data cases, or
stopping execution of the EMTP (see Section 4.1). Of the two interpretations shown below, the
16 - 1

first is for a "BEGIN A NEW DATA CASE" card, and the second is for the blank card to end an EMTP
data case.
MARKER CARD PRECEDING NEW DATA CASE.
BLANK TERMINATION-OF-RUN CARD.
Section 4.1.4 and Section 4.1.8 show how the user can re-define built-in program parameters
"LNPIN", "SZPLOT, "NSMTH", "EPSILN", "KPARTB" and "STATFR". Interpretation for these special
request records is as follows:
123456789 123456789 123456789 123456789 123456789 123456789
NEW PRINTER SPACING, LINES/DISTANCE = [LNPIN]
NEW PLOTTER PAPER-HEIGHT LIMIT.
[SZPLT]
CHANGE SUCCESSIVE OSCILLATION LIMIT.

MISC. DATA CONSTANT "EPSILN"
[EPSILN]
EXPONENT OF HIGH RESISTANCE.

REDEFINED POWER FREQUENCY =
[NSMTH]
R =
[10KPARTB]
[STATFR]
Section 4.1.4 explains usage of the "REPLOT" feature, the leading request card for which is
interpreted as follows:
REQUEST TO RE-PLOT OLD PLOT DATA
Section 4.1.8 explains how installation-dependent module "MIDOV1" of overlay number one can be
called on demand by the user. Interpretation of the special request record reading "FILE REQUEST"
is as follows:
CALL SUBROUTINE "MIDOV1"
Section 4.1.4 explains how six-character alphanumeric identification of the user can be specified,
principally for purposes of EMTP plot identification. Interpretation of the special request record
reading "USER IDENTIFICATION" is as follows:
USER IDENTIFICATION.
[USERID]
Section 4.1.1 explains how to disable the execution of the EMTP data case (one of several to be
solved). Interpretation of the special request record reading "ABORT DATA CASE" is as follows:
REQUEST TO ABORT THIS DATA CASE.
16 - 2

Section 4.1.2 explains how to print a range of EMTP error messages (KILL codes, beginning with
KILL1 and continuing through KILL2). Interpretation of the special request record reading "KILL
CODES" is:
LISTING OF ERROR MESSAGES.
[KILL1]
[KILL2]
Section 4.1.2 explains how the EMTP output vectors can be averaged over successive time steps.
Interpretation of the special request record reading "AVERAGE OUTPUT" is as follows:
REQUEST TO AVERAGE (SMOOTH) OUTPUT.
Section 4.1.5 explains how the total TACS working space of List 19 can be allocated among the 15
different TACS tables. First, there is interpretation for the three "ABSOLUTE TACS DIMENSIONS"
cards:
SET ABSOLUTE TACS LIST SIZE LIMITS.
S1ST TACS DIMENSIONS CARD.
2ND TACS DIMENSIONS CARD.
[LT1]
[LT11]
[LT2]
[LT12]
[LT3]
[LT13]
Second, there is interpretation for the three data cards associated with the special request record
"RELATIVE TACS DIMENSIONS":
PROPORTIONAL ALLOCATION OF TOTAL TACS STORAGE.
RELATIVE LIST SIZES.
[F1]
[F2]
[F3]
RELATIVE LIST SIZES.
[F11]
[F12]
[F13]
Section 4.1.3 explains how the results of two or more "STATISTICS" solutions can be combined for
purposes of overvoltage tabulation. First, there is interpretation for the special request record
which reads "TABULATE ENERGIZATION RESULTS":
REQUEST FOR "STATISTICS" TERMINATION.
Next comes interpretation for the one or more data cards which give the integers that characterize
the data files in question:
FILE NAMES.
[JF1] [JF2] [JF3] [JF4] [JF5] [JF6] [JF7] [JF8] [JF9]
Section 4.1.3 explains how a "STATISTICS" simulation can be saved for later statistical tabulation,
possibly in combination with other such partial solutions. Interpretation of the special request
record reading "STATISTICS OUTPUT SALVAGE" is as follows:
DISK STORAGE OF ENERGIZATION RESULTS.
[JFLSOS]
Section 4.1.3 explains how the user can omit running the base case solution of a "STATISTICS" or
"SYSTEMATIC" data case. The special request record "OMIT BASE CASE" is interpreted as follows:
OMIT BASE CASE IF STATISTICS/SYSTEMATIC CASE.
16 - 3

Section 4.1.1 explains how the user can have the familiar floating-point and integer miscellaneous
data cards (Section 4.2) read. Interpretation of the miscellaneous data cards themselves, as well as
any extensions, are as expected, so shall not be repeated here. Interpretation of the special request
record reading "MISCELLANEOUS DATA CARDS" is as follows:
REQUEST RECORD BEFORE MISC. DATA CARDS.
Section 4.1.2 explains how the user can request a special reading of the familiar data card which
provides for variation of the time-step printout frequency (see Section 4.3.2). Because the data
card itself and its interpretation are unchanged, they shall not be repeated here. Interpretation of
the special request record reading "CHANGE PRINTOUT FREQUENCY" is as follows:
REQUEST RECORD BEFORE PRINTOUT FREQUENCIES.
Section 4.1.2 explains how the user can manually define the time at which the search for extrema is
to begin. Interpretation of the special request record reading "BEGIN PEAK VALUE SEARCH" is as
follows:
EXTREMA CALC. BEGINS AT
[BEGMAX]
SECONDS.
Section 4.1.3 explains how the user can define the table-saving time in the case of "STATISTICS"
data case. Interpretation of the special request record "TIME OF DICE ROLL" is as follows:
STATISTICS TABLE-SAVING TIME
= [TENERG] SEC
Section 4.1.2 explains how the user can request output of the peak node voltage for the entire
network (peak over all nodes). The special request record "PEAK VOLTAGE MONITOR" is interpreted
as:
OVERALL PROBLEM PEAK NODE VOLTAGE
Section 4.1.6 explains how the total U.M. working space of List 25 can be allocated among the four
different U.M. tables. First, there is interpretation of the "ABSOLUTE U.M. DIMENSIONS" card:
U.M. TABLE SIZES.
[NCLFIX]
[NUMFIX]
[IOTFIX]
[IBSFIX]
An alternative is provided by the special request "RELATIVE U.M. DIMENSIONS", which is interpreted
as follows. Note that the derived absolute sizes, not the proportions, are printed:
DERIVED U.M. SIZES.
[NCLFIX]
[NUMFIX]
[IOTFIX]
[IBSFIX]
Section 4.1.1 explains how the user can restart a previously-halted simulation (which had MEMSAV
= 1 flag set). There are several data cards associated with this operation. The first of these is the
special request record "START AGAIN", which is interpreted as:
CONTINUE PARTIALLY-COMPLETED DATA CASE.
16 - 4

Next comes an indeterminate number of switch cards (switch number, new switch closing time),
each of which is interpreted as follows:
ALTERED SWITCH "TCLOSE".
[J]
[TCLOSEJ]
Finally, there is the terminator for such cards ("9999" punched in columns 58), which is given
the interpretation that follows:
TERMINATOR FOR SWITCH CLOSING TIMES.
Section 4.1.1 explains how the user can transfer control directly to the time-step loop.
Interpretation of the special request record "TIME STEP LOOP" is as follows:
TRANSFER CONTROL TO TIME-STEP LOOP.
Section 4.1.5 explains how the user can postprocess an old plot file using TACS. This begins with
an installation-dependent $-card ($OLDFILE), which shall not be documented here. But the key
special request record is "POSTPROCESS PLOT FILE", which is interpreted as follows:
POSTPROCESS.
IPLOT =
[IPLOT]
Section 4.1.2 explains how Karrenbauer node voltage output can be requested for a single
distributed line. Interpretation of this special request record is as follows:
REQUEST FOR TRICKY KARRENBAUER OUTPUT.
[MODOUT] [MODES]
Section 4.1.1 explains that the intended control of Type 1-10 EMTP sources from FORTRAN
subroutine "ANALYT" must be declared early, by means of a card bearing the key word "ANALYTIC
SOURCES". Interpretation of this special-request record is as follows:
REQUEST FOR USE OF ANALYTIC TYPE 1-10 SOURCES.
Section 4.1.5 explains that Type 1-10 EMTP sources can be controlled by TACS variables which are
declared on a "TACS EMTP SOURCES" card. Interpretation of this special-request record is as
follows:
TACS NAMES CONTROLLING TYPE 1-10 EMTP SOURCES.
Section 4.1.1 explains that the EMTP will automatically loop over steady-state solutions of
different frequencies, in response to a "FREQUENCY SCAN" card. Interpretation of this specialrequest record is as follows:
F-SCAN.
[fmin]
[f]
[fmax]
[NPD]
16 - 5

Section 4.1.8 explains that transient network node renumering can be bypassed by means of a
special request card which bears the key-word text "RENUMBER BYPASS". Interpretation is as
follows:
BYPASS OF TRANSIENT NETWORK RENUMBERING
Section 4.1.8 provides for user-redefinition of the two characters which are used in conjuction with
free-format data input. "CSEPAR" is the separator character between data fields, while "CHCONT" is
the character which requests a continuation card. Interpretation of this special-request record
which bears the key word "FREE FORMAT" is as follows:
FREE-FIELD CHARACTERS
[CSEPAR]
[CHCONT]
Section 4.1.2 explains that EMTP diagnostic printout can be controlled selectively, overlay by
overlay, using a card which bears the key-word text "DIAGNOSTIC". Interpretation of this specialrequest record is as follows:
DIAGNOSTIC PRINTOUT CODES. [ov1]
[ov2]
[ov3]
[ov3]
[ov4]
[ov5]
where [ov#] represents the overlay number whose output is being controlled.
The floating-point and integer miscellaneous data cards of Section 4.2 are interpreted as follows:
MISC. DATA.
[t]
[Tmax]
[XOPT]
MISC. DATA [IOUT][IPLOT]

[IDOUBL][KSSOUT][MAXOUT][IPUN][ICAT][MEMERG][IPRSUP]
16.1.3
Specially-Requested Extensions to Miscellaneous Data Cards
This class of data consists of cards which are only present in a data case if certain miscellaneous
data parameters of Section 4.2.2 take on special, characteristic values.
For a "STATISTICS" or "SYSTEMATIC" data case, "NENERG" is punched nonzero (with absolute value
equal to the number of energizations), and an extra statistical-overvoltage miscellaneous data card
is appended, as per Section 4.3.1. Interpretation of this record is as follows:
STATISTICS DATA.
[ISW]
[ITEST]
[IDIST]
[AINCR]
If the user wants to vary the printout frequency during the solution, "IPUN" is punched as -1, and
the special printout-frequency-change-card of Section 4.3.2 is appended. Interpretation of this
record is as follows:
PRINTOUT
[KCHG1]
[MULT1]
[KCHG2]
[MULT1]
16 - 6
[KCHG3]
[MULT3]
16.1.4
TACS Data
If a data case contains TACS modeling (see Section 14), such cards precede the first branch card
(and follow the last extension, if any, to the miscellaneous data cards).
First, comes special-request card with the keyword "TACS HYBRID" or "TACS STAND ALONE" (see
Section 4.4). Interpretation is as follows:
TACS SETUP.
DATA CARDS FOLLOW.
Next come TACS function and the zero-th order blocks of Section 14.6.1. The first card of each
function block is interpreted as follows:
TACS FUNCTION.
[output name]
ORDER
[N]
[GAIN]
Here "N" is the order of the block, as read from columns 1-2; "output name" is the block name, as
read from columns 3-8; and the block gain "GAIN" was read from columns 5156.
For a dynamic block (N > 1), the polynomial coefficients of "s" of the transfer function follow, on
two or more cards. The numerator coefficient precede the denominator coefficients. These two
interpretations follow:
NUMER
[N0]
[N1]
[N2]
[N3]
DENOM
[D0]
[D1]
[D2]
[D3]
Next come TACS signal source cards of Section 14.6.3. Interpretation is as follows, where "A",
"B", and "C" are read from columns 11-40 using 3E10.0 format:
TACS SOURCE.
[A]
[B]
[C]
Next come TACS supplemental variables. Three types can now be defined:
"VARIABLE":
old fixed-format expression
"FORTRAN EXPRESSION":
new free-format logical/algebraic FORTRAN expression
"DEVICE TYPE NN":
a TACS device type- 50, 51, 52, etc. ...
Each of the above may also belong to one of the three groups of supplemental variables: "inside",
"input" and "output". The interpretation is then as follows:
TACS SUPPLEMENTAL {INPUT/OUTPUT/INSIDE} {VARIABLE/FORTRAN EXPRESSN/DEV

TYPE}[CODE]
16 - 7

Certain TACS supplemental device cards may be immediately followed by an indeterminate
number of associated data cards, each of which bears one or two floating-point numbers.
Examples are the Type-55 digitizer, the Type-56 point-by-point nonlinearity, and the Type-57
time-sequenced switch. The end of such cards is marked by a "9999" card (punched in columns
13-16). Interpretations are as follows:
DATA VALUES.
[X]
[Y]
END OF DATA VALUES FOR LAST-READ DEVICE.
Next come the cards (or single card) which request TACS output variables, to be used for printing
and/or plotting. For selectively-specified outputs, interpretation is:
TACS VARIABLES FOR EMTP OUTPUT VECTOR.
The output of all TACS variables can be requested by a single card with a "1"-punch in column 2.
Interpretation is then as follows:
REQUEST FOR OUTPUT OF ALL TACS VARIABLES.
Finally come initial condition specification cards, if any. Interpretation is as follows:

INIT. COND.
16.1.5
[NAME]
[INITIAL CONDITION]
Lumped Series R-L-C Branches
Each lumped series R-L-C branch card has the following interpretation:
SERIES R-L-C.
16.1.6
[R]
[C or C]
[L or L]
Pi-Equivalents
Here a distinction is made between the card for the first conductor of a piequivalent and possible
succeeding cards. The first card has the following interpretation:
1ST OF PI-CKT.
[R11]
[L11]
[C11]
The first card of succeeding conductors of a multi-phase pi-equivalent bear no descriptive text,
with only the first five numbers of the card being printed out. Assuming conductor k,
[Rk1]
[Lk1]
[Ck1]
[Rk2]
[Lk2]
16 - 8

The fourth and later conductors of course require continuation cards, since all parameters for row
k cannot be punched on a single card. For such cards, either three or five of the punched numbers
are printed:
If the card has one triplet of numbers (Rk,m, Lk,m, Ck,m):
[Rkm]
[Lkm]
[Ckm]
[Rkm]
[Lkm]
If the card has two triplets, for columns m and m+1; or

If the card has three triplets, for columns m, m+1, m+2:
[Rkm]
[Lkm]
[Ckm]
[Rk,m+1]
[Lk,m+1]
The preceding description assumes that the older, narrow formats are being used ($VINTAGE, 0). If
instead the new, wider format is being used ($VINTAGE, 1), then any one data card only bears one
triplet of numbers Rk,m, Lk,m and Ck,m. The first data card is interpreted no differently than with
the narrow formats (see top of section, "1ST OF PI-CKT."):
1ST OF PI-CKT.
[R11]
[L11]
[C11]
But the second and later cards (if any; if two or more phases) use a unique format. Interpretation
of the data card which begins a new row K of the matrix (for K=2, 3, ...) is as follows:
PHASE
[K]
[Rk1]
[Lk1]
[Ck1]
All other cards of the row (column M, for M=2, 3, ...) are interpreted as follows:
CONT
[Rkm]
[Lkm]
[Ckm]
The user might have occasion to employ the alternate formulation using matrices [R] and [B]
instead of the usual [R] and [L]. In this case, he will use the special formulation-change request
records reading "USE RB" and "USE RL", which are interpreted as follows:
BEGIN COUPLED, LUMPED ELEMENTS USING (R), (B).
BEGIN COUPLED, LUMPED ELEMENTS USING (R), (L).
16 - 9
16.1.7
Coupled R-L Branches
First consider the specification of the R-L branches in phase coordinates.

A distinction is made between the card for the first conductor of a coupled RL group, and all
succeeding cards. For the first-conductor card:
1ST OF COUPLED R-L.
[R11]
[L11]
All succeeding cards simply have printout of the 1, 2, or 3 pairs of R-L punched on the card, in
order of increasing column number. For the extreme case having three pairs in row k, columns m
through m+2, the interpretation is:
[Rk,m] [Rk, m] [Rk,m+1]
[Lk,m+1]
[Rk, m+2]
Should the user input zero and positive sequence values rather than the normal phase values, then
interpretation of the three cards is as follows:
First Card (with zero sequence values):

1ST OF COUPLED R-L.
[Ro]
[Lo]
Second Card (with positive-sequence values):

[R1]
[L1]
[0.0]
[0.0]
[0.0]
Third Card:
The third card gives the third row of the matrices as if the input had been made in phase quantities;
it gives the phase equivalent of the preceding sequence values:
[Rm]
[Lm]
[Rs]
[Lm]
[Rs]
Just as with pi-circuits of Section 5.2, it is possible to employ the alternate formulation using
matrices [R], [B] instead of the usual [R] and [L]. In this case, the user inputs special request
records reading "USE RB" and "USE RL", which are interpreted as shown at the end of Section
16.1.6.
16.1.8
"CASCADED PI" Steady-State Feature
Data input for "branch" which is to be defined using the "CASCADED PI" feature of Section 5.6
begins with the header card (Class 1 input), which is interpreted as follows:
CASCADED-PI HEADER CARD.
[NPHCAS]
[FREQCS]
16 - 10

Class 2 data defines the terminal node names at the sending and the receiving ends, as well as
matrices [R], [L] and [C] for the sending-end line section. Interpretation for these data cards is
identical to that used for the conventional pi-circuit component of Section 5.2. For convenience,
this interpretation explanation is reproduced immediately below.
Here a distinction is made between the card for the first conductor of a piequivalent and possible
succeeding cards. The first card has the following interpretation:
1ST OF PI-CKT.
[R11]
[L11]
[C11]
The first card of succeeding conductors of a multi-phase pi-equivalent bears no descriptive text,
with only the first five numbers of the card being printed out. Assuming conductor k,
[Rk1]
[Lk1]
[Ck1]
[Rk2]
[Lk2]
The fourth and later conductors of course require continuation cards, since all parameters for row
k cannot be punched on a single card. For such cards, either three or five of the punched numbers
are printed:
If the card has one triplet of numbers (Rk,m, Lk,m, Ck,m):
[Rkm]
[Lkm]
[Ckm] [Rkm]
[Lkm]
If the card has two triplets, for columns m and m+1; or

If the card has three triplets, for columns m, m+1, m+2:
[Rkm]
[Lkm]
[Ckm]
[Rk,m+1]
[Lk,m+1]
Class 3 or Class 4 data is for the line-position (transposition) card, which is interpreted as follows:
LINE POS. [DSECTJ] [MULTIP] [MSER] [MBR] [MSECT] [MAPCAS1] [MAPCAS2]
[MAPCAS3]
Actually, if the line in question consists of more than 14 conductors, one or more "continuation"
cards is needed for the input of MAPCAS(I) which was begun on the just-interpreted card. Each
such extra card ("K" equal to 14 for the first; "K" equal to 28 for the second; etc.) would be
interpreted as follows, if it existed:
[MAPCASk+1][MAPCASk+2][MAPCASk+3][MAPCASk+4][MAPCASk+5][MAPCASk+6][MAPCASk+
7][MAPCASk+8][MAPCASk+9]...[MAPCASk+10][MAPCASk+11][MAPCASk12]
Class 5 data is for the input of series R-L-C branches, each card of which bears the following
interpretation:
SERIES R-L-C.
[Rseries]
[Lseries]
[Cseries]
16 - 11

Class 6 data is for the input of shunt R-L-C branches, each card of which bears the following
interpretations:
SHUNT
[N1]
[N2]
[R]
[L]
[C]
Class 7 data is interpreted identically to Class 2 data.

Class 8 data consists of the single "STOP CASCADE" record, which is interpreted as follows:
TERMINATION OF CASCADED PI.
16.1.9
Saturable Transformer Component
See (Section 5.4)

The first card of a single-phase saturable transformer component bears the special request word
"TRANSFORMER", which is interpreted as follows:
SAT. XFORMER.
[isteady]
[steady]
[Rmag]
Actually, this interpretation assumes that the reference component feature has not been used to
define this unit (i.e., field "BUS3" of columns 1520 is blank). If the reference component feature
is being used, interpretation then consists of only the following:
TRANSFORMER COPY USING REFERENCE NAME
[BUS3]
Assuming no use of the reference component features, the "TRANSFORMER" request card is
followed by current-flux breakpoints (if any) which define the magnetization characteristic.
Interpretation of all such records, if any, is as follows:
BREAKPOINT
[Current]
[Flux]
With no reference component usage, the preceding data class (whether present or absent) is
terminated by a 9999-card, which is interpreted as follows:
SPECIAL TERMINATION OF POINTS CARD.
Next come the transformer winding cards, which are interpreted as follows assuming no usage of
the reference-component capability:
WINDING [ITYPE]
[Rk]
[Lk]
[Volts]
If reference component usage were being employed for this unit, columns 14-49 of the above
interpretation for a winding card would have been left blank. This is because the winding card in
question bears no floating-point parameters in such a case.
16 - 12
16.1.10 Distributed-Parameter Transmission Line

The first two cards of a distributed-parameter line carry the zero and positive-sequence
parameters, respectively. Interpretation consists of three principal floating-point values read from
the card, plus the characteristic impedance Zo (computed neglecting line loss) and the travel
time t:
[R]
[A]
[B]
[ Zo =
L
---- ]
C
[ = l LC ]
If the line is continuously transposed (Section 7.3), then the third and any later cards only define
the terminal node names, and are therefore only interpreted by the message:
3RD OR LATER UNIF.-TRANSPOSED DISTRIBUTED COND.
But if instead the component is a double-circuit distributed-parameter line with zero-sequence

inter-circuit coupling (Section 7.3.4), then interpretation of the third card is identical to the first
two; in this case, the third card has parameters for the "IL" mode. The fourth through sixth
conductor cards then bear the following interpretation:
4TH OR LATER DOUBLE CKT DISTRIBUTED CONDUCTOR.
16.1.11 Multi-Phase ZnO Surge Arresters

Section 8.9 describes the various data cards which are associated with multi-phase, compensationbased, ZnO surge arrester modeling. These begin with the definition of a subtype of Type-92
nonlinear resistance. This first data card bears the following interpretation:
GAP, TYPE 92.
[0.0]
[0.0]
[ZnO-CODE]
Then follows a single card specifying general arrester data which bears the following
interpretation:
VREF, VGAP, VINIT
[VREF]
[VFLASH]
[VZERO]
The actual arrester characteristic will follow the above card. There can be any number of cards
with each one of them bearing the following interpretation:
BREAKPOINT
[COEF]
[EXPON]
[VMIN]
Recall that there is an indeterminate number of these cards terminated by a 9999 card. This
termination card bears the interpretation:
SPECIAL TERMINATION-OF-POINTS CARD
16 - 13

The characteristic before flashover appears first, followed by the characteristic flashover (if the
arrester is equipped with a gap). The second part of the arrester characteristic, if any, is
also terminated with a 9999-card.
16.1.12 Branch Element Types 91-93 and 96-99

For purposes of documenting the interpretation, it is convenient to lump together all remaining
nonlinear and pseudo-nonlinear elements. These have type codes 91, 92, 93, 96, 97, 98, and 99, as
described in Sections 8.2 onward.
Definition of each of these elements begins with a first data card which specifies the terminal code
names, among other things. Assuming no usage of the reference branch procedure, interpretations
are as follows:
Staircase time-varying R, Type-97:

TYPE-97 R(T)
[Vflash]
[Tdelay]
Pseudo-nonlinear R, Type-94:
See Section 8.3
TYPE-99 NONLINEAR R.
[Vflash]
[Tdelay]
[JUMP]
Pseudo-nonlinear L, Type-98:
See Section 8.4
TYPE-98 PSEUDO-NONLINEAR L.
[isteady]
[steady]
Pseudo-nonlinear hysteretic reactor, Type-96 :

See (Section 8.5)
HYSTERESIS
[isteady]
[steady]
[resid]
Time-varying R, Type-91 (Section 8.7):

TIME VARING R. TYPE 91.
[0.0]
[R-CODE]
The above card is followed by a single card specifying the starting (breakdown) voltage VSTART.
This card bears the following interpretation:
VSTART=
[VSTART]
16 - 14

The actual characteristic R = f(tR) is specified on an indeterminate number of cards. Each of these
cards bears the following interpretation:
BREAKPOINT
[R(tr)]
[tR]
The specification of the resistance values is terminated by a 9999-card which bears the following
interpretation:
SPECIAL TERMINATION-OF-POINTS CARD.
Piecewise-linear, continuous R, Type-92:

See (Section 8.8)
GAP, TYPE 92.
[0.0]
[0.0]
[R-CODE]
Then follows a single card specifying general resistor data which bears the following
interpretation:
RLIN,VFLASH,NFLASH
[RLIN]
[VFLASH]
[NFLASH]
The actual resistor characteristic will follow the above card. There can be any number of cards
with each of them bearing the following interpretation:
BREAKPOINT
[CUR]
[VOLT]
The indeterminate number of characteristic cards is terminated with a 9999-card which bears the
following interpretation:
16.1.13 Branches Using the Reference-Branch Idea

When the reference-branch concept is used, fields BUS3 and BUS4 (columns 1520 and 21-26,
respectively of data card) carry the node names of the preceding branch which is being referred to.
The first card of a coupled group using this capability bears the following interpretation:
REFERENCE BRANCH.
COPY
[BUS3]
TO
[BUS4]
All later branches of the coupled group (if any) bear the following abbreviated interpretation:
REFERENCE BRANCH.
COPY CONT.
16 - 15
16.1.14 Switches and Switched Elements

See Section 9
The "ordinary" switches of Section 9.3 are EMTP switch components which have no linear
resistance or inductance element associated with them, and which are not valves, diodes, or TACScontrolled. Generally these switches are type-code zero (as punched in columns 1-2), although
type-code '76' is also possible for a Class 3 switch. Ordinary switches come in five classes, with
the first four interpreted as per the following format:
Class 1:
SWITCH
[Tclose]
[Topen]
[I]
[Tclose]
[Tdelay]
[I]
[Vflash]
[T or TB]
[ or B]
[I]
[Vflash]
[t]
[NSTEP]
Class 2:
SWITCH
Class 3:
SWITCH
Class 4 independent:
SWITCH
[Tmid or Tbeg]
Class 4 dependent:
SWITCH
[T,offset]
The final one, for Class 5 switches ("MEASURING" switches), bears the following interpretation:
PERMANENTLY CLOSED SWITCH USED FOR METERING.
For diode or TACS-controlled valve operation, the Type-11 switch of Section 9.5.2 is used.
Parameter "GRID" is read from the A6 field in columns 6570. Parameter "CLAMP" is read from the
A6 field in columns 7176.
For diode or TACS-controlled valve operation, the Type-12 switch of Section 9.5.4 is used.
Parameter "SPARK" is read from the A6 field in columns 6570. Parameter "CLAMP" is read from
the A6 field in columns 7176.
For simple TACS-controlled switch operation, the Type13 switch of Section 9.5.6 is used.
Parameter "SIGNAL" is read from the A6 field in columns 7176.
16 - 16

Interpretation is as follows:
DIODE, NO TACS GRID
VALVE, TACS GRID = [VALUE]
TACS CLAMP = [VALUE].
GAP, NO TACS SPARK

GAP, TACS SPARK = [VALUE]
VALVE, TACS GRID = [VALUE]
SWITCH, TACS CONTROL SIGNAL = [VALUE]
16.1.15 Sources
See Section 10
Source cards are interpreted as follows:
For sources Type 115:
SOURCE.
[AMPL]
[f ;]
[TO;.;]
[Tstart]
For Type16 dc-simulator sources, the first card is interpreted as:

SOURCE
[K]
[init; iset]
[T1]
[iinitial]
The second and final card then bears only the label:
SECOND DC SIMULATOR CARD.
The Type17 EMTP source component provides for representation of a zinc-oxide surge arrester
that has constraint equation i = A * (v/vref)G. Interpretation is as follows:
SOURCE
[A]
[G]
[Vref]
Type 110 sources have the functions f(t) read in off cards in point-by-point fashion, as the
transient solution progresses. Assuming that the user is printing out solution results, the input card
images may be periodically separated by such printed output. In any case, the interpretation of the
source cards of type 110 is as follows:
ANOTHER INPUT CARD FOR TYPE 1-10 SOURCES.
16 - 17

If all such source definitions are to terminate before the last time-step of the EMTP simulation, then
a bounding record with "9999" punched in columns 58 is to be used. This last card then bears the
added column 4548 interpretation as shown below.
END.
Dynamic Synchronous Machines:

Data input for dynamic synchronous machines (S.M.) EMTP source components is described in
Section 10.11. The type-code which is punched in columns 12 will be equal to 59, in this case.
Each dynamic S.M. requires a number of data cards, the interpretation of which shall now be
described in order of data input.
The S.M. cards begin with Class 1 data, for which there are three cards, one for each armature
connection. The first of these is interpreted as follows:
SOURCE.
[VOLT]
[FREQ] - [ANGLE]
The second Class 1 card bears the interpretation:

2ND PHASE OF S.M.
[VOLT] - [ANGLE]
Here the printed angle is in degrees. Class 1 data is then finished by a comparable interpretation
for the third card (which is for phase "c"):
3RD PHASE OF S.M.
[VOLT] - [ANGLE]
Class 2 S.M. data (if it exists) consists of up to three optional special-request cards, of which the
order is immaterial. One of these represents a request for a delta connection of the S.M. armature
windings. Interpretation of this record which bears the key-word text "DELTA CONNECTION" is as
follows:
NOTIFICATION OF DELTA-CONNECTED ARMATURE.
The second optional special-request card is used to redefine one or more tolerances or iteration
limits which are used in the S.M. solution process. Interpretation of this record which bears the
key-word text "TOLERANCES" is as follows:
[EPSILON]
[EPSUBA]
[EPOMEG]
[EPDGEL]
[NIAMAX]
[NIOMAX]
The third and final optional special-request card is used to obtain the mathematical fitting of the
S.M. parameters of Park's equations to the available data. Interpretation of this record which bears
the key word "PARAMETER FITTING" is as follows:
OPTIMIZE PARK DATA.
[FM]
16 - 18

Class 3 S.M. data consists of either three or four cards, the first of which bears the following
interpretation:
4TH S.M. CARD.
[NUMAS]
[KMAC]
[KEXC]
[NP]
[SMOUTP]
[SMOUTQ]
The saturation data on this card applies to the direct axis. Additional saturation data will be
specified on a supplemental card following the "4TH S.M. CARD". It will be interpreted as follows:
Q-AXIS SATURATE
[AGLQ]
[S1Q]
[S2Q]
The remaining cards of Class 3 S.M. data depend upon whether standard manufacturer-supplied
data are being used. If "PARAMETER FITTING" card used, then two additional cards which are
interpreted as follows complete the Class 2 data:
5TH S.M. CARD.
6TH S.M. CARD.
[Ra] [Xl] [Xd]

[Td] [Tqo] [Tdo]
On the other hand, should the user have chosen to describe the machine by means of per unit
reactances and resistance matrices (no "PARAMETER FITTING" card), then the Class 2 S.M. data is
completed with three cards which are interpreted as follows:
5TH S.M. CARD.
[Xf]
[Xaf]
[Xfkd]
6TH S.M. CARD.
[Xq]
[Xag]
[Xgkq]
7TH S.M. CARD.
[Xo]
[Ra]
[Rf]
Class 4 S.M. data consists of mass cards, one for each mass of the shaft system of the rotor.
Interpretation of each such card is as follows:
MASS CARD
[CARD #]
[MASS #]
[EXTRKS]
[HJ] [D]
Here the "MASS #" is the number of the mass (variable "ML") as read from columns 1-2 of the card.
The "CARD #" is simply a counter which is equal to unity for the first such mass card, equal to two
for the second, etc.
Mass cards are terminated by a special terminator card which bears the following interpretation:
BLANK CARD TERMINATING MASS CARDS.
Class 5 S.M. data consists of an indeterminate number of output-variable specification cards. Each
request card bears the following interpretation:
OUTPUT REQUEST CARD FOR CLASS
[CLASS]
16 - 19

The output-variable specification cards are terminated witha special terminator cad which bears the
following interpretation:
BLANK CARD TERMINATING OUTPUT REQUESTS.
Universal Machine (U.M.) Model:

See Section 10.12
The U.M. data cards are announced as an EMTP source of Type-19 (punched in columns 12).
Nothing else is required on this lead card, whose interpretation shows the number of cells of List
25 storage which are actually being used by the U.M. tables. Recall that "ABSOLUTE U.M.
DIMENSIONS" of Section 4.1.6 determines this:
U.M. DATA BEGINS.
LIST-25 CELLS USED =
[L25]
Next comes the data card bearing nothing but data flag INPU, which chooses between per unit and
physical units for the input data:
BEGIN U.M.
INPU =
[INPU]
Next comes the "machine table", which will have a triplet of data cards for each U.M. component
that is being used. Symbols JSD and JSQ are abbreviations for JSATD and JSATQ, respectively. The
three interpretations are as follows:
FIRST OF CARD SPLIT.
[JTYPE]
2ND OF SPLIT.
[OMEGM]
3RD OF SPLIT.
[THETAM]
[NCLD]
[LMUD]
[LMUQ]
[NCLQ]
[JSD]
[JSQ]
Such triplets of machine-specifications continue until they are terminated by a blank card, which is
interpreted as follows:
BLANK CARD ENDING MACHINE TABLE.
Then come the coil cards, one for each coil of all U.M. components, in order. Using the
abbreviation IO = CUROUT for the output request flag, interpretation of the card for coil KK is as
follows:
[COIL]
[KK]
[RESIS]
[LLEAK]
[IO]
The last such U.M. coil card is followed by a terminating blank card, which is interpreted as
follows:
BLANK CARD ENDING COIL TABLE.
This completes the U.M. data input, and control returns to "OVER 5" source input.
16 - 20
16.1.16 Overriding Initial Conditions

Because the initial condition capability of the EMTP is far from complete, the explanation provided
here amounts to little more than a documentation of those initial condition cards which are
honored.
First, there are the original three classes of initial condition input, as per the subdivisions under
Section 11.3 Rule 3. These are node voltages, linear branches, and nonlinear branches (type code
ID = 2, 3, and 4, respectively):
Cards for inputting node voltages come first (ID=2 punched in columns 1-2), and are interpreted as
follows. Only the first such card actually has the third field (frequency f) actually printed.
NODE VOLT INIT COND.
[Re {E}]
[Im {E}]
[f]
Next come the cards specifying currents in linear branches, with ID of 3. For lumped series R-L-C,
for pi-equivalents, the interpretation is as follows:
LINEAR I.
[A]
[B]
[C]
Finally, there is the interpretation of the cards for nonlinear and time-varying branches (ID = 4,
punched in columns 1-2 of the card):
NONLIN. BRANCH INIT COND.
[Ikm]
[km]
The Type96 hysteretic inductor is an exception to this general rule for nonlinear elements in that
there are two additional cards which follow the initial one just described. Call these cards 2 and 3.
Card 2 is punched with two integers and four floating point numbers, reading from left to right.
The integers and first two floating point numbers appear in the interpretation of card 2:
TYPE-96.
[N15]
[N16]
[VCHAR(N17)]
[VCHAR(N17+1]
Card 3 is punched with one integer and four floating point numbers; the integer and first two
floating-point numbers are part of the interpretation, which appears as follows:
TYPE-96.
[N20]
[VCHAR(N18)]
[VCHAR(N19]
No further explanation of this Type96 hysteretic inductor shall be provided, since such initial
conditions should always be EMTP-punched rather than user-punched, and understanding is not
actually required.
In the case of EMTP-punched terminal conditions which are re-used as EMTP initial conditions,
there also is an extra class of cards (in addition to those listed above), for switches. There is one
such card for each switch, with type code ID = 5. Each such card carries two bus names (2A6
16 - 21

format, to identify switch), three integers (3I4), and four floating-point variables (4E13.6).
Interpretation only uses the first three integers and the first floating-point variable.
SWITCH INIT. COND.
[N1]
[N2]
[N3]
[D1]
16.1.17 Output Specifications Cards

See Section 12
Solution node voltages which are to be printed or plotted as functions of time are defined on one or
more cards as per Section 12.1. If the user requests all node voltage outputs by means of a 1
punch in column 2, the interpretation is:
REQUEST FOR OUTPUT OF ALL NODE VOLTAGES.
But for the usual case of selective voltage output, each card is interpreted by:
CARD OF BUS NAMES FOR NODE-VOLTAGE OUTPUT.
The preceding records were for deterministic (i.e., non-statistical) data cases, in accord with
Section 12.1. But for a "STATISTICS" data case, the separate, special specification of Section 12.2
applies. Here only selective output requests are allowed. Depending upon whether node voltage
or voltage difference ("branch voltage" for short) outputs are being requested, interpretation is as
per one of the following:
Node voltage requests, "STATISTICS":

STATISTICAL OUTPUT OF NODE VOLTAGES.
[BASEV]
Voltage-difference requests, "STATISTICS":

STATISTICAL OUTPUT OF BRANCH VOLTAGES.
[BASEV]
16.1.18 Cards for Batch-Mode Plotting

The card of 78character case-title text, Subsection 15.2.1, is interpreted as follows:
PLOT CASE-TITLE TEXT.
Cards defining the graph subtitle text of Subsection 15.2.2 are interpreted as follows, assuming that
the user has not attempted to input more than the limit of four for any one plot:
PLOT SUBTITLE CARD.
16 - 22

Excess graph subtitle-text cards (in excess of four) are simply ignored by the EMTP; they are given
the following interpretation:
OVERFLOW SUBTITLE CARD DISCARDED.
A plot specification card of Subsection 15.2.3 is interpreted as follows:

** PLOT CARD.
[HPI]
[HMIN]
[HMAX]
The extra "continuation card" of Subsection 15.2.4, for the input of three or four branch-variable
node-pair identifications (six or eight node names), is interpreted as follows:
CONTINUATION TO READ BRANCH NODE-PAIR NAMES.
The extra "continuation card" of Subsection 15.2.5, for re-reading the five floating-point
parameters of the plot-specification card, is interpreted as follows:
RE-READ OF FLOATING-POINT FIELDS FOR ACCURACY.
The card for Postscript graph-size adjustment, bearing key word "HEIGHT" in columns 2530 as
per Subsection 15.2.6, is interpreted as follows:
GRAPH SIZE ADJUSTMENT CARD.
[BEGIN]
[SPAN]
The card for redefinition of the smoothing tolerance "EPS", bearing key word "SMOOTH" in
columns 2530 as per Subsection 15.2.8, is interpreted as follows:
REDEFEINITION OF SMOOTHING TOLERANCE.
[EPS]
Batch-mode plotting can be done on either the line printer, a Postscript file, or both, for any given
plot. The three request cards of Subsection 15.2.9 allow for changing this graph mode selection
by use of the three key words "PRINTER PLOT", "POSTSCRIPT PLOT", and "POSTSCRIPT PRINTER".
Interpretation for these three cards is as follows:
REQUEST FOR POSTSCRIPT PLOT.
REQUEST FOR POSTSCRIPT AND LINE PRINTER PLOTS.
With Postscript plotting, there is a choice of several different plotter pens, and the EMTP allows the
user to specify these both for the background grid and also for the first four curves of a graph.
This is via the "PEN CHOICE" request of Subsection 15.2.11, which reads five pen numbers. All
five of these integers are confirmed in the data card interpretation:
GRID & PEN CHOICES.
[KGR]
[KP1]
[KP2]
16 - 23
[KP3]
[KP4]

For printer (character) plots, there is redefinition of the line limit for sparse plots via the "PLOT LINE
LIMIT" request of Subsection 15.2.12. Interpretation confirms the new line limits "LINLIM":
LINE LIMIT FOR SPARSE PRINTER PLOTS =
[LINLIM]
For Postscript plots, there is the option of superimposing more than one plot on top of each other,
thanks to the "SUPERIMPOSE" request of Subsection 15.2.13. The request card bears three integers,
which are all echoed in the interpretation:
GRAPH SUPERPOSITION.
[MPL1]
[MPL4]
[MPL5]
For Postscript plots, it is possible to plot one EMTP variable against another, thanks to the "X-Y
PLOT" request of Subsection 15.12.15. Three data cards are involved. The first of these bears the
just-stated request word, along with a 24-character horizontal axis label, which is echoed in the
data card interpretation:
PLOT X VS. Y.
[label]
Horizontal axis
Next comes a data card for the horizontal ("X") axis, bearing three floating point numbers, which
are echoed in the interpretation:
X-AXIS
[LENX]
[XMIN]
[XMAX]
Finally there is a corresponding card for the vertical ("Y") axis, though it also bears fourth and fifth
parameters which do not appear in the interpretation:
Y-AXIS
[LENY]
[YMIN]
[YMAX]
16.1.19 Blank Termination Cards

Various classes of data are terminated by blank cards. Any blank card so read is interpreted as
such, with an appropriate message telling what the blank card has signaled the end of. All of these
are shown below, in the order that they would be encountered in a data case. Note that:
1)
(A)
The first five are associated with TACS data (if any).
(B)
Number 9 will not be present of the user requests the output of all node
voltages (by means of a "1"-punch in column 2). For a "STATISTICS" or
"SYSTEMATIC" data case, this card would pertain to the base-case solution only,
in any event.
(C)
Number 11 exists only for a "STATISTICS" or "SYSTEMATIC" data case.
(D)
Number 12, which stops execution of the EMTP, is actually the first card of the
following (nonexistent data case).
BLANK CARD TERMINATING TACS FUNCTION BLOCKS.
16 - 24
2)
BLANK CARD TERMINATING TACS SOURCE CARDS.
3)
BLANK CARD TERMINATING TACS SUPPL. VAR. CARDS.
4)
BLANK CARD TERMINATING OUTPUT VARIABLE REQUESTS.
5)
BLANK CARD TERMINATING TACS INIT. CONDIT. CARDS.
6)
BLANK CARD TERMINATING BRANCH CARDS.
7)
BLANK CARD TERMINATING SWITCH CARDS.
8)
BLANK CARD TERMINATING SOURCE CARDS.
9)
BLANK CARD ENDING NODE NAMES FOR VOLTAGE OUTPUT.
10)
BLANK CARD TERMINATING PLOT SPEC. CARDS.
11)
BLANK CARD TERMINATING STATISTICS OUTPUT CARDS.
12)
16.2 Sinusoidal Steady-State Solution Output

The sinusoidal steady-state output can be obtained:
1.
For an AC steady-state solution at one or more frequencies, whether followed by

a transients study or not. In this case, the data case must contain at least one
Type-14 source with TSTART < 0 or at least one Type59 S.M. model.
2.
For an AC-steady state solution with the frequency of the sources varied in steps
f from fmin to fmax (frequency scan).
If the user is interested in printout of the steady-state solution, he has two options:
(A)
Complete Solution:
By punching a "1" in column 32 of the Integer Miscellaneous Data Card

(Section 4.2.2), complete printout of the flows in all network branches is produced.
The printout includes complex currents and powers, as well as terminal voltages for
the two ends of each branch, in both rectangular and polar form. Included are also real
and reactive power losses in all branches.
(B)
Transients Variables Only:

16 - 25

By setting Tmax of the Floating-Point Miscellaneous Data Card to be non-positive (see
Section 4.2.1; columns 916), the EMTP will never reach the transients mode in solving
the data case. Rather, the steady-state solution is found, and then a special phasor
printout of only those variables normally printed during the transient run (see Section
12) occurs. The case is then terminated following this printout.
Note that following the output of Option A), a transient solution to the network is allowed. Also,
either, both, or neither steady-state printout can be requested, since they are independent. See
Section 16.2.2 for sample output of this type.
16.2.1
Definition of Power and Power Losses in AC Steady-State Solution
The formula used for the power flow calculation is P+jQ=VI*/2. Hence it is implicitly assumed
that sources are specified in peak, and not in rms terms. Such usage of peak quantities is consistent
with source-input rules of Section 10. For example, a nominal wye-connected voltage source for
500kV would have amplitude of 500 2
--- = 408,25 kV. With sources given in volts and amps,
3
power then comes out in watts or vars.

The complete steady-state solution output provides printout on the extreme right of the page for
power loss, branch by branch (see example of Section 16.2.2). Assuming that the branch in
question connects nodes k and m as shown, then by definition, the program prints out the
following:
Ploss + j Qloss = (Pkm + Pmk) + j (Qkm + Qmk)

k
Pmk
Qmk
Pkm
Qkm
Figure 16.1: Branch (k,m)

For simple series R-L-C branches, this is indeed the true loss figure for the branch the heating
loss (considering real power only) due to resistance of the branch.
But for distributed-conductor or pi-equivalent branches, the loss figure so printed is not really loss
at all; it is simply the sum of the power inputs at the two ends.
16 - 26
Beware: Conservation of energy dictates that power in equals power out, but only if terminals
k and m were the only connection to branch (k,m) would the printout actually give
true branch loss. Mutual coupling (capacitive and/or inductive) between branch (k,m)
and other branches provides additional paths for power entry and/or exit (see Figure
16.2); thus adding the k-to-m and the m-to-k powers does not provide loss.
mutual
coupling
Figure 16.1:
Mutual coupling with branch (k,m) may be present.
Conclusion: With distributed or pi-equivalent branches, do not interpret the printed loss
figure (P or Q) of the complete steady-state solution printout as loss attributable
uniquely to the branch in question. The one exception to this warning is in the singlephase (uncoupled) case, where there is no problem.
16.2.2
Example of a Complete Steady-State Solution Printout
A sample of the complete steady-state solution printout is shown below. This begins with branch
flows, as follows:
SINUSOIDAL STEADY STATE SOLUTION, BRANCH BY BRANCH.
ALL FLOWS ARE AWAY FROM BUS, AND REAL PART, MAGNITUDE, OR P
IS PRINTED ABOVE THE IMAGINARY PART, THE ANGLE, OR Q.

FIRST SOLUTION FREQUENCY =
0.600000000E+02
HERTZ.
BUS K
NODE VOLTAGE
BRANCH CURRENT
POWER FLOW
BUS M
RECTANGULAR
POLAR
RECTANGULAR
POLAR
P AND Q
NAVH A
0.5429179E+05 0.4168487E+06
-0.2514881E+03 0.1348524E+04
0.2669555E+09
-0.4132981E+06
-82.5163
-0.1324866E+04
-100.7481
0.8793446E+08
MCC1 A
-0.1599845E+06 0.3973212E+06
0.2514881E+03 0.1348524E+04
-0.2610362E+09
-0.3636882E+06
-113.7444
0.1324866E+04
79.2519
0.6024744E+08
NAVH B
-0.3850725E+06 0.4168487E+06
-0.1021624E+04 0.1348524E+04
0.2669555E+09
0.1596310E+06
157.4837
0.8802283E+03
139.2519
0.8793446E+08
MCC1 B
-0.2349709E+06 0.3973212E+06
0.1021624E+04 0.1348524E+04
-0.2610362E+09
0.3203947E+06
126.2556
-0.8802283E+03
-40.7481
0.6024744E+08
16 - 27
POWER LOSS
P AND Q
0.5919274E+07
0.1481819E+09
0.5919274E+07
0.1481819E+09

Note that the first branch connects node "NAVH A" with node "MCC1 A". Using "k" and "m" as
abbreviated subscripts for those, the following diagram shows which quantities are printed out for
this branch:
Pkm + jQkm
Pmk + jQmk
node "k"
node "m"
Ikm
Imk
In mathematical notation, the following answers will be read from the printout:
Vk
54291.79
j413298.1
= 416848.7 -82.5163o
Vm
159984.5
j363688.2
= 397321.2 113.7444o
Ikm
-251.488
j1324.866
= 1348.524 100.7481o
Imk
251.488
+ j1324.866
= 1348.524 79.2519o
Pkm
jQkm
= (266.95555
+ j87.93446) . 106
Pmk
jQmk
= (-261.0362
+ j60.24744) . 106
Plos
jQloss
= sum of above
= (5.919274 + 148.1819) . 106
After the last such branch-flow printout, there will be a total network loss "ploss" and switch flow printout as shown below:
TOTAL NETWORK LOSS "PLOSS" BY SUMMING NODAL INJECTIONS =
0.1376853967E+01
NODE-K
NODE-M
I-REAL
I-IMAG
I-MAGN
DEGREES
THR
FOU
0.27537079E+01
-0.27776735E+01
0.39113140E+01
-45.2482
POWER
-0.55511151E-16
REACTIVE
-0.91810422E+00
After the last such switch-flow printout, there will be an injection printout for all non-ground nodes
which are connected to voltage sources during the steady-state phasor solution. The general
format is similar to that which has just been illustrated for branches.
SOLUTION AT NODES WITH KNOWN VOLTAGE.
NODES SHORTED TOGETHER BY SWITCHES ARE SHOWN AS A GROUP OF NAMES, WITH
THE PRINTED RESULT APPLYING TO THE COMPOSITE GROUP.
THE ENTRY 'MVA' IS SQRT(P**2 + Q**2) IN UNITS OF POWER,
WHILE 'P.F.' IS THE ASSOCIATED POWER FACTOR.
NODE
SOURCE NODE VOLTAGE
INJECTED SOURCE CURRENT
INJECTED SOURCE POWER
NAME
RECTANGULAR
POLAR
RECTANGULAR
POLAR
P AND Q
MVA AND P.F.
ONE
0.1000000E+01
0.0000000E+00
0.1000000E+01
0.0000
0.2753708E+01
-0.2777673E+01
16 - 28
0.3911314E+01
-45.2482
0.1376854E+01
0.1388837E+01
0.1955657E+01
0.7040365E+00

Assuming that we are considering node "k", then the next to the last column gives the injection pk
+ jQk positive if from the source into the network. The final column shows the magnitude of
this complex number, plus the power factor (the cosine of the angle which has tangent Qk / Pk).
One special feature of the injection printout concerns the effect of closed switches. There will
only be one double line of injection printout for each point of known, equal voltage in the
network. For example, suppose that three nodes happen to be shorted together by closed switches
during the steady-state phasor solution, as per the sketch below. Then there would only be one
line of injection printout for these three EMTP nodes. All three node names in question would be
listed continuously in the "NODE NAME" column, however, as indication of the node combination
or shorting which has occurred:
RBG1
RBG
RBG2
The user should be warned that the aforementioned steady-state phasor branch-flow printout
shows an entry for each component in the EMTP branch table, in order. Since the branch table is
not sorted, the printout is in the order of data input to the branch table. Nonetheless, there can be
confusion in some cases, due to the internally-defined nodes and branches. The following
comments document the way data can thus be disguised:
1.
The saturable "TRANSFORMER" of Section 5.4 may contribute a number of

entries to the EMTP branch title. See the list of these elements under Message 48
of Section 17.1. At least in this case all nodes are distinct, and user-defined (so
that there is no ambiguity).
2.
For the Type16 EMTP source component of Section 10, the EMTP internally
defines two branches (both resistors) and two nodes ("TYPE-16 and "......").
There will be two entries in the steady-state branch-flow printout for each Type
16 component, then.
3.
Each Type98 pseudo-nonlinear inductor (Section 8.4) and Type-36 hysteretic

reactor (Section 8) is represented by a linear inductor, for purposes of steadystate solution. The same holds true for each Type93 (true) nonlinear inductor
of Section 8.11.
16 - 29

4.
Flashover branches are all open circuits during the steady-state phasor solution.
They are not in the branch table (List #2) at all, and have no associated steadystate printout. Included in this category are:
(A)
Type91 time-varying resistance of Section 8.6.5;
(B)
Type92 (true) nonlinear resistance of Section 8.8, Section 8.9 and

Section 8.10;
Optional Type59 S.M. Steady-State Solution Printout
16.2.3
The requests for steady-state (initial condition) printout for Type59 S.M. model is handled
separately as part of the S.M. data. See "Class 5 s.m. Data" of Section 10.11. Note that such
printout is separately controlled for each machine, at present. Anyway, should such printout be
requested for one more machines, the injection printout of Section 16.2.2 will be followed by
initial condition printout for each machine, an example of which is shown below:
First comes the tabulation of negative and zero sequences armature currents and a listing of
electrical and mechanical parameters of the machine:
TABULATION OF NEG. AND ZERO SEQU. CURRENTS

NEG. SEQ.
ZER. SEQ.
MACH 1
' MACA1'
UNIT
0.9758467E+03
0.3880511E-10
0.2050407E+01
0.3141593E+01
DATA PARAMETERS AND INITIAL CONDITIONS OF NEXT MACHINE FOLLOW.

1
-----------------------------------------------
MACHINE REACTANCES AND RESISTANCES, IN OHMS (QUANTITIES LABELED AS INDUCTANCES ARE ACTUALLY REACTANCES).
0.1471482E+03
LF
= D-AXIS FIELD SELF INDUCTANCE
0.1485714E+02
LAF = D-AXIS FIELD-ARMATURE MUTUAL INDUCTANCE
0.1362746E+03
LFKD = D-AXIS FIELD-DAMPER MUTUAL INDUCTANCE
0.1797673E+01
LD
= D-AXIS ARMATURE SELF INDUCTANCE (SYNCHRONOUS REACTANCE)
0.1485714E+02
LAKD = D-AXIS ARMATURE-DAMPER MUTUAL INDUCTANCE
0.1484620E+03
LKD = D-AXIS DAMPER SELF INDUCTANCE
0.1800781E+03
LG
= Q-AXIS CIRCUIT 1 SELF INDUCTANCE
0.1425599E+02
LAG = Q-AXIS CIRCUIT 1-ARMATURE MUTUAL INDUCTANCE
0.1307606E+03
LGKQ = Q-AXIS CIRCUIT 1-CIRCUIT 2 MUTUAL INDUCTANCE
0.1732133E+01
LQ
= Q-AXIS ARMATURE SELF INDUCTANCE
0.1425599E+02
LAKQ = Q-AXIS CIRCUIT 2-ARMATURE MUTUAL INDUCTANCE
0.1372782E+03
LKQ = Q-AXIS CIRCUIT 2-SELF INDUCTANCE
0.1123576E+00
L0
= ZERO SEQUENCE REACTANCE
0.1613972E-02
R0
= ZERO SEQUENCE RESISTANCE
0.8561412E-01
RF
= RESISTANCE OF FIELD WINDING
0.1610972E-02
RA
= ARMATURE RESISTANCE
0.1401913E+01
RKD = D-AXIS DAMPER RESISTANCE
0.5612917E+00
RG
= Q-AXIS CIRCUIT-1 RESISTANCE
0.1802885E+01
RKQ = Q-AXIS CIRCUIT-2 RESISTANCE
MECHANICAL PARAMETERS OF GENERATOR, IN PHYSICAL UNITS AS SHOWN BY COLUMN HEADINGS.
MOMENT OF INERTIA
OF ROTOR MASS
MILLION (N-M)/(RAD/SEC**2)
0.3666265E-01
SELF-DAMPING COEFFICIENTS OF MASS

SPEED-DEVIATION
ABSOLUTE-SPEED
MILLION (N-M)/(RAD/SEC)
0.1356306E-09
0.0000000E+00
MUTUAL-DAMPING COEFF.
(WITH FOLLOWING MASS)
MILLION (N-M)/(RAD/SEC)
0.0000000E+00
TORSIONAL SPRING CONSTANT

(WITH FOLLOWING MASS)
MILLION (N-M)/(RAD)
0.0000000E+00
The above printout is followed by a listing of different machine variables calculated by the
initialization procedure:
16 - 30
TOTAL CURRENT INJECTED INTO NETWORK AT GENERATOR BUS, IN PHASE COORDINATES.

FOR A DUAL-MACHINE BUS,
THIS IS THE TOTAL INJECTION ( 'A' + 'B' ).
THE FIRST LINE DISPLAYS THE CURRENTS AS FOUND BY THE PHASOR NETWORK
SOLUTION (WHICH MAY BE UNBALANCED, IF THE NETWORK IS).
THE 2ND LINE SHOWS ONLY THE POSITIVE-SEQUENCE COMPONENT,
MAGNITUDES OF THE CURRENTS ARE IN UNITS OF (AMPS) .
PHASE 'A' INJECTION
PHASE 'B' INJECTION
PHASE 'C' INJECTION
MAGNITUDE
DEGREES
MAGNITUDE
DEGREES
MAGNITUDE
DEGREES
0.9143855E+05
44.8346908
0.9223755E+05
-77.2146592
0.8898150E+05
163.3599268
+ACTUAL
POS. SEQ.
0.9087547E+05
43.6601114
0.9087547E+05
-76.3398886
0.9087547E+05
163.6601114
ARMATURE CURRENTS OF GENERATOR IN ROTATING REFERENCE FRAME (D-Q-0 COORDINATES), IN UNITS OF
ID
IQ
I0
-0.5319401E+05 0.1195111E+05 -0.6721242E-10
(AMPS) .
POSITIVE-SEQUENCE COMPONENT OF GENERATOR ARMATURE CURRENT IN PHASE COORDINATES, IN UNITS OF (AMPS) .

ARMATURE OF PHASE 'A'
ARMATURE OF PHASE 'B'
ARMATURE OF PHASE 'C'
MAGNITUDE
DEGREES
MAGNITUDE
DEGREES
MAGNITUDE
DEGREES
0.4543774E+05
43.6601114
0.4543774E+05
-76.3398886
0.4543774E+05
163.6601114
FIELD CURRENT OF GENERATOR IN UNITS OF (AMPS) .
TOTAL
DC-COMPONENT
0.7706384E+04
0.7761695E+04
ELECTROMECHANICAL TORQUE OF GENERATOR, IN UNITS OF
TOTAL
DC-COMPONENT
0.3626624E+01
0.3730468E+01
MILLION (N - M).
CRITICAL LEVEL OF TOTAL AIR GAP MMF AT WHICH SATURATION BEGINS, IN UNITS OF
0.000000000E+00
0.000000000E+00
MECHANICAL ANGLES OF ROTOR MASSES, IN UNITS OF ( DEGREES)
210.5916874
'THETA' FOR MASS NO. 1
(AMPERES) .
ANGULAR VELOCITIES OF ROTOR MASSES, IN UNITS OF (RAD/SEC) .

376.9911184
'OMEGA' FOR MASS NO. 1
16.2.4
Abbreviated "FREQUENCY SCAN" Printout
In addition to the already-described full branch-flow and injection printout, only node-voltage
output is presently available in the case of "FREQUENCY SCAN" data cases. Any column80
punches for branch or switch variables (voltage differences, branch currents, powers and energies)
will be disregarded by the EMTP. Node voltages are outputted in both polar (magnitude, angle in
degrees) and rectangular (real, imaginary) forms.
For the LUNIT6 line printer output, there really are two component output vectors; first comes the
polar part, and second (beginning at the left on a new line) comes the rectangular part. For the
LUNIT4 plot file, the two are concatenated into one double-size output vector. An example of the
LUNIT6 output is displayed on the following page. This output is extracted from the he final line of
heading (before step1 output) shows that there are four complex output quantities, or 10 numbers
total in both the polar output and the rectangular output.
For purposes of plotting, all output variables are given type code "9" (branch currents). A pair of
names is associated with each number. Suppose that NODNAM is the name of the node which is of
immediate interest. Then plot the "current" with names:
(NODNAM, MAG)
for the magnitude of node voltage NODNAM;
(NODNAM, ANG)
for the associated angle in degrees.

16 - 31

(NODNAM, REAL)
for the real part of node voltage NODNAM;
(NODNAM, IMAG)
for the associated imaginary part.
__The FREQUENCY
SCAN output vector has the
following format :
MAG___________
REAL__________
ANGLE_________
IMAG__________
Variables are treated as though they were BRANCH CURRENTS plot type-code
The suffix MAG, ANGLE, REAL or IMAG follows the
actual node name of the output variable.
_Requested output of node voltages :
FREQ
LOAD
0.6000000E+02
0.1000000E+03
0.0000000E+00
0.1000000E+03
0.0000000E+00
0.8000000E+02
0.1000000E+03
0.0000000E+00
0.1000000E+03
0.0000000E+00
0.1000000E+03
0.1000000E+03
0.0000000E+00
0.1000000E+03
0.0000000E+00
0.1200000E+03
0.1000000E+03
0.0000000E+00
0.1000000E+03
0.0000000E+00
MID
0.9357152E+00
0.8755630E+00
0.8934764E+00
0.7983000E+00
0.8467330E+00
0.7169568E+00
0.7984712E+00
0.6375562E+00
-0.2065600E+02
-0.3300795E+00
-0.2668661E+02
-0.4012694E+00
-0.3214191E+02
-0.4504772E+00
-0.3701564E+02
-0.4807060E+00
TRAN
0.9357152E+00 -0.2065600E+02
0.8755630E+00 -0.3300795E+00
0.8934764E+00 -0.2668661E+02
0.7983000E+00 -0.4012694E+00
0.8467330E+00 -0.3214191E+02
0.7169568E+00 -0.4504772E+00
0.7984712E+00 -0.3701564E+02
0.6375562E+00 -0.4807060E+00
CUR
0.9357152E+02
0.8755630E+02
0.8934764E+02
0.7983000E+02
0.8467330E+02
0.7169568E+02
0.7984712E+02
0.6375562E+02
-0.2065600E+02
-0.3300795E+02
-0.2668661E+02
-0.4012694E+02
-0.3214191E+02
-0.4504772E+02
-0.3701564E+02
-0.4807060E+02
Such plotting is illustrated by BENCHMARK DC-51, where printer plots of both polar and
rectangular quantities are illustrated. For easy reference, the two plot cards of that data case are
reproduced below. All plotting is for steady-state voltage components versus frequency at node
"CUR ":
PRINTER PLOT
19630. 60.150.
CUR
19630. 60.150.
CUR
MAG
REAL
CUR
CUR
ANGLE
IMAG
If the user needs branch currents, or voltage differences, it is of course possible to add one or more
measuring transformers for this purpose (see Section 5.4). A two-winding transformer which
draws no magnetizing current and has small leakage impedance should of course be used, to ensure
accurate measurement (just as in the real world of instrumentation). For a voltage difference, the
primary would be connected across the two desired nodes; the turns ratio would be unity, and the
secondary would be left open circuited, with one terminal grounded. See Figure 16.3.
BUSK
Actual
Network
1:1
VKM
BUSM
Figure 16.2:
Measuring voltage difference

16 - 32

The just-described sampler of voltage difference can be easily extended to measure current. If
nodes "K" and "M" are the terminals of a resistor which carries the desired current, then the
measured voltage will equal the current if one uses a turns ratio R:1 (see Figure 16.4).
K
R:1
IKM
ikm
Figure 16.3:
Current Sampling
16.3 Record Of Transient Results

Results can be printed, saved on file for plotting, or both, either at each time step or at every K-th
time step.
16.3.1
Printed Transients Solution Variables
As controlled by integer miscellaneous data parameter "IOUT" (see Section 4.2.2), selected
variables of the transient solution are outputted to the line printer every IOUT-th time-step point.
This frequency can be altered as the solution progresses, if desired (see Section 4.3).
Characteristics of this time-step loop output include the following:
Variables:
Variables are printed out in a fixed order, time-step by time-step, as the solution is advanced.
Column headings at the start identify all output variables which will be printed.
Time-Step Number:
The time-step number and also the time corresponding to each batch of values begins on the
extreme left of the page, beneath the headings "STEP" and "TIME".
16 - 33
Node Voltages:
Immediately thereafter, on the same line, all node voltages requested for output (see Section 12)
are printed. These are identified by a heading bearing the node names in question. In number,
such columns are noted by the message
"FIRST XX OUTPUT VARIABLES ARE ELECTRIC-NETWORK NODE VOLTAGES (WITH
RESPECT TO LOCAL GROUND);"
which precedes the heading. Here "XX" is an integer. If over 9 node voltage outputs exist, printing
continues on the line or lines immediately following.
Branch Voltage and Power:

Immediately to the right of the last node voltage will come all branch voltages or branch powers
(as requested by column-80 punches on the branch cards in question), identified by the pair of
terminal node names. The two node names are printed one above the other; if K is the top node,
and M is the lower node, then three types of output are possible within this class:
(A)
Voltage difference, vkm = vk - vm.
(B)
For switches, "power output" always represents power flow through the switch,
from K to M : P = Pkm.
(C)
For all nonswitch branches, "power output" always represents the loss, input,
or dissipation. This is P = Pkm + Pmk.
In number, such columns are noted by the message

"NEXT XX OUTPUT VARIABLES ARE BRANCH VOLTAGES (VOLTAGE OF UPPER NODE
MINUS VOLTAGE OF LOWER NODE);"
which appears immediately below the similar message for node voltages. As with node voltages,
continuation on as many lines as necessary is automatically provided.
Branch Current and Energies:

Immediately to the right of the last branch voltage or power will come all branch currents or
energies. The sign on flows is as just mentioned under Branch Voltage and Power with again three
types of output within this class:
(A)
Branch current ikm (from K to M, measured at end K).
16 - 34

(B)
For switches, "energy output" always represents the flow of energy through
the switch, from K to M, since time zero:
E( t) =
(C)
O Pkm ( u )du
(16.1)
For all nonswitch branches, "energy output" always represents the loss, input,
or dissipation. This is
E( t) =
O ( Pkm ( u ) + Pmk ( u ) ) du
(16.2)
In number, such columns are noted by the message

"NEXT XX OUTPUT VARIABLES ARE BRANCH CURRENTS (FOLLOWING FROM THE
UPPER EMTP NODE TO THE LOWER);"
which appears immediately below the similar message for branch voltages and powers. As with
preceding variables, continuation on as many lines as necessary is automatically provided.
Synchronous Machine Variables:

Immediately to the right of the last branch current or energy will come all dynamic synchronous
machine (S.M.) and universal machine (U.M.) output variables. Various types of variables are
possible (angles, currents, torques, etc.), as fully explained in Section 10.11 under Class 5 data and
Section 10.12. A pair of 6character names is used for identification, with the first (upper) one
always identifying the machine of interest (e.g., "MACH 3" stands for the third S.M. component, in
order of input). The second (lower) name identifies the variable type (e.g., "ANG 7" stands for the
angle of the 7th mass of the shaft system). In number, such columns (of S.M. variables) are noted
by the message:
"NEXT XX OUTPUT VARIABLES PERTAIN TO DYNAMIC SYNCHRONOUS MACHINES,
WITH NAMES GENERATED INTERNALLY;"
which appears immediately below the similar message for branch currents. As with preceding
variables, continuation on as many lines as necessary is automatically provided for.
TACS Output Variables:

Immediately to the right of the last S.M. variable will come all TACS output variables. These were
requested using data cards as are described in Section 14.7. A pair of 6character alphanumeric
names is used for identification of each variable, with the first (upper) name always being "TACS".
16 - 35

The second (lower) name is simply the name of the TACS variable in question. In number, such
TACS variable columns are noted by the message:
"FINAL XX OUTPUT VARIABLES BELONG TO 'TACS' (NOTE INTERNALLY-ADDED
UPPER NAME OF PAIR)."
which appears immediately below the similar message for dynamic S.M. variables. As with
preceding variables, continuation on as many lines as necessary is automatically provided for.
16.3.2
Sample Illustrative Output

4
first
2 output variables are electric-network node voltages (with respect to local
ground)
next
2 output variables are branch currents (flowing from the upper node to the
lower)
STEP
0
1
2
3
4
5
10
15
20
40
60
80
100
TIME
GENA
GENB
TERRA
TERRA
GENA
GENB
0.000000 0.377000E+03 0.377000E+03 0.000000E+00 0.000000E+00
0.000050 0.376933E+03 0.376933E+03-0.523565E+00-0.523565E+00
0.000100 0.376732E+03 0.376732E+03-0.104694E+01-0.104694E+01
0.000150 0.376397E+03 0.376397E+03-0.102375E+01-0.102375E+01
0.000200 0.375929E+03 0.375929E+03-0.104552E+01-0.104552E+01
0.000250 0.375327E+03 0.375327E+03-0.106725E+01-0.106725E+01
0.000500 0.370322E+03 0.370322E+03-0.117521E+01-0.117521E+01
0.000750 0.362031E+03 0.362031E+03-0.128124E+01-0.128124E+01
0.001000 0.350526E+03 0.350526E+03-0.138440E+01-0.138440E+01
0.002000 0.274821E+03 0.274821E+03-0.175063E+01-0.175063E+01
0.003000 0.160519E+03 0.160519E+03-0.200558E+01-0.200558E+01
0.004000 0.236720E+02 0.236720E+02-0.347031E+01-0.347031E+01
0.005000-0.116499E+03-0.116499E+03-0.276576E+01-0.276576E+01
16 - 36
Section 17
17.1
Informative (Nonfatal) Messages ........................................................ 17-1
17.2
Connection Discontinuities During Solution ...................................... 17-29
17.3
Printout Of Storage And Running-time Statistics ............................... 17-30
17.4
Printer And Plotter Output of Sample EMTP Test Problem ............... 17-31
17.5
EMTP Error-Message Terminations ................................................... 17-39
Section 17
The printed EMTP output may include various messages which:
(A)
Remind the user of special situations which occurred during the solution
(nonfatal, informative);
(B)
Signal to the user data errors or misunderstandings which resulted in the

termination of a data case (fatal error messages).
17.1 Informative (Nonfatal) Messages

Situations that lead to warning or informative messages can reflect improper model usage, and
assumptions. Sometimes these messages are reminder of EMTP-specific limitations. In any event
it is always advisable to examine the output file for such messages, even after a successful
simulation.
The list presented below is not complete (program development often precedes documentation).
However, EMTP and warning messages are intended to be self-explanatory, and such omissions
should not be a problem.
Message 1:
----- WARNING.
NONZERO MISC. DATA, PARAMETER "XOPT" DIFFERS FROM THE POWER
FREQUENCY OF "XREF" .
THIS IS UNUSUAL. A VALUE OF "XOPT" WAS READ FROM
COLUMNS 17-24 OF THE DATA CARD JUST READ.
EXECUTION WILL CONTINUE USING THIS
VALUE, AS SUSPICIOUS AS IT SEEMS TO THE EMTP.
Because any given country generally has only one synchronous power frequency (60 Hz for the
USA and Canada), nonzero miscellaneous data parameter "XOPT" will typically equal this power
frequency. Message number 1 simply reminds the user of any other nonzero usage of floatingpoint miscellaneous data parameter "XOPT". The reference value "XREF" is assigned within
system-dependent module "SYSDEP" of overlay 1, and hence this value may vary. In particular,
European users will generally want to set this reference value to 50.0 (for 50 Hz).
17 - 1
Message 2:
----- WARNING.
NONZERO MISC. DATA,PARAMETER "COPT" DIFFERS FROM THE,POWER
FREQUENCY OF, "REF". THIS IS UNUSUAL. A VALUE OF, "COPT" WAS READ FROM COLUMNS
25-32 OF THE DATA CARD JUST READ.
EXECUTION WILL CONTINUE USING THIS VALUE, AS
SUSPICIOUS AS IT SEEMS TO THE EMTP
This explanation is identical to that for Message 1, only here the EMTP is concerned with
miscellaneous data parameter "COPT" rather than "XOPT".
Message 3:
BUS1 (A6), BUS2 (A6), R (E6.1)
HIGH RES. ADDED BY EMTP ACROSS TYPE-99 ELEM. "BUS1", "BUS2", "R"
If the user has not paralleled each Type-99 (Type-98) pseudo-nonlinear resistance (inductance)
element by a lumped linear element, then the EMTP will internally connect very high impedance
resistors across such elements. Such a branch should not significantly affect the solution. The user
is informed of such an automatic addition by means of Message 3 printout, where the 80-column
card image listing and interpretation are provided almost as though the user had inputted the
branch himself. The card-image listing will show the terminal node names "BUS1" and "BUS2" of
the Type-99 nonlinear element (printed in columns 314); the resistance value "R" will appear in
columns 2732.
TYPE
99
Figure 17.1: Addition of a Parallel Resistor

The internally-defined resistor has an ohmic value of .01/FLZERO, where "FLZERO" is the EMTP
variable of /BLANK/ which gives a measure of the floating-point precision limit of the computer
being used.
It should be mentioned that the reference branch procedure is used for these internally-defined
high-resistance elements. Only the first one will show "R" as being punched in columns 27-32.
All later branches will have columns 1526 referencing the first such branch. In this way, a small
amount of List 3 storage space is saved in the EMTP.
17 - 2
Message 4:
REMEMBER ---- WHAT ARE LABELED AS THE INITIAL "N1" BRANCH-OUTPUT CURRENTS ARE IN REALITY
MODAL VOLTAGES AT THE TWO ENDS OF THE LAST DISTRIBUTED-PARAMETER LINE
OF THE DATA CASE BEING SOLVED. THE FIRST "MODOUT" MODE VOLTAGES AT THE
BUS1'END ALL COME FIRST, FOLLOWED BY ALL THE CORRESPONDING ENTRIES FOR THE
'BUS2' END OF THE LINE.
This message is simply a reminder that the EMTP output vector may not be quite what labelling
would imply, if the user has requested the special "MODE VOLTAGE OUTPUT" option described in
Section 4.1.2. Recall the Karrenbauer mode voltages at both ends of one transmission line are
available for output purposes, by means of this option. The user inputs variable "MODOUT" (I8
information read from columns 3340), and "N1" is always twice this value. Message 4 will
always be printed for a data case which uses the "MODE VOLTAGE OUTPUT" option, immediately
before the output-variable headings for the time-step loop.
Message 5:
NOTE ---- THE PRECEDING PRINTOUT SHOWS THAT GROUND WAS NOT RENUMBERED FIRST IN
THE TRANSIENT-NETWORK RENUMBERING. WE WILL SWAP THE NEW NUMBERS BETWEEN
GROUND AND WHATEVER NODE WAS RENUMBERED FIRST, IN ORDER TO GET A LEGITIMATE
NUMBERING FOR THE REST OF THE EMTP CALCULATIONS.
Immediately preceding the three lines of diagnostic text of Message 5 will be found printout of
(NORDER(I), I=1, NTOT), which is the renumbering map from original numbers (assigned
according to order of input) to transient node numbers (internal node numbering, as used for the
network solution within the time-step loop). Ground is always given number one, and must be
preserved this way by renumbering. Yet if the renumbering routine itself (module "OVER7") does
not do this, the following EMTP code must manually make the adjustment as per the printed
message. The user should not be bothered by this message, which really is just a public notice of
some internal adjustments which are being made. This diagnostic seems to generally occur only
for very small test problems.
Message 6:
NONLINEAR AND TIME-VARYING RESISTANCES IGNORED IN STEADY STATE SOLUTION.
NONLINEAR INDUCTANCES INCLUDED WITH LINEAR PART
If nonlinear, time-varying, or pseudo-nonlinear elements are present in the user's data case, a
reminder is issued indicating that the phasor (sinusoidal steady-state) solution for initial conditions
requires a linear representation.
Specifically, resistance elements are all assumed to be open circuits (i.e., they are "ignored"),
while inductance elements are represented by linear inductances.
17 - 3
Message 7:
COMMENT ---- NO SINUSOIDAL SOURCES REQUESTED FOR STEADY-STATE SOLUTION.
THIS SOLUTION IS BYPASSED.
THUS
The typical data case will generally involve one or more Type14 source elements which are
present as part of the phasor (sinusoidal steady-state) solution for initial conditions. But if no such
sources are present, there is no need to perform a phasor solution, since it is known to be
identically zero. It is of this latter situation that the user is reminded by Message 7.
Message 8:
NOTICE. ---- TWO OR MORE SINUSOIDAL VOLTAGE SOURCES ARE PRESENT ON NODE DURING
THE STEADY-STATE SOLUTION.
AS PER THE USER'S MANUAL, THESE VOLTAGES WILL ALL
BE ADDED TOGETHER TO GET A TOTAL NODE VALUE.
As per Section 3.3, multiple voltage sources on a single node are all interpreted as being connected
in series. Message 8 is simply a reminder to the user of all such situations, for Type14 sources
which are present during the phasor solution for initial conditions.
Message 9:
NOTE. ---- NODE "BUS k" HAS BOTH VOLTAGE AND CURRENT SOURCES ON IT.
THE
CURRENT SOURCE HAS NO EFFECT ON THE SOLUTION, THEN, AND COULD BE OMITTED.
If a current source and a voltage source are paralleled, then the current source has no effect on the
network solution. (The A6 name is read from columns 38 of the source card). At least for the
phasor (sinusoidal steady-state) solution for initial conditions, to which Message 9 applies, such
extraneousness is pointed out to the user.
Message 10:
CAUTION. ---- DURING Y-MATRIX ELIMINATION FOR STEADY-STATE SOLUTION VOLTAGES, A
NEAR-ZERO DIAGONAL ELEMENT FOR NODE "BUS K" EXISTS JUST BEFORE RECIPROCATION.
USING MAGNITUDES SQUARED FOR ALL 3 QUANTITIES, WE HAVE ORIGINAL DIAGONAL VALUE
= 0.213E+03 QUESTIONABLE VALUE = 0.49E-31 TOLERANCE RATIO = 0.10E-15 THE NODE IN
QUESTION MAY BE CONNECTED TO OTHER NODES, FORMING A SUBNETWORK.
BUT THE
SUBNETWORK HAS NO (OR VERY WEAK) PATH TO GROUND OR OTHER KNOWN-VOLTAGE NODE IN
THE STEADY-STATE.
SOLUTION VOLTAGES OF THIS SUBNETWORK WILL ALL BE SET TO
ZERO.
This warning message is printed out when the elimination process on the phasor (sinusoidal
steady-state) network equation set [Y] V = I has been temporarily suspended due to the occurrence
of a near-zero diagonal element just before that element is to be reciprocated. Of course all
computers have limited precision for their floating-point representation, so it is impossible to say
whether in theory an exact zero (as opposed to just a very small) diagonal element has been
encountered.
17 - 4

If the phasor network consists of two disconnected subnetworks, the associated complex nodal
admittance matrix [Y] is block diagonal, as shown in Figure 17.2 below. Ground is the reference
node for the equations, which for purposes of this explanation, may be assumed to be in the
subsystem B without any loss of generality. Then subsystem A has no path to ground, nor to any
non-ground node of subsystem B. Submatrix [Yaa] is then singular, which will show up during
triangularization as the elimination on the last row of [Yaa] is being completed; after elimination to
the left of the diagonal, a zero diagonal element will be observed -- identically zero in theory, a
"near-zero" floating-point number in practice.
Yaa
Ybb
Subsystem
Subsystem
ground
Figure 17.2:
System Consisting of Two Disconnected Subnetworks
The tolerance which is used for near-zero checking of diagonal elements just before reciprocation
is floating-point miscellaneous data parameter "TOLMAT". While the user can define (input)
whatever positive value he wants for this variable, it is common to use the EMTP default value.
The fourth printout variable of Message 10 is D3 = TOLMAT**2.
The "original diagonal value" which is mentioned by the message text is the value of the diagonal
element of the matrix for the row which is causing trouble, before any elimination has begun. The
"questionable value" which is mentioned by the message text is the value of this same diagonal
element just before the reciprocation which is not being allowed. Note that "magnitudes squared"
are used for all quantities. This is because the matrix [Y] is complex (has both real and imaginary
parts), and yet all the user wants is a single real number which gives an indication of how small the
diagonal element in question is. The magnitude or modulus of the numbers would be the natural
choice, though the squares have been used for convenience.
If the subnetwork in question really is disconnected from the rest of the system and ground, then
the stated EMTP identification and associated corrective action (setting subnetwork voltages to
zero) was of course correct. In this case, the user should be happy.
17 - 5

On the other hand, if ties do exist from subsystem A to subsystem B (see Figure 17.3 below), then
these ties are just too weak, and the EMTP has been fooled. In this case, either the network
parameter values or "TOLMAT", or both, should be modified, since the solution with zero voltages
for subsystem A nodes will generally be grossly incorrect. The resulting transient simulation may
well have a substantial extraneous discontinuity or shock at time zero, because of the erroneous
initial-condition voltages. For this reason, the user should always pay attention to Message 10, and
verify that it only appears for cases where the subsequent zero-voltage solution is acceptable and
correct.
weak
A
ground
Figure 17.3:
System with Weak Tie to Ground
One special problem concerning Message 10 has to do with phasor solutions at frequencies that
differ drastically from the synchronous power frequency. For example, DC initial conditions
representing trapped charge on an unenergized line may commonly be generated by a phasor
solution using frequency 0.001 Hz. But in such a case, the impedance of small "isolation"
reactances becomes extremely small while the impedance of capacitors becomes very large. What
might well have been good strong coupling at 60 Hz can become very weak coupling at such a
drastically altered frequency. For such usage, small isolation resistors may then be preferable to
isolation reactors, since the impedance of a resistor is frequency invariant.
Message 11:
REQUESTED OUTPUT FOR NONEXISTENT NODE "BUS K" WILL BE IGNORED
In case program execution is to be stopped after the steady-state phasor solution is complete
(which is the case if floating-point miscellaneous data parameter "TMAX" is punched nonpositive), the steady-state phasor solution for transients variables is always provided. This is in
addition to, and subsequent to, the possible printout of the complete steady-state solution which is
controlled by integer miscellaneous data parameter "KSSOUT". Selective node voltage output
requests are read as per the description of Section 12. Should one or more of these A6 node names
be illegal (i.e., non-existent in the user's previously-defined network), then the text of Message 11
17 - 6

will result. This is usually the result of a spelling error, which includes the proper positioning of
blank characters in the field of width six.
Message 12:
INITIAL FLUX IN COIL "NODE K" TO "NODE M" = 0.451E+02
WARNING. ASSUMPTION THAT AC STEADY STATE HAS FUNDAMENTAL FREQUENCY ONLY IS
QUESTIONABLE WITH PRECEDING FLUX OUTSIDE LINEAR REGION)
The initial-flux message (one of the first two lines of printout) will be outputted by the EMTP for
every inductance element of the data case which is other than perfectly linear. Included, then, are:
1.
Type98 pseudo-nonlinear reactors of Section 8.4. Recall that the saturable

transformer component of Section 5.4 has a built-in Type98 element for the
magnetizing branch, provided the characteristic is truly nonlinear (two or more
points used to define it).
2.
Type93 true nonlinear inductance element of Section 8.13.
3.
Type-96 hysteretic reactor of Section 8.5.
The last of the two printed lines is only outputted if the flux in question is "out of bounds". It is
important to realize what such a message means, and also what it does not mean.
First, what Message 12 does not mean. Absence of the message does not mean that sinusoidal
steady-state peak phasor flux is within the bound which is being checked. It is only the flux value
at time zero - which is to be used as an initial condition for the subsequent transient simulation that is checked. This is the projection of the complex phasor y on the real axis. It is possible for
|y| to be way out of bounds, and yet y(0) might be small or even zero. Absence of such a warning
message thus does not imply that the true steady-state, periodic solution is sinusoidal; it may in
fact be highly nonsinusoidal.
Imaginary
( t ) = Re { e
jt
Real
(0)
Figure 17.4:
Definition of Initial Flux

17 - 7

Second, there is the limit value which (0) is checked against. For the true nonlinear and the
pseudo-nonlinear elements, the assumed limit is ysteady as read from columns 3338 of the first
card defining the component in question.
Finally, there is the potential numerical complication which may result from initial flux being
"outside the linear region".
1.
The Type93 true nonlinear inductance element, on the first time step operation
will transfer to the nonlinear characteristic (see Figure 17.5 below). This will
generally be a nearly horizontal jump (constant flux trajectory), which applies a
sudden change i in the current. The result may well be an extraneous shock to
the system, and a spurious transient. For example, if the current jump i is
supplied through a linear inductor L in the time-step of size t, a spurious voltage
change on the order of V=Li/t will appear across that element. For small time
step t, this may be astronomical.
steady-state
operating point
Jump
(0)
operating point
1st time step
i(0)
i1
Note: The common choice of having slope

steady / isteady equal to the slope of
segment number 1 is assumed here.
Figure 17.5:
Initial Flux Outside the Linear Region
As a remedy to the just-started difficulty, the user is free to pick isteady and steady so
that the steady-state line passes through (or almost through) the transient operating
point. This is shown in the sketch below. If done carefully, the shock at time zero can
be effectively eliminated, generally.
17 - 8
Slope of line = steady-state

linear inductance, chosen to
approximately equal
(0) / i1
(0)
steady-state
operating point
i
i0
Figure 17.6:
ti m
e-s
te p
th r
tim
e-s
tep
ze
tim
ro
e- s
t ep
on
e
tim
e-s
tep
tw
o
(0)
ee
For the Type98 pseudo-nonlinear inductance element, the preceding general

idea of a current jump i at time zero is applicable. However, present Type98
element logic allows only one segment change per time step, with operation
always beginning on the first segment, at time zero. Hence, on the first timestep operation might switch to segment number 2, on the second step to segment
number 3, etc., as shown below. It thus may require more than one time-step for
operation to move to the nonlinear characteristic.
2.
Approximately Correct Steady-State Initialization
i
i1
Figure 17.7:
Problem of Segment-Change Logic of Type-98 Element

17 - 9

As a general remedy for the preceding complication, the user can often advantageously rotate
source angles, for a balanced 3phase system. This does not affect the problem physically, of
course, since one is thereby simply redefining the time instant which one chooses to call zero. The
optimal situation is to have fluxes (currents) situated at 30o, 150o and 270o in the complex plane
3
--- times the peak value.
2
This gains 13.3% compared with the worst choice, which is frequently enough to eliminate
Message 12 texts.
(phasor diagram), so that all three initial values are limited to .866 =
a
30
a ( 0 ) =
3
---
2
b
Figure 17.8:
Optimal Rotation of Initial Fluxes
Another remedy for Message 12 problem can be obtained via the special request word
"INITIALIZATION WITH HARMONICS" (see Section 4.1.1). The user eliminates the problem
entirely by requesting a solution with harmonic distortion due to saturation effects.
Message 13:
5 OUT OF 50 SWITCH CLOSING TIMES FALLING BEYOND 4.000 TIMES THE STANDARD
DEVIATION
This message can appear only during execution of the special test procedure for the EMTP randomnumber generator. Recall that if integer miscellaneous data parameter "NENERG" (see
Section 4.2.2) is positive, and if "ISW" of the following "STATISTICS" miscellaneous data card has
value 4444, then such a special test will occur. Switch closing times for the EMTP data case are
generated, but no subsequent simulations are actually run. Instead, such switch closing times are
simply tabulated for user scrutiny. The user specifies "SIGMAX" as the number of standard
deviations on each side of the mean over which this tabulation is to be produced. If this is 3.0 or
4.0, most likely none of sample switch closing time will fall outside the range of the tabulation, and
no such Message 13 text will be seen. But should N1>0 sample switch closing times (out of the
"NENERG" which were requested by the user) fall outside the tabulated range, this Message 13 text
will be outputted immediately after the closing-time tabulation.
17 - 10
Message 14:
TROUBLE AT 73912 ON TYPE-99 OR 98 ELEMENT NUMBER 1 -0.2E+01
0.320E+01
0.500E+03
Message 14 is associated with pseudo-nonlinear element trouble. Specifically, operation was on

some segment other than that passing through the origin, when a sign change of the ordinate (v or
) has been found. See Figure 17.9 below, where operation on segment number 2 has been
assumed for purposes of illustration. While operating on the infinite extension to segment number
2, a solution with negative ordinate value has been calculated. Clearly we are not moving slowly
up and down the characteristic, then, which was one of the underlying assumptions behind
successful pseudo-nonlinear element usage. Once operation ceases to track the characteristic, the
simulation becomes invalid, of course, from that point on.
Infinite extension of
segment 2 is actual
assumed by EMTP
modelling constraint
at the time instant
of interest.
Calculated operating point

which represents a sign
change on V.
Figure 17.9: Incorrect Zero-Crossing

It is important for the user to realize that the Message 14 text is sufficient to show erroneous
pseudo-nonlinear element usage, but that its appearance is not necessary for such erroneous
operation. Quite erratic, jerky, unacceptable tracking of the characteristic will go without any
warning message, should such occur without a sign change of the ordinate variable (v or ).
Type98 elements should always be well behaved, with the possible exception of the
discontinuity at time zero. This is because the ordinate variable y is the integral of another
network solution variable, = v dt Changes in from one time-step to another can be made
arbitrarily small, then, by decreasing the time-step size "DELTAT" (floating-point miscellaneous
data parameter). Appearance of Message 14 for a Type98 element is usually an indication that
17 - 11

the simulation is diverging numerically (i.e., blowing up). In this case, voltage magnitudes
become arbitrarily large, and flux changes over one time-step are no longer small.
It is Type99 elements which can "get into trouble", if not connected across a network element
which maintains continuity of voltage either directly or indirectly. Connection across a capacitor
should make them perfectly well behaved, since capacitor voltage is inherently continuous. At the
other extreme, connection across an inductor can frequently cause trouble, since such voltage is
capable of drastically changing from one time step to another
Message 15:
REQUESTED OUTPUT FOR NONEXISTENT NODE "BUS K" WILL BE IGNORED
Identical in meaning to Message 11. Only this particular output is for cases which do involve
transients simulation; here the end-time of the study, floating-point miscellaneous data parameter
"TMAX", is positive.
Message 16:
**** WARNING **** NO CURRENTS WILL BE PUNCHED FOR THE DISTRIBUTED LINE
CONNECTING NODES "BUS K" AND "BUS M"
Integer miscellaneous data parameter "IPUN" provides for the option of punching terminal
conditions of the study, if so desired. But this only works for data cases which have no distributed
lines. If the user tries to exercise this punch option for a data case having distributed parameter
lines, the punching of the distributed line data will be skipped and Message 16 printed.
Message 17:
************************************* LOOK, LOOK
***********************************
DURING THIS RUN, A TOTAL OF 10 RANDOM SWITCH CLOSINGS LESS THAN TIME ZERO WERE
GENERATED BY THE RANDOM NUMBER GENERATOR.
BUT THE EMTP HAS NO WAY OF HANDLING
SUCH CLOSURES.
ALL SUCH ILLEGAL CLOSING TIMES WERE CONVERTED TO TIME ZERO
(THEY SHOULD SHOW UP IN THE PRINTOUT THAT WAY) FOR SIMULATION PURPOSES.
THE
IMPLICATIONS OF THIS MODIFICATION SHOULD BE UNDERSTOOD BY THE USER.
IF IN ANY
DOUBT, THE USER IS STRONGLY ADVISED TO SEEK EXPERIENCED COUNSEL ON THIS SUBJECT.
************************************* LOOK, LOOK
***********************************
When using a Gaussian distribution for random switch closing times of a statistical overvoltage
study, there is always a finite probability of generating negative closing times, regardless of the
prescribed mean closing time T and standard deviation . This is shown in the sketch of the
density function below, where the shaded area under the left tail gives the probability of a negative,
illegal switch closing time. This is for each switch, for each energization (each time we roll the
dice). Since the simulation begins at time zero, the EMTP has no way of representing such negative
17 - 12
f(t)
T SIGMAX
f(t)
+ SIGMAX
closing times, and this is why they are called "illegal". The best EMTP can do is close as soon as
possible, namely at time zero.
2
0
t
T
Figure 17.10: "Illegal" switch closing times

The text of Message 17 is simply informative, to remind the user that the internally-generated
random numbers have been distorted so as to fit into the underlying EMTP framework. If a
significant fraction of the closing times show up as having been converted to zero, the user is
advised to appropriately increase the associated mean closing time T thereby eliminating the
problem. Recall that time is relative anyway, with the instant of time which is called zero an
arbitrary choice of the user.
Message 18:
THE SPECIFIED GRAPH SEPARATION OF 1 INCHES IS CONSIDERED TOO SMALL AND IS RESET
TO 3 INCHES
This message pertains to the old Calcomp plots specification and it is probably obsolete. It could
be triggered by an invalid usage of Postscript options described in Section 15.1
Message 19:
**** WARNING **** REQUESTED SMOOTHING TOLERANCE OF 0.500e+00 MAY CAUSE
INACCURATE PLOT
Section 15.1.3 shows how the user can redefine the smoothing tolerance which is used for
Postscript batch mode plotting, by means of a special request card which bears the key word
"SMOOTH" in columns 2530. While any positive value for the new tolerance "EPS" (read from
columns 16-20) is permitted, large values will lead to rather jagged looking plots. The text of
Message 19 is outputted to warn the user, should he use a tolerance of 0.1 inches or more.
17 - 13
Message 20:
NOTE ---- THE REQUESTED BOTTOM MARGIN OF 0.500E+01 AND GRAPH HEIGHT OF 0.200E+02
REQUIRES A TOTAL PAPER HEIGHT OF 0.20E+02. THIS IS GREATER THAN THE HEIGHT
SPECIFIED IN "CALL PAPRSZ", WHICH IS 0.120E+02 THE REQUESTED VALUES WILL BE
IGNORED AND THE LAST SPECIFIED (OR DEFAULT IF NO HEIGHT VALUES WERE EVER GIVEN)
WILL BE USED.)
This message is self explanatory. Refer to Section 15.1.3 for more details.
Message 21:
NOTE ----- A TIME SCALE OF 0.20E+01 WAS READ FROM COLUMNS 5-7 OF THE PRECEDING
PLOT-REQUEST CARD TOGETHER WITH THE REQUESTED MINIMUM OF 0.000 UNITS (READ FROM
COLUMNS 8-11) AND THE REQUESTED MAXIMUM OF 0.200E+03 UNITS (READ FROM COLUMNS
12-15), THIS IMPLIES A PLOT OF LENGTH ,0.100E+03 WHICH EXCEEDS THE CURRENTLY
IMPOSED FLAT-BED LIMIT OF 0.720E+02 INCHES. THE SCALE ON THE TIME AXIS WILL BE
CHANGED BY THE EMTP SO THAT THE SPECIFIED RANGE COVERS EXACTLY 12 INCHES OF
PAPER.
IF THE USER HAS QUESTIONS, CALL PROGRAM MAINTENANCE.
This message pertains to the old Calcomp plots specification and it is probably obsolete. It could
be triggered by an invalid usage of Postscript options described in Section 15.1.
Message 22:
ACTUALLY, THE JUST-QUOTED PAPER HEIGHT IS THE PAPER HEIGHT MINUS THE OFFSET
HEIGHT OF 0.20E+01 FOR THE ONE OR MORE GRAPHS WHICH HAVE ALREADY BEEN DRAWN
VERTICALLY BELOW THE UPCOMING PLOT.
THE USER SHOULD NOT INCREASE VERTICAL
DIMENSIONS UNLESS HE IS AT THE BOTTOM OF THE PAPER, OR UNLESS HE IS SURE THAT HE
HAS SUFFICIENT VERTICAL SPACE LEFT ON THE PAPER FOR AT LEAST ONE MORE PLOT.
THE SIZE ADJUSTMENT REMAINS CANCELLED.
This message will only appear immediately after Message 20, as a qualification.
Message 23:
THE NUMBER SPECIFYING THE UNITS ON THE HORIZONTAL SCALE, PUNCHED IN COLUMN 4 OF
THE PLOT REQUEST CARD, MUST BE BETWEEN 1 AND 7 (INCLUSIVE). THE NUMBER READ
FROM THE LAST SUCH CARD WAS THIS PLOT REQUEST IS CANCELLED.
This message is self-explanatory. Refer to Section 15.1 for more details.
Message 24:
BUS NAME "BUS K" OF THE USERS LAST-READ PLOT CARD IS NOT THE NAME OF A BUS
HAVING NODE VOLTAGE OUTPUT.
HENCE THIS FIELD WILL BE IGNORED BY THE EMTP
(TREATED AS IF IT HAD BEEN BLANK THE USER IS REMINDED THAT CORRECT SPELLING AND
THE CONSISTENT POSITIONING OF ALL BLANKS WITHIN THE DATA FIELDS OF WIDTH 6 FOR
ALL BUS NAMES IS
17 - 14

Message 25:
THE USERS LAST-READ PLOT CARD REQUESTS A PLOT FOR A BRANCH-VARIABLE WHICH IS
IDENTIFIED BY TERMINAL NAMES = "NODE K" AND = "NODE M".
BUT THE EMTP CANNOT
FIND THIS REQUESTED VARIABLE IN THE LIST OF OUTPUT VARIABLES, SO THIS
PARTICULAR PLOT REQUEST MUST BE IGNORE
ALSO, THE USER SHOULD BE REMINDED THAT BRANCH-OUTPUT REQUESTS ARE MADE USING
COLUMN-80 PUNCHES ON THE BRANCH CARDS IN QUESTION.
THE USER SHOULD DOUBLE-CHECK THAT HE REALLY HAS REQUESTED THE OUTPUT VARIABLE
WHICH HE IS TRYING TO PLOT (AND WHICH GOT HIM IN TROUBLE).
ON COMMON ERROR IS
TO REQUEST ONLY BRANCH-CURRENT OUTPUT.
A 1-PUNCH IN COLUMN 80) AND THEN TRY
TO PLOT BRANCH VOLTAGE ---- OR VICE VERSA.
FINALLY, THE USER SHOULD CHECK
THAT BRANCH OUTPUT IS EVEN POSSIBLE FOR THE COMPONENT IN QUESTION, SINCE
COLUMN-80 PUNCHES MAY BE IGNORED IF THE COMPONENT IN QUESTION DOES NOT PROVIDE
FOR SUCH OUTPUT.
ANY BRANCH-OUTPUT REQUEST FOR A MULTI-PHASE DISTRIBUTED LINE
FALLS INTO THIS CLASS, IT WILL BE NOTED.
Message 26:
PLOT CARD ERROR.
SKIPPED
TIME-AXIS SPECIFICATION IS ILLEGAL. THE PLOT REQUEST IS BEING

XSCALE
XMIN
XMAX
The three variables which are printed out as part of the message are read from columns 57, 811,
and 1215 of the plot card in question. See Section 15.1.3 for further explanation. One of the
three illegal conditions listed below has been noted:
A)
XSCALE < 0.0
The scale of the time axis of the plot can not be

negative or zero.
B)
XMAX < 0.0
The ending time "HMAX" cannot be negative or

zero.
XMAX < 0.0
It is impossible for the plot to end (at end-time

"HMAX") before it begins (at beginning-time
"HMIN").
C)
Message 27:
NO VALID PLOT-VARIABLE NAME OR NAMES WAS PUNCHED ON THE USER=S LAST-READ PLOT
CARD.
HENCE THE EMTP WILL IGNORE THIS PLOT CARD COMPLETELY, AND GO ON TO READ
THE NEXT ONE (FUN AND GAMES CONTINUE)
17 - 15
Message 28:
*** THE PLOTTING ARRAY EV HAS BEEN DIMENSIONED SUCH THAT THE WORKING SIZE IS ,
I10, 20H. THIS IS LESS THAN 2500 WHICH IS THE SIZE REQUIRED FOR THIS PLOT.
THIS REQUEST IS CANCELLED.
Plotting overlay 31 is variably dimensioned automatically, whenever the user redimensions the
EMTP as per Section 2.5. But because the computer code (machine instructions) for plotting is
longer than that of the longest solution overlay, not all of /LABEL/ is available for dimensioned
working space within plotting. It is module "VDOV31", generated by the variable-dimensioning
program, which reserves the working space for plotting - a block of memory equal to the length of
/LABEL/ minus a constant offset (which approximately compensates for the extra length of the
code of this overlay). If this resulting space (never allowed to be less than unity by the variabledimensioning program) is insufficient for plotting purposes, the plot request is simply cancelled, as
per the text of Message 28.
The solution to the dilemma is to simply redimension the EMTP as per Section 2.5. Since it is only
the total length of /LABEL/ which matters, it is immaterial which of the independent lists are used
to provide for the increased storage.
It should be emphasized that the number of points to be plotted on any one graph is not directly
related to the appearance of Message 28. The EMTP is capable of plotting an infinite number of
points on any one graph, since the working space is used only as a buffer, to be dumped onto
logical unit number 9 if it should ever become full. Message 28 indicates a more fundamental
problem: the EMTP does not even have enough working space to set up the just-mentioned buffer.
The program version which generated Message 28 must really be dimensioned way down, to run in
a very small memory partition.
Message 29:
NO PLOT POINTS FOUND BETWEEN THE REQUESTED TIME LIMIT IS 0.8000E+01
0.1000E+02.
THIS PLOT REQUEST CANCELLED.
TO
Self-explanatory. The EMTP simulation begins at time zero, and continues through time "TMAX"
(floating-point miscellaneous data parameter). But the user's last-read plot-request card wanted a
plot to begin and end at the time which are incorporated into the message text; these times (as read
from columns 811 and 1215, respectively) define an interval which does not include any times
for which the solution has been found.
Message 30:
WHILE SCANNING THE DATA POINTS FOR CURVES OF THE LAST-READ PLOT CARD, A NEED FOR
SMOOTHING OF CURVE NUMBER 4 BEYOND TIME 0.4500E-3 SECONDS HAS BEEN DETERMINED.
AT THIS POINT, 50 SUCCESSIVE, UNINTERRUPTED ALTERNATING RELATIVE MAXIMA AND
RELATIVE MINIMA HAVE BEEN OBSERVED.
THIS IS TAKEN AS A SIGN OF A SPURIOUS
MATHEMATICAL OSCILLATION, SOMETHING WHICH SHOULD NOT EXIST PHYSICALLY (AT LEAST
NOT FOR AN INTELLIGENT USER WHO HAS PICKED THE TIME-STEP SIZE "DELTAT" AND THE
OUTPUT FREQUENCY "IPLOT" PROPERLY).
AT THIS POINT, THE OMNISCIENT AND
OMNIPOTENT EMTP (OTHERWISE AFFECTIONATELY REFERRED TO AS "BIG BROTHER" BY THOSE
17 - 16

USERS WHO ARE ACCUSTOMED TO
CURVE FOR ALL LATER TIME.
SUCCESSIVE ORDINATES IN THE
BEFORE PLOTTING BEYOND THIS
HIS MODUS OPERANDI) HAS DECIDED TO SMOOTH THIS

THIS SMOOTHING INVOLVES SIMPLY THE AVERAGING OF
OUTPUT FILE OF PLOT-VARIABLE POINTS FOR THIS CURVE
POINT IN TIME.
This message signals the commencement of the averaging of successive plot points (for purposes
of plotting only), for one of the curves of the last-read plot request card. The limiting number of
oscillations (before such averaging is instituted) can be chosen at will by the user, as per
Section 4.1.2.
Message 31:
****THE NUMBER OF LINES PRINTED FOR THIS PLOT REQUEST NOW EXCEEDS THE TOTAL
NUMBER OF DATA POINTS
****SINCE POSTSCRIPT PLOT WAS ALSO SPECIFIED, ONLY PRINTER PLOT IS CANCELLED
FOR THIS REQUEST
Message 32:
****THE NUMBER OF LINES PRINTED FOR THIS PLOT REQUEST NOW EXCEEDS THE TOTAL
NUMBER OF DATA POINTS
****SINCE PRINTER PLOT WAS ALSO SPECIFIED, ONLY PRINTER PLOT IS CANCELLED FOR
THIS REQUEST
The user is not allowed to stretch the time axis of a printer plot out so as to cover more lines on the
paper than there are data points in the curve being plotted. If such an illegally-long printer-plot
request is made, the printer plot will be truncated at the aforementioned limit, as per the text of
Message 31 or Message 32.
As for the rationale behind this line limit on printer plots, it just seemed like a good way of at least
bounding the amount of wasted paper, in case the user makes a gross scaling error on his time
axis. Some computers or compilers may not allow a line limit on the printed output, so hundreds
or even thousands of pages of extraneous output could conceivably be generated if one were not
careful. Yet, by bounding the number of printed lines by the number of time steps of the
simulation, at least a meaningful amount of computation goes along with each line of output; an
astronomical number of output lines (e.g., 100,000) then is not possible, because the user could
never afford to run such a simulation that long to generate the raw data points! Further, such a
limit should in no way interfere with production plotting usage, for it makes no sense for the user
to plot fewer than one point of a curve per printer line. Maximum time-axis resolution is reached
with one output plot point per printer line; it would only look worse, and confusing, if fewer than
one point of the curve were outputted for each printer line.
Message 33:
***WARNING TSSOL*** LIMITS OF BLOCK "BUS 1" HAVE CRISS-CROSSED. EXPECT PUZZLING
RESULTS. THIS WARNING MESSAGE WILL NOT BE REPEATED
17 - 17

This message will be printed for TACS function block "BUS1" if there is trouble with the limits
which have been specified by the user. See Section 14.3 for rules about limits. In the case of two
variable limits as in Figure 17.11 ("LOW" and "HIGH" are themselves TACS variables), the whole
concept becomes inconsistent if the numerical value of "LOW" should ever exceed that of "HIGH".
Such a contradictory situation is referred to as "criss-crossing of the limits". Rather than terminate
the simulation with a fatal error stop in this case, Message 33 will appear as a warning. The user is
advised to rethink the problem formulation, should this message appear. A fatal EMTP error stop
really might have been more appropriate, because at the instant this message is printed, the
problem as posed by the user has no solution.
HIGH
HIGH
criss-cross
X(S)
TACS
function
block
BUS1
feasible
region
LOW
time
LOW
Figure 17.11:
Incorrect Limiter Operation
Message 34:
***WARNING TSSOL*** LIMITS OF BLOCK "BUS 1" HAVE CRISS-CROSSED. EXPECT PUZZLING
RESULTS. THIS WARNING MESSAGE WILL NOT BE REPEATED
The same as Message 33, except for the steady-state conditions.
Message 35:
THE USER HAS REQUESTED A TACS VARIABLE NAMED "BUS K" FOR OUTPUT PURPOSES. bUT
THIS BUS IS A NON-EXISTANT TACS VARIABLE, SO THE REQUEST WILL BE DISREGARDED
This message indicates an error in preparation of the TACS output-variable specification card of
Section 14.8. The 6-character alphanumeric name "BUS1" is somehow in error. There is no such
TACS variable, so the output request is being ignored (no such variable will be printed or plotted).
17 - 18
Message 36:
***WARNING INSUDC*** Block "BUS K" has its limiter operating during the TACS dc
steady-state solution. Double-check the program output at T = 0.0 for any
misunderstanding between the program and the user.
First, read Section 14.3 on limiters. The TACS-calculated dc steady-state calculation is a one-pass
operation. Depending on the ordering of the blocks, a limit may be superimposed on a block
either after or (preferably) before its value is used as input to another block. During the
calculation at t > 0.0, the effect of the limit is at worst delayed by one time-step, as discussed in
Section 14.3. But for the steady-state initialization at t=0.0, there is no second chance, and should
a limiter be found to operate at t=0.0, the user has to make sure that all other blocks depending on
the output of this limiter have indeed been calculated using the limited value.
Message 37:
******** WARNING ******** REQUEST FOR NONEXISTENT OUTPUT CLASS 7 NEGLECTED
This message implies an erroneous punch in column number 3 of a Class 5 synchronous machine
(S.M.) data card, as described in Section 10.11.2. Presumably the user has punched the wrong
column of the data card.
Message 38:
******** WARNING ******** REQUEST FOR NONEXISTENT VARIABLE 45 IN CLASS 2 OF
S.M. NO. 3 DISCARDED, IN THIS CLASS THE ABOVE S.M. CAN HAVE NUMBERS BETWEEN 1
(ONE) AND 14.
Message 38 implies an erroneous request for variable number "N1" within output class number
"N2" of synchronous machine number "N3" (in the order of input). For example, if a machine has
one mass, a request for mechanical torque between masses 1 and 2 is erroneous.
The allowable range of output variables within the output class in question is "1" and "N4".
Message 39:
NOTE ---- THE LAST-READ DATA CARD BELONGS TO A DYNAMIC SYNCHRONOUS MACHINE
FOR WHICH THE PARAMETERS ARE TO BE CALCULATED WITHIN THE MODULE "SMPFIT"
OFOVERLAY NUMBER 5. THE EXACT SOLUTION HAS FAILED., 32H THE EMTP LOGIC WILL NOW
RECOVER AND USE AN APPROXIMATE FORMULA.
Further comments about the "PARAMETER FITTING" option which has experienced trouble will be
found in Section 10.11.2, where the synchronous machine (S.M.) component is fully described;
see Class 2 data cards, Point c. Message 39 might possibly be printed twice for each physical
generating unit (four times for a dual machine), since d-axis parameters are fitted separately from
q-axis parameters.
17 - 19

As explained in Section 10.11.2, execution will continue without parameter fitting for this axis of
this machine. The subsequent simulation should be generally valid -- at least as valid as the input
data.
Message 40:
THE NUMBER OF ENERGIZATIONS FOR THIS STATISTICS RUN IS 1.
DEVIATION CANNOT BE CMPUTED
VARIANCE AND STANDARD
This message will be printed only for a "STATISTICS" or "SYSTEMATIC" data case (see Section
9.2.3 and Section 9.2.4), in which there is just one energization. This situation is user-controlled,
with integer miscellaneous data parameter "NENERG" (see Section 4.2.2) equal to unity. Of course
no sort of meaningful tabulation is possible, for the case of just a single energization. Yet the
EMTP will perform all of the conventional postprocessing of overlay #29 (module "SUBR29"), all
except for the variance (and standard deviation), calculation. Division by NENERG - 1 = 0 would
be required, and this is being bypassed; Message 40 is being printed instead.
Message 41:
WARNING ---- NODE RENUMBERING OF THE TRANSIENTS NETWORK HAS BROKEN DOWN,
PRESUMABLY DUE TO TABLE OVERFLOW (I.E., THE NETWORK IS TOO BIG AND/OR TOO DENSE
FOR THE PRESENT EMTP DIMENSIONING).
THE NEXT TIME THAT THE USER REDIMENSIONS
THE EMTP, HE IS ADVISED TO INCREASE THE SIZE OF LIST NUMBER 5 AND/OR 8 .
BOTH OF THESE LISTS CONTRIBUTE FULLY (100 PER CENT) TO DEPENDENT LIST NUMBER 99
WHICH IS WHAT HAS ACTUALLY OVERFLOWED AT THIS POINT.
149 NODES OUT OF TOTAL OF 180 WERE RENUMBERED BEFORE BREAKDOWN IN THE
RENUMBERING OVERLAY.
HAD WE MADE IT TO 170 NODES, THE OPERATION WOULD HAVE TERMINATED NORMALLY (SINCE
THE REMAINING ONES ARE ALWAYS FORCED LAST WITHOUT REGARD TO SPARSITY
CONSIDERATIONS)
ANYWAY, THE EMTP WILL TRY TO CONTINUE WITH EXECUTION OF THIS DATA CASE, AS BEST
IT CAN.
NODES WHICH WERE NOT RENUMBERED BEFORE THE OVERFLOW LIMIT WAS REACHED
WILL NOW SIMPLY BE RENUMBERED IN THEIR ORIGINAL RELATIVE ORDER, WITHOUT REGARD
TO SPARSITY CONSIDERATIONS.
RECALL THAT THE ORIGINAL NODE ORDER COMES FROM THE
ORDER OF DATA INPUT (THE ORDER IN WHICH NODE NAMES ARE ENCOUNTERED, AS THE EMTP
DATA CARDS ARE READ).
THE FINAL RENUMBERING MAP
(NORDER(I), I=1, NTOT)
1 WILL THEN APPEAR AS FOLLOWS ....
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30 31 32 33 34
THE K-TH SUCH ENTRY GIVES THE ROW (AND COLUMN) POSITION OF OLD VARIABLE K IN
THE REORDERED MATRIX.
This message is largely self-explanatory. It should be emphasized that it is the renumbering

attempt for the transient network (as used in the time-step loop) which has broken down, not the
renumbering attempt for the steady-state network. Should the same storage inadequacy occur
during renumbering for the steady-state phasor solution, a fatal EMTP error stop (with a "KILL"
code) will result.
17 - 20

Note that the EMTP is recovering from the storage inadequacy, as per the second paragraph of
Message 41. Yet this may be only a temporary reprieve. There is a fundamental storage
inadequacy, and a subsequent EMTP error stop during either the steady-state renumbering or the
Y-matrix factorization or formation (for the time-step loop) is not unlikely. It is dependent
List #99 (of the summary statistics which terminate each case; see Section 17.3 and 17.4) which is
of inadequate size for the user's data case. EMTP redimensioning of Section 3.3 is recommended,
with large numbers for List #5 and/or List #8. Both of these lists contribute fully (100%) to List
#99.
For illustrative purposes only, the renumbering map NORDER(I) has been shown as NORDER(I) =
I, for I=I, NTOT. Of course this is completely problem-dependent.
Message 42:
WARNING. TACS VARIABLE "BUS K" IS ALLOWED TO OPERATE OUTSIDE ITS LIMITS DURING
STEADY-STATE SOLUTION
This situation may occur after superimposing the DC and AC TACS-calculated steady-state
initialization solutions. In this case, neither DC or AC contributions to the output of this block
involved the operation of the associated limiter. But when adding the two, it is found that the
output value exceeds either lower or higher limits of the block. However, the value of other
blocks depending on this output has already been calculated. Read the interpretation of Message
36 for more details.
Message 43:
NOTICE ----- OVERVOLTAGE TABULATION FOR THIS VOLTAGE VARIABLE CAN NOT CONTINUE,
DUE TO INSUFFICIENT WORKING
SPACE.
STATISTICS MISCELLANEOUS DATA PARAMETER
"MAXMX' HAS BEEN EXCEEDED BY THE PEAK PER UNIT OVERVOLTAGE (ACTUALLY, EXCEEDED
BY 5.0 TIMES OR MORE
Something is seriously wrong, if the user sees Message 43. Either one or more solutions is
garbage, or the user has made an error in specifying his base voltage, or both! Recall that
"XMAXMX" of the "STATISTICS" or "SYSTEMATIC" miscellaneous data card (see Section 4.3.1) is
supposed to be a voltage level (in per unit) which should never be exceeded during any of the
energizations. Then, just to be double sure, the EMTP multiplies this user-supplied figure by five.
Message 43 then results when the user-supplied bound is exceeded by a factor of five or more! So
what is wrong? Is "XMAXMX" reasonable? Are the base voltages as defined on the "STATISTICS"
or "SYSTEMATIC" output cards (see Section 12.3.3) reasonable? Are peak voltages for each
energization reasonable? An inconsistency exists in one or more of these three cards.
17 - 21
Message 44:
ONE OR MORE NONEXISTENT NODE NAMES IN VOLTAGE DIFFERENCE "BUS K" TO "BUS M" .
THIS REQUEST IGNORED
The EMTP data case which is now being solved is a "STATISTICS" or "SYSTEMATIC" one (see
integer miscellaneous data parameter "NENERG" of Section 4.2.2). The base case (with all
variances temporarily set to zero) has already been solved and plotted. The EMTP is now reading
and processing the output requests for statistical overvoltage tabulation (see Section 12.3.3). Any
pair of legal node names is allowed, in making a voltage-difference request. But one or more of
the user's names ("BUS1" and "BUS2" of the message) does not correspond to any node of the user's
network. Is there spelling trouble? Or is there trouble with the positioning of blank characters
within the alphanumeric field of width six?
Message 45:
WARNING. ---- FREQUENCY SENSOR "BUS K" HAS ZERO CROSSING AT 0.2500E-04 SEC.
BUT NEW FREQUENCY OF 0.6500E+03 HZ
DIFFERS BY OVER FIFTY PERCENT FROM THE OLD
FREQUENCY OF 0.6000E+02 HZ .
REJECT IT.
The data case being solved involves TACS modeling, as described in Section 14. Included is a
frequency sensor having name "BUS1", a Type50 supplemental device (see Section 14.4.). Recall
that this device merely detects the time interval between successive zero-crossings of the input
signal, and reciprocates twice this to give an estimate of the instantaneous frequency (whatever
that is). But the new frequency which has just been calculated differs from the old one (calculated
at the preceding zero-crossing; or possibly left over from the initial condition) by at least fifty
percent. Such drastic changes in frequency are not allowed, for they are an indication that the
solution is not varying smoothly. Possibly there was a spurious zero-crossing due to "hash" or
"noise", and rejection of this latest estimate is based on this assumption. As stated in Section 14.4,
"the device will automatically reject higher-frequency non-characteristic oscillations appearing on
the main signal." Frequency of the "main signal" is of course defined by the user as the initial
frequency which is punched in columns 5156 of the device data card.
Message 46:
WARNING. ---- VALUE OF DELAY EXCEEDED ,
MAX. DELAY VALUE OF 0.18400E-02 FOR
"BUS K" AT TIME = 0.15000E-01. THIS MESSAGE WILL NOT BE REPEATED.
The data case being solved includes TACS modeling, as described in Section 14. Included is a
transport-delay block bearing output name "BUS1" a Type53 supplemental device of
Section 14.4. Recall that the delay can be a variable (dynamically-changing) quantity, since it is
the sum of a fixed delay and a named delay. Well, this total delay was never supposed to exceed
the user-supplied bound of "D1". But it did. The transport-delay block needs more storage, but
cannot get it (dynamic expansion of the tables is not possible at this time). The simulation will
continue, but is erroneous in that delays are strictly limited to time "D1". Note that only this first
instance of such saturation or clipping will be noted on the line printer. A plot would reveal the
full story (see Figure 17.12).
17 - 22
problem
limit D1
desired delay (t)
simulation time
Figure 17.12:
Inadequate Storage for Transport-Delay Block
Message 47:
WARNING. ---- THE PULSE FREQUENCY AT THE PULSE TRANSPORT DELAY "BUS K" IS TOO
FAST FOR THE PRESENT DELAY OF 0.83333E-03 SEC AT SIMULATION TIME, 0.15000E-1
SEC. USE DEVICE TYPE 53 INSTEAD OF TYPE 54 .
******** THE ANSWER MAY BE WRONG LATER ********
The data case being solved involves TACS modeling, as is described in Section 14. Included is a
pulse transport delay block bearing the output name "BUS1", a Type54 supplemental device of
Section 14.4. Recall that the delay must never exceed the time between successive changes in the
input signal. Phrased another way, never more than one variable change can be stored within the
history of the delay; it only has scalar storage (unlike the Type53 block which has vector
storage). The last change has to propagate through the delay block and out the other end before a
new change is allowed in the front end. But this limitation has not been observed in the user's
case.
17 - 23
1.0
etc.
1.0
1.0
etc.
time
For illustrated train of pulses, the delay must

not exceed this minimum time between changes.
Figure 17.13:
Requirements on Delay-Time Specification
If the user sees Message 47, possibly it is an indication of trouble elsewhere in the system. That is,
problems elsewhere make for an erroneous input signal to the block, which in turn triggers this
warning message. After all, garbage in, garbage out. On the other hand, if the input signal and all
other parameters of the block seem to be valid, then perhaps the user must consider switching from
the Type54 device to the Type53 device.
Message 48:
WARNING. ---- VALLUE OF TDELAY BECAME NEGATIVE FOR "BUS K" AT TIME = 0.15000E-01 BUT
LOWER LIMIT VALUE = 0.0. THIS MESSAGE WILL NOT BE REPEATED
Similar to Message 46. In this case, the dynamically-changing value of the delay has become
negative. So as not to transform this Type53 device into a crystal ball, the lower bound on the
variable delay value is 0.0.
Message 49:
NOTE ---- FOR THE TYPE-96 HYSTERETIC INDUCTOR WHICH CONNECTS BUS "BUS K" TO
"BUS M" AN ATTEMPT WAS DETECTED TO OPERATE OUTSIDE THE MAJOR HYSTERESIS LOOP THE
TRAJECTORY HAS BEEN MODIFIED TO PREVENT THIS. ACCURACY OF THE RESULTS SHOULD BE
UNAFFECTED. T = 0.15000E-01
This message is associated with the way minor loop trajectories are created for a Type96
hysteretic inductor. A minor loop trajectory is specified in terms of the major hysteresis loop
trajectory by constraining how the distance between the trajectories (along a line of constant
current) varies as operation proceeds along the minor loop. This is shown below.
17 - 24
point 1 (i1, 2)
point 2 (i2, 2)
point 3 (isat, sat)
D1
(i , )
D2
D()
In this figure the solid line from pt. 1 to the saturation point indicates a minor loop trajectory,
which beings at pt. 1 and extends out into the saturation region. Notice that as increases from
1 to sat, D() decreases from D1 to 0. This decrease occurs linearily and can be represented in
the D()- plane as shown.
D()
D1
sat
There are other instances in which it is known (from a knowledge of past behaviour of the core)
that the minor loop trajectory should pass through point 2. This requires the D() function show
below.
17 - 25
D()
D1
D2
sat
Notice that in the first instance the minor loop joined the major loop (D() became zero) right at
the saturation point, while in the second instance the minor loop joined the major loop at the point
(i', '). This is all fine. However, consider the following case:
D()
D1
D2
sat
Suppose point 2 was located as shown in the above figure. Following the same algorithm as
previously used would lead to a function D(y) as shown by the dashed line. Notice that now ' >
sat, which indicates that the minor and major loop trajectories would not intersect until
somewhere beyond the saturation point. This is impossible, however, since beyond the saturation
point there is no difference between major and minor loop trajectories (because of the single
valued nature of hysteresis in the saturation region). In this case a different algorithm is chosen,
which leads to the trajectory shown as a solid line. For this trajectory the major and minor loop
will intersect right at the saturation point.
17 - 26

The occurrence of Message 49 indicates that this last case described has occurred. It is anticipated
that this case should occur very seldom. Should it occur the results obtained may be considered as
accurate as any other results obtained using Type96 hysteresis modeling. Unless you receive this
message quite often it may simply be ignored. If however you do receive this message quite often
please inform program maintenance. It may be necessary to rethink some aspects of the hysteresis
algorithm.
Message 50:
NOTE ---- NONLINEAR ELEMENT NUMBER 1 IS A TYPE-96 HYSTERETIC INDUCTOR WHICH IS
CONNECTED BETWEEN BUSSES "BUS K" AND "BUS M".
THE INITIAL FLUX-CURRENT
POINT AS FOUND BY THE PHASOR STEADY-STATE SOLUTION HAS BEEN OBSERVED TO LIE
OUTSIDE THE USER-DEFINED MAJOR HYSTERESIS LOOP, HOWEVER THE INITIAL FLUX IS
0.6264E+02 AND THE INITIAL CURRENT IS 0.15000E+04.
THE EMTP SHALL NOW ALTER
THIS JUST-PRINTED FLUX SO AS TO MAKE IT LEGAL, WHILE HOLDING THE CURRENT
CONSTANT.
THE LINE OF CONSTANT CURRENT INTERSECTS THE USER-SUPPLIED MAJOR
HYSTERESIS LOOP AT TWO POINTS (POSSIBLY EQUAL, IF THE CURRENT IS LARGE ENOUGH).
THE 'UPPER' IS CUT AT FLUX VALUE, 0.6000E+02 AND THE 'LOWER' AT FLUX VALUE
0.58000E+02 THE INITIAL FLUX SHALL BE TAKEN BY THE EMTP TO BE THE AVERAGE OF
THESE, WHICH HAS FLUX VALUE 0.59000E+02.
This message indicates a discontinuity in the solution of a problem involving a Type96 hysteretic
reactor at time zero. It is a result of two different methods used to model the Type96 flux-current
behaviour. During a steady-state solution the Type96 is represented as a linear element, defined
by the point isteady, steady (see Section 8.5) as shown below. It is possible for the initial fluxcurrent point to lie anywhere on the line, even outside of the major hysteresis loop, as shown by
point A. The difficulty arises at time zero when the transient solution begins. During the transient
solution the actual hysteresis behaviour is modelled, and operation outside of the major loop is
never allowed. Therefore an initial point, calculated by a steady-state solution, which lies outside
of the major loop violates this basic rule. In such a case it is necessary to either move the initial
point inside of the loop or terminate the run at this point. It was decided to move the point within
the major loop, print Message 50 and leave the user with the responsibility of deciding for himself
whether the changes made are acceptable or not.
17 - 27
(isteady, steady)
Figure 17.14:
Steady-State Point Outside the Major Loop
The method used to move this point within the major loop is the following. The line of constant
current is drawn through the initial point, and its intersections with the upper and lower half of the
major hysteresis loop are calculated (points B and C). The average of these points (point D) is then
taken as the initial point. Notice that if the value of initial current is large enough (> isat) then
points B, C and D are all the same point, namely the new initial point. This same discussion holds,
of course, for an initial point in the third quadrant (see Figure 17.15).
1
a
b
d
c
i
Figure 17.15: Redefinition of the Steady-State Initial Conditions

17 - 28

As mentioned earlier, it is the user's responsibility to decide if the changes made are acceptable in
this case. It is very strongly suggested that the user make legitimate changes which will cause all
such messages to disappear. It is important to realize that transients due to sudden changes in flux
(like those being made here) may die out very slowly. A further discussion of this problem and
some suggestions for correcting it are contained in Section 8.5.
17.2 Connection Discontinuities During Solution

The closing and/or opening of switches and flashover elements is noted by means of printed
messages, as the solution proceeds. Such messages thus interrupt the regular time-step printout of
solution variables (see Section 16.3). The message texts are as follows, for the different element
types which have such printout. Where sample node names are needed, the names "NODE1" and
"NODE2" are used. Where a floating-point number is printed, the FORTRAN format specification is
used.
Ordinary Switch: (See Section 9)

The message for closure:
SWITCH 'NODE 1' TO 'NODE 2' CLOSED AFTER
E12.5
SEC.
The message for opening:

SWITCH 'NODE 1' TO 'NODE 2' OPEN AFTER
E12.5
SEC.
Diode, Valve or TACs-Controlled Switch: (See Section 9.5)

Messages for the opening and closing of a Type11, Type12, or Type13 TACS-controlled
switch are similar to those for the "Ordinary Switch" just described. But one difference is that
here such messages must be explicitly requested by the user by the column79 parameter "ECHO".
The general form of the message is:
(DIODE VALVE) 'NODE 1' TO 'NODE 2' (OPENING/CLOSING) AFTER
(GAP SWITCH) 'NODE 1' TO 'NODE 2' (OPENING/CLOSING) AFTER
E12.5 SEC.
E12.5 SEC.
Pseudo-Nonlinear Resistance: (See Section 8.3)

The same message text applies both to flashover and to opening, with only the segment number
"XX" differing:
17 - 29
Flashover:
Segment number XX = JUMP > 0

(see columns 39-44 of input data card)
Opening:
Segment number XX = 0
The message itself reads as follows:

TYPE99 N.L. ELEM 'NODE 1' to 'NODE 2' AT T =
SEGMENT XX.
E12.4 BEGINS OPERATION ON
Nonlinear Resistance: (See Section 8.8)

As per Section 8.8, the piecewise-nonlinear resistance element can flashover either once or many
times. The flashover message reads as follows:
TYPE92 NONLINEAR VI FROM FROM 'NODE 1' TO 'NODE 2' FIRES AT TIME
E14.6 SEC.
17.3 Printout Of Storage And Running-time Statistics

At the conclusion of the solution of each data case, a half page of printout provides a summary of
storage and running-time requirements. Refer to the sample output of Section 17.4 In summary,
the following will be found:
1.
Current program dimensioned limits for all tables. These are identified by "list
number", row by row, as per Section 2.5.1.
2.
Actual number of cells of these tables which the current problem has used.
3.
Computer running time for this case. CPU, input-output (I/O), and total (CPU + I/
O) are all three printed, for each of five major subdivisions of the program. Sums
of these are also provided, giving the total computer time attributable to
execution of the case just finished. Yet caution should be used in interpreting
these figures, due to variations between different computer installations. Some
computers may not have a CPU figure available, in which case zero contribution
for this component will generally be observed.
17 - 30
17.4 Printer And Plotter Output of Sample EMTP Test

Problem
The next few pages show the line printer output which resulted from the EMTP solution of typical
benchmark Minor variations in the appearance of the results might be expected for a different
computer and different plotter.
EMTP begins:
Done reading disk file into EMTP cache. There are NUMCRD = 106 cards.
ELETROMAGNETIC TRANSIENT PROGRAM (EMTP) - DCG/EPRI VERSION 3.0F MS WINDOWS TRANSLATION
RELEASE 01.08.96
User Support & Maintenance Centre - Ontario Hydro, Canada.
phone:
(800) 245-8791 or (416) 592-4793
fax:
(416) 592-9448
e-mail: technical.support@emtp96.com
RUN DATE (MM/DD/YY) AND TIME (HH.MM.SS.)= 4/11/97 10.17.30
NAME OF PLOT DATA FILE (IF ANY) =D:\EMTPV3\BENCH\EMTP\dc32.pl4
--------------------------------------------------+------------------------------------------------------------------------DESCRIPTIVE INTERPRETATION OF NEW-CASE INPUT DATA 1 INPUT DATA CARD IMAGES PRINTED BELOW, ALL 80 COLUMNS, CHARACTER BY CHARACTER.
0
1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
0
--------------------------------------------------+------------------------------------------------------------------------MARKER CARD PRECEDING NEW DATA CASE.
1BEGIN NEW DATA CASE
Comment line.
1C
BENCHMARK DC-32
Comment line.
1C
Comment line.
1C ----dt<---Tmax<---Xopt<---Copt<-Epsiln<-Tolmat<-Tstart
MISC. DATA.
0.200E-04
0.100E-01
0.000E+00 1 .000020
.010
Comment line.
1C .....^.......^.......^.......^.......^.......^.......^
Comment line.
1C -Iprnt<--Iplot<-Idoubl<-Kssout<-Maxout<---Ipun<-Memsav<---Icat<-Nenerg
MISC. DATA.
1
1 1 1 1 -1 1 2 0 0
1
1
1
1
1
1
-1
1
2
Comment line.
1C .....^.......^.......^.......^.......^.......^.......^.......^.......^.......^
Comment line.
1C <--Isw<--Itest<--Idist<--Aincr<-Xmaxmx<-Degmin<-Degmax<-Statfr<-Sigmax<--Nseed
Comment line.
1C
Comment line.
1C .....^.......^.......^.......^.......^.......^.......^.......^.......^.......^
Comment line.
1C --Kchg<---Mult<---Kchg<---Mult<---Kchg<---Mult<---Kchg<---Mult<---Kchg<---Mult
PRINTOUT :
5
5
10
10
50
50
1
5
5
10
10
50
50
Comment line.
1C .....^.......^.......^.......^.......^.......^.......^.......^.......^.......^
Comment line.
1C
Comment line.
1C
THREE 2-WINDING TRANSFORMERS, WITH TURNS RATIOS 1 TO 1 ,
Comment line.
1C
CONNECTED IN DELTA ON THE SOURCE SIDE, AND GROUNDED WYE ON THE
Comment line.
1C
RECTIFIER SIDE.
MAGNETIZING BRANCH IS NONLINEAR AND LOSSY.
1 TRANSFORMER
.01
.05TRANA
1.E5
1
_SAT. XFORMER. :
0.100E-01 0.500E-01 0.100E+06
BREAKPOINT.
0.10000E-01
0.50000E-01
1
.01
.05
BREAKPOINT.
0.20000E-01
0.60000E-01
1
.02
.06
BREAKPOINT.
0.50000E-01
0.70000E-01
1
.05
.07
BREAKPOINT.
0.15000E+00
0.80000E-01
1
.15
.08
1
9999
WINDING 1.
0.6000E-01 0.1000E-01 0.1150E+03 1 1GENA GENB
.06
.01 115.
WINDING 2.
0.6000E-01 0.1000E-01 0.1150E+03 1 2SECA
.06
.01 115.
TRANSFORMER COPY USING REFERENCE NAME 'TRANA '.
1 TRANSFORMER TRANA
TRANB
WINDING 1.
1 1GENB GENC
WINDING 2.
1 2SECB
TRANSFORMER COPY USING REFERENCE NAME 'TRANA '.
1 TRANSFORMER TRANA
TRANC
WINDING 1.
1 1GENC GENA
WINDING 2.
1 2SECC
Comment line.
1C
SMALL IMPEDANCE BRANCHES USED TO ISOLATE SWITCHES FROM EACH OTHER AT
Comment line.
1C
'POLEA' NODE.
RESISTANCE OF 5.0 OHMS WAS USED.
SERIES R-L-C.
0.500E+01 0.000E+00 0.000E+00 1 CATH1 POLEA
5.0
REFERENCE BRANCH.
COPY 'CATH1 ' TO 'POLEA '
1 CATH2 POLEA CATH1 POLEA
REFERENCE BRANCH.
COPY 'CATH1 ' TO 'POLEA '
1 CATH3 POLEA CATH1 POLEA
Comment line.
1C
SMALL IMPEDANCE BRANCHES USED TO ISOLATE SWITCHES FROM EACH OTHER AT
Comment line.
1C
'POLEB' NODE, AND ALSO AT THE TRANSFORMER SECONDARY.
Comment line.
1C
HALF OF THE VALUE USED FOR 'POLEA' WAS USED HERE, FOR BALANCE.
SERIES R-L-C.
0.250E+01 0.000E+00 0.000E+00 1 CATH4 SECC
2.5
REFERENCE BRANCH.
COPY 'CATH4 ' TO 'SECC '
1 CATH5 SECA CATH4 SECC
REFERENCE BRANCH.
1 CATH6 SECB CATH4 SECC
REFERENCE BRANCH.
1 AN4
POLEB CATH4 SECC
REFERENCE BRANCH.
1 AN5
POLEB CATH4 SECC
REFERENCE BRANCH.
1 AN6
POLEB CATH4 SECC
Comment line.
1C
2-CONDUCTOR DISTRIBUTED-PARAMETER LINE CONNECTING 'POLE' WITH 'LOAD' .
Comment line.
1C
THE INDUCTANCE PARAMETERS ARE 5.8 AND 1.6 MH FOR THE GROUND AND SKY
Comment line.
1C
MODES.
LENGTH OF THE LINE IS FIVE MILES.
Comment line.
1C
MODAL LINE CONSTANTS ARE TYPICAL OF 500-KV OVERHEAD LINE OF BPA.
0.300E+00 0.580E+01 0.120E-01 0.695E+03 0.417E-041-1POLEA LOADA
0.3
5.8 .012
5.0
0.300E-01 0.160E+01 0.190E-01 0.290E+03 0.276E-041-2POLEB LOADB
.03
1.6 .019
5.0
Comment line.
1C
LARGE STRAY CAPACITANCE TO TIE THE DC SIDE TO GROUND.
SERIES R-L-C.
0.000E+00 0.000E+00 0.100E+03 1 POLEA
100.
REFERENCE BRANCH.
COPY 'POLEA ' TO '
'
1 POLEB
POLEA
Comment line.
1C
LOAD ON DC SIDE IS 200 OHM RESISTANCE PLUS 100 MICROFARAD
Comment line.
1C
SHUNT CAPACITANCE.
SERIES R-L-C.
0.200E+03 0.000E+00 0.000E+00 1 LOADA LOADB
200.
17 - 31

SERIES R-L-C.
0.000E+00 0.000E+00 0.100E+03
Comment line.
SERIES R-L-C.
0.100E+01 0.000E+00 0.000E+00
BLANK CARD TERMINATING BRANCH CARDS.
Comment line.
SWITCH.
0.10E+01 0.20E+01 0.00E+00 0.00E+00
Comment line.
Comment line.
Comment line.
DIODE, NO TACS GRID
DIODE, NO TACS GRID
DIODE, NO TACS GRID
DIODE, NO TACS GRID
DIODE, NO TACS GRID
DIODE, NO TACS GRID
BLANK CARD TERMINATING SWITCH CARDS.
Comment line.
SOURCE.
0.94E+02 0.40E+03 -0.30E+02 -0.10E+01
SOURCE.
0.94E+02 0.40E+03 -0.15E+03 -0.10E+01
SOURCE.
0.94E+02 0.40E+03 0.90E+02 -0.10E+01
BLANK CARD TERMINATING SOURCE CARDS.
1 LOADA LOADB
100.
1C
ONE-OHM GROUNDING RESISTANCE FOR LATER FAULT DURING DC-49 .
1 LOADA FAULT
1.0
1BLANK CARD ENDING BRANCH CARDS
1C
FIRST SWITCH IS TO BE THE FAULT SWITCH (OPEN HERE; SEE DC-49)
1 FAULT
1.0
2.0
1C
BRIDGE CIRCUIT CONSISTING OF SIX UNCONTROLLED VALVES (DIODES).
1C
DIODES 1, 2, AND 3 FEED 'POLEA' ,
AND MAKE IT POSITIVE WITH RESPECT
1C
TO GROUND.
DIODES 4, 5, 6 FEED 'POLEB' ,
MAKING IT NEGATIVE.
111SECB CATH1
13
111SECA CATH2
CLOSED
13
111SECC CATH3
13
111AN4
CATH4
CLOSED
13
111AN5
CATH5
13
111AN6
CATH6
CLOSED
13
1BLANK CARD ENDING SWITCH CARDS
1C
BALANCED 3-PHASE SOURCE OF 400 HZ IS CONNECTED TO XFORMER DELTA.
114GENA
93.897
400.
-30.
-1.
114GENB
93.897
400.
-150.
-1.
114GENC
93.897
400.
90.
-1.
1BLANK CARD ENDING SOURCE CARDS
LIST OF INPUT ELEMENTS CONNECTED TO EACH BUS.

1) ONLY THE PHYSICAL CONNECTIONS OF MULTIPHASE LINES ARE SHOWN (CAPACITIVE AND INDUCTIVE COUPLING IGNORED)
2) REPEATED ENTRIES IMPLY PARALLEL CONNECTIONS
3) SOURCES ARE OMITTED, ALTHOUGH SWITCHES ARE INCLUDED;
4) U.M. USAGE PRODUCES EXTRA, INTERNALLY-DEFINED NODES "UM????" (1ST 2 LETTERS "UM").
FROM BUS NAME 1 NAMES OF ALL ADJACENT BUSSES
--------------+---------------------------------------------------------------------------------------------------------TRANA 1GENA *GENB *GENB *GENB *
GENA 1TRANA *TRANC *TRANC *TRANC *
GENB 1TRANA *TRANA *TRANA *TRANB *
SECA 1TERRA *CATH2 *CATH5 *
TRANB 1GENB *GENC *GENC *GENC *
GENC 1TRANB *TRANB *TRANB *TRANC *
SECB 1TERRA *CATH1 *CATH6 *
TRANC 1GENA *GENA *GENA *GENC *
SECC 1TERRA *CATH3 *CATH4 *
CATH1 1SECB *POLEA *
POLEA 1TERRA *CATH1 *CATH2 *CATH3 *LOADA *
CATH2 1SECA *POLEA *
CATH3 1SECC *POLEA *
CATH4 1SECC *AN4
*
CATH5 1SECA *AN5
*
CATH6 1SECB *AN6
*
AN4
1CATH4 *POLEB *
POLEB 1TERRA *AN4
*AN5
*AN6
*LOADB *
AN5
1CATH5 *POLEB *
AN6
1CATH6 *POLEB *
LOADA 1POLEA *LOADB *LOADB *FAULT *
LOADB 1POLEB *LOADA *LOADA *
FAULT 1TERRA *LOADA *
TERRA 1SECA *SECB *SECC *POLEA *POLEB *FAULT *
--------------+---------------------------------------------------------------------------------------------------------PI-EQUIV BRANCHES OF DISTRIB LINES IN TR, TX, ETC. BETWEEN LIMITS
12
14
NONLINEAR AND TIME-VARYING RESISTANCES IGNORED IN STEADY STATE SOLUTION. NONLINEAR INDUCTANCES INCLUDED WITH LINEAR PART
SINUSOIDAL STEADY STATE SOLUTION, BRANCH BY BRANCH. ALL FLOWS ARE AWAY FROM BUS, AND REAL PART, MAGNITUDE, OR P
IS PRINTED ABOVE THE IMAGINARY PART, THE ANGLE, OR Q.
FIRST SOLUTION FREQUENCY =
0.400000000E+03
HERTZ.
BUS K
NODE VOLTAGE
BRANCH CURRENT
POWER FLOW
BUS M
RECTANGULAR
POLAR
RECTANGULAR
POLAR
P AND Q
GENA
TRANA
TRANA
GENB
SECA
TERRA
TRANA
GENB
GENB
POWER LOSS
P AND Q
0.8131719E+02
-0.4694850E+02
0.9389700E+02
-30.0000
0.1687481E+02
0.1531615E+02
0.2278911E+02
42.2280
0.3265708E+03
-0.1018857E+04
0.8068964E+02
-0.4829158E+02
0.9403666E+02
-30.8999
-0.1687481E+02
-0.1531615E+02
0.2278911E+02
-137.7720
-0.3109905E+03
0.1025383E+04
0.8068964E+02
-0.4829158E+02
0.9403666E+02
-30.8999
0.1620068E-02
-0.1343079E-04
0.1620124E-02
-0.4750
0.6568565E-01 0.1312401E+00
-0.3857596E-01 -0.6938894E-17
-0.8131719E+02
-0.4694850E+02
0.9389700E+02
-150.0000
-0.1620068E-02
0.1343079E-04
0.1620124E-02
179.5250
0.1613797E+03
-0.2686895E+01
0.1614021E+03
-0.9539
-0.1687330E+02
-0.1532906E+02
0.2279667E+02
-137.7455
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
0.1687330E+02
0.1532906E+02
0.2279667E+02
42.2545
0.8068964E+02
-0.4829158E+02
0.9403666E+02
-30.8999
0.1687330E+02
0.1532906E+02
0.2279667E+02
42.2545
0.3106178E+03 0.1356500E+04
-0.1025867E+04 -0.1253037E+04
-0.8131719E+02
-0.4694850E+02
0.9389700E+02
-150.0000
-0.1687330E+02
-0.1532906E+02
0.2279667E+02
-137.7455
0.1045883E+04
-0.2271700E+03
-0.8131719E+02
-0.4694850E+02
0.9389700E+02
-150.0000
-0.2959540E+01
-0.3402403E+02
0.3415251E+02
-94.9713
0.9190194E+03
-0.1313896E+04
17 - 32
0.1558031E+02
0.6526266E+01
0.6555442E-01
0.3857596E-01
-0.1340910E+04 -0.1340910E+04
0.1259568E+04 0.1259568E+04
0.0000000E+00
0.0000000E+00
0.3499181E+02
0.1465734E+02
TRANB
TRANB
GENC
SECB
TERRA
TRANB
GENC
GENC
TRANC
TRANC
GENA
SECC
TERRA
TRANC
GENA
CATH1
POLEA
CATH2
POLEA
CATH3
POLEA
CATH4
SECC
CATH5
SECA
CATH6
-0.8199473E+02
-0.4483268E+02
0.9345108E+02
-151.3313
0.2959540E+01
0.3402403E+02
0.3415251E+02
85.0287
-0.8840276E+03
0.1328554E+04
-0.8199473E+02
-0.4483268E+02
0.9345108E+02
-151.3313
-0.8199473E-03
-0.1387297E-02
0.1611492E-02
-120.5848
0.6471379E-01 0.1298453E+00
-0.3849530E-01 -0.6938894E-17
0.5749343E-14
0.9389700E+02
0.9389700E+02
90.0000
0.8199473E-03
0.1387297E-02
0.1611492E-02
59.4152
-0.8267312E+02
-0.1366138E+03
0.1596815E+03
-121.1806
0.2947680E+01
0.3402917E+02
0.3415660E+02
85.0493
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
-0.2947680E+01
-0.3402917E+02
0.3415660E+02
-94.9507
-0.8199473E+02
-0.4483268E+02
0.9345108E+02
-151.3313
-0.2947680E+01
-0.3402917E+02
0.3415660E+02
-94.9507
0.5749343E-14
0.9389700E+02
0.9389700E+02
90.0000
0.2947680E+01
0.3402917E+02
0.3415660E+02
85.0493
0.1597619E+04
0.1383892E+03
0.5749343E-14
0.9389700E+02
0.9389700E+02
90.0000
-0.2397538E+01
0.2099098E+02
0.2112745E+02
96.5159
0.9854948E+03
-0.1125608E+03
0.6714130E+00
0.9269780E+02
0.9270023E+02
89.5850
0.2397538E+01
-0.2099098E+02
0.2112745E+02
-83.4841
-0.9721037E+03
0.1181701E+03
0.6714130E+00
0.9269780E+02
0.9270023E+02
89.5850
-0.8064577E-03
0.1396463E-02
0.1612601E-02
120.0064
0.6445379E-01
-0.3784723E-01
0.8131719E+02
-0.4694850E+02
0.9389700E+02
-30.0000
0.8064577E-03
-0.1396463E-02
0.1612601E-02
-59.9936
0.6557036E-01
0.3784723E-01
-0.7997394E+02
0.1384478E+03
0.1598863E+03
120.0127
0.2407845E+01
-0.2098316E+02
0.2112086E+02
-83.4539
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
-0.2407845E+01
0.2098316E+02
0.2112086E+02
96.5461
0.0000000E+00
0.0000000E+00
0.6714130E+00
0.9269780E+02
0.9270023E+02
89.5850
-0.2407845E+01
0.2098316E+02
0.2112086E+02
96.5461
0.9717381E+03
-0.1186451E+03
0.8131719E+02
-0.4694850E+02
0.9389700E+02
-30.0000
0.2407845E+01
-0.2098316E+02
0.2112086E+02
-83.4539
0.5904635E+03
0.7966235E+03
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
-0.2842171E-14
0.2842171E-14
0.4019437E-14
135.0000
-0.2221802E-12 0.5048710E-28
0.3295470E-14 -0.1262177E-28
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
0.2842171E-14
-0.2842171E-14
0.4019437E-14
-45.0000
0.2221802E-12
-0.3295470E-14
0.1613797E+03
-0.2686895E+01
0.1614021E+03
-0.9539
0.1687330E+02
0.1532906E+02
0.2279667E+02
42.2545
0.1340910E+04
-0.1259568E+04
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
-0.1687330E+02
-0.1532906E+02
0.2279667E+02
-137.7455
-0.4168940E+02
0.1259568E+04
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
-0.2842171E-14
0.2842171E-14
0.4019437E-14
135.0000
-0.2221802E-12 0.5048710E-28
0.3295470E-14 -0.1262177E-28
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
0.2842171E-14
-0.2842171E-14
0.4019437E-14
-45.0000
0.2221802E-12
-0.3295470E-14
-0.7395433E+02
0.8598992E+02
0.1134174E+03
130.6967
0.2407845E+01
-0.2098316E+02
0.2112086E+02
-83.4539
-0.9912055E+03
-0.6723726E+03
-0.7997394E+02
0.1384478E+03
0.1598863E+03
120.0127
-0.2407845E+01
0.2098316E+02
0.2112086E+02
96.5461
0.1548819E+04
0.6723726E+03
0.1613797E+03
-0.2686895E+01
0.1614021E+03
-0.9539
-0.1136868E-13
0.1776357E-15
0.1137007E-13
179.1048
-0.9175760E-12
0.9398350E-15
0.1613797E+03
-0.2686895E+01
0.1614021E+03
-0.9539
0.1136868E-13
-0.1776357E-15
0.1137007E-13
-0.8952
0.9175760E-12
-0.9398350E-15
-0.7530392E+02
-0.5154091E+02
0.9125319E+02
-145.6108
0.2947680E+01
0.3402917E+02
0.3415660E+02
85.0493
-0.9879332E+03
0.1205302E+04
17 - 33
0.6513150E-01
0.3849530E-01
-0.2446275E+04 -0.2446275E+04
0.1205302E+04 0.1205302E+04
0.0000000E+00
0.0000000E+00
0.8836566E+03 0.2481275E+04
-0.1329030E+04 -0.1190641E+04
0.1339108E+02
0.5609241E+01
0.1300241E+00
0.0000000E+00
-0.1548819E+04 -0.1548819E+04
-0.6723726E+03 -0.6723726E+03
0.1562202E+04
0.6779783E+03
0.1299220E+04
0.2273737E-12
0.5576135E+03
0.0000000E+00
0.2019484E-27
0.0000000E+00
0.1458342E+04
0.2273737E-12

SECB
AN4
POLEB
AN5
POLEB
AN6
POLEB
POLEA
LOADA
POLEB
LOADB
POLEA
TERRA
POLEB
TERRA
LOADA
LOADB
LOADA
LOADB
LOADA
FAULT
TRANA
GENB
TRANB
GENC
TRANC
GENA
-0.8267312E+02
-0.1366138E+03
0.1596815E+03
-121.1806
-0.2947680E+01
-0.3402917E+02
0.3415660E+02
-94.9507
0.2446275E+04
-0.1205302E+04
-0.7395433E+02
0.8598992E+02
0.1134174E+03
130.6967
-0.2407845E+01
0.2098316E+02
0.2112086E+02
96.5461
0.9912055E+03
0.6723726E+03
-0.6793472E+02
0.3353202E+02
0.7575963E+02
153.7294
0.2407845E+01
-0.2098316E+02
0.2112086E+02
-83.4539
-0.4335920E+03
-0.6723726E+03
-0.6793472E+02
0.3353202E+02
0.7575963E+02
153.7294
0.5684342E-14
-0.2842171E-14
0.6355287E-14
-26.5651
-0.2407339E-12
-0.1237308E-14
-0.6793472E+02
0.3353202E+02
0.7575963E+02
153.7294
-0.5684342E-14
0.2842171E-14
0.6355287E-14
153.4349
0.2407339E-12
0.1237308E-14
-0.7530392E+02
-0.5154091E+02
0.9125319E+02
-145.6108
-0.2947680E+01
-0.3402917E+02
0.3415660E+02
-94.9507
0.9879332E+03
-0.1205302E+04
-0.6793472E+02
0.3353202E+02
0.7575963E+02
153.7294
0.2947680E+01
0.3402917E+02
0.3415660E+02
85.0493
0.4704085E+03
0.1205302E+04
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
-0.3065059E+01
-0.4026472E+01
0.5060342E+01
-127.2794
0.4168940E+02
0.2766247E+03
-0.3583831E+01
-0.1707245E+02
0.1744455E+02
-101.8553
0.3074553E+01
0.4034841E+01
0.5072753E+01
52.6926
-0.3995165E+02
-0.1901498E+02
-0.6793472E+02
0.3353202E+02
0.7575963E+02
153.7294
0.3071991E+01
0.4027845E+01
0.5065635E+01
52.6677
-0.3681652E+02
0.1883203E+03
0.1270722E+02
-0.2898161E+02
0.3164501E+02
-66.3246
-0.3074553E+01
-0.4034841E+01
0.5072753E+01
-127.3074
0.3893358E+02
0.7018854E+02
0.7701321E+02
-0.7933219E+02
0.1105651E+03
-45.8498
0.1993835E+02
0.1935553E+02
0.2778803E+02
44.1502
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
-0.1993835E+02
-0.1935553E+02
0.2778803E+02
-135.8498
-0.6793472E+02
0.3353202E+02
0.7575963E+02
153.7294
-0.8427516E+01
-0.1707386E+02
0.1904047E+02
-116.2706
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
0.8427516E+01
0.1707386E+02
0.1904047E+02
63.7294
0.0000000E+00
0.0000000E+00
-0.3583831E+01
-0.1707245E+02
0.1744455E+02
-101.8553
-0.8145524E-01
0.5954579E-01
0.1008992E+00
143.8323
-0.3623353E+00
0.8020212E+00
0.1270722E+02
-0.2898161E+02
0.3164501E+02
-66.3246
0.8145524E-01
-0.5954579E-01
0.1008992E+00
-36.1677
0.1380401E+01
-0.8020212E+00
-0.3583831E+01
-0.1707245E+02
0.1744455E+02
-101.8553
-0.2993098E+01
-0.4094387E+01
0.5071749E+01
-126.1677
0.1270722E+02
-0.2898161E+02
0.3164501E+02
-66.3246
0.2993098E+01
0.4094387E+01
0.5071749E+01
53.8323
-0.4031398E+02
-0.6938652E+02
-0.3583831E+01
-0.1707245E+02
0.1744455E+02
-101.8553
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
0.0000000E+00
0.0000000E+00
-0.3583831E+01
-0.1707245E+02
0.1744455E+02
-101.8553
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.0000
0.0000000E+00
0.0000000E+00
0.8068964E+02
-0.4829158E+02
0.9403666E+02
-30.8999
-0.1068789E-03
-0.1289209E-01
0.1289254E-01
-90.4750
0.3069778E+00
0.5227098E+00
-0.8131719E+02
-0.4694850E+02
0.9389700E+02
-150.0000
0.1068789E-03
0.1289209E-01
0.1289254E-01
89.5250
-0.3069778E+00
0.5216655E+00
-0.8199473E+02
-0.4483268E+02
0.9345108E+02
-151.3313
-0.1103976E-01
0.6524933E-02
0.1282384E-01
149.4152
0.3063358E+00
0.5149760E+00
0.5749343E-14
0.9389700E+02
0.9389700E+02
90.0000
0.1103976E-01
-0.6524933E-02
0.1282384E-01
-30.5848
-0.3063358E+00
0.5183000E+00
0.6714130E+00
0.9269780E+02
0.9270023E+02
89.5850
0.1111270E-01
0.6417587E-02
0.1283267E-01
30.0064
0.8131719E+02
-0.4694850E+02
0.9389700E+02
-30.0000
-0.1111270E-01
-0.6417587E-02
0.1283267E-01
-149.9936
17 - 34
0.5576135E+03
0.0000000E+00
0.1009742E-27
0.1262177E-28
0.1458342E+04
0.2273737E-12
0.1737757E+01
0.2576097E+03
0.2117061E+01
0.2585088E+03
0.0000000E+00 0.0000000E+00
-0.1536192E+04 -0.1536192E+04
0.0000000E+00
0.0000000E+00
0.0000000E+00 0.0000000E+00
-0.7212496E+03 -0.7212496E+03
0.1018066E+01
0.0000000E+00
0.4031398E+02 0.0000000E+00
0.1821296E+02 -0.5117356E+02
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.1044375E+01
0.1110223E-15
0.1033276E+01
0.3011787E+00 -0.1110223E-15
0.5129070E+00 0.1034699E+01
-0.3011787E+00
0.5217923E+00

TOTAL NETWORK LOSS "PLOSS" BY SUMMING NODAL INJECTIONS =
0.5464331532E+04
NODE-K
NODE-M
I-REAL
I-IMAG
I-MAGN
DEGREES
FAULT
OPEN
OPEN
OPEN
SECB
CATH1
OPEN
OPEN
OPEN
SECA
CATH2
0.16873295E+02
0.15329059E+02
0.22796670E+02
42.2545
SECC
CATH3
OPEN
OPEN
OPEN
AN4
CATH4
0.24078446E+01
-0.20983161E+02
0.21120861E+02
-83.4539
AN5
CATH5
OPEN
OPEN
OPEN
AN6
CATH6
0.29476803E+01
0.34029172E+02
0.34156601E+02
85.0493
POWER
OPEN
OPEN
0.13409098E+04
OPEN
-0.99120545E+03
OPEN
-0.98793323E+03
REACTIVE
OPEN
OPEN
-0.12595678E+04
OPEN
-0.67237260E+03
OPEN
0.12053019E+04
SOLUTION AT NODES WITH KNOWN VOLTAGE.

NODES SHORTED TOGETHER BY SWITCHES ARE SHOWN AS A GROUP OF NAMES, WITH
THE PRINTED RESULT APPLYING TO THE COMPOSITE GROUP.
THE ENTRY 'MVA' IS SQRT(P**2 + Q**2) IN UNITS OF POWER,
WHILE 'P.F.' IS THE ASSOCIATED POWER FACTOR.
NODE
SOURCE NODE VOLTAGE
INJECTED SOURCE CURRENT
INJECTED SOURCE POWER
NAME
RECTANGULAR
POLAR
RECTANGULAR
POLAR
P AND Q
MVA AND P.F.
GENA
0.8131719E+02
-0.4694850E+02
0.9389700E+02
-30.0000
0.1927235E+02
-0.5674822E+01
0.2009047E+02
-16.4073
0.9167987E+03
-0.2216736E+03
0.9432174E+03
0.9719909E+00
GENB
-0.8131719E+02
-0.4694850E+02
0.9389700E+02
-150.0000
-0.1983435E+02
-0.4934019E+02
0.5317759E+02
-111.8997
0.1964661E+04
-0.1540506E+04
0.2496608E+04
0.7869320E+00
GENC
INITIAL FLUX IN COIL
WARNING. ASSUMPTION
WARNING. ASSUMPTION
0.5749343E-14
0.9389700E+02
'TRANA ' TO 'GENB '
'TRANB ' TO 'GENC '
THAT AC STEADY STATE
'TRANC ' TO 'GENA '
THAT AC STEADY STATE
0.9389700E+02
0.5620016E+00 0.5501788E+02
0.2582872E+04 0.2583007E+04
90.0000
0.5501501E+02
89.4147
0.2638513E+02 0.9999478E+00
= -0.53439E-03
= -0.55199E-01
HAS FUNDAMENTAL FREQUENCY ONLY IS QUESTIONABLE WITH PRECEDING FLUX OUTSIDE LINEAR REGION
= 0.55563E-01
HAS FUNDAMENTAL FREQUENCY ONLY IS QUESTIONABLE WITH PRECEDING FLUX OUTSIDE LINEAR REGION
1 1
_Request for output of all node voltages

36
These are ordered according to the
output-variable classes:
first
next
next
STEP
***
***
***
0
10
20
TIME
FAULT
LOADA
LOADB
AN6
AN5
AN4
CATH6
CATH5
CATH4
CATH3
CATH2
CATH1
TRANC
TRANB
TRANA
POLEB
POLEA
SECC
SECB
SECA
GENA
GENB
GENC
SECB
CATH1
SECA
CATH2
SECC
CATH3
AN4
CATH4
AN5
AN6
SECB
SECA
SECC
AN4
AN5
AN6
TRANA
CATH5
CATH6
CATH1
CATH2
CATH3
CATH4
CATH5
CATH6
GENB
PHASOR I(0) = 0.1687330E+02
SWITCH "SECA " TO "CATH2 " CLOSED AFTER 0.00000E+00 SEC.
PHASOR I(0) = 0.2407845E+01
SWITCH "AN4
" TO "CATH4 " CLOSED AFTER 0.00000E+00 SEC.
PHASOR I(0) = 0.2947680E+01
SWITCH "AN6
" TO "CATH6 " CLOSED AFTER 0.00000E+00 SEC.
0.000000-0.358383E+01-0.358383E+01 0.127072E+02-0.753039E+02-0.679347E+02-0.739543E+02-0.753039E+02 0.161380E+03-0.739543E+02
0.770132E+02 0.161380E+03 0.770132E+02 0.671413E+00-0.819947E+02 0.806896E+02-0.679347E+02 0.770132E+02-0.799739E+02
-0.826731E+02 0.161380E+03 0.813172E+02-0.813172E+02 0.574934E-14-0.159686E+03 0.000000E+00-0.156987E+03 0.000000E+00
-0.229314E+03 0.000000E+00 0.000000E+00 0.168733E+02 0.000000E+00 0.240784E+01 0.000000E+00 0.294768E+01-0.106879E-03
0.000020-0.272194E+01-0.272194E+01 0.141467E+02-0.726171E+02-0.695333E+02-0.781821E+02-0.726171E+02 0.161311E+03-0.781821E+02
0.809010E+02 0.161311E+03 0.809010E+02-0.398955E+01-0.796339E+02 0.830141E+02-0.695333E+02 0.809010E+02-0.868310E+02
-0.757010E+02 0.161311E+03 0.835734E+02-0.788556E+02-0.471779E+01-0.156602E+03 0.000000E+00-0.167732E+03 0.000000E+00
-0.230844E+03 0.000000E+00 0.000000E+00 0.160820E+02 0.000000E+00 0.345955E+01 0.000000E+00 0.123354E+01 0.540874E-03
DIODE 'AN6
' TO 'CATH6 ' OPENING AFTER 0.40000E-04 SEC.
0.000040-0.185269E+01-0.185269E+01 0.155508E+02-0.697508E+02-0.709561E+02-0.822095E+02-0.697508E+02 0.160834E+03-0.822095E+02
0.845844E+02 0.160834E+03 0.845844E+02-0.863199E+01-0.770835E+02 0.851288E+02-0.709561E+02 0.845844E+02-0.934629E+02
-0.685455E+02 0.160834E+03 0.856185E+02-0.761948E+02-0.942366E+01-0.153130E+03 0.000000E+00-0.178047E+03 0.000000E+00
-0.231791E+03 0.000000E+00 0.000000E+00 0.152500E+02 0.000000E+00 0.450137E+01 0.000000E+00-0.482105E+00 0.118726E-02
0.000060-0.978850E+00-0.978850E+00 0.169154E+02-0.724163E+02-0.724163E+02-0.861725E+02-0.565515E+02 0.159952E+03-0.861725E+02
0.880541E+02 0.159952E+03 0.880541E+02-0.132952E+02-0.719961E+02 0.870285E+02-0.724163E+02 0.880541E+02-0.999287E+02
-0.565515E+02 0.159952E+03 0.874472E+02-0.733415E+02-0.141057E+02-0.144606E+03 0.000000E+00-0.187983E+03 0.000000E+00
-0.232368E+03-0.158647E+02 0.000000E+00 0.143796E+02 0.000000E+00 0.550249E+01 0.000000E+00 0.000000E+00 0.183065E-02
0.000080-0.102611E+00-0.102611E+00 0.182371E+02-0.740715E+02-0.740715E+02-0.901191E+02-0.542315E+02 0.158665E+03-0.901191E+02
0.913013E+02 0.158665E+03 0.913013E+02-0.179296E+02-0.716448E+02 0.887084E+02-0.740715E+02 0.913013E+02-0.106167E+03
-0.542315E+02 0.158665E+03 0.890551E+02-0.703030E+02-0.187522E+02-0.145533E+03 0.000000E+00-0.197468E+03 0.000000E+00
-0.232737E+03-0.198400E+02 0.000000E+00 0.134728E+02 0.000000E+00 0.641904E+01 0.000000E+00 0.000000E+00 0.246941E-02
0.000100 0.576538E+00 0.576538E+00 0.193150E+02-0.758555E+02-0.758555E+02-0.939890E+02-0.410482E+02 0.156978E+03-0.939890E+02
0.943176E+02 0.156978E+03 0.943176E+02-0.225210E+02-0.657383E+02 0.901642E+02-0.758555E+02 0.943176E+02-0.112122E+03
-0.410482E+02 0.156978E+03 0.904381E+02-0.670868E+02-0.233512E+02-0.135366E+03 0.000000E+00-0.206440E+03 0.000000E+00
-0.232833E+03-0.348073E+02 0.000000E+00 0.125320E+02 0.000000E+00 0.725339E+01 0.000000E+00 0.000000E+00 0.310194E-02
0.000200-0.265434E+00-0.265434E+00 0.197264E+02-0.861164E+02-0.861164E+02-0.111791E+03-0.608845E+01 0.142711E+03-0.111791E+03
0.105682E+03 0.142711E+03 0.105682E+03-0.444093E+02-0.499848E+02 0.939727E+02-0.861164E+02 0.105682E+03-0.137466E+03
-0.608845E+01 0.142711E+03 0.938764E+02-0.486412E+02-0.452352E+02-0.111771E+03 0.000000E+00-0.243149E+03 0.000000E+00
-0.228827E+03-0.800279E+02 0.000000E+00 0.740568E+01 0.000000E+00 0.102699E+02 0.000000E+00 0.000000E+00 0.611584E-02
DIODE 'SECA ' TO 'CATH2 ' OPENING AFTER 0.34000E-03 SEC.
0.000400-0.853686E+01-0.853686E+01 0.989650E+01-0.107019E+03-0.107019E+03-0.134234E+03 0.726611E+02 0.850994E+02-0.134234E+03
0.109187E+03 0.109187E+03 0.109187E+03-0.787552E+02-0.527988E+01 0.821893E+02-0.107019E+03 0.109187E+03-0.161449E+03
0.726611E+02 0.850994E+02 0.832119E+02-0.393200E+01-0.792799E+02-0.365258E+02-0.240875E+02-0.270636E+03 0.000000E+00
-0.192119E+03-0.179680E+03 0.000000E+00 0.000000E+00 0.000000E+00 0.108859E+02 0.000000E+00 0.000000E+00 0.140811E-01
DIODE 'SECB ' TO 'CATH1 ' CLOSING AFTER 0.50000E-03 SEC.
17 - 35

30 0.000600-0.121705E+02-0.121705E+02-0.105384E+01-0.117206E+03-0.117206E+03-0.131462E+03 0.133784E+03 0.816423E+01-0.131462E+03
0.105099E+03 0.105099E+03 0.133784E+03-0.937305E+02 0.409058E+02 0.509361E+02-0.117206E+03 0.105099E+03-0.145718E+03
0.133784E+03 0.816423E+01 0.519618E+02 0.417499E+02-0.937117E+02 0.000000E+00-0.969352E+02-0.250817E+03 0.000000E+00
-0.125370E+03-0.250990E+03 0.573686E+01 0.000000E+00 0.000000E+00 0.570237E+01 0.000000E+00 0.000000E+00 0.320734E-01
DIODE 'AN4
40 0.000800 0.204919E+01 0.204919E+01 0.185997E+00-0.108373E+03-0.108373E+03-0.108373E+03 0.160780E+03-0.712946E+02-0.918810E+02
0.109528E+03 0.109528E+03 0.160780E+03-0.844873E+02 0.764559E+02 0.683081E+01-0.108373E+03 0.109528E+03-0.918810E+02
0.160780E+03-0.712946E+02 0.785710E+01 0.771034E+02-0.849605E+02 0.000000E+00-0.180822E+03-0.201409E+03-0.164916E+02
-0.370780E+02-0.269152E+03 0.102503E+02 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.179946E-01
DIODE 'AN5
' TO 'CATH5 ' CLOSING AFTER 0.90000E-03 SEC.
50 0.001000 0.191835E+02 0.191835E+02-0.807289E+00-0.940707E+02-0.111849E+03-0.940707E+02 0.148195E+03-0.111849E+03-0.160599E+02
0.110128E+03 0.110128E+03 0.148195E+03-0.547147E+02 0.931876E+02-0.372205E+02-0.940707E+02 0.110128E+03-0.160599E+02
0.148195E+03-0.129627E+03-0.381914E+02 0.933826E+02-0.551913E+02 0.000000E+00-0.239754E+03-0.126187E+03-0.780108E+02
0.000000E+00-0.242266E+03 0.761347E+01 0.000000E+00 0.000000E+00 0.000000E+00 0.711120E+01 0.000000E+00 0.751555E-02
DIODE 'SECB ' TO 'CATH1 ' OPENING AFTER 0.12200E-02 SEC.
DIODE 'SECC ' TO 'CATH3 ' CLOSING AFTER 0.12800E-02 SEC.
DIODE 'AN6
DIODE 'AN5
100 0.002000 0.806006E+02 0.806006E+02-0.733875E+02-0.110154E+03-0.636715E+02-0.636715E+02-0.110154E+03 0.474938E+02 0.108800E+03
0.108800E+03 0.707438E+02 0.707438E+02 0.892929E+02-0.685641E+02-0.209035E+02-0.636715E+02 0.707438E+02 0.108800E+03
-0.156636E+03 0.474938E+02-0.195223E+02-0.697791E+02 0.893014E+02-0.227380E+03-0.232500E+02 0.000000E+00-0.172471E+03
-0.111165E+03 0.000000E+00 0.000000E+00 0.000000E+00 0.761119E+01 0.000000E+00 0.000000E+00 0.185929E+02-0.249477E-01
DIODE 'SECA ' TO 'CATH2 ' CLOSING AFTER 0.20600E-02 SEC.
DIODE 'SECC ' TO 'CATH3 ' OPENING AFTER 0.21600E-02 SEC.
DIODE 'AN4
DIODE 'AN6
150 0.003000 0.128808E+03 0.128808E+03-0.142567E+03-0.787354E+02-0.787354E+02-0.118242E+03 0.106622E+03 0.515522E+02-0.118242E+03
0.657217E+02 0.657217E+02 0.106622E+03-0.886284E+02 0.184181E+02 0.704238E+02-0.787354E+02 0.657217E+02-0.157749E+03
0.106622E+03 0.515522E+02 0.697791E+02 0.195223E+02-0.893014E+02 0.000000E+00-0.141695E+02-0.223471E+03 0.000000E+00
-0.130288E+03-0.185358E+03 0.818012E+01 0.000000E+00 0.000000E+00 0.158028E+02 0.000000E+00 0.000000E+00 0.207336E-01
DIODE 'AN4
DIODE 'AN5
200 0.004000 0.148416E+03 0.148416E+03-0.163419E+03-0.114316E+03-0.123230E+03-0.114316E+03-0.140728E+02-0.123230E+03 0.147083E+03
0.147083E+03 0.100753E+03 0.100753E+03 0.544519E+02 0.396517E+02-0.936654E+02-0.114316E+03 0.100753E+03 0.147083E+03
-0.140728E+02-0.132143E+03-0.933826E+02 0.381914E+02 0.551913E+02-0.114826E+03-0.232896E+03 0.000000E+00-0.261399E+03
0.000000E+00-0.100243E+03 0.000000E+00 0.000000E+00 0.926607E+01 0.000000E+00 0.356532E+01 0.000000E+00-0.768386E-02
DIODE 'AN5
DIODE 'AN6
DIODE 'AN6
250 0.005000 0.146420E+03 0.146420E+03-0.147567E+03-0.142525E+03-0.142525E+03-0.142525E+03-0.786609E+02 0.162349E+03-0.796512E+02
0.142092E+03 0.162349E+03 0.142092E+03 0.835646E+00-0.799934E+02 0.811794E+02-0.142525E+03 0.142092E+03-0.796512E+02
-0.786609E+02 0.162349E+03 0.813172E+02-0.813172E+02-0.261697E-11-0.220753E+03 0.000000E+00-0.221743E+03-0.628742E+02
-0.304875E+03-0.638645E+02 0.000000E+00 0.405156E+01 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00-0.101526E-03
DIODE 'AN4
DIODE 'AN4
300 0.006000 0.135243E+03 0.135243E+03-0.134568E+03-0.157139E+03-0.157139E+03-0.157139E+03 0.147937E+03-0.133024E+03-0.183443E+02
0.157838E+03 0.157838E+03 0.157838E+03-0.558558E+02 0.930548E+02-0.389142E+02-0.157139E+03 0.157838E+03-0.183443E+02
0.147937E+03-0.133024E+03-0.381914E+02 0.933826E+02-0.551913E+02-0.990163E+01-0.290863E+03-0.176183E+03-0.138794E+03
-0.241143E+02-0.305075E+03 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.750712E-02
DIODE 'AN5
DIODE 'AN5
DIODE 'AN6
350 0.007000 0.136143E+03 0.136143E+03-0.133988E+03-0.152276E+03-0.146004E+03-0.146004E+03-0.152276E+03 0.514719E+02 0.108110E+03
0.147532E+03 0.147532E+03 0.147532E+03 0.889495E+02-0.695221E+02-0.189141E+02-0.146004E+03 0.147532E+03 0.108110E+03
-0.158549E+03 0.514719E+02-0.195223E+02-0.697791E+02 0.893014E+02-0.306081E+03-0.960601E+02-0.394222E+02-0.254114E+03
-0.197476E+03 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.250902E+01-0.260375E-01
DIODE 'AN6
DIODE 'AN4
400 0.008000 0.139057E+03 0.139057E+03-0.144715E+03-0.137684E+03-0.137684E+03-0.148319E+03 0.107753E+03 0.512870E+02-0.148319E+03
0.132324E+03 0.132324E+03 0.132324E+03-0.892299E+02 0.189825E+02 0.702916E+02-0.137684E+03 0.132324E+03-0.158953E+03
0.107753E+03 0.512870E+02 0.697791E+02 0.195223E+02-0.893014E+02-0.245706E+02-0.810365E+02-0.291277E+03 0.000000E+00
-0.188971E+03-0.245437E+03 0.000000E+00 0.000000E+00 0.000000E+00 0.425387E+01 0.000000E+00 0.000000E+00 0.209734E-01
DIODE 'AN4
DIODE 'AN5
DIODE 'AN5
450 0.009000 0.139656E+03 0.139656E+03-0.149528E+03-0.138485E+03-0.138485E+03-0.138485E+03-0.192647E+02-0.130079E+03 0.146931E+03
0.146931E+03 0.131256E+03 0.131256E+03 0.543765E+02 0.370554E+02-0.926333E+02-0.138485E+03 0.131256E+03 0.146931E+03
-0.192647E+02-0.130079E+03-0.933826E+02 0.381914E+02 0.551913E+02-0.150520E+03-0.261335E+03 0.000000E+00-0.285415E+03
-0.840516E+01-0.119220E+03 0.000000E+00 0.000000E+00 0.313501E+01 0.000000E+00 0.000000E+00 0.000000E+00-0.769615E-02
DIODE 'AN6
DIODE 'AN6
500 0.010000 0.141410E+03 0.141410E+03-0.142699E+03-0.141958E+03-0.141958E+03-0.141958E+03-0.834915E+02 0.162269E+03-0.823104E+02
0.142310E+03 0.162269E+03 0.142310E+03-0.492786E+00-0.824084E+02 0.811390E+02-0.141958E+03 0.142310E+03-0.823104E+02
-0.834915E+02 0.162269E+03 0.813172E+02-0.813172E+02 0.166104E-10-0.225802E+03 0.000000E+00-0.224621E+03-0.596472E+02
-0.304227E+03-0.584660E+02 0.000000E+00 0.399182E+01 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00-0.105816E-03
----- "MEMSAV = 1
REPRESENTS REQUEST FOR TABLE DUMPING ON DISK.
17 - 36

:\EMTPV3\BENCH\EMTP\dc32.bin
SUCCESSFUL SAVING OF EMTP TABLES AS FILE :

D:\EMTPV3\BENCH\EMTP\dc32.bin
SIZE OF COMMON LABCOM LTLABL =
6000
__Maxima and minima which occurred during the simulation follow.

The order and column positioning are the same as for the regular printed output vs. time.
STEP
TIME
FAULT
LOADA
LOADB
AN6
AN5
AN4
CATH6
CATH5
CATH4
CATH3
CATH2
CATH1
TRANC
TRANB
TRANA
POLEB
POLEA
SECC
SECB
SECA
GENA
GENB
GENC
SECB
CATH1
SECA
CATH2
SECC
CATH3
AN4
CATH4
AN5
AN6
SECB
SECA
SECC
AN4
AN5
AN6
TRANA
CATH5
CATH6
CATH1
CATH2
CATH3
CATH4
CATH5
CATH6
GENB
VARIABLE MAXIMA :
0.161778E+03 0.161778E+03 0.202728E+02-0.557365E+02-0.576214E+02-0.609992E+02 0.162583E+03 0.162349E+03 0.162368E+03
0.162368E+03 0.162349E+03 0.162583E+03 0.942715E+02 0.942204E+02 0.945734E+02-0.576214E+02 0.157838E+03 0.162368E+03
0.162583E+03 0.162349E+03 0.938764E+02 0.938962E+02 0.938896E+02 0.675797E+01 0.102225E+02 0.799664E+01 0.545240E+01
0.441774E+01 0.557027E+01 0.158740E+02 0.191583E+02 0.168070E+02 0.180334E+02 0.134790E+02 0.188458E+02 0.325985E-01
TIMES OF MAXIMA :
0.432000E-02 0.432000E-02 0.140000E-03 0.258000E-02 0.262000E-02 0.186000E-02 0.584000E-02 0.500000E-02 0.666000E-02
0.666000E-02 0.500000E-02 0.584000E-02 0.938000E-02 0.854000E-02 0.522000E-02 0.262000E-02 0.600000E-02 0.666000E-02
0.584000E-02 0.500000E-02 0.770000E-02 0.604000E-02 0.938000E-02 0.810000E-02 0.206000E-02 0.894000E-02 0.526000E-02
0.334000E-02 0.936000E-02 0.326000E-02 0.244000E-02 0.166000E-02 0.284000E-02 0.124000E-02 0.204000E-02 0.562000E-02
VARIABLE MINIMA :
-0.132335E+02-0.132335E+02-0.163730E+03-0.157139E+03-0.159691E+03-0.157263E+03-0.154313E+03-0.159941E+03-0.157263E+03
0.541934E+02 0.654657E+02 0.556600E+02-0.945019E+02-0.943138E+02-0.945070E+02-0.157139E+03 0.556600E+02-0.162506E+03
-0.162245E+03-0.162547E+03-0.938962E+02-0.938764E+02-0.938896E+02-0.307901E+03-0.319515E+03-0.314746E+03-0.314645E+03
-0.305068E+03-0.319399E+03-0.896250E+00-0.452485E+00-0.741820E+00-0.597760E+00-0.844743E+00-0.108747E+01-0.355220E-01
TIMES OF MINIMA :
0.560000E-03 0.560000E-03 0.398000E-02 0.600000E-02 0.624000E-02 0.546000E-02 0.710000E-02 0.616000E-02 0.546000E-02
0.216000E-02 0.206000E-02 0.224000E-02 0.564000E-02 0.480000E-02 0.646000E-02 0.600000E-02 0.224000E-02 0.542000E-02
0.708000E-02 0.624000E-02 0.896000E-02 0.980000E-02 0.562000E-02 0.706000E-02 0.624000E-02 0.544000E-02 0.666000E-02
0.502000E-02 0.584000E-02 0.372000E-02 0.340000E-03 0.216000E-02 0.816000E-02 0.172000E-02 0.258000E-02 0.688000E-02

** PLOT CARD.
0.100E+01 0.000E+00
4/11/97 10.17.30
PLOT TYPE
4
NODE NAMES
LOADA
0.100E+02
1 PRINTER PLOT
1 144 1. 0.0 10.
LOADA LOADB
1
LOADB
MILLISECONDS
( X 10**( 2) )
-1.637 -1.387
-1.137
-0.886
-0.636
-0.385
-0.135
0.115
0.366
0.616
0.867
1.117
1.367
1.618
----------------------------------------------------------------AA----BBBB-------------------------------------------------------AA
B
AAAA
BBBBB
AAA BBBBB
AAAAB
A**AAA
1.000
BBBBB
AAAAA
BBBBBBBB 1
AA
BBBBBBB
1
AA
BBBB
1
AAAAAAA
B
1
AAAAAAAAA
BBBBB
1
AAAAA
2.000
BBBBBBBBB
1
AA
BBBBBBB
1
AAA
BBB
1
AAAAAA
B
1
AAAAAAAAA
BBBB
1
AAAA
BBBBBB
1
A
BBBBBB
1
A
B
1
AAAA
BBBB
1
AAAAAAA
BB
1
AA
BBBB
1
AAA
BBBB
1
AAA
BBB
1
AAA
BBBB
1
AAA
17 - 37

BB
BBB
B
BB
BBB
BB
B
BB
B
B
6.000 BB
B
B
B
B
B
7.000 BB
BB
B
BB
B
BBB
BBB
BB
BB
BB
BB
BBB
BB
BBB
BB
BBB
BB
BB
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
AA
AAAA
AAA
AA
A
AA
AA
AA
AA
A
AA
A
AA
AA
A
AA
AA
A
AA
AAA
AA
AA
AA
AA
AAA
AA
AAA
AA
AAA
AA
AA
AAA
AAA
AA
A
END OF GRAPH.
BLANK CARD TERMINATING PLOT SPEC. CARDS.
1BLANK CARD ENDING PLOT CARDS
CORE STORAGE FIGURES FOR PRECEDING DATA CASE NOW COMPLETED. --------------------------------------A VALUE OF -9999 INDICATES DEFAULT, WITH NO FIGURE AVAILABLE.
SIZE LIST 1.
NUMBER OF NETWORK NODES.
SIZE LIST 2.
NUMBER OF NETWORK BRANCHES.
SIZE LIST 3.
NUMBER OF DATA VALUES IN R, L, C TABLES.
SIZE LIST 4.
NUMBER OF ENTRIES IN SOURCE TABLE.
SIZE LIST 5.
STORAGE FOR (Y) AND TRIANGULARIZED (Y).
NO. TIMES = 112
FACTORS =
46
SIZE LIST 6.
NUMBER OF ENTRIES IN SWITCH TABLE.
NO. FLOPS =
51
SIZE LIST 7.
NUMBER OF TOTAL DISTINCT ALPHANUMERIC (A6) PROGRAM NAMES
SIZE LIST 8.
NUMBER OF PAST HISTORY POINTS FOR DISTRIBUTED LINES.
SIZE LIST 9.
NUMBER OF NONLINEAR ELEMENTS.
SIZE LIST 10. NUMBER OF POINTS DEFINING NONLINEAR CHARACTERISTICS.
SIZE LIST 11. NUMBER OF BRANCH OR SELECTIVE-NODE-VOLTAGE OUTPUTS.
SIZE LIST 12. NUMBER OF OUTPUT QUANTITIES (LIMITED ONLY WHEN PRINTING MAX ABSOLUTE VALUES).
SIZE LIST 13. NUMBER OF 'WEIGHTING' FREQUENCY-DEPENDENT LINE MODES.
SIZE LIST 14. NUMBER OF CELLS USED TO STORE FREQ.-DEPENDENT NETWORK EQUIVALENTS
SIZE LIST 15. NUMBER OF CELLS USED FOR EXPONENTIAL-TAIL LINE-HISTORY STORAGE.
SIZE LIST 16. TOTAL NUMBER OF TYPE-59 S.M. MASSES.
SIZE LIST 17. NUMBER OF DYNAMIC SYNCHRONOUS MACHINES.
SIZE LIST 18. NUMBER OF BRANCH POWER-AND-ENERGY OUTPUTS.
SIZE LIST 19. FLOATING-POINT WORKING SPACE FOR ALL TACS ARRAYS.
SIZE LIST 20. RECURSIVE CONVOLUTION PARAMETER STORAGE FOR NON-COPIED BRANCH COMPONENTS.
SIZE LIST 21. TOTAL STORAGE CELLS FOR MODAL-PHASE TRANSFORMATION MATRICES.
SIZE LIST 22. NUMBER OF CELLS FOR CONVOLUTION HISTORY.
SIZE LIST 23. GIANT ARRAYS FOR RENUMBERING AND STEADY-STATE SOLUTION CALCULATIONS.
SIZE LIST 24. NUMBER OF PHASES OF COMPENSATION, BASED ON MAXIMUM NODES.
SIZE LIST 25. FLOATING-POINT WORKING SPACE FOR U.M. ARRAYS.
SIZE LIST 26. SQUARE OF MAXIMUM NUMBER OF COUPLED PHASES.
SIZE LIST 27. NUMBER OF CELLS FOR FREQ. DEP. EQUIVALENTS.
ADDITIONAL STORAGE FIGURES FOR FREQUENCY-DEPENDENT LINE/CABLE MODELS OF TYPES -3 AND -4
SIZE LIST 28. MAXIMUM NUMBER OF FDQ LINE/CABLE DEVICES WITH/WITHOUT CONSTANT Q.
SIZE LIST 29. NUMBER OF FDQ LINE/CABLE DEVICES WITH CONSTANT Q.
SIZE LIST 30. MAXIMUM NUMBER OF PHASES ALLOWED IN A GIVEN FDQ LINE/CABLE DEVICE
SIZE LIST 31. MAXIMUM ORDER OF ANY RATIONAL-FUNCTIONS APPROXIMATION.
SIZE LIST 32. AVERAGE ORDER OF ALL RATIONAL-FUNCTIONS APPROXIMATIONS.
SIZE LIST 33. AVERAGE NUMBER OF HISTORY TERMS FOR A GIVEN FDQ MODE/BRANCH.
SIZE LIST 34. AVERAGE NUMBER OF PHASES FOR A GIVEN FDQ DEVICE
TIMING FIGURES (DECIMAL) CHARACTERIZING CASE SOLUTION SPEED. ------------------------------------DATA INPUT, SORTING, AND RENUMBERING (PRE STEADY STATE STUFF) .....
STEADY-STATE (S.S.) SOLUTION CALCULATIONS .....
POST-S.S. TO PRE-INTEGRATION-SETUP CALCULATIONS .....
INTEGRATION CALCULATION (TIME IN TIME-STEP LOOP) .....
COMPUTER TIME IN PLOTTING OR STATISTICS TERMINATION OVERLAY .....
'DELTAT'-CHANGE RESTART TIME .......
TOTALS
ELETROMAGNETIC TRANSIENT PROGRAM (EMTP) - DCG/EPRI VERSION 3.0F MS WINDOWS TRANSLATION

RELEASE 01.08.96
User Support & Maintenance Centre - Ontario Hydro, Canada.
17 - 38
PRESENT
FIGURE
24
28
17
3
112
7
2
7
3
4
13
36
0
0
0
0
0
0
137
0
8
0
59
0
-9999
-9999
-9999
PROGRAM
LIMIT
2002
3900
5000
1000
15000
1500
5000
20000
800
2000
500
500
1
180000
1
400
50
500
200000
30000
3000
2000
140000
100
15000
20000
100
(NAME)
(LBUS)
(LBRNCH)
(LDATA)
(LEXCT)
(LYMAT)
(LSWTCH)
(LSIZE7)
(LPAST)
(LNONL)
(LCHAR)
(LSMOUT)
(LSIZ12)
(LFDEP)
(L27DEP)
(LTAILS)
(LIMASS)
(LSYN)
(MAXPE)
(LTACST)
(LFSEM)
(LFD)
(LHIST)
(LSIZ23)
(NCOMP)
(LSPCUM)
(LSIZ26)
(LSIZ27)
120 (NFDUNT)
1 (NFDUCQ)
12 (NFDPH)
50 (NFDPOL)
35 (NFDPAV)
100 (NFDHAV)
6 (NFDPHA)
CP SEC
I/O SEC
SUM SEC
7.090
0.000
7.090
0.870
0.000
0.870
0.500
0.000
0.500
1.590
0.000
1.590
28.290
0.000
28.290
0.000
0.000
0.000
----------------------------38.340
0.000
38.340

phone:
fax:
e-mail:
(800) 245-8791 or (416) 592-4793

(416) 592-9448
technical.support@emtp96.com
RUN DATE (MM/DD/YY) AND TIME (HH.MM.SS.)= 4/11/97 10.18. 8
NAME OF PLOT DATA FILE (IF ANY) =D:\EMTPV3\BENCH\EMTP\dc32.p14
--------------------------------------------------+------------------------------------------------------------------------------DESCRIPTIVE INTERPRETATION OF NEW-CASE INPUT DATA 1 INPUT DATA CARD IMAGES PRINTED BELOW, ALL 80 COLUMNS, CHARACTER BY CHARACTER.
0
1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
0
--------------------------------------------------+------------------------------------------------------------------------------MARKER CARD PRECEDING NEW DATA CASE.
1BEGIN NEW DATA CASE
1BLANK
___Normal end of emtp run__
17.5 EMTP Error-Message Terminations

All sorts of user data errors or misunderstandings can produce situations where the EMTP will
terminate execution of a data case being solved by means of the printing out of an appropriate
error message. A sample is shown below, for which the following points might be made:
(A)
"KILL CODE NUMBER" is just a characteristic number assigned to the particular

error message in question; it has no special significance to the user, as such.
Each different error message has such a unique number associated with it.
(B)
"OVERLAY NUMBER" refers to the overlay number where the error condition
has been detected. This localizes the point of trouble to within one of the
FORTRAN modules of that overlay:
Number "-1" is for the "main" code (principal mode "MAIN00" which is always
in core.
Number "0" is for the first primary level overlay (principal module "MAIN10"),
which calls all solution overlays (numbers 120).
Positive integers are associated with overlays whose principal modules have
names beginning with the four letters "OVER", and which end with the one or
two digits of the overlay number. For example, overlay 13 has principal
module "OVER13".
The EMTP FORTRAN is ordered from front to back in order of increasing
overlay number; this makes it easy to locate the overlay in question.
(C)
"NEARBY STATEMENT NO." is the statement number of a FORTRAN statement

near the point in the code where the error condition was detected. This is
useful in tracing program bugs, when reference to the FORTRAN coding is
required. This statement number should always be unique. Having visually
located the statement number in question (within a module of the overlay in
question) control has then passed through a nearby statement which stores this
17 - 39

number in LSTAT(19). Shortly before or after this, the kill code number should
also be assigned, to variable "KILL".
(D)
Following the above three numbers is a brief explanation of conditions which

led to program termination.
(E)
The user will note the "case-recovery" attempt by the EMTP, after the error
message has been completed. Upon the encounter of a "BEGIN NEW DATA
CASE" record, case summary statistics are being printed.
(F)
Core-storage figures which follow the error message may not all (or even
"any", in extreme cases) be valid. A value of "-9999" will be found wherever
the figure in question is not known by the EMTP at this point of the execution
where the decision to kill the case has been made.
----------------------------------------------------------------------------------------------------------ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/
ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/
-----------------------------------------------------------------------------------------------------------the 80-column card image
that caused an error is the last one printed out before this
termination message.
a copy follows....,
KILL CODE NUMBER
OVERLAY NUMBER
NEARBY STATEMENT NO.
191
13
4590
THE EMTP FINDS THE USER'S DATA CASE TO BE OBNOXIOUSLY DEGENERATE, AND REFUSES TO CONTINUE WITH THE
SIMULATION.
THE ELECTRIC NETWORK HAS NO SOURCES AND NO DYNAMIC SYNCHRONOUS MACHINES, SO ONLY A NATURAL
(UNFORCED) SOLUTION IS CALLED FOR.
BUT THE USER HAS FAILED TO INPUT ANY NONZERO INITIAL CONDITIONS.
HENCE THE
SOLUTION WILL BE IDENTICALLY ZERO FOR ALL TIME.
THERE IS NO NEED TO CONTINUE WITH THE SOLUTION, THEN.
----------------------------------------------------------------------------------------------------------ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/
ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/ERROR/
-----------------------------------------------------------------------------------------------------------CARD IGNORED IN SEARCH FOR NEW-CASE BEGINNING.
CARD IGNORED IN SEARCH FOR NEW-CASE BEGINNING.
BEGIN SUPPRESSION OF SKIPPED-RECORD PRINTOUT.
**** ****
DATA CRISIS.
LAST LINE HAS BEEN READ.
NUMCRD =
93
ERROR STOP IN "STOPTP".
NCHAIN, LASTOV =
55
54
LAST-READ CARD IMAGE ABUFF FOLLOWS ....
___Catastrophic end of emtp run__
CORE STORAGE FIGURES FOR PRECEDING DATA CASE NOW COMPLETED. --------------------------------------A VALUE OF -9999 INDICATES DEFAULT, WITH NO FIGURE AVAILABLE.
SIZE LIST 1.
NUMBER OF NETWORK NODES.
SIZE LIST 2.
NUMBER OF NETWORK BRANCHES.
SIZE LIST 3.
NUMBER OF DATA VALUES IN R, L, C TABLES.
SIZE LIST 4.
NUMBER OF ENTRIES IN SOURCE TABLE.
SIZE LIST 5.
STORAGE FOR (Y) AND TRIANGULARIZED (Y).
NO. TIMES =
1
FACTORS =
2
SIZE LIST 6.
NUMBER OF ENTRIES IN SWITCH TABLE.
NO. FLOPS =
1
SIZE LIST 7.
NUMBER OF TOTAL DISTINCT ALPHANUMERIC (A6) PROGRAM NAMES
SIZE LIST 8.
NUMBER OF PAST HISTORY POINTS FOR DISTRIBUTED LINES.
SIZE LIST 9.
NUMBER OF NONLINEAR ELEMENTS.
SIZE LIST 10. NUMBER OF POINTS DEFINING NONLINEAR CHARACTERISTICS.
SIZE LIST 11. NUMBER OF BRANCH OR SELECTIVE-NODE-VOLTAGE OUTPUTS.
SIZE LIST 12. NUMBER OF OUTPUT QUANTITIES (LIMITED ONLY WHEN PRINTING MAX ABSOLUTE VALUES).
SIZE LIST 13. NUMBER OF 'WEIGHTING' FREQUENCY-DEPENDENT LINE MODES.
SIZE LIST 14. NUMBER OF CELLS USED TO STORE FREQ.-DEPENDENT NETWORK EQUIVALENTS
SIZE LIST 15. NUMBER OF CELLS USED FOR EXPONENTIAL-TAIL LINE-HISTORY STORAGE.
SIZE LIST 16. TOTAL NUMBER OF TYPE-59 S.M. MASSES.
SIZE LIST 17. NUMBER OF DYNAMIC SYNCHRONOUS MACHINES.
SIZE LIST 18. NUMBER OF BRANCH POWER-AND-ENERGY OUTPUTS.
SIZE LIST 19. FLOATING-POINT WORKING SPACE FOR ALL TACS ARRAYS.
17 - 40
PRESENT
FIGURE
4
2
2
1
7
1
20
0
0
0
6
9
0
0
0
0
0
0
188371
PROGRAM
LIMIT
2002
3900
5000
1000
15000
1500
5000
20000
800
2000
500
500
1
180000
1
400
50
500
200000
(NAME)
(LBUS)
(LBRNCH)
(LDATA)
(LEXCT)
(LYMAT)
(LSWTCH)
(LSIZE7)
(LPAST)
(LNONL)
(LCHAR)
(LSMOUT)
(LSIZ12)
(LFDEP)
(L27DEP)
(LTAILS)
(LIMASS)
(LSYN)
(MAXPE)
(LTACST)
Section 18
EMTP Data Modules
18.1
Introduction ......................................................................................... 18-1
18.2
Syntax of EDM ................................................................................... 18-2

18.2.1
Preamble or Declaration ................................................. 18-2
18.2.2
Body ............................................................................... 18-5
18.2.3
Recommendations .......................................................... 18-9
18.3
Conversion from a Module to an Include File .................................... 18-9
18.4
Advanced Features .............................................................................. 18-11

18.4.1
Calculation Mode ........................................................... 18-11
18.4.2
Default Value Specification ........................................... 18-12
18.4.3
Language Syntax for the Calculation Section of EDM .. 18-13
Section 18
EMTP Data Modules
This chapter describes EMTP Data Modules (EDM). EMTP data modules allow encapsulation of
complex devices or subnetworks into reusable modules with features such as argument
substitution and automatic generation of internal node names.
18.1 Introduction
EMTP Data Modules (EDM) allow an EMTP user to group or encapsulate input data into modules.
A module has two parts: a declaration section and a body. In the declaration section, external and
internal variables are specified. In the body, EMTP data cards describing a given model are
collected in data groups such as branch, source or TACS data. The data groups within a module
need not follow the normal EMTP input data structure which requires that TACS data precede
branch data, and so forth. The data cards will however be sorted in the right order by EDM before
all input data are submitted to the EMTP for simulation.
Starting in version 3, EDM allows simple calculations to be performed as part of a module. To this
effect, the declaration section has been enhanced and a calculation section has been added to the
module structure.
Advantages of EDM are summarized below:
(A)
Natural Grouping of Input Data - EDM helps to organize an input data file
for readability. For example, to create an EMTP input data deck to simulate a
two-terminal DC system is a massive task. This task can be divided into a
number of subtasks, such as a network module, a converter module, and a
control module with the help of EDM. Each of these modules can be further
subdivided into sub-modules. Linking of sub-modules of a module is
achieved by the "nesting" concept in EDM. Up to six levels of nesting are
allowed. A module may be built using either models from branch data or from
source data or models from TACS, etc. It does not require hierarchical datagroup order, i.e., TACS data must precede branch data, and so forth. EDM sorts
each module to form hierarchical data groups. Furthermore, the modular
structure allows each module to be tested separately before the final assembly.
This reduces errors and facilitates debugging.
(B)
Argument Substitution - EDM provides the means to substitute variables in a

module. These variables can be either character strings, i.e., a type ARG, or
18 - 1
EMTP Data Modules

numeric strings, i.e., a type ARG with NUM qualifier, or dummy variables,
i.e., a type DUM. These type declarations are in Section 18.2.1 below.
(C)
Transportability and Reliability - Each module resides in a different

computer file. Once a module is tested and working, it requires no
maintenance. The possibility of an unknown error to creep in is reduced.
These modules can be reused for a number of studies of a similar nature.
(D)
Ease of Use - Each module can be easily accessed through $INCLUDE cards,
and can be called many times within a data case.
Creating and using EDM is a three-step operation. First, the user writes modules with proper
syntax. Second, these modules are converted to an Include-File" format? by running the EMTP
using MODULE command. Third, these Include files are called from an input file using
$INCLUDE cards. The first two steps are required only once, unless the module needs to be
modified.
18.2 Syntax of EDM

A module has essentially two parts, a preamble and a body:
18.2.1
Preamble or Declaration
In this section discussion will be limited to the preamble part of a module. A simple example
(Figure 18.1) will be used to illustrate the syntax of the preamble. A module preamble can have
three types of declarations, namely, ARG, NUM and DUM. Type ARG means arguments, type
NUM means numeric constants, and type DUM refers to internal variables or dummy variables.
These type declarations are specified is the following order: type ARG must be declared first,
followed by type NUM, with type DUM last.
18 - 2
EMTP Data Modules
C MODULE ACFAULT.MOD
ARG FAULT,EBUS##, - ; electrical nodes
R_____,LPHASE,RNEUTL - ; numeric constants
?,@ ; output requests
NUM R_____ ,LPHASE,RNEUTL
DUM NEUTRL ; internal variables
C
C BELOW IS THE BODY OF THE MODULE
/BRANCH
FAULTANEUTRL
R______LPHASE
FAULTBNEUTRLFAULTANEUTRL
FAULTCNEUTRLFAULTANEUTRL
NEUTRL
RNEUTL
/SWITCH
EBUS FAULTA
/ENDMODULE
$EOF
C End Body
Figure 18.1: Sample Module

Syntax of type declarations is shown below.
Type
name,name,name, name,name
; Comment
; Comment
For example, valid declarations are

ARG
NUM
DUM
FAULT,EBUS##, R___,LPHASE,RNEUTL, ?,@

R___,LPHASE,RNEUTL
NEUTRL
; Electrical Nodes
; Numeric Constants
; Output Requests
In the above examples, there are special characters, such as #, -, ?, @, ; and -. The
meaning of these special symbols is as follows:
-
A continuation character for type declaration.
To indicate that rest of the line is a comment.
To indicate imbedded blanks, i.e., EBUS##?

means #EBUS .
To indicate imbedded blanks in a numeric field,

i.e., R___? means R .
18 - 3
EMTP Data Modules

?,@
To change the branch output request in a module,

i.e., one-character substitution in the 80th column.
The total number of arguments and dummy variables substitution in a module is installation
dependent.
The following indicates the valid data types associated with EDM.
(A)
Type ARG
All external variables (Numeric or Character string) replacing either electrical nodes or
TACS variables or numeric fields must be declared as ARG. Furthermore, variables
representing numeric fields must be again declared as type NUM (see Point B below).
The variable type ARG is neither right nor left justified: it replaces the pattern as
given in the argument list. A variable can be padded with blanks using the special
character #.
Since the program replaces a pattern and not a 6 character wide argument, it can be
used to generate 3-phase node names. For example (refer to Figure 18.1), the
argument FAULT is a part of names referring to nodes FAULTA, FAULTB, and
FAULTC. Therefore, with one 5-character variable, the user can generate three node
names. This scheme has been tested but the pattern must be unique to the nodes which
one desires to be altered; it would fail otherwise. Users are advised to be cautious.
(B)
Type NUM
The type NUM is a subset of type ARG. This declaration informs the program to
expect a numeric string instead of an alphanumeric string. These variables replace
numeric fields in the EMTP data. When specifying a numeric variable, users must refer
to an actual EMTP card image and specify the width of the variable to be the full width
of the numeric field. If the field width required is 10 characters long, then the variable
representing that field must have a 10 character width. Users may use the special
character "_" to indicate imbedded blanks in a numeric field. In an $Include card,
users may use a numeric string of width less than the specified field width. If this
occurs, the program right justifies the supplied numeric string in the specified field.
(C)
Type DUM
These dummy variables are internal variables of a module. They are replaced uniquely
by the program. Variables of type DUM:
(i)
must have a SIX character width, and
(ii)
the first character should be a letter
18 - 4
EMTP Data Modules

An internal variable is replaced with a unique name created by the program. The
unique name is composed by suffixing a number (from 0000? to 9999) to the two
character word ZZ. For example, the first dummy variable will be named as
ZZ0000 and 10000th dummy variable will be named as ZZ9999. When the
program exhausts the first set of 10000 dummy variables, it starts with the next set
using a different prefix than ZZ, as specified by the $DUMMY command.
18.2.2
Body
This section discusses the rules for writing a module. There are two different sets of cards,
namely /-cards and $-cards. The /-cards inform the program about the type of data. The $cards request the program either to
1.
read another input file, or
2.
set file attributes, such as directory, etc., or
3.
modify the contents of card images using argument list, or
4.
change the seed key of internal variables.
The following rules must be observed when writing a module:

Rule 1:
Each data block must start with a proper / card.
Rule 2:
A $ card can be a part of a data block.
Rule 3:
If a $INCLUDE card is embedded in a data block then it must be followed by a proper

/-card.
/-Card
Functional Description
/REQUEST
to request special-request data including

miscellaneous cards
/TACS
to indicate TACS data
/BRANCH
to indicate branch data
/SWITCH
to indicate switch data
/SOURCE
to indicate source data (except machines)
/OUTPUT
to indicate source data
/PLOT
to indicate plot data
18 - 5
EMTP Data Modules
/-Card
Functional Description
/LOAD FLOW
to indicate power constrain data for the EMTP

load-flow
/STATISTICS
to indicate statistics request cards
/INITIAL
to indicate user-supplied initial condition cards
/MACHINE
to indicate machine data
/ENDMODULE
to indicate end of module or end of current data

class
Rule 4:
The first eleven characters of a line in a module should not be BLANK ENDS. This
pattern is a keyword. Exception: This pattern is used to terminate a data type, e.g.,
BLANK ENDS BRANCH or BLANK BRANCH terminates branch data. See point C
below.
Rule 5:
The body of a module must end with a /ENDMODULE or END card followed by a
$EOF card.
Some remarks will clarify the above rules:
(A)
/-Cards
Each of these cards describes a class of EMTP data. The order in which they are used is
of no importance as they are internally sorted into a valid sequence.
(B)
$-Cards
The valid $ cards are: $INCLUDE, $SUFFIX, $PREFIX,$DUMMY,$EOF
$INCLUDE CARD
This card requests the program to read another input file. The format of the call is
given below:
$INCLUDE
File-Name Argument,Argument,Argument,Argument
; Comments
; Comments
For example, the include statement to call the module shown in Figure 18.1 will be
$INCLUDE Module YOLD1,0.01,10.0,15. ; AC fault at System 1
where Module is the name of the file. YOLD1 will replace the pattern FAULT.
Numeric values 0.01, 10.0, and 15. will replace numeric fields represented by
RPHASE, LPHASE, and RNEUTL. The rest of the line after ; is a comment.
18 - 6
EMTP Data Modules

Special Case: If a module does not have an argument but has dummy variables, then a
special format of an $INCLUDE card has to be used. The format is:
$INCLUDE File-Name, ; Comment
$SUFFIX and $PREFIX Cards

The formats of these cards are
$SUFFIX
$PREFIX
Suffix-Name
Prefix-Name
These cards are used to provide a common file attribute to the File-Name associated
with a $INCLUDE card. The variable associated with the File-Name of an include card
is always modified by a FORTRAN statement
File-Name = Prefix-Name//File-Name//Suffix-Name
Therefore, these cards need to be called only once to add a Prefix-Name and/or to add
a Suffix-Name to file names in different include calls.
$DUMMY Card
The format of the card is
$DUMMY XXnnnn
where
XX
is any two-character dummy word
nnnn
represents the starting numeric number
e.g. $DUMMY HQ0101

The first dummy variable will be named HQ0101 after this card has been read. Users
should avoid using this card inside a module. This card is typically used in the top
level of input data.
$EOF Card
This card indicates the end of the file to the program. The rest of the file is ignored.
(C)
Blank Cards
Blank EMTP data cards are mandatory to terminate the different EMTP data classes in a
standard EMTP data case. A new keyword BLANK ENDS has been used to identify a
blank EMTP data card for the purpose of terminating an EMTP data class. This was
necessary because input data for the Type-19 U.M. and Type-59 S.M. use blank cards
to terminate different classes of data cards.
18 - 7
EMTP Data Modules

The new keyword followed by the name of an EMTP data class will form a unique
blank card, which assists the sorting logic in locating the position for placement of
records associated with /-cards. For example, the card BLANK ENDS BRANCH
(BLANK BRANCH is also allowed) represents the blank card terminating the branch
data. The sorting logic identifies this card and places data associated with all /
BRANCH-cards before the appropriate blank card. The following rule is used to form
blank cards associated with EMTP data classes:
Rule:
BLANK ENDS Class-Name (or in short BLANK Class-Name)

e.g.BLANK ENDS BRANCH (to end branch data)
BLANK ENDS TACS (to end TACS data)
BLANK ENDS SOURCE (to end source data)
Users should note that there are no associated blank cards either for LOAD FLOW, or
for REQUEST or for ENDMODULE. The BLANK ENDS-cards are typically used in
the top level of input data.
(D)
Style of Documentation
All arguments of a module can be classified under 5 headings, namely:

1.
Electrical Nodes,
2.
TACS Output Variables,
3.
TACS Input Variables,
4.
Numeric Constants,
5.
Initialization Constants, and
6.
Special Output Requests.
It is suggested that, while declaring the arguments for a module, users follow the
specified order as shown above. This order will give a sense of organization and an
implicit documentation. Users may label each argument as shown in Example 18.1
using the in-line comment character ;. Both the order of argument declaration and
the in-line comment will provide a good documentation style.
This style of documentation can be extended to document include cards in a module or
in an input data file.
18 - 8
EMTP Data Modules
18.2.3
Recommendations
Here are a few recommendations for modules:

1.
A / card should immediately follow declaration statements;
2.
Document a module as clearly as possible;
3.
Promptly convert a module to an include file;
4.
Keep a correlation between the name of a module file and the name of its include
file;
5.
Check the date and time of conversion of an include file before using it.
Now, here are a few practices to avoid:

1.
Avoid using BLANK ENDS card in a module; and
2.
Avoid using $DUMMY card in a module.
18.3 Conversion from a Module to an Include File

In order to use the $INCLUDE card, a module has to be converted to an include file. This is done
using the MODULE command in a data file executed in batch mode. The conversion of a .MOD
file to a .INC file requires using the two commands MODULE and STOP in the input file. Two file
names are to be supplied with each usage of MODULE
(A)
module file (input)
(B)
include file (output)
The MODULE command can be repeated to process as many files as necessary. The STOP
command ends the processing of modules.
MODULE
MODULE_1.MOD
MODULE_1.INC
MODULE_2.MOD
MODULE_2.INC
STOP
18 - 9
EMTP Data Modules

The above-described conversion is done to reduce the burden of search at simulation time. The
include file has 5 vectors inserted at the beginning of the file. These vectors inform the program
about:
the number of substitutions in a module;
the number of arguments in a module;
the number of dummy variables in a module;
the cards and columns where arguments and dummy variables have to be replaced;
the mode of substitution (i.e., numeric or alphanumeric).
These vectors are read with a FORMAT (4X, 15I5) in the order as follows, where J is the
substitution index.
KARD(J)
gives the number of a non-commented data card, where the argument given by KARG(J) has to be replaced;
KARG(J)
gives the number of the argument; if the number is negative it is

a 6 character dummy variable;
DBEG(J)
provides the starting column number for substitution of the variable given by KARG(J);
KEND(J)
provides the ending column number for substitution of the variable given by KARG(J);
KTEX(J)
denotes the nature of the substitution (numeric=0 and alphanumeric=1);
The following additional information is given for module compilation which a user is not required
to know:
At the compiling the calculation section code is transferred to reversed polish notation and written at the end of the include file after the module body.
Before the calculation code the number of default parameters (if any) are given. The
default values are given together with the position in the compiled code variable/constant list.
The first line of compiled code includes the total lines of calculation code and the
sum of constants and internal calculation section variables, i.e. the number of undeclared variables constants in the calculation section.
Then follows the code in reversed polish notation. There each variable (external and
18 - 10
EMTP Data Modules

internal) and constant is replaced by a number in a list. At the left of each line the
number of continuation lines are given. At the right the line references for next
instruction are given. The second reference is for a negative IF-statement or after fulfilled IF-THEN-ELSE sections.
After the code a line with the number of constants are given. Then each position, in
the compiled code variable/constant list, for the constants are given. Following that
the value for each constant are given.
At the end some statistics and a list of ARG, DUM and NUM declarations are given.
Only comments given on the same line as the declaration is written. Comments
longer than one line can be used if a continuation sign "-" is used as shown in the following example:
ARG DCBUS
1)
;
ARG TAPPOS, TAPSIZ
Node name for connection (the sixth

character is the Converter id = ARG 10)
2)
Position and stepsize of tap changer
18.4 Advanced Features

It is possible to perform some simple calculations using EDM, it is also possible to define default
parameters.
18.4.1
Calculation Mode
This is an advanced feature that allows the substitution of internally-calculated numeric values,
thereby enhancing usage of the data type 'ZZ'. This is explained by an example of unit conversion
problem. Suppose two quantities in a module are known in per-unit but the EMTP needs them in
ohms. An obvious way to pass correct values to the EMTP is to use hand calculations for the
necessary unit conversion, and pass two ohmic values through two external numeric variables
defined for the module, e.g. 'ROHMS1' and 'ROHMS2'. The enhanced EDM provides an advanced
way to substitute numeric values. Now, per-unit values are supplied to the module through two
external numeric variables R_' and 'RNEUTL'. The two variables 'ROHMS1' and 'ROHMS2' are
defined as internal variables. EDM computes the values for 'ROHMS1' and 'ROHMS2' from the perunit values obtained through 'R_' and 'RNEUTL' respectively, and then substitutes the computed
values in the body of the module before the module is included as a part of the EMTP input data
stream.
An internal variable may be qualified simultaneously as 'DUM' and 'NUM' types. If an internal
variable is defined as 'NUM', EDM expects that a calculation section to be present and the numeric
18 - 11
EMTP Data Modules

internal variables to be defined in the calculation section. An internal variable with a 'NUM'
qualifier can have a length between 1 and 20 numeric characters; and an internal variable without
the NUM qualifier must be of length 6. The two internal variables 'ROHMS1' and 'ROHMS2' of the
example are of course declared as both 'DUM' and 'NUM' types.
18.4.2
Default Value Specification
In addition to the existing data types - 'ARG', 'NUM', and 'DUM', a data type DEF is also available
starting in version 3. When a variable is defined as 'DEF', it is given a value in a DEF declaration
statement. This value serves as the default value of the variable when it is not specified in the
$INCLUDE statement. The syntax of a 'DEF declaration statement is
DEF variable = value [, variable = value]...
Example:
ARG RESIST, CAPACS, ZBASE
NUM RESIST, CAPACS, ZBASE
DEF ZBASE=100.0,RESIST=10.0
An example of a module using these two advanced features is shown below.
18 - 12
EMTP Data Modules
C MODULE ACFAULT.MOD WITH ADVANCED FEATURES

ARG FAULT,EBUS##, - ; electrical nodes
R_____,LPHASE,RNEUTL,ZBASE - ; numeric constants
?,@ ; output requests
NUM R_____ ,LPHASE,RNEUTL, ZBASE, ROHMS1, ROHMS2
DUM NEUTRL, ROHMS1, ROHMS2 ; internal variables
C
DEF ZBASE = 100.
C
BEGIN CALCULATION MODE
ROHMS1 = R_____* ZBASE
ROHMS2 = RNEUTL * ZBASE
END CALCULATION MODE
C BELOW IS THE BODY OF THE MODULE
/BRANCH
FAULTANEUTRL
ROHMS1LPHASE
FAULTBNEUTRLFAULTANEUTRL
FAULTCNEUTRLFAULTANEUTRL
NEUTRL
ROHMS2
/SWITCH
EBUS FAULTA
/ENDMODULE
$EOF
C End Body
Figure 18.2:
18.4.3
A module with advanced features
Language Syntax for the Calculation Section of EDM
The main purpose of the Calculation Section of EDM is to perform simple calculations, such as
calculating initial conditions, etc. This section is not active in the time step loop. It can send and
receive data inside a module. This section is not meant for writing large programs and does not
have Input/Output and file handling capabilities.
The calculation section can also be used for checking constants or fixed values inputted to the
module. This feature with input of fixed values via the calculation section is that all fixed values
are concentrated in the module and will substitute uniform at multi-locations.
Language for the Calculation Section and Its Limitation: The language
syntax for EDM is very similar to the FORTRAN 77 language syntax. However, the language does
not support all the features of FORTRAN 77. The language for the calculation section supports the
following features:
Data Types and Operators

18 - 13
EMTP Data Modules

(i)
All variables of type NUM are considered to be of REAL*4. Integer,

character, complex, and logical variables are not supported inside the
calculation section.
(ii)
Exponentiation, arithmetic, logical and relation operations, and intrinsic

functions are supported.
Expressions and Assignments

(i)
There is no type declaration statement inside the calculation section.

However, there is a type declaration section which precedes the
calculation section. There are three type declaration statements which
are (a) ARG, (b) NUM, and (c) DUM. All external variables to the
module must be declared as type ARG. All internal variables of the
module must be declared as type DUM. All variables which contain
numeric values must be declared simultaneously as type NUM. An
variable can be a character string or numeric.
Notes:
1.
The use of the characters "_" (underscore) and "#" in node-names is not
recommended because these characters are used to indicate embedded blanks.
There is one important distinction between these characters. The character "_"
(underscore) in an argument specification is always treated as an embedded blank
character when it is substituted inside a module. However, the character # in an
argument is not treated as an embedded blank character when it is substituted on
an embedded $INCLUDE line in a module.
2.
Text which is not used as node names (e.g., filenames or plottext) can be
declared as NUM and will then be maximized to 20 characters and right adjusted.
(i)
There can be undeclared variables in the calculation section. However,

only the variables which are declared as type 'DUM' will be searched for
in the body of a module and will be replaced.
(ii)
All variables must begin with an letter and may have a length of 1 to 20
except when the variable is declared as DUM without NUM qualifier. All
variables declared as DUM without NUM qualifier must have 6
characters (e.g., dummy node names or dummy TACS variables).
Internal calculation section variables may not be declared: the first
character has to be an letter.
(iii)
Arrays of any kind are not supported by this language.
(iv)
All expressions are evaluated from left to right except in the case of
exponentiation. Expressions in parentheses are evaluated first.
18 - 14
EMTP Data Modules

(v)
A general assignment statement is in the form:

Variable Name = Expression
Variable Name - Internal module variable of type DUM with NUM
qualifier or internal calculation section variable (no declaration
necessary)
Expression - Any arithmetic, relational and logical expressions;
An expression may contain internal and external variables.
An assignment statement is written between columns 7 and 72. The
assignment statement may be continued in the next line if a nonblank character is punched in column 6 and the columns 1 to 5 are
left blank. A comment line is indicated by entering the character C
in the column 1. Text including and to the right of; is handled as
comments in the same way as in the module body.
(vi)
Assign statement is not supported.
(vii)
The transfer of arguments and declarations can be summarized as

follows:
ARG: fixed node-names or fixed TACS-variable, transferred to the
module body from the EMTP-main file or higher level module.
Cannot be transferred to the calculation section. Maximum 6
characters. The external variable (argument) must have the same
length as the internal variable. Could also be used as part of a node
name, i.e. as index when standardized node names are used.
ARG+NUM: External numerical value transferred to the calculation
section and/or module body from the EMTP-main file or higher level
module. Can not be modified in the calculation section. Can also be
a text string (i.e. filenames or plottext) to the module body. Can be
1 to 20 characters, is replaced right adjusted. The external variable
(argument) must have the same length or shorter as the internal
variable, i.e. the accuracy is limited by the length of the internal
variable.
If the EMTP assumes a real (not integer) value a decimal point must
be included in the external value. To avoid this all arguments can
be transferred to the calculation section as ARG+NUM and then
passed to the module body as a variable declared DUM+NUM.
18 - 15
EMTP Data Modules

DUM: Internal module node-names or TACS variables. The names
are generated by the program. Can not be used in the calculation

mode. Must be 6 characters
DUM+NUM: Internal variable to be transferred from the calculation
section to the module body. Can be 1 to 20 characters. When replaced

the maximum number of digits are used, i.e. the accuracy is limited
by the length of the internal variable.
No declaration: Internal calculation section variables. Must begin
with a letter and have a length of 1-20 characters.
Control Statements and Subprograms

(i)
The following IF control statements are supported.
IF (relational and logical expression) THEN

......variable = expression
END IF
ELSE
END IF
ELSE IF (relational and logical expression) THEN
ELSE
END IF
IF (relational and logical expression) GOTO <label>
IF (relational and logical expression) variable = expression
(ii)
GOTO statement is supported.
(iii)
Loops of any kind are not supported (except via GOTO statement).
(iv)
CALL, RETURN, PAUSE, STOP and END control statements are not
supported. However, WARNING and STOP statements are handled as
intrinsic functions.
(v)
Subprograms such as user defined functions, subroutines, and block

data are not supported.
18 - 16
EMTP Data Modules
Structure of the Calculation Section

(i)
The calculation section must follow the type declaration section of a

module.
(ii)
The calculation section communicates with other parts via argument

type DUM. Values of the declared internal variables are only
substituted.
(iii)
A calculation section begins with the keyword "BEGIN CALCULATION

MODE" and ends with the keyword "END CALCULATION MODE".
These keywords must be start in column 1. There is an option to output
the values of all variables in the calculation section by beginning the
calculation section with the keyword BEGIN CALCULATION MODE/
LIST instead of "BEGIN CALCULATION MODE". The values will be
printed in the listing file. These values can be used to debug the
calculation section. Due to the relative primitive nature of the
calculation section code, each value is given with an integer value as
reference to the reversed polish notation code in the include file.
An alternative to debug the calculation code is to use a dummy
warning function, as the following example.
IF (1.GT.0) X = WARNING(2, TAPPOS)
The actually calculated value of TAPPOS will then be printed in the

listing file (result file) together with the code 2 and a warning message
(to be neglected).
(iv)
The above keywords bound the calculation section. The calculation

section may contain a number of assignment statements and control
statements. The total number of these statements may not exceed 200.
This limit can be changed inside the program if desired.
(v)
An example of a calculation section is shown in Figure 18.3.
Operators
The following arithmetic, relational, and logical operators are supported:
Operator
Function
18 - 17
Precedence
EMTP Data Modules

**
Exponentiation
First
* and /
Multiplication and Division
Second
+ and -
Addition and Subtraction
Third
.LT.
Less than
Fourth
.LE.
Less than or equal to
Fourth
.GT.
Greater than
Fourth
.GE.
Greater than or equal to
Fourth
.NE.
Not equal to
Fourth
.EQ
Equal to
Fourth
. NOT.
Logical negation
Fifth
.AND.
Logical conjunction
Sixth
.OR.
Logical disjunction
Seventh
.EQV. AND .NEQV.
Logical equivalence and exclusive
Eighth
18 - 18
EMTP Data Modules
Intrinsic Functions
The following standard FORTRAN 77 intrinsic functions are supported:
Function Description
Name
Square Root
SQRT()
Natural Logarithm
ALOG()
Common Logarithm (Base 10)
ALOG10()
Exponential
EXP()
Sine
SIN()
CoSine
COS()
Tangent
TAN()
Arc Sine
ASIN()
Arc CoSine
ACOS()
Arc Tangent (one argument)
ATAN()
Arc Tangent (two arguments)
ATAN2()
Hyperbolic Sine
SINH()
Hyperbolic Cosine
COSH()
Hyperbolic Tangent
TANH()
Positive Difference (two arguments)

absolute value of (a1 - a2)
DIM(,)
Remainder (two arguments)

a1 - a2*(lNT(a1/a2))
AMOD(,)
Transfer Sign (two arguments)

la1lSign(a2)
SIGN(,)
18 - 19
EMTP Data Modules
BEGIN CALCULATION MODE

ALPHA = SQRT(2.0) * UAC * CO ;(THETA)
PHI = 5.0
C FIND WHICH VALVES ARE CONDUCTING - SET VALVES TO ZERO
FIRST
VALVE 1 = 0.0
VALVE2 = 0.0
VALVE3 = 0.0
VALVE4 = 0.0
VALVES = 0.0
VALVE6 = 0.0
POINTER = PHI + ALPHA
IF( POINTER .GE. 0.0 .AND. POINTER .LE. 120.0) THEN
VALVE 1 = 1.0
END IF
IF(POINTER .GE. 60.0 .AND. POINTER .LE. 180.0) THEN
VALVE2 = 1.0
END IF
IF( POINTER . GE . 120.0 . AND . POINTER .LE. 240.0)
THEN
VALVE 3 = 1.0
END IF
IF( POINTER .GE. 180.0 .AND. POINTER .LE. 300.0) THEN
VALVE4 = 1.0
END IF
IF(POINTER .GE. 240.0 .AND. POINTER .LE. 360.0) THEN
VALVE5 = 1.0
END IF
IF( POINTER .GE. -60.0 .AND. POINTER .LE. 60.0) THEN
VALVE 6 = 1.0
END IF
END CALCULATION MODE
Figure 18.3:
A typical calculaiton section of a module
Complex Mathematics: At present, the EMTP requires only real values, therefore complex
mathematics are supported in a limited way by using non-standard FORTRAN intrinsic functions. If
complex expressions are required in the future, the program can be modified to incorporate such
needs. The following intrinsic functions can be used for complex mathematics.
Complex Multiplication
Two intrinsic functions, CMULR and CMULI are provided for multiplying two complex
numbers. The output of these intrinsic functions are real quantities. For example if
need arises to multiply two complex numbers X (Xr + jXi) and Y (Yr +jYi). The
product Z (Zr + jZi) can obtained by
Zr = CMULR(Xr, Xi, Yr, Yi)
18 - 20
EMTP Data Modules

Zi = CMULI (Xr, Xi, Yr, Yi) .
Where
zr- real part of the product
zi- imaginary part of the product
xr- real part of the first complex number
xi- imaginary part of the first complex number
yr- real part of the second complex number
yi- imaginary part of the second complex number
Complex Division
Two intrinsic functions, CDIVR and CDIVI are provided for dividing two complex
numbers. The output of these intrinsic functions are real quantities. For example if
need arises to divide the complex number X (Xr + jXi) by the complex number Y (Yr
+jYi). The quotient Z (Zr + jZi) can obtained by
Zr = CDIVR(Xr, Xi, Yr, Yi)
Z; = CDIVI((Xr, Xi, Yr, Yi)
Where
zr- real part of the quotient
zi- imaginary part of the quotient
xr- real part of the dividend
xi- imaginary part of the dividend
yr- real part of the divisor
yi- imaginary part of the divisor
Rectangular Coordinates to Polar Coordinates Conversion

Two intrinsic functions, RTOPM and RTOPA are provided for converting a complex
number in rectangular coordinates to polar coordinates. The output of these intrinsic
18 - 21
EMTP Data Modules

functions are real quantities. The complex number X (Xr + jXi) which is in rectangular
coordinates can be converted to polar coordinates by
Xm = RTOPM((Xr, Xi) Xn = RTOPA((Xr, X;).
Where
xm - magnitude of the complex number
xa - angle of the complex number in degrees
xr - real part of the complex number
xi - imaginary part of the complex number
Polar Coordinates to Rectangular Coordinates Conversion

Two intrinsic functions, PTORR and PTORI are provided for converting a complex
number in polar coordinates to rectangular coordinates. The output of these intrinsic
functions are real quantities. The complex number X (Xm, Xa) which is in polar
coordinates can be converted to rectangular coordinates by
Xr = PTORR((Xm, Xa) X; = PTORI(Xm, Xa).
Where
xm - magnitude of the complex number
xa - angle of the complex number in degrees
xr - real part of the complex number
xi - imaginary part of the complex number
Warning and Stop: Warning and stop are handled as intrinsic functions and preferable in
combination with an IF statement.
As the first argument to the intrinsic function a code number is given. This code can be used in
combination with the users module description to explain the reason for the Warning/Stop since
the calculation mode do not handle strings.
As the second argument to the intrinsic function the variable that initiated the warning or stop is
given.
A dummy variable (internal calculation section variable, no declarations) is given the value of the
intrinsic function. This variable is necessary just to simplify the FORTRAN-Coding of the
18 - 22
EMTP Data Modules

calculation code. The same code as handling ordinary intrinsic is used. After the intrinsic request,
this variable has the same value as the second argument.
Example:
IF (TAPPOS.LT.-25) X = STOP (1, TAPPOS)
IF (TAPPOS . LT. -15) X = WARNING (2, TAPPOS)
18 - 23
Section 19
High Voltage DC
19.1
General Comments .............................................................................. 19-1
19.2
Applications ........................................................................................ 19-2
19.3
Computer Aspects ............................................................................... 19-2
19.4
19.3.1
Time Step ....................................................................... 19-2
19.3.2
TACS Working Space ..................................................... 19-2
19.3.3
Absolute TACS Dimensioning ........................................ 19-3
Structure .............................................................................................. 19-3

19.4.1
Subsystems ..................................................................... 19-3
19.4.2
Standardized EMTP Names ............................................. 19-5
19.5
Main Data File ..................................................................................... 19-9
19.6
Bipolar HVDC Station Module (DET-STATION) .................................. 19-10
19.7
19.8
19.9
19.6.1
Module Description ........................................................ 19-10
19.6.2
Arguments ...................................................................... 19-12
19.6.3
Module Constants .......................................................... 19-13
19.6.4
Calculation Mode ........................................................... 19-14
19.6.5
Warning and Stop ........................................................... 19-16
12-Pulse HVDC Converter Module (DET-POLE) .................................. 19-16

19.7.1
Module Description ........................................................ 19-16
19.7.2
Arguments ...................................................................... 19-18
19.7.3
Module Constants .......................................................... 19-21
19.7.4
External Control and Protection Signals ........................ 19-21
19.7.5
Reduced Control System ................................................ 19-22
Converter Control System Model ....................................................... 19-23

19.8.1
Voltage Dependent Current Order Limiter (VDCOL) ..... 19-24
19.8.2
Constant Current Control (AMPLIFIER) ......................... 19-27
19.8.3
Converter Firing Control ................................................ 19-31
19.8.4
External Control and Protection Signals ........................ 19-31
DC Line Module .................................................................................. 19-32
19.9.1
Arguments ...................................................................... 19-33
19.9.2
DC Resistance ................................................................. 19-34
Section 19
19.10
19.11
AC and DC Filter Modules .................................................................. 19-35
19.10.1
Arguments ..................................................................... 19-36
19.10.2
Internal Node Names ..................................................... 19-40
AC System Modules ............................................................................ 19-41
19.11.1
19.12
19.13
19.14
19.16
6-Pulse HVDC Converter Module (SIXPULSEBR) .............................. 19-47

19.12.1
Arguments ..................................................................... 19-47
19.12.2
Converter Transformer Modules ........................................................ 19-50

19.13.1
Arguments ..................................................................... 19-52
19.13.2
Smoothing Reactor (SMOOTR) ........................................................... 19-55

19.14.1
19.15
High Voltage DC
Arguments ..................................................................... 19-55
Some Comments About Modules ....................................................... 19-56

19.15.1
Converting a Module ..................................................... 19-56
19.15.2
Processing ...................................................................... 19-57
19.15.3
Special Characters ......................................................... 19-57
19.15.4
FORTRAN Format ........................................................... 19-58
TACS Signal Names ............................................................................ 19-58
Section 19
High Voltage DC
19.1 General Comments
The HVDC EMTP model is a detailed TACS-based representation of an HVDC converter terminal
consisting of a 12-pulse converter bridge, the associated Y-Y and Y-D converter transformers, and
the current and firing control systems for a single converter. Several converter models are used in
combination with models for the AC and DC systems to form a complete HVDC transmission
system. The model is intended primarily for 12-pulse operation and is constructed in such a way
that operation is independent of the number of poles, operating mode, and rectifier or inverter
operation. 6-pulse, series-connected 12-pulse, and 24-pulse operation are possible with
modification of the modules. Higher level control systems, such as real and reactive power
controls, are not included but can be added as required. The time constants of these control
systems are typically large and beyond the scope of most transient studies.
The HVDC model is built in a modular structure utilizing existing EMTP models. This modularized
structure lends itself to modification to suit the specific modelling needs of HVDC control system
studies. High level modules define application-specific hardware and control system parameters
and establish the overall structure of the model. Low level modules represent single apparatus or
subfunctions of the control system. Whereas the high level modules will typically be modified by
the user to define the system under study, the low level modules will rarely require modification.
However, the low level modules can be modified to investigate alternative control strategies, etc.
The HVDC model is completely self-initializing when used with the simplified AC network
module supplied. The calculation mode of EMTP data modules (EDM) is used to initialize the
HVDC system and includes the main circuit calculations, active and reactive power balance
calculations, and source voltage calculations for a simple AC network equivalent. The AC and DC
systems are electrically isolated during the steady-state phasor solution and are initialized by
separate steady-state phasor solutions. The AC system is initialized with the fundamental
frequency while the DC system steady-state phasor solution is computed with an initial frequency
of 0.01 Hz. Once the time solution begins, the two systems are connected. The procedure used
results in almost no transient when the two systems are connected.
19 - 1
High Voltage DC
19.2 Applications
The detailed HVDC model is intended for the following applications:
Studying the electromagnetic dynamics of a combined AC - DC system.
Studying the electromechanical dynamics of an AC system.
Studying arrester stresses in AC and DC systems.
General dynamic performance evaluation of HVDC controls in the time range of

about 1.5 seconds.
Limitations of the detailed HVDC model include:
The model is not suitable for conducting complete dynamic performance studies
because the model does not run in real-time.
The valves (thyristors) in this model are represented by TACS-controlled switches,

and the turn-on and turn-off characteristics of a valve are not modelled. The model is
therefore not suitable for designing the snubber circuit, for optimizing the valve reactor, or for studying the turn-on and turn-off losses of a valve. However, valve sensitivity to commutation voltage distortions which lead to commutation failures can be
represented reasonably well by the EMTP switch parameter TDEION, the deionization
time representing the recovery time for the valve.
19.3 Computer Aspects

19.3.1
Time Step
The maximum time step recommended for use with the detailed HVDC EMTP model, as based on
the converter firing control angle, is about 0.5 degrees or 25 s at 60 Hz. The time step can be
further reduced for higher frequency transient phenomena.
19.3.2
TACS Working Space
For a two station bipolar HVDC system, a "Floating-point working space for all TACS arrays" (List
19) of 100,000 is recommended.
19 - 2
High Voltage DC
19.3.3
Absolute TACS Dimensioning
For a two station bipolar HVDC system, the following "Absolute TACS Dimensioning" is
recommended.
LT1
1,000
LT5
2,000
LT2
1,000
LT6
40,000
LT3
2,000
LT7
4,500
LT4
800
LT8
4,000
19.4 Structure
19.4.1
Subsystems
The detailed HVDC system model is divided into three subsystems, namely:
1.
AC system, including AC filters
2.
DC system, including DC filters
3.
Converter system, including the converter control system
With the exception of the nodes representing the terminals of the HVDC converter, the AC and DC
systems are generally left to the user's discretion. Typical AC and DC system network modules are
included with the model and are documented in Section 19.10.
The converter system includes:
Valve bridge, converter transformers, and smoothing reactor
Converter control system
Initialization sources
The valve bridge and converter transformers are represented in the EMTP using the Type 11 TACScontrolled switch model and the single-phase saturable transformer model. The 12-pulse valve
bridge module consists of two 6-pulse converter modules. The converter transformer module
consists of Y-Y and Y-D connected transformers. The magnetizing branches of the converter
transformers are not included in the model, but can be added as an external element. Generally, a
magnetizing branch is connected on the primary side of a converter transformer.
19 - 3
High Voltage DC
The control principle used in the model is equidistant firing as described in Reference 10 and is
implemented using TACS. The control system includes constant current control, rectifier operation
on minimum delay angle, and inverter operation on minimum extinction angle for both rectifier
and inverter operation. The logical sub-function modules of the converter system are supplemented
with typical default values for control parameters. Inter-station communication has not been
implemented in the model, but can be added to existing modules if required. An overview of the
control structure of the HVDC model is shown in Figure 19.1.
EMTP Data File
DC Line
DET-STATIION
2
DC
5
AC
DET-POLE
Filters
Systems
Measuring
Device
Trafo
Smoothing
Reactor
Valves
AC
Filters
Initialization
Control
System
Figure 19.1: Overview of EMTP HVDC Converter Model

At the top of the structure lies the EMTP main data file. For the typical bipolar or monopolar HVDC
transmission system, this data file includes two DET-STATION modules and a DC line module. In
turn, each DET-STATION module includes two DET-POLE modules for bipolar operation, the AC
system and filter modules, and the DC filter modules. One DET-POLE module is automatically
disabled if monopolar operation is specified. The DET-POLE module consists of:
Valve Modules
Transformer Module
Smoothing Reactor Module

19 - 4
High Voltage DC
Measuring Device Modules
Control System Modules
Initialization Source Modules.
By using this structure (main data file, DET-STATION, DET-POLE, and low level modules), one
DET-POLE module can represent both poles in a bipolar station and also both stations if they are
identical. In addition, all initialization calculations can be performed inside the modules using the
CALCULATION MODE OF EDT. This is possible as long as the AC network representation can be
made relatively simple, which is normally the case.
The main data file and the modules DET-STATION and DET-POLE are application dependent.
Typically, they would be copied from a system library and then modified in the user's directory for
a specific application. The converter and control system modules are described in more detail in
Reference 10.
19.4.2
Standardized EMTP Names
Node names and TACS variable names have been standardized in this model to minimize the
number of arguments transferred between modules and to allow, at least in principle, an unlimited
number of converter stations in an EMTP data case. Standardized node names also help to avoid
erroneous connections in relatively large EMTP cases. The node and TACS variable names can be
divided into two basic categories, the electrical network and the control system.
Electrical Network:
The AC and DC bus node names used in the HVDC converter model determine the interconnection
of the AC and DC systems under study. All electrical network node names have been standardized
using the six character positions of the name as follows:
The first two positions are reserved for the area and module type.
AS for AC system
AC for AC bus
FA for AC filter
TR for transformer
VB for valve bridge
FD for DC filter
DC for DC bus
DL for DC line
The third position of the node name is reserved for the station or pole identifier.
The station identifier is used in all modules at the station level, i.e., the AC
system, AC filters, commutation voltage measurement, and the neutral bus.
19 - 5
High Voltage DC
The initial character in the actual station name is recommended, R for Rihand
as an example.
The pole identifier is used in all modules at the pole level, i.e., the converter
transformer, initialization sources, smoothing reactors, valve bridges,
measuring devices, and controls. A number is recommended, e.g., 1 for the
rectifier cathode pole, 2 for the rectifier anode pole, 3 for the inverter cathode
pole, and 4 for the inverter anode pole. If more than nine poles are used,
letters may be used.
The fourth position is reserved for the AC and DC filter bank identifiers within one
station.
The sixth position is reserved for phase identifiers (A, B or C), valve identifier (1 to
6), or pole-line identifiers (1, 2, N or E).
The remaining positions (4-5 for AC node names, 4-6 for DC bus node names, and 5 for filter bus
node names) are used as described below to make the node name unique within a module.
AC System Bus:
BU is the recommended designation for the AC system bus to which the HVDC converter is
connected. If the user is modelling multiple HVDC systems, or if the converter is connected to
more than one AC node, it is recommended that the characters used in positions 4 and 5 be
modified in order to avoid erroneous connections, e.g. B1, B2, etc.
Examples:
One AC node in station X (phases A - C),
ACXBUA
ACXBUB
ACXBUC
ACXB1A
ACXB1B
ACXB1C
ACXB2A
ACXB2B
ACXB2C
More than one AC node in station X,
DC System Bus:
The characters BUS in positions 4-6 designate a single DC converter bus. For multiple DC nodes,
BUA, BUB, BUC, BUD, etc. are recommended.
19 - 6
High Voltage DC
Examples:
One DC bus in pole 1,
DC1BUS
More than one DC bus in pole 1,

DC1BUA, DC1BUB, . . . .
DC Neutral Bus:
The neutral bus designation, NTR, is recommended for character positions 4-6 in the neutral bus
node name. With more than one node on the neutral bus, NTA, NTB, etc. can be used. Position 3 of
the name can be the station identifier if this node is common to both poles of a bipolar system, or
the pole identifier when the neutral bus node names are unique to each pole of a bipolar station.
However, the latter is possible only if the pole identifier is not identical to the station identifier. It
is recommended that the station identifiers be limited to characters while numbers are to be used
exclusively for the pole identifiers.
Examples:
One neutral bus in station X,
DCXNTR
More than one neutral bus in station X,

DCXNTA, DCXNTB, . . . .
Neutral bus in pole 1,

DC1NTR (DC1NTA, DC1NTB ...)
19 - 7
High Voltage DC
DC Line Nodes:
Positions 3-5 are reserved for geographical location. Position 6 is ordinarily assigned as follows:
1
for pole 1
for pole 2
for electrode line
for separate metallic return conductor, if used
Examples:
LOCATION:
001
002
003
NODES:
DL0011
DL0021
DL0031
DL001E
DL002E
DL003E
DL001M
DL002M
DL003M (IF USED)
DL0022
DL0032
Node Names In Other Modules:

The name conventions for local nodes within the various modules are described in the respective
module descriptions.
Control System Signal Names:

Named control signals are used for interconnection of the low level control modules in the model.
The TACS signal names for the control system are standardized as follows:
Position 1
Area identifier or type of control. For pole control, C is

used. For other controls, an alternate character is selected
for use in position 1.
Position 2
Converter identifier.
Position 5
6-pulse bridge identifier (Y or D) when position 6 is occupied by valve identifier.
Position 6
Valve identifier (1-6), or 6-pulse bridge identifier (Y or D).
The remaining positions are used to describe the control signal.
19 - 8
High Voltage DC
Control signals internal to a module are declared as dummy (DUM) and the names generated by the
EMTP. The signal names are listed in Section 19.16.
Output and Plot:

Output and batch-mode plotting cards can be inserted in the main data file, in the DET-STATION
module, or the DET-POLE module. Node voltages or TACS outputs can be specified using the
standard name syntax. Valve, smoothing reactor, and filter breaker outputs are available via predefined column 80 punches. Output of these variables are enabled by an argument in the
$INCLUDE command for the DET-STATION and DET-POLE modules.
Internal nodes or TACS signals specified as dummy variables (DUM) within a module are not
available as output in the higher level modules. If an output other than those described above is
desired, the user must modify the appropriate module to allow for that specific output.
It should be noted that the number of output variables available are limited by the EMTP. The
number of "standard" output variables from the HVDC model should be specified with care.
19.5 Main Data File

The main data file consists of standard EMTP cards, such as:
Miscellaneous data cards
TACS HYBRID card
TACS output requests
EMTP branch, switch, and source cards
Node voltage output requests
Batch mode plotting cards
BLANK ENDS cards separating the different data cards
In addition, the file includes the following module requests to define the overall structure of the
HVDC transmission system under study:
two DET-STATION modules
the DC line module
Figure 19.2 illustrates the module connections in EMTP data file.
19 - 9
High Voltage DC
DET-STATION
DC1BUS
Cid=1
ACXBUA
ACXBUB
ACXBUC
Cid=2
DET-STATION
DC Line
(Station ID X)
DL0011

DL0021
DCXNTR
DL001E

DL002E

DL0012

DL0022
DC2BUS
(Station ID Y)
DC4BUS
DCYNTR
DC3BUS
Cid=4
ACYBUA
ACYBUB
ACYBUC
Cid=3
Cid = Pole Identifier

1 = Cathode Pole
2 = Anode Pole
3 = Cathode Pole
4 = Anode Pole
Figure 19.2: Module Connections in EMTP Data File

Arrester models, fault switches, and other case dependent cards or modules can be included in the
main data file as required. Standard output such as valve voltages and valve currents that are
expected to be of interest are included in the main data file.
19.6 Bipolar HVDC Station Module (DET-STATION)

19.6.1
Module Description
The module DET-STATION defines the base bipolar station to be used in the HVDC model and
establishes the station-level control system for the station. The module can then be used for either
bipolar or monopolar, rectifier or inverter operation, as specified by the arguments to the module.
Two poles are included in each module, cathode and anode, and the two poles together form a
bipolar station. For monopolar operation, one pole in each station is automatically disabled via the
module arguments. For normal operation, the cathode is connected to the positive pole in rectifier
operation while the anode is connected to the positive pole in inverter operation. In other words,
rectifier pole 1 is connected to inverter pole 4 via the DC line, while rectifier pole 2 is connected to
inverter pole 3, as shown in Figure 19.2.
19 - 10
High Voltage DC
The module calls the following sub modules:
commutation voltage measurement
DET-POLE
DC filters
AC filters
AC system
Figure 19.3 below illustrates the internal module connections in the DET-STATION module.
DC1BUS
Internal AC
System Nodes:
ASXyyA
ASXyyB
ASXyyC
P,Q
Det-pole
Nodes:
ACXBUA
ACXBUB
ACXBUC
P,Q
DC
Filter
AC
System
Q
DCXNTR
Det-pole
DC
Filter
P,Q
AC
Filter
AC
Filter
AC
Filter
DC2BUS
Bank Identifier:
Internal Node Name: FAXAzA

FAXAzB
(z = internal)
FAXAzC
Figure 19.3:
FAXBzA
FAXBzB
FAXBzC
FAXCzA
FAXCzB
FAXCzC
Station Identifier X
Cathode Pole Identifier 1
Anode Pole Identifier 2
HVDC Station Module, DET-STATION
19 - 11
High Voltage DC
19.6.2
Arguments
The $INCLUDE statement for each HVDC converter station included in the EMTP data file is shown
below.
$INCLUDE DET-STATION
RECTIFIER
;
;
;
;
1)
1 (> 0.5) = rectifier,

0 ( 0.5) = inverter
2 (> 1.5) = reduced rectifier
-1 (< 0.5) = inverter, reduced
control system
UDCinit
;
;
2)
Initial DC bus voltage in rectifier,

(kV) always positive
IDCinit
3)
Initial DC bus current (Amp), always

positive
DCresist
4)
DC resistance in the dc circuit
UACxpunit
;
;
;
5)
AC bus voltage (initial) (pu), in
6)
Station identification, single character
7)
Converter identification, Cathode
"
8)
Converter identification, Anode
POLES
;
;
;
;
9)
Number of used poles

1 monopolar, positive voltage
2 bipolar
-1 monopolar, negative voltage
inverter operation might be modified

by initialization routine
The "-" in the $INCLUDE statement allows the statement to be continued on the next card. The ";"
indicates that the remainder of the card is to be interpreted as a comment. The module requires the
arguments listed in Table 19.1.
ARGUMENT TYPE
1
NC
WIDTH UNITS
9
DESCRIPTION
RECTIFIER, Identifier for operating mode
1 (> 0.5) = rectifier operation

0 ( 0.5) = inverter operation
2 (> 1.5) = rectifier, reduced control system
-1 (< 0.5) = inverter, reduced control system
2
NC
kV
UDCinit, initial DC bus voltage at rectifier. Always
given as the rectifier value and always positive.
Table 19.1: DET-STATION Arguments
19 - 12
High Voltage DC
3
NC
IDCinit, initial DC bus current (absolute value)
NC
DCresist, DC resistance in the DC circuit. See SecUACxpunit, initial AC bus voltage.
NC
pu
Node
tion 1.6.2
%, Single character station identification. See
Section 19.4.2
7
Node
!,
Single character pole identification, Cathode.
Node
",
Single character pole identification, Anode.
NC
POLES,
Number of poles used.

1 = Monopolar operation, positive voltage
2 = Bipolar operation
-1 = Monopolar operation, negative voltage
Node = Node or signal name characters

NC
= Numerical constant
Table 19.1: DET-STATION Arguments
19.6.3
Module Constants
Several project or station dependent constants are specified inside the DET-STATION module.
With modification to the argument list of the module, these constants can also be supplied as
arguments to the module if the user so desires. The constants are shown below in Table 19.2
below.
VARIABLE
UNIT
DXnom
pu
DX, relative inductive voltage drop. normally transformer reactance/2
DRnom
pu
DR, relative resistive voltage drop.
UDCnom
kV
Nominal DC bus voltage. Always positive and common to both stations.

Inverter value is based on this value, the DC current, and the DC line
resistance.
IDCnom
Nominal DC bus current.
udi0nom
kV
Nominal, no-load, fullwave rectified DC bus voltage per 6-pulse group.

The valve-side AC nominal voltage (UVnom) is calculated as PI/
3*udi0nom/SQRT(2).
UACnom
DESCRIPTION
kV, rms Nominal AC bus voltage (3-phase). Together with UVnom, defines the
nominal transformer turns ratio.
Table 19.2: DET-STATION Module Constants
19 - 13
High Voltage DC
ALPHAlim
deg.
Initial minimum alpha limit. Used as the minimum value when calculating TAPPOS. The initial alpha will slightly exceed this value depending
on the tap changer step.
ALPHAmin
deg.
Extreme minimum value of alpha. Passed to the DET-POLE module and

used for UMIN-firing (normally 5).
GAMMAlim
deg.
Initial gamma. Used as the minimum value when calculating TAPPOS.

Passed to the DET-POLE module as the initial value (normally 17). The
AC bus voltage is adjusted to this value in inverter operation.
GAMMAmin
deg.
Extreme minimum value of gamma. Passed to the DET-POLE module and

used for GMIN firing (predicted area minimum, AMIN). Can be 16 or the
same as GAMMAlim.
TAPPOS
Position of the tap changer. Start value in the iteration to find the
balance between the AC bus voltage, current, alpha, and tap
changer position at steady-state.
TAPSIZ
pu
Size of tap changer step (normally 0.0125).
PI
deg.
The value of pi (4*ATAN(1)).
THETAinitA
deg.
Phase angle in phase A at time = 0. Used for calculating the time reference for initialization of several modules.
FREQnom
Hz
Nominal frequency of the AC system.
FREQinit
Hz
Initial frequency of the AC system.
QfiltA
MVAR
MVAR size of filter bank A (3-phase). This or CfiltA must be given.
CfiltA
Capacitance of filter bank A. The value is used for calculating the

steady-state reactive power generation of the AC filters.
TINfiltA
sec.
Filter connection time (<0 used, >1 not used). The value is
used for calculating the steady-state reactive power generation of
the AC filters. If additional filters are used, add more variables
Qfilt/Cfilt and TINfilter, declared as both DUM and NUM, in the
calculation section and the appropriate module calls.
DELTA
Small value used to avoid singularity points by modifying initial phase

angle or alpha, if critical.
Table 19.2: DET-STATION Module Constants
19.6.4
Calculation Mode
For proper initialization of the AC and DC systems, the module includes the following steady-state
calculations in the calculation mode:
main circuit calculations
19 - 14
High Voltage DC
power and reactive power balance calculations
AC system active and reactive power supply calculations
The initial conditions calculated assume the same operating conditions for both poles in a bipolar
station. If this is not the case, the user will be required to modify the module accordingly. Most of
the steady-state calculations performed in the calculation mode are self-explanatory. The more
involved calculations are discussed in detail below.
Reactive Power Generation:

The reactive power generation in the AC filters is calculated using the total filter capacitance (C),
the AC bus voltage, and the initial frequency if the filters are specified by capacitance. If the filter
sizes are specified by reactive power generation (Q), then the total Q is simply the sum of
individual Q's and is adjusted according to the initial AC bus voltage. The total C and/or Q is
calculated using the Cfilt, Qfilt and TINfilter constants, Table 19.2. If additional filter banks or
shunt reactors are used, the same procedure must be followed.
Main Circuit Calculation:

The main circuit calculation determines the AC bus voltage, tap changer position, firing angle, and
DC current and voltage. The module assumes that the firing angle, alpha, is unknown. If this is not
the case, the user can modify the module accordingly.
Based on the parameters specified, alpha is calculated. If alpha is too low, the tap changer is
moved one step from the initial value. If alpha is to large, a warning message is printed in the
listing file. If alpha is much too large, the calculation is terminated with an error message. (Please
refer to the "Warning and Stop" section below.) Once alpha is acceptable, the overlap angle and
extinction angle, gamma, are determined.
For inverter operation, the same calculations are used to determine gamma. Normally, constant
gamma operation is used in the inverter and only the calculated tap position is required. The AC
bus voltage is adjusted slightly to match the selected gamma and tap position and then alpha is
determined. In a real plant, the AC bus voltage is given and the DC bus voltage is adjusted. In this
model, the DC bus voltage and direct current are the common initialization parameters for the two
station modules (rectifier and inverter). With this method, the AC bus voltage can then differ from
the value specified by as much as one tap step. If this is critical for the user, the DC bus voltage
must be specified to give the desired AC bus voltage "exactly".
Finally, the real and reactive power consumption (reference direction is from AC to DC) is
calculated and the AC system supply is initialized according to the power balance calculations.
Overlap Check:
During the steady-state initialization of the DC system, conduction overlap of the thyristor valves
is not considered. An overlap check is therefore performed to avoid starting the time simulation
19 - 15
High Voltage DC
during an overlap period. If t = 0 is found to be inside a conduction overlap period, the initial
phase angle for the AC bus is shifted to give t = 0 just after the overlap is completed. If the overlap
angle is greater then 30 degrees (very rare), this shift is not performed.
Singularity Check:
Several control signals are not continuous and can be triggered by zero crossings or thyristor firing
pulses. These triggering points must be avoided during initialization. A singularity check is
therefore performed and if a zero crossing or firing instant is detected at t=0, a small offset is added
to the phase angle of the AC bus.
19.6.5
Warning and Stop
Stop messages and codes generated in the DET-STATION module are as follows:
1.
The tap position is less than 25.
2.
Alpha (in rectifier operation) or gamma (in inverter operation) exceeds 90 in the
main circuit calculations.
3.
Overlap exceeds 90 in the main circuit calculations.
4.
One filter is specified by both reactive power generation and capacitance.
Warning messages similar to codes 2 and 3 above are generated when alpha and overlap exceeds
50, respectively. The messages are printed in the listing file.
19.7 12-Pulse HVDC Converter Module (DET-POLE)

19.7.1
Module Description
The module DET-POLE defines the basic 12-pulse converter unit and pole-level control system used
in the detailed HVDC model. The module can be used for either rectifier or inverter operation as
determined by the arguments specified in the main data file for the DET-STATION module.
External signals are provided to represent the actions of higher level control and/or protection
systems not explicitly modelled. For certain applications, selected features of the control system
can be disabled via the module arguments, thereby reducing the simulation time with no significant
loss in accuracy.
The DET-POLE module calls the following modules:
19 - 16
High Voltage DC
Valve bridge
Initialization sources
Initialization isolating switches
Control functions, several modules
Converter transformer
Smoothing reactor
Measuring devices
Connections follow the standard node name specification and are shown in Figure 19.4.
SMOOTHR
Station Identifier = X
Cathode Polle Identifier = 1
Anode Pole Identifier = 2
VB1TYA
VB1TYB
VB1TYC
VB1VYA
VB1VYB
VB1VYC
IPHMEASURE
UDCMEASURE
SIXPULSEBR D
VB1TDA
VB1TDB
VB1TDC
VB1VDA
VB1VDB
VB1VDC
+
VB1BOT
IDCMEASURE
ACXBUA
ACXBUB
ACXBUC
t=0
t=0
+
V
-
VB1MID
YY12D11
SIXPULSEBR Y
DC1BUS
VB1TOP
t=0
VBXBUA
VBXBUB
VBXBUC
DCXNTR
IDCMEASURE
VB2TDA
VB2TDB
VB2TDC
ACSYSINIT
YY12D11
INITCONVDC
VB2VDA
VB2VDB
VB2VDC
IPHMEASURE
VB2BOT
VB2MID
SIXPULSEBR Y
VB2VYA
VB2VYB
VB2VYC
UDCMEASURE
HVDC Pole Module, DET-POLE
19 - 17
+
V
-
DC2BUS
VB2TOP
SMOOTHR
Figure 19.4:
SIXPULSEBR D
VB2TYA
VB2TYB
VB2TYC
High Voltage DC
19.7.2
Arguments
The $INCLUDE statement for each pole of a converter station is shown below.
$INCLUDE DET-POLE CATHDY, ANODEY
- ; 1-2)
Nodes for Y-valve connection
CATHDD, ANODED
- ; 3-4)
Nodes for D-valve connection
SMOVLV, SMOPOL
- ; 5-6)
Nodes for smoothing reactor
UMEPOS, UMENEG
- ; 7-8)
- ;
Nodes for voltage measurement
IMEPOS, IMENEG
- ; 9-10)
- ;
Nodes for current measurement
VB%BU
- ; 11)
- ;
Nodes for AC connection,

Phase identification A, B, or C
- ; 12-14) Commutation voltages for Y/Y

- ;
C%U14D, C%U36D, C%U52D - ; 15-17) Commutation voltages for Y/D
- ;
C%Udi0
- ; 18)
Udi0, unfiltered (pu)
RECTIFIER
1 (>0.5) = rectifier
0 (0.5) = inverter
2 (>1.5) = reduced rectifier
-1 (<-0.5) = reduced inverter
UDCpunit, UDCnom
- ; 20-21) Initial DC bus voltage in pu and

- ;
kV
IDCpunit, IDCnom
- ; 22-23) Initial DC current in pu and

- ;
Ampere
UACpunit, UACnom
- ; 24-25) AC voltage (pu and kV 3-ph

- ;
RMS)
Udi0punit
- ; 26)
Initial Udi0 (pu)
UVnom
- ; 27)
- ;
Valve side AC voltage (kV 3ph RMS)
TAPPOS, TAPSIZ
- ; 28-29) Tap position and size
THETAinitA
- ; 30)
- ;
FREQnom, FREQinit
- ; 31-32) Nominal & initial frequency

- ;
(Hz)
19 - 18
; 19)
;
;
;
Phase A angle at time=0 (deg.)
High Voltage DC
ALPHAinit, GAMMAinit
- ; 33-34) Initial alpha and gamma (deg.)
ALPHAmin, GAMMAmin
- ; 35-36) Alpha and GAMMAmin limits

- ;
(deg.)
DXnom
- ; 37)
Converter dx (pu)
DRnom
- ; 38)
Converter dr (pu)
Pconverter
- ; 39)
- ;
Real power/pole from bus to

converter (MW)
Qconverter -
- ; 40)
- ;
Reactive power/pole from bus

to converter (MVAr)
- ; 41)
Converter identification
- ; 42)
Valve output (1=current, etc...)
indicates that the remainder of the card is to be interpreted as a comment. The "$INCLUDE DETPOLE" statement is found in the DET-STATION module and under normal circumstances should not
require modification. The arguments are described in Table 19.3 which follows.
ARGUMENT
1
TYPE
Node
WIDTH
6
UNITS
-
Node
Node
Node
Node
Node
Node
Node
Node
DESCRIPTION
CATHDY, Node name for connection of Y-valve,
Cathode.
ANODEY, Node name for connection of Y-valve,
Anode.
CATHDD, Node name for connection of D-valve,
Cathode.
ANODED, Node name for connection of D-valve,
Anode.
SMOVLV, Node name for valve-side connection of
smoothing reactor.
SMOPOL, Node name for line side connection of
smoothing reactor.
UMEPOS, Node name for cathode connection of voltage measuring device.
UMENEG, Node name for anode connection of voltage
measuring device.
IMENEG, Node name for connection of current measuring device (Positive).
Table 19.3: DET-POLE Module Arguments
19 - 19
High Voltage DC
10
Node
11
12
Node
1
5
6
pu
13
pu
14
pu
15
pu
16
pu
17
pu
18
19
6
9
pu
-
NC
20
21
22
23
24
25
26
27
NC
28
29
30
31
32
33
34
35
36
37
38
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
8
6
8
6
8
6
9
5
pu
kV
pu
A
pu
kV,rms
pu
kV,RMS
6
6
10
8
8
7
9
8
8
11
11
deg.
Hz
Hz
deg.
deg.
deg.
deg.
pu
pu
IMEPOS, Node name for connection of current measuring device (Negative).

VB%BU, Node name for connection to AC system.
C%U14Y, Commutation voltage for valve 1 + 4,
Y-bridge.
Y-bridge.
Y-bridge.
C%U14D, Commutation voltage for valve 1 + 4,
D-bridge.
D-bridge.
D-bridge.
C%UDI0, Unfiltered ideal direct voltage, a=0.
Identifier for operating mode.
1 (or >0.5) = rectifier operation
0 (or 0.5) = inverter operation
2 (or >1.5) = rectifier, reduced controls
1 (or <0.5) = inverter, reduced controls
UDCpunit, Initial DC bus voltage.
UDCnom, Nominal DC bus voltage.
IDCpunit, Initial DC current.
IDCnom, Nominal DC current.
UACpunit, Initial AC bus voltage.
UACnom, Nominal AC bus voltage. (3-phase)
Udi0punit, Initial ideal direct voltage, a=0.
UVnom, Nominal valve-side AC bus voltage.
(3-phase)
TAPPOS, Position of tap changer.
TAPSIZ, Size of tap changer step.
THETAinitA, Phase angle in phase A at t=0.
FREQnom, Nominal frequency.
FREQinit, Initial frequency.
ALPHAinit, Initial value of alpha.
GAMMAinit, Initial value of gamma.
ALPHAmin, Minimum extreme level of alpha.
GAMMAmin, Minimum extreme level of gamma.
DXnom, Relative inductive voltage drop.
DRnom, Relative resistive voltage drop.
19 - 20
High Voltage DC
39
NC
10
MW
40
NC
10
MVAr
41
Node
1
NC
42
1
NC = Numerical Constant
I
= Input signal to module
Pconverter, Power/pole (12-pulse bridge) from bus to

converter.
Qconverter, Reactive power/pole (12-pulse bridge)
from bus to converter.
!, Pole identification, single character.
?, Valve output (1=current, 2=voltage etc...)
19.7.3
Module Constants
The main control parameters at the pole level are defined in the calculation mode area of this
module. Remaining control parameters are given default values in the sub modules.
19.7.4
External Control and Protection Signals
The external signals listed below represent the actions of higher level control systems and/or
protection systems not explicitly included in the HVDC model. The actions of the associated
control/protection systems during system disturbances can be expected to vary from application to
application. It is therefore left to the user to implement the necessary control functions, either by
user-created modules or simply defined as TACS-supplemental cards. All signals, with the
exception of IORD, are of the logical type, i.e., 0.0 or 1.0.
IORD:
IORD is an "analog" signal representing the current order in pu,
i.e., 1.0 pu for normal operation. The signal can be supplied by the
user as a constant or a function of time, or created by a higher
level control function.
AL90:
Alpha 90 firing is ordered when this logical signal is high.
XBLK:
When the XBLK "logical" signal is high, X Blocking is ordered, i.e., all
firing pulses are interrupted.
ZBLK:
When this "logical" signal is high, Z Blocking is ordered, i.e., the

converter is blocked with a bypass pair. This signal is normally
activated by a higher level function or protection system. In the
HVDC model, the signal can be defined by the user as shown in the
examples below, or defined by a user-created function.
19 - 21
High Voltage DC
RETA:
The retard order is a "logical" signal activated by higher level

functions and protection systems. In the HVDC model, the user
must define the signal or create a higher level function for this
purpose. Suggested guidelines for the use of the RETA signal are
as follows.
For rectifier operation, the retard order is normally activated by
the DC line protection system. In the HVDC model, the signal
RETA should be set by the user during DC line faults
For inverter operation; the retard order is activated by the
subfunction 0 start for low DC bus voltage and other special
conditions. The signal remains active until the direct voltage is
restored (above a reference value).
REST:
The restart order for the rectifier should be high after DC line faults
until either the direct current is restored or a specified period of
time has elapsed (100 ms). The signal forces a to a
predetermined value specified in the model.
Examples:
99C1XBLK = 1.0*(TIMEX.GT.50.E-3)
99C4RETA = 0.0
99C3REST = 1.0*((TIMEX.GT.50.E-3) .AND. (TIMEX.LT.150.E-3))
In example 1, XBLK is activated after 50 ms for the pole 1.
In example 2, RETA is never activated for the pole 4.
In example 3, REST is activated between 50 and 150 ms for the pole 3.
19.7.5
Reduced Control System
For many applications, the complete HVDC control system is not required for the phenomenon
under investigation. The control system can then be reduced, resulting in a substantial savings in
computation time with no significant loss in accuracy. The disabling of selected control system
features is accomplished automatically via argument 1 of the DET-STATION module, or argument
19 of the DET-POLE module, as noted in Section 19.6 and Section19.7.
Reduced Rectifier:
For rectifier operation, the control system is reduced by disabling the modules associated with the
commutation margin reference and bypassing the double overlap (very rare) protection. By setting
the argument RECTIFIER equal to 2 (common to both the DET-STATION and DET-POLE modules),
the applicable modules are $DISABLED using the keywords "ARPcontrol" and "REDRECTcontrol".
In addition, some internal control signals are assigned dummy values.
19 - 22
High Voltage DC
Reduced Inverter:
For inverter operation, there is feedback of the filtered predicted area error in the commutation
margin control. The time constant associated with this feedback is in the range of 100-200 ms.
During large AC system disturbances, the area error calculation is disabled by the error counter
signals. For some applications, this area error feedback can be $DISABLED. By setting the
argument RECTIFIER equal to -1 (common to both the DET-STATION and DET-POLE modules), the
applicable modules are $DISABLED using the keywords "AERcontrol" and "REDINVcontrol". The
AERR signal is then also given a dummy value.
19.8 Converter Control System Model

As noted above, the control principle used in the detailed HVDC model is equidistant firing as
described in Reference 10. The primary functions of the control system include:
Voltage dependent current order limiter
Constant current control
Rectifier operation on minimum delay angle
Inverter operation on minimum extinction angle
An overview of the HVDC control system is shown below in Figure 19.5.
Parameters
min, min, etc.
Udms
Iord
VDCOL
Cord
Voltage Dependent
Current Order Limiter
CCA
ord
CFC
CPY1 ... CPD6

Current Control
Amplifier
Converter Firing
Control
Idms
Figure 19.5: Overview of HVDC Control System

The complete control system is provided with default parameter values. These values are
applicable for typical HVDC studies. The selection of the optimal set of control parameters is
performed at the detailed design stage of an HVDC system. All control system signal names are
summarized in Section 19.6.
19 - 23
High Voltage DC
19.8.1
Voltage Dependent Current Order Limiter (VDCOL)
The function of the VDCOL is to reduce the station current order when the direct voltage has
dropped below a prescribed value. This is necessary to:
Avoid an AC voltage instability which can occur if a high current is forced into a
weak inverter AC network because of a low voltage,
Attain a controlled recovery after disturbances and to, as far as possible, avoid commutation failures during recovery, and
Assist the AC system in recovering from faults by limiting the current which must be
commutated between valves during inverter operation.
The EMTP module VDCOL is placed ahead of the AMPLIFIER module in the path of the current
order. There are two inputs, current order (defined by the user) and measured DC voltage. The
module output is a modified current order for input to the current control amplifier. Figure 19.6
illustrates the block diagram for the VDCOL module.
UDFLT3
Z-1
A
A>B
B
UDCmax
Udms
1
1 + sT1
O
O
1
1 + sT2
IOabsmax
UDFLT2
UDFLT1
UDFLT4
1.0
IOabsmax
UDCbreak
Output from this

filter follows the
input when A is
high, i.e., T2 is 0.
MAXCOR
IOabsmax
Iord
1.0
CORMAX
1.0
COMAMI
IOlowlim
IOabsmax
IORMAX
MAX
IORMIN
Cord
MIN
IOminlim
Figure 19.6:
Voltage Dependent Current Order Limiter (VDCOL)

19 - 24
High Voltage DC
The module has two distinct sections. Section 1 filters the DC voltage through an asymmetrical
filter to obtain UDFLT2. Section 2 compounds the filtered DC voltage UDFLT2 with an input
current order to obtain a dynamic output current order CORD. The VDCOL generates the
characteristic shown in Figure 19.7 below. The characteristic consists of a family of curves with
each curve representing a particular input current order, IORD.
Cord
Maximum limit
IOabsmax
Cord
dependent on
Iord
IOlowlim
Minimum limit
IOminlim
UDCbreak
Udms
Figure 19.7: VDCOL Characteristic
Arguments:
The $INCLUDE statement for the VDCOL module is shown below.
$INCLUDE VDCOL C!CORD
- ; 1)
Limited current order
C!UDMS
2)
Measured DC bus voltage
C!IORD
3)
Unlimited current order
UDCmax
4)
Maximum DC bus voltage limit (pu)
Timedown
5)
Filter timeconstant, down (sec.)
TimeupR
6)
Filter timeconstant, up Rectifier (sec.)
19 - 25
High Voltage DC
TimeupI
7)
Filter timeconstant, up Inverter (sec.)
IOabsmax
8)
Maximum current order (pu)
UDCbreak
9)
Break point, DC bus voltage (pu)
IOlowlim
;
;
10)
Maximum current order at low voltage

(pu)
IOminlim
11)
Minimum current order (pu)
RECTIFIER
12)
RECTIFIER if > 0
UDCpunit
13)
Initial DC bus voltage (pu)
IDCpunit
14)
Initial direct current (pu)
indicates that the remainder of the card is to be interpreted as a comment. The "$INCLUDE
VDCOL" statement is found in the DET-POLE module and under normal circumstances should not
require modification. The arguments are described in Table 19.4.
Argument Type
1
Width Units
6
Deg
Variable
CORD
Description
Default
Current order from

VDCOL
pu
UDMS
Measured direct voltage
Deg
IORD
Current order from

higher level control
NC
pu
UDCMAX
Maximum DC bus voltage limit
NC
Sec.
Timedown
Down filter time constant 0.005

(down)
NC
Sec.
TimeupR
Up filter timeconstant
(Rectifier)
0.070
NC
Sec.
TimeupI
Up filter timeconstant
(Inverter)
0.055
NC
pu
IOabsmax
Maximum current order
1.5
NC
pu
UDCbreak
DC bus voltage break
0.6
pu
10
NC
2.0
point,
IOlowlim
Maximum current order

at low voltage
Table 19.4: VDCOL Module Arguments
19 - 26
0.45
High Voltage DC
11
NC
pu
IOminlim
Minimum current order
12
NC
RECTIFIER
Identifier for operating

mode
1 (or > 0.5) = rectifier
operation
0 (or 0.5) = inverter
operation
13
NC
pu
UDCpunit
Initial DC bus voltage
14
NC
pu
IDCpunit
Initial DC bus current

(absolute value)
0.2

I
Table 19.4: VDCOL Module Arguments
19.8.2
Constant Current Control (AMPLIFIER)
The constant current control is performed by a Current Control Amplifier (CCA). The CCA
includes a PI-controller as well as current margin and dynamic limitations.
Current Control Amplifier (CCA):

The CCA is modelled by a classic PI-controller in cascade with a low pass filter. The transfer
function G(s) below describes the amplifier characteristic. The firing angle output is related to the
current error input by G(s).
1 + sT 2
G ( s ) = --------------------------------------sT 1 ( 1 + sT 3 )
where T3 < T2. The block diagram

of the CCA is shown in Figure 19.8.
19 - 27
High Voltage DC
RETA
REST
1
0
CCAconstant
ALORDP
AOMA
MaxDif
Cord
Idms
1
1 + sT1
IDMSFI
CURDIF
AORD
AOMI
MinDif
AIMA
0
CCAcurrIMRG
O
O
K1
s
DIFMRG
CCAcurrMRG
ALORDI
AIMI
RECTIFIER
The limitations AIMA and AIMI are created by a submodule, which includes the dependence of
the signals RETA, REST, IORD and UDIO.
The limitations AOMA and AOMI are also created by a submodule.
Figure 19.8:
Current Control Amplifier (CCA)
Current Margin:
To achieve a proper operating point, the inverter CCA must be provided with a current margin.
This margin is normally 0.0 pu for rectifier operation and 0.1 pu for inverter operation. The
implementation of the current margin is shown in Figure 19.8.
Dynamic Limitations:
The alpha order output from the CCA is provided with dynamic limitations. Using dynamic limits,
alpha 90, retard, and restart functions can be performed. In addition, the inverter characteristic can
be modified to provide a positive slope on the voltage-current characteristic in the current margin
area for better stability. The dynamic limitations are calculated in the EMTP modules AORDLIM and
AORDILIM, which are supplied with default values for all control parameters.
The retard and restart functions are performed in the AORDILIM module by modifying the
limitations AIMI and AIMA respectively. The AORD signal will then be determined by the integrator
in the CCA, with the proportional part of the CCA temporarily set to zero when the retard or restart
order is activated as shown in Figure 19.8.
19 - 28
High Voltage DC
Arguments:
The $INCLUDE statement for the AMPLIFIER module is shown below.
$INCLUDE AMPLIFIER
C!AORD
1)
Alpha order (deg.)
C!CORD
;
;
2)
Current order and measured current

(pu)
C!IDMS
;
;
3)
Current order and measured current

(pu)
C!RETA
4)
Retard and restart order (logical)
C!REST
5)
Retard and restart order (logical)
C!AIMI
6)
Alpha limits, integrator part (deg.)
C!AIMA
7)
Alpha limits, integrator part (deg.)
C!AOMI
8)
Alpha limits, total (deg.)
C!AOMA
9)
Alpha limits, total (deg.)
Tlpass
;
;
10)
Time constant T1, lowpass filtering

(sec.)
MinDif
11)
Min difference, Cord-Idms (pu)
MaxDif
12)
Max difference, Cord-Idms (pu)
CCAconstant
;
;
13)
Constant in the proport. part(deg/

pu)
CCAcurrImrg
14)
Current margin, inverter (pu)
CCAIconstant
;
;
15)
Constant in the integer. part(deg/

(pu*sec.))
IDCpunit
16)
Initial DC current (pu)
ALPHAinit
17)
Initial alpha (deg.)
RECTIFIER
18)
Rectifier > 0.5
indicates that the remainder of the card is to be interpreted as a comment. The "$INCLUDE
AMPLIFIER" statement is found in the DET-POLE module and under normal circumstances should
not require modification. The arguments are described in Table 19.5.
19 - 29
High Voltage DC
Argument
Type
Width
Units
Variable
DESCRIPTION
DEFAULT
Deg
AORD
Alpha order
pu
CORD
Current order from VDCOL
pu
IDMS
Measured direct current
Logical
RETA
Retard order
Logical
REST
Restart order
Deg
AIMI
Alpha minimum limit (integrator)
Deg
AIMA
Alpha maximum limit (integrator)
Deg
AOMI
Alpha minimum limit (output)
Deg
AOMA
Alpha maximum limit (output)
10
nc
Sec.
Tlpass
Timeconstant of low pass fil- 0.006

ter (amplifier input)
11
nc
pu
MinDif
Minimum limit on difference

between idms and cord
-5.0
12
nc
pu
MaxDif
Maximum limit on difference between IDMS and
5.0
CORD
13
nc
11
deg/pu
CCAconstant
Amplifier gain (proportional) 100.0
14
nc
11
pu
CCAcurrlmrg
Current margin for inverter
0.10
15
nc
12
CCAlconstant
Amplifier constant (integrator)
2000.
16
nc
pu
17
nc
Deg
18
nc
deg/(pu x
s)
IDCpunit
Initial DC bus current
ALPHAinit
Initial firing angle alpha ()
RECTIFIER
Identifier for operation mode

1 (or > 0.5) = rectifier operation
0 (or 0.5) = inverter operation

I
Table 19.5: AMPLIFIER Module Arguments
19 - 30
High Voltage DC
19.8.3
Converter Firing Control
Minimum Delay Angle Control (UMIN Firing):

The delay (or firing) angle, a, is normally controlled during rectifier operation to indirectly
regulate the DC current to the requested current order via the rectifier DC voltage. During
disturbances resulting in reduced commutation voltage at the rectifier AC bus, the delay angle is
reduced to increase the rectifier DC voltage and maintain the requested (or VDCOL-limited) current
level. However, firing can not occur until the minimum delay angle limit, amin, is reached. After
firing in the amin position, a symmetrizing function is activated to maintain equidistant firing for
the next eleven valves. If, during an interval of symmetrical firing, the current control requests an
increase in a, normal firing control will resume.
The control function just described for minimum delay angle firing at amin is typically called
UMIN firing due to the way the function is performed. The UMIN function is performed by
comparing the commutation voltage and a reference value corresponding to amin and nominal
commutation voltage. The amin value is normally 5 but can be defined by the user and passed to
the EMTP module UMINAMIN.
Extinction Angle Control (AMIN Firing):

The extinction angle gamma, g, is controlled primarily during inverter operation to directly
regulate the system DC voltage. The CCA alpha order is overridden if the minimum extinction
angle, gmin, is reached. After firing in the gmin position, a symmetrizing function is activated to
keep the firing equidistant for the next eleven valves. If the CCA requests a decrease in a, the
current control will override the symmetrization.
The control function just described for minimum extinction angle control is called AMIN firing due
to the way the function is performed. The AMIN function is performed by predicting the
remaining voltage-time area of the commutation voltage and comparing the predicted area with a
reference value. This reference value is calculated by
sin ( t ) dt
from
min
t = ---------------------o
to
t = ------o
which corresponds to gmin and nominal commutation voltage. The gmin value is normally 17
but can be user defined and passed to the EMTP module COMMARG.
19.8.4
External Control and Protection Signals
The HVDC control system requires additional signals defined by the user to represent the actions of
higher level control and/or protection systems not explicitly included in the model. The actions of
19 - 31
High Voltage DC
the associated control/protection systems during system disturbances can be expected to vary from
application to application. It is therefore left to the user to implement the necessary control
functions, either by user-created modules or simply defined as TACS-supplemental cards. All
signals, with the exception of IORD, are of the logical type, i.e., 0.0 or 1.0.
IORD:
IORD is an "analog" signal representing the current order in pu,

i.e., 1.0 pu for normal operation. The signal can be supplied by
the user as a constant or a function of time, or created by a
higher level control function.
AL90:
Alpha 90 firing is ordered when this logical signal is high.
XBLK:
When the XBLK "logical" signal is high, X Blocking is ordered, i.e.,

all firing pulses are interrupted.
ZBLK:
When this "logical" signal is high, Z Blocking is ordered, i.e.,

the converter is blocked with a by-pass pair. This signal is
normally activated by a higher level function or protection
system. In the HVDC model, the signal can be defined either by
the user or by a user-created function.
RETA:
The retard order is a "logical" signal activated by higher level

functions and protection systems. In the HVDC model, the user
must define the signal or create a higher level function for this
purpose. Suggested guidelines for the use of the RETA signal
are as follows.
For rectifier operation; the retard order is normally activated by
the DC line protection system. In the HVDC model, the signal
RETA should be set by the user during DC line faults.
For inverter operation; the retard order is activated by the
subfunction g0 start for low DC bus voltage and other special
conditions. The signal remains active until the direct voltage is
restored (above a reference value).
REST:
The restart order for the rectifier should be high after DC line
faults until either the direct current is restored or a specified
period of time has elapsed (100 ms). The signal forces a to a
predetermined value specified in the model.
19.9 DC Line Module

The DC line module is application dependent and is normally created from the DC line
configuration using the AUX support program. (Rulebook 2, Section 9.4.6 and "FD-Line Model" in
Section 9.6.1) If the transient phenomena to be examined requires the use of a frequencydependent line model (FD-LINE), the line parameters should be determined using a low base
19 - 32
High Voltage DC
frequency (FMATRX) such that the steady-state DC voltage drop across the line is acceptable.
Sections of the DC line with the electrode line on the same tower as the pole conductors must be
handled as untransposed lines. If electrode resistances are significant, they should also be
included in the model. Normally, the DC line model is included in the main data file.
19.9.1
Arguments
The $INCLUDE statement for a sample DC line module is shown below.

$INCLUDE DCLINE
Bipolar DC line
DC1BUS, DL0011, DL0021, DL0031, DC4BUS
Pole 1
DC2BUS, DL0012, DL0022, DL0032, DC3BUS
Pole 2
DC1NTR, DC2NTR
Neutral Electrodes
indicates that the remainder of the card is to be interpreted as a comment. The arguments to the
DC line module are generally left to the user's discretion. Arguments for the sample module
provided with the model are shown in Table 19.6. Note that the anode and cathode node names
are reversed for inverter operation, as shown in Figure 19.2.
Argument Type
Width
Units
1
Node
6
2
Node
6
3
Node
6
4
Node
6
5
Node
6
6
Node
6
7
Node
6
8
Node
6
9
Node
6
10
Node
6
11
Node
5
12
Node
6
I
Description
Node name of cathode pole, station A
Internal node name, location 1, pole 1
Node name of anode pole, station B
Node name of anode pole, station A
Node name of cathode pole, station B
Node name of neutral bus, station A
Node name of neutral bus, station B
Table 19.6: DC Line Module Arguments
19 - 33
High Voltage DC
19.9.2
DC Resistance
In the DET-STATION module, the DC line resistance is required as an argument to the module and is
used in the initialization procedure for the model. For bipolar operation, the value required is the
resistance of one of the two poles. For monopolar metallic return operation, the resistance required
is that of both poles connected in series, i.e., twice the bipolar resistance. Finally, for monopolar
ground return operation, the resistance to be specified includes the pole resistance and that of the
electrode lines and electrodes. If the DC resistance is not specified properly, the initialization will
be unstable and additional time will be required for the model to reach steady-state.
For line parameters calculated at low frequency (0.01 Hz is the frequency used as "low" frequency
in HVDC model initialization), the DC resistance is given as:
R dc = ---- L
A
where:
Resistivity
Conductor cross sectional area
Conductor length
If the line parameters are calculated at higher frequencies, the DC resistance will differ from the
true DC resistance. The user can still determine the resistance in the line model by the following
method.
1.
In the DC line module, connect one pole to neutral at one end for ground return
operation, or connect both poles to ground at one end for bipolar or metallic
return operation. At the other end of the line model, connect two 0.01 Hz current
sources at +1.0 A and -1.0 A, respectively. For monopolar ground return
operation, connect the sources to the pole and neutral conductors. Otherwise,
connect the sources to the two pole conductors.
2.
Run an EMTP steady state solution with the specified current sources and the DC
line model, refer to Section 5.2.2 (KSSOUT) and Section 16.2. The DC resistance
for the line model can then be determined from the voltages at the source end of
the line. For bipolar operation, the pole voltage is equal to the DC resistance of
the pole. For monopolar operation, the DC resistance is equal to the sum of the
absolute values of the two pole voltages.
19 - 34
High Voltage DC
19.10 AC and DC Filter Modules

Typical AC and DC filter configurations for an HVDC station, along with the module names for the
various configurations, are shown in Figure 19.9. The AC filter modules supplied with the HVDC
model are three-phase models while the DC filter modules are single phase. Component values for
the filters will vary with application and are left to the discretion of the user.
*DBLE-BP
*DBLE-BP-R1
XXXBUS
*DBLE-BP-R1R2
XXXBUS
2
L1
L1
L2
C2
L1
R2
BOTTOM
L1
R1
L2
R2
BOTTOM
Figure 19.9:
* : Either "AC -"or
"DC - "
L2
where:
NN - area identifier, FA for ac filter
and FD for dc filter
x - station (ac) or pole (dc) ID
(Arg1)
y - bank ID (Arg 2)
z - Internal node ID (numeric)
A, B, C - phase identifier
D - sixth character for dc filter
All ac filters are three phase.

* : Either "AC -"or
BOTTOM
AC and DC Filter Modules
19 - 35
NNxyzA, NNxyzB, NNxyzC, NNxyzd
XXXBUS - dc or ac bus (Arg 3)

BOTTOM - filter neutral node
(Arg 4)
C2
R2
Internal Node Names:
XXXBUS
L2
R2
L1
where:
(Arg1)
y - bank ID (Arg 2)
BOTTOM
XXXBUS
C2
C2
L2
*DBLE-BP-RC-R1
R1

(Arg 4)
C2
*DBLE-BP-RC
XXXBUS
L1
BOTTOM
*DBLE-BP-R2
R1
L2
C2
BOTTOM
XXXBUS
"DC - "
High Voltage DC
*BP
*BP-QC
*BP-C
XXXBUS
*2ND-HP
*2nd-HP-QC
*2nd-HP-C
*MOD-HP
XXXBUS
XXXBUS
2
L1
L1
R1
C2

(Arg 4)
R12

* : Either "AC -"or
R2
"DC - "
L2
BOTTOM
BOTTOM
Figure 19.9:
19.10.1

where:
(Arg1)
y - bank ID (Arg 2)
1
C
BOTTOM
AC and DC Filter Modules
Arguments
The format of the $INCLUDE statement for several of the AC and DC filter configurations shown in
Figure 19.9 are illustrated below.
$INCLUDE DC-Dble-BP-RC
1)
Converter identifier
"
;
;
2)
Bank identifier (normally A, B, C,

D, E...)
DCxBUS
;
;
3)
Top of filter connection node, normally DC!BUS
BOTTOM
;
;
;
;
;
4)
Bottom of filter connection node,

for DC filter this argument will normally be the neutral bus node name
DC%NTR where % is station identifier.
TIMECLOSED
;
;
;
;
5)
The closing time of the filter

breaker, shall be negative if the filter is connected at steady state calculations.
TIMEOPENED
;
;
6)
The opening time of the filter

breaker
19 - 36
High Voltage DC
$INCLUDE AC-2ND-HP
;
;
;
;
;
7)
The output code (column 80 punch)

of filter bank breaker, 0=no output,
1=current,
2=voltage, 3=current+voltage,
4=power+energy
#CAPA1
8)
Main capacitor (F)
#INDU1
9)
Main reactor (mH)
#CAPA2
10)
Low voltage capacitor (F)
#INDU2
11)
Low voltage reactor (mH)
#RESI2
12)
Low voltage resistor ()
1)
Station identifier
"
;
;
2)
ACxBU
;
;
;
3)
Top of filter connection node, normally AC%BU for AC bus (phase

identifier A, B, C by def.)
BOTTOM
;
;
;
;
;
;
;
4)

for AC filters direct grounded this
argument shall be ______ (six
underscore). For AC filters with isolated neutral the argument will normally FAxyNE, where x is station ID
and y is bank identifier.
TIMECLOSED
;
;
;
;
5)

TIMEOPENED
;
;
6)

breaker
;
;
;
;
;
7)

1=current,
4=power+energy.
#CAPA1
8)
Main capacitor (F)
#INDU1
9)
Main reactor (mH)
19 - 37

D, E...)
High Voltage DC
#RESI1
$INCLUDE AC-MOD-HP
10)
Main resistor ()
1)
Station identifier
"
2)

D, E...)
ACxBU
;
;
;
3)
Top of filter connection node, normally AC%BU for AC bus (phase

identifier A, B, C by def.)
BOTTOM
;
;
;
;
;
;
;
4)

for AC filters direct grounded this
argument shall be ______ (six
underscore). For AC filters with isolated neutral the argument will normally FAxyNE, where x is station ID
and y is bank identifier.
TIMECLOSED
;
;
;
;
5)

TIMEOPENED
;
;
6)

breaker
;
;
;
;
;
7)

1=current,
4=power+energy.
#CAPA1
8)
Main capacitor (F)
#CAPA2
9)
Low voltage capacitor (F)
#INDU2
10)
Low voltage reactor (mH)
11)
Resistor ()
#RES12
The "-" in the $INCLUDE statements allows the statement to be continued on the next card. The ";"
indicates that the remainder of the card is to be interpreted as a comment.
The first seven arguments are common to all filter modules and define the filter connections to the
AC or DC system, closing time, opening time, etc. The final arguments define the filter component
values and vary with the filter configuration. The filter "$INCLUDE" statements are found in the
DET-STATION module and typically require modification of filter parameters and switching times
for the system under study. The order of the arguments can be found in Tables 19.7 through 19.9.
19 - 38
High Voltage DC
Argument Type
Width
Units
Node
2
3
Node
Node
1
5 AC
6 AC
Node
NC
10
sec.
NC
10
sec.
Description
Station ID for AC filters (%) and converter ID for
DC filters (!)
Bank ID (A, B, C, D, etc.) (")
Top filter connection node. Normally AC%BU
for AC bus (phase ID A, B, C by default), where %
= station ID.
Top filter connection node. Normally DC!BUS
for DC bus where ! = station ID.
BOTTOM, Bottom filter connection node.
For solidly grounded AC filters, this argument
will be ______ (six underscores). For
ungrounded AC filters, this argument will
normally be FAxyNE, where x is the station ID
and y is the bank ID. For DC filters, this argument
will be the neutral bus node name, normally
DC%NTR, where % is the station ID.
TIMECLOSED is the closing time of the filter
breaker. Negative if the filter is connected during
steady state conditions.
TIMEOPENED is the opening time of the filter
breaker.
? is the output code (column 80 punch) of the
filter bank breaker, i.e., 0=no output, 1=current,
2=voltage, 3=current & voltage, 4=power &
energy.

Table 19.7: AC and DC Filter Module Arguments
Parameter
Type
Width
Unit
C1
NC
6
F
Qfilter
NC
7
MVAr
UACnom
NC
6
kV
FREQnom
NC
7
Hz
Harmonic
NC
8
n
Qfactor
NC
7
-
BP-QC(ac)
10
11
12
2nd-HP-QC(ac)
10
11
12
BP-C (dc)
10
11
2nd-HP-C(dc)
11
11
Node = Numerical constant
Table 19.8: AC and DC Filter Modules Arguments - Basic Filter Types

19 - 39
High Voltage DC
Parameter
Type
Width
Unit
C1
NC
6
F
L1
NC
6
mH
R1
NC
6
W
C2
NC
7
Hz
L2
NC
6
mH
R2
NC
6
W
R12
NC
6
W
BP
10
2nd-HP-QC(ac)
10
11
BP-C (dc)
10
11
12
2nd-HP-C(dc)
10
11
12
10
11
12
13
10
11
12
10
11
12
13
10
10
11
NC = Numerical constant
Table 19.9: AC and DC Filter Modules Arguments - Detailed Filter Types
19.10.2 Internal Node Names

The internal filter node names follow the node name convention outlined in Section 19.4.2.
NNXYZA
NNXYZB
for AC filter phases A, B & C
NNXYZC
NNxyzD for DC pole filter
where:
NN
area identifier, FA for AC filter, FD for DC filter
station identifier for AC filters and pole identifier for DC filters
bank identifier (A, B, C, D,..)
internal node identifier (1,2,3,4..) (See Figure 19.9)
Using the internal node names, it is possible to connect additional components in parallel with the
filter components, i.e. DC filter arresters.
19 - 40
High Voltage DC
19.11 AC System Modules

The different AC system equivalents available are shown in Figures 19.10 through 19.13.
Although only one phase is shown in the figures, all AC system modules are represented internally
with three phases. The arguments to the module can be found in the accompanying tables. The
$INCLUDE statement for the "DETAILED" AC system representation module, found in DETSTATION, is shown below. Note that the "-" in the $INCLUDE statement allows the statement to be
continued on the next card. The ";" indicates that the remainder of the card is to be interpreted as
a comment.
$INCLUDE DETAILED
???
- ; 1)
AC system identifier
(AS1...AS9)
ACBUS
- ; 2)
- ;
AC bus node, excluding

phase identifier
BUSVOLTMAG
- ; 3)
- ;
Amplitude of bus voltage

(kV)
FREQinit
- ; 4)
- ;
Frequency of bus voltage

(Hz)
ANGLPHASEA
- ; 5)
- ;
Phase angle of A-phase

(B=A+240, C=A+120)
PfromACtoHVDC
- ; 6)
- ;
Active power from AC-network to bus (MW)
QfromACtoHVDC
- ; 7)
- ;
Reactive power from AC-network to bus (MVAr)
??RES1, ??RES2
- ; 8-9)
- ;
R2 ()
??INDL, ??CAPC
- ; 10-11)
- ;
Inductance L and capacitance

C (mH/F)
??RESN, ??INDN
- ; 12-13)
- ;
Resistance RN and inductance LN (/mH)
19 - 41
Resistance of resistor R1 AND
High Voltage DC

The internal node names follow the node name convention outlined in Section 19.4.2
NNXYYA
NNXYYB
for AC filter phases A, B & C
NNXYYC
where:
NNx
= AC system identifier, AS1 for AC system at station 1
yy
= internal node identifier (see Figures 19.10 through 19.13)
Using the internal node names, it is possible to specify internal module nodes as output variables
for printing and plotting. Connecting additional components in the positive sequence network from
outside the module is not recommended unless they are switched after t=0. Otherwise the internal
initialization calculation will fail. Switching events involving AC system equivalent components
after t=0 are permitted.
19 - 42
High Voltage DC
SQA
Phase A
0.001 mH
R2
1:1
0.001 mH
R1
Phase B
Phase C
RN
EQA
EQN
zzzzzA
where zzzzz is
argument 2
AQA

NNxyyA, NNxyyB, NNxyyC, NNxyyN
where: NNx - system identifier
yyA - internal node id (EQA, AQA, etc.)
LN
Argument
1
Type
Node
Width
3
Units
-
Description
AC system identifier. ASX for AC system at station
X
2
3
4
5
Node
4
10
8
10
kV
Hz
deg.
6
7
8
NC
MW
MVAr
NC
13
13
6
Active power from AC network to bus.

Reactive power from AC network to bus.
Resistance of R1
NC
Resistance of R2
NC
NC
NC
NC
AC bus node name at the station connected to the AC
Amplitude of AC bus voltage. (rms., three phase)

Initial frequency of the network
Phase angle of phase A at t=0 (cosine), B=A+240
NC
10
6
mH
Inductance of L
NC
11
6
mF
Capacitance of C
NC
W
12
6
Resistance of RN
NC
13
6
mH
Inductance of LN
NOTE: For values 8-13, a decimal must be given, due to EMTP FORTRAN fixed
format input
FIGURE 19.10: AC System Equivalent Module, DETAILED
19 - 43
High Voltage DC
Phase A
RS
SQA
zzzzzA
where zzzzz is
argument 2
EQA
RP
Argument
Type
Width
Units
Description
1
2
3
4
Node
Node
kV
Hz
AC system identifier. ASX for AC system at station x
NC
3
5
10
8
NC
10
deg.

NC
13
MW
7
8
NC
MVAr
NC
13
6
Resistance of RS
NC
Resistance of RP
10
NC
mH
Inductance of L
NC
format input
Figure 19.11:
AC System Equivalent Module, BASIC
19 - 44
High Voltage DC
Phase A
L1, L0
R1S, R0S
SQA
zzzzzA
where zzzzz is
argument 2
EQA
R1P, R0P
Argument
1
2
3
4
Type
Node
Node
NC
Width
3
5
10
8
Units
kV
Hz
NC
10
deg.

NC
13
MW
7
8
NC
MVAr
NC
13
6
Resistance of R1S
NC
Resistance of R1P
10
NC
mH
11
NC
12
NC
Inductance of L
Capacitance of C
Resistance of R0S
13
NC
mH
Inductance of R0P
NC
Description
AC system identifier. ASX for AC system at station x
format input
Figure 19.12:
AC System Equivalent Module, BASIC-ZERO (EMTP Type 51, 52, 53
Branches)
19 - 45
High Voltage DC
Phase A
L1 , L0
R1S, R0S
SQA
zzzzzA
where zzzzz is
argument 2
EQA
R1P, R0P
Argument
1
Type
Node
Width
3
Units
-
Node
Description
AC system identifier. ASX for AC system at station X
AC bus node name at the station connected to the
AC
3
4
5
6
7
8
9
10
11
NC
NC
10
8
8
10
13
13
10
8
7
kV,rms
-
12
NC
13
NC
14
NC
15
NC
16
NC
17
NC
NC
NC
NC
NC
NC
NC
NC
kV
Hz
Hz
deg.
MW
MVAr
MVA

Nominal frequency of the network
Short Circuit Capacity of network
Nominal voltage (3-phase)
"Quality factor" of network at nominal
frequency, X1/R1.
"Quality factor" of network at harmonic
frequency n,
The harmonic number, n, where the quality
factor above
The ratio between X0 and X1 (X0/X1). For
transmission
Zero sequence "Quality factor" of network at
nominal
Zero sequence "Quality factor" of network at
harmonic
The harmonic number, n, where the zero
sequence
NC
= Numerical Constant
Figure 19.13: AC System Equivalent Module, BASIC-ZERO-SK (EMTP Type 51, 52, 53 Branches)
19 - 46
High Voltage DC
19.12 6-Pulse HVDC Converter Module (SIXPULSEBR)

The module SIXPULSEBR models the basic six-pulse converter bridge unit used in the HVDC
model. Two six-pulse bridges are connected in series to obtain the 12-pulse unit used in the
model. The thyristor valves of the bridge are represented in the EMTP by the Type 11 TACScontrolled switch. The module is shown in Figure 19.14.
VB1TOP
INDUCT = Total
inductance for di/dt
limitation
5
INDUCT/2
VB1VYC
VB1VYA
VB1CY5
VB1VYB
Type 11
switch
VB1AY5
RESIST
CAPAC1
INDUCT/2
VB1MID
Node names shown for

valve bridge Y in pole 1
Figure 19.14: Six-pulse Bridge Module, SIXPULSEBR
19.12.1 Arguments
The $INCLUDE statement format for module SIXPULSEBR is illustrated below.
$INCLUDE SIXPULSEBR
VBxTY
;
;
5 character three VBxTY AC terminal

nodes
CATHOD
DC terminal node - cathode
ANODE#
DC terminal node - anode
CxCPY
Gate signals for CxCPY 1 to 6 (input)
Voltageig
;
;
Minimum ignition voltage (Volt), Default

=0
Iholdmin
;
;
Minimum holding current (Ampere),

Default = 0
TimeDEION
;
;
Minimum deionizing time (sec), Default

= 0.5 ms
19 - 47
High Voltage DC
RESIST
Snubber circuit resistance
CAPAC#
Snubber circuit capacitance
INDUCT
Inductance for di/dt limiter
ALPHAinit
;
;
Initial value of firing angle, alpha, in deg.
THETAinitA
Phase A angle at time=0
SHIFT
;
;
VBxTY phase shift (0 for Y-bridge, 30
;
;
Converter identification, single character
&
Bridge identifier, Y or D
Valve output (1=current, 2=voltage etc...)
for D-bridge)
The argument list for the module is shown in Table 19.10. The AC system connection for the
bridge is specified by argument 1. The node names supplied should preferably be of the form
VB!V&A, VB!V&B and VB!V&C, where ! and & represent the converter and six pulse bridge
identifiers (Y or D), respectively. The Type 11 switch parameters Vig, Ihold and tDEION are also
passed as arguments to the module. The de-ionization time, tDEION, in particular, can be used to
obtain a reasonable approximation of the commutation failure sensitivity of the valve.
Number
Type
Width
Units
Variable
Description
Node
VBxV%
Node name for connection to AC phases.

Sixth character in node
name will be phase
identifier.
Node
CATHOD
Node name for cathode

connection
Node
ANODEN
Node name for anode

connection
Node
CxCP%
Signal name for firing

pulses. Sixth character
in signal name will be
valve identifier (1-6).
Table 19.10: SIXPULSBR Module Arguments
19 - 48
Default
High Voltage DC
5
NC
Voltageig
Minimum ignition voltage.
0.0
NC
Iholdmin
Minimum holding current
0.0
NC
Sec.
TimeDEION
NC
RESIST
Resistance in snubber
circuit.
NC
mF
CAPACI
Capacitance in snubber
circuit.
10
NC
mH
INDUCT
Inductance for di/dt

limit.
11
NC
Deg
ALPHAinit
Initial value of alpha.
12
NC
10
Deg
THETAinitA
Phase A angle at
time=0 (cosine ref).
13
NC
SHIFT
Phase shift (0 for Y, 30

for Delta).
14
Node
Pole identification (single character).
15
Node
&
Bridge identifier (Y or
D)
16
NC
Valve output (1=current, 2=voltage)
Minimum deionizing
time.

Table 19.10: SIXPULSBR Module Arguments

The internal node names follow the node name convention outlined in Section 19.4.2.
VB!C&n for the cathode valve.
VB!A&n for the anode valve.
where ! = converter identifier and & = six pulse bridge identifier (Y or D)
19 - 49
0.5E-3
High Voltage DC
Using the internal node names, it is possible to connect arresters inside the converter bridge.
Figure 19.15 illustrates suggested arrester connections for the DC valve arresters.
VB1TOP
Measuring
Resistance (0.001 )
VB1ZY5
ZnO
VB1OY5
Measuring
Resistance (0.001 )
VB1VYC
Figure 19.15:
Suggested Arrester Connection for DC Valve Arresters
19.13 Converter Transformer Modules

There are four converter transformer modules used in the HVDC model. Each module represents
two 3-phase transformers, one connected Y-Y and the other Y-D, and differ only in the connection
of the primary neutrals. The transformer modules are valid for both single phase units or 3-phase
units with 5-leg cores. No mutual impedance is represented between the Y-Y and Y-D windings,
as is normally the case. The description of the four types of converter transformers are as follows:
Yy12d11
Both the primaries are Y connected and solidly grounded.
Yy12d11ug
Both the primaries are Y connected with the neutrals connected

to each other. The primaries are ungrounded.
Yy12ug-d11ug
Both the primaries are Y connected with the neutrals not

connected to each other. The primaries are ungrounded.
19 - 50
High Voltage DC
Yy12-d11ug
Both the primaries are Y connected. The Y-Y connected transformer is solidly grounded while the Y-D connected transformer
is ungrounded.
The magnetizing branch (saturation) is not represented in the module. However, it can easily be
added as an external element and must be connected on the AC side of the initialization switches
shown in Figure 19.4. If not, the magnetizing branch will be initialized with dc, resulting in large
initial currents at t=0.
The basic transformer module, Yy12d11, is shown in Figure 19.16. Figure 19.17 illustrates the
neutral connections for the transformer modules, Yy12d11, Yy12d11ug, Yy12ug-d11ug, and
Yy12-d11ug.
ex
TR11YA
VB1BUA
TR11YB
VB1BUB
TR11YC
VB1BUC
VB1TYA
VB1TYB
106
VB1TYC
TR11DA
VB1TDA
TR11DB
VB1TDB
TR11DC
VB1TDC
UNrms3PHline x (1 + TAPlevel x TAPPOS)
Yy12d11
Node names shown for transformer TR1 connected to the valve bridge in pole 1.
Figure 19.16:
er
Transformer Module, Yy12d11
19 - 51
UNrms3PHvalve
High Voltage DC
TR1Y0
TR1YY0
106
TR1YD0
106
Yy12d11ug
TR1YD0
106
Yy12ug-d11ug
106
Yy12-d11ug
Node names shown for transformer TR1 connected to the valve bridge in pole 1.
Figure 19.17: Transformer Neutral Connections - Yy12d11ug, Yy12ug-d11ug, and

Yy12-d11ug
19.13.1 Arguments
The $INCLUDE statement for the converter transformer modules Yy12d11, Yy12d11ug, Yy12d11ug, and Yy12ug-d11ug is illustrated below.
$INCLUDE Yy12d11
???
Transformer identifier
PRIM1
Primary node name
SECN1
;
;
Y-connected winding secondary node

name
SECN2
;
;
D-connected winding secondary node

name
UNrms3PHline
;
;
Nominal line side voltage (phase-phase,

rms)
UNrms3PHvalve
;
;
Nominal valve side voltage (phasephase, rms)
19 - 52
High Voltage DC
SN3ph3WIND
;
;
Nominal transformer power (3 phase, 3winding)
EXpercent
;
;
Relative short circuit reactance of transformer (%)
ERpercent
;
;
Relative short circuit resistance of

transformer (%)
TAPlevel
Tap changer position
TAPSIZ
Tap changer step size (pu)
FREQnom
Nominal frequency (Hz)
The description of the argument list can be found in the Table 19.11. The valve-side connections
of the transformer (secondary) are specified by arguments 3 and 4. The node names supplied
should preferably be of the form VB!T&A, VB!T&B and VB!T&C, where ! and & represent the
converter and six pulse bridge identifiers (Y or D), respectively.
Number
Type
Width
Units
Variable
Description
Node
???
Transformer identifier.
Normally TR%, where
% is the station id.
Node
PRIM1
Primary node name.

The sixth character in
the name will be internally set to the phase id.
Node
SECN1
Y-connected winding
secondary node name.
the name will be set to
the phase id.
Node
SECN2
D-connected winding
secondary node name.
the name will be set to
the phase id.
Default
Table 19.11: Transformer Module Arguments - Yy12d11, Yy12d11ug, Yy12ug-d11ug, &

Yy12-d11ug
19 - 53
High Voltage DC
5
NC
12
kV
UNrms3PHline
Nominal line side voltage (phase-phase, rms)
NC
11
kV
UNrms3PHvalv
Nominal valve side

voltage (phase-phase,
rms)
e
7
NC
10
MVA
SN3ph3WIND
Nominal transformer
power (3 phase, 3winding)
NC
EXpercent
Relative short circuit

reactance of transformer
NC
ERpercent
Relative short circuit

resistance of transformer
10
NC
TAPlevel
Tap changer position
11
NC
pu
TAPSIZ
Tap changer step size
0.0125
12
NC
Hz
FREQnom
Nominal frequency

Table 19.11: Transformer Module Arguments - Yy12d11, Yy12d11ug, Yy12ug-d11ug, &
Yy12-d11ug

The internal node names follow the node name convention outlined in Section19.4.2.
TRx1%A
TRx1%B
for the EMTP transformer node BUSTOP
TRx1%C
TRxYT0
for the valve-side wye neutral (always grounded)
TRxY0
for the primary wye neutral if ungrounded (Yy12d11ug)
TRxYY0
for the primary wye-wye neutral if ungrounded (Yy12ug - d11ug
19 - 54
High Voltage DC
TRxYD0
for the primary wye- neutral if ungrounded (Yy12 - d11ug, Yy12ug -
d11ug)
where:
Trx
transformer identifier (argument 1)
six pulse bridge identifier (Y or D)
By using the internal node names, it is possible to connect a neutral arrester inside a module with
an ungrounded primary neutral. The internal node names are shown in Figures 19.16 and 19.17.
19.14 Smoothing Reactor (SMOOTR)

The smoothing reactor is modelled by an inductance in parallel with a damping resistor. The
series resistance of the inductor is not modelled since the contribution of the smoothing reactor to
the resistive voltage drop (dr) is lumped with the transformer resistance and modelled in the
transformer. The module SMOOTR is shown in Figure 19.18.
RESIST
VB1TOP
DC1BUS
INDUCT
Node names shown for a smoothing reactor connected to the valve bridge in pole 1
Figure 19.18: Smoothing Reactor Module, SMOOTR
19.14.1 Arguments
The $INCLUDE statement for the module SMOOTR is illustrated below.
$INCLUDE SMOOTR
BRIDGE
Valve-side node 1 connection
DCPOLE
INDUCT
Inductance
RESIST
Parallel resistance
Reactor output, (1=current, 2=voltage, etc...)
19 - 55
High Voltage DC
indicates that the remainder of the card is to be interpreted as a comment. The description of the
argument list can be found in the Table 19.12.
Number
Type
Width
Units
1
2
3
4
Node
Node
mH
BRIDGE
NC
6
6
6
6
NC
NC
Variable
DCPOLE
INDUCT
RESIST
Description
Inductance
Parallel resistance
Reactor output, (1=current,
2=voltage, etc...)

Table 19.12: SMOOTR Module Arguments
19.15 Some Comments About Modules

The module handling features of the EMTP are described in detail in Section 19 of this rulebook.
The present section highlights the features most relevant to the use and modification of the HVDC
model.
19.15.1 Converting a Module

A module created or modified following the rules of EMTP data modules must be compiled before
it can be included in an EMTP data file. The conversion is most easily performed using the EMTP
interactively with the EMTP command, MODULE. The command requests both the name of the
input module file and the name of the output $INCLUDE file. The output file can then be included
directly in the EMTP data file. The conversion process can also be done in batch mode with an input
file of the form:
MODULE
xxx.MOD
xxx.INC
MODULE
yyy.MOD
yyy.INC
STOP
where xxx and yyy are the module names and .MOD and .INC are the typical file extensions used
for the module and $INCLUDE versions of the file.
19 - 56
High Voltage DC
19.15.2 Processing
When running an EMTP data case with modules calls ($INCLUDE), a special subroutine is
executed. This subroutine will:
Replace all variables declared as ARG with the actual node names.
Replace all variables declared as ARG and NUM with values.
Calculate all values in the CALCULATION MODE.
Replace all variables declared as NUM and DUM.
Replace all names declared as DUM by dummy node names (internal nodes).
Insert default values if the corresponding argument is not given.
Assemble all modules required to complete the EMTP data file.
Sort module sections according to EMTP card types (TACS, Branches, Switches,
Sources, etc.).
The module specifies which of the various data cards are contained therein by the keywords /TACS
DATA, /BRANCHES, /SWITCHES, /SOURCES, /OUTPUT, AND /PLOT. Any or all of the card types
may be used in any module file in any order as long as they are properly delineated. The main
data file must then contain BLANK cards written in the following format:
BLANK ENDS TACS
BLANK ENDS BRANCH
BLANK ENDS SWITCH
BLANK ENDS SOURCE
BLANK ENDS OUTPUT
BLANK ENDS PLOT
BLANK END CASE
(note without ENDS)
These blank cards are used as flags by the sorting subroutine. If they are not given as shown,
sorting can not be completed.
19.15.3 Special Characters

Two special characters can be used in modules. The characters # and _ indicate embedded blanks
in an argument. The _ character given in an argument will be replaced by a blank at all locations
in the module. The # character, on the other hand, will be replaced by a blank at all locations in
the module except where it appears in an $INCLUDE statement. All such arguments will be passed
19 - 57
High Voltage DC
to lower level modules unchanged, and the # will not be replaced until the lowest module is
processed.
These characters are thus mainly used to pass node names shorter than that requested by the
module. The argument ###### or ______ will be replaced by six blanks, i.e., ground.
19.15.4 FORTRAN Format

All constants passed to or defined in the CALCULATION MODE are handled as free format, i.e., the
decimal point is not required. The EMTP main program, on the other hand, reads input data with a
fixed format, normally E6.2 in six character numeric fields. Thus, a value written as 685 in the
input data file will be interpreted as 6.85 by the program using an E6.2 format. This can be
verified in the echo of the input data in the EMTP output file.
It is therefore recommended that arguments passed directly to the EMTP be given with the decimal.
The subroutine handling argument substitution makes no distinction between integer or real values
and does not remove the decimal from arguments during the processing of the module.
19.16 TACS Signal Names

All TACS external signal names are listed below, together with an explanation of their use. These
names should be considered as reserved. As written here, the converter identifier is x and the
station identifier is y. Signals recommended for output and plotting are noted in the table.
Signal Name
CyU14Y...CyU52D (14,36,52,... &
Y,D)
CxPRTR
CxTSYN
CxPTFA
CxPTSL
CxISY1...CxISD5 (1,3,5 & Y,D)
CxAMES
CxZCTI
CxZCXT
CxPZY1...CxPZD2 (1,2 & Y,D)
CxSCY1...CxSCD2 (1,2 & Y,D)
Description
Commutation voltage. (14 = valves 1 and
4)
Period time ramp
Trig for synchronization, voltage zeros
Period time "fast", unfiltered, repr. 180 deg.
Period time "slow", filtered, repr. 180 deg.
Individual trig for synchronization, valve
voltage zeros
Alpha, measured
Time to zero crossing
Time to next zero crossing
Predicted zero crossing
Selected clock pulse
19 - 58
Plot
High Voltage DC
CxAMRF
CxUCFI
CxPTZC
CxGAMI
CxAPRD
CxERRC
CxERR1
CxERR3
CxER11
CxER12
CxIORD
CxCORD
CxUDMS
CxIDMS
CxIPYA...CxIPDC (A,B,C & Y,D)
CxALCF
CxAORD
CxAIMI
CxAIMA
CxAOMI
CxAOMA
CxAPCO
CxAPCB
CxTPOR
CxTPOT
CxUMIN
CxBMAX
CxALMN
CxAMIF
CxUMIF
CxVFRY
CxVFRD
Area minimum reference, Commutation

voltage
Commutation voltage for valve just to be
fired
Predicted time to zero crossing
Gamma minimum firing
Area predicted, Commutation voltage
Error counter
Error signal 1
Error signal 3
Error signal 11
Error signal 12
Current order, to VDCOL
Current order, to Amplifier
Measured DC voltage
Measured DC current
Measured valve side transformer phase
current
Alpha limitation under commutation failure
Alpha order
Dynamic low limit for Current Control
Amplifier, integrator part
Dynamic high limit for Current Control
Amplifier, integrator part
Dynamic output low limit for Current
Control Amplifier
Dynamic output high limit for Current
Control Amplifier
Alpha, phase controlled oscillator
Alpha, phase controlled oscillator, buffered
Trig Pulse Oscillator (TPO) reference
Trig pulse to Trig Pulse Oscillator (TPO)
U minimum reference
Beta maximum reference
Alpha minimum reference
Area minimum firing, 11 following firing
U minimum firing, 11 following firing
Valve fired Y-bridge
Valve fired D-bridge
19 - 59
*
*
*
*
*
*
High Voltage DC
CxOLPY
CxOLPD
CxAEY1...CxAED2 (1,2 & Y,D)
CxAERR
CxCFLG
CxEMGC
CxEMGT
CxCPTR
CxCPTB
CxCPY
CxCPD
CxCPY1...CxCPD6 (1-6 & Y,D)
CxCBY
CxCBD
CxCBY1...CxCBD6 (1-6 &Y,D)
CyUDI0
CxUSYP
CxRETL
CxADVL
CxAL90
CxXBLK
CxZBLK
CxRETA
CxREST
Overlap Y-bridge
Overlap D-bridge
Area error in prediction, 1=1,3,5; 2=2,4,6
Filtered area error
Commutation failure flag
Emergency trig constant
Emergency trig
Clock pulse trig
Clock pulse trig, buffered
Clock pulse argument between modules,
bridge Y
Clock pulse argument between modules,
bridge D
Clock pulses for individual valves,
transferred between modules with
arguments above
Clock pulse, buffered, argument between
modules, bridge Y
Clock pulse, buffered, argument between
modules, bridge D
Clock pulses, buffered, for individual
valves, transferred between modules with
arguments above
Udi0, measured, (NOTE: number 0, not
alphabetic O)
Unsymmetry, predicted
Retard limit, DALIM
Advance limit, DALIM
Alpha 90 operation
Signal for X-block
Signal for Z-block (with bypass pair)
Signal for ordering retard
Signal for ordering restart
19 - 60
*
*
*
*
Section 20
20.1
General Comments .............................................................................. 20-1
20.2
Applications ........................................................................................ 20-1
20.3
Computer Aspects ............................................................................... 20-2
20.4
20.3.1
Time Step ....................................................................... 20-2
20.3.2
Dimensioning ................................................................. 20-2
Structure .............................................................................................. 20-2

20.4.1
Overview ........................................................................ 20-2
20.4.2
Internal EMTP Node Names ........................................... 20-3
20.4.3
Control System Signal Names ....................................... 20-4
20.4.4
Output and Plot .............................................................. 20-4
20.5
Main Data File ..................................................................................... 20-5
20.6
SVC Module Files ................................................................................ 20-6
20.6.1
Overview ........................................................................ 20-6
20.6.2
Top Level Modules - SVCA or SVCB .............................. 20-6
20.6.3
Distribution Unit, ALLOCATOR ...................................... 20-10
20.6.4
Flip-Flop, BCOUNTER .................................................... 20-12
20.6.5
Binary Function with Hysteresis, BFHYST ..................... 20-13
20.6.6
Damped Filter, DAMPFILT ............................................. 20-14
20.6.7
Transformation from 3-Phase to

(d,q)-components, DQTRNS ............................................ 20-15
20.6.8
TCR Current Measurement, ITCRMEAS ......................... 20-16
20.6.9
Measurement Circuit Type-A, MEASUREA .................... 20-17
20.6.10
Measurement Circuit Type-B, MEASUREB .................... 20-19
20.6.11
Phase Comparator Type-A, PHCOMPA .......................... 20-21
20.6.12
Phase Comparator Type-B, PHCOMPB ........................... 20-22
20.6.13
Proportional and Integral Regulator, PIREG ................... 20-24
20.6.14
Phase Locked Loop for One Phase, PLL1 ...................... 20-25
20.6.15
Phase Locked Loop for a 3-Phase System

Type-A, PLLA ................................................................. 20-27
Section 20
20.6.16

Phase Locked Loop for a 3-Phase System
Type-B, PLLB ................................................................ 20-28
20.6.17
Potential Transformer, PT .............................................. 20-29
20.6.18
Measurement of a 3-Phase Voltage or Current,

RMSMETER .................................................................... 20-30
20.6.19
Single Tuned Filter, STFILT ........................................... 20-31
20.6.20
Six-Pulse Delta-Connected TCR Unit, TCR6 ................. 20-32
20.6.21
Six-Pulse Delta-Connected TSC Unit, TSC6 .................. 20-34
20.6.22
TCR Firing Angle Calculation, TCRCONV ..................... 20-35
20.6.23
TCR Overcurrent Control, TCROVCUR .......................... 20-36
20.6.24
Firing Pulse Generator for a Six-Pulse

TCR Unit, TCRPULS ....................................................... 20-38
20.6.25
Firing Pulse Generator for a Six-Pulse

TSC Unit, TSCPULS ........................................................ 20-39
20.6.26
Secondary Overvoltage Limiter, UOVLIM ..................... 20-41
20.6.27
Voltage Response Filter Type-A, URFILTA ................... 20-42
20.6.28
Voltage Response Filter Type-B, URFILTB ................... 20-43
20.6.29
Undervoltage Strategy, UVSTRATEGY .......................... 20-43
20.6.30
Voltage Controlled Oscillator, VCO .............................. 20-45
20.6.31
Voltage Regulator, VREG .............................................. 20-47
20.6.32
Three-Phase Two-Winding Transformer, XFORM2W ... 20-49
20.6.33
Three-Phase Three-Winding Transformer, XFORM3W . 20-50
20.7
Initialization ........................................................................................ 20-51
20.8
Some Comments About Modules ....................................................... 20-52
20.9
TACS Signal Names ............................................................................ 20-52
Section 20
This chapter describes EMTP models for a representation of SVC configurations employing
Thyristor-Controlled Reactors (TCR) and Thyristor-Switched Capacitors (TSC) in six-pulse or
twelve-pulse arrangements with typical SVC control systems.
20.1 General Comments

The SVC model provides a representation of SVC configurations employing Thyristor-Controlled
Reactors (TCR) and Thyristor-Switched Capacitors (TSC) in six-pulse or twelve-pulse
arrangements with typical SVC control systems.
To maximize modelling flexibility, the model has been implemented as a set of modules or
building blocks, each representing a main circuit component or a particular control function.
Various SVC configurations can be simulated by combining these modules in a suitable way. This
modular approach requires the assistance of the EMTP data modules (EDM).
The SVC model represents the SVC main circuit components and various control functions. The
main circuit components are: step-down transformer, TCR unit, TSC unit, and harmonic filters.
These components are modelled with EMTP branch and switch cards. Thyristors are represented
by TACS controlled switches. Essential control circuits are: phase-locked-loop (PLL), voltage
regulator, measurement circuit, allocator, linearizer, and TCR and TSC firing circuits. Additional
control functions included in the model are: TCR overcurrent control, secondary overvoltage
limiter, and undervoltage strategies. Higher level controls such as power and reactive power
modulations are not included. The characteristics of higher level controls are normally unique for
each specific application and can easily be accomplished by the user through TACS.
20.2 Applications
The detailed SVC model is intended for the following applications:
Studying electromagnetic dynamics of a power system.
Studying electromechanical dynamics of an AC system.
General dynamic performance evaluations in the time range of up to 2 seconds.
20 - 1

The limitations of the detailed SVC model include:
The valves (thyristors) in this model are represented by TACS controlled switches and
the turn-on and turn-off characteristics are not modelled. The model is therefore not
suitable for designing the snubber circuit or for studying unbalances that may exist in
the actual circuits.
The model does not represent the stray capacitances and inductances and as such is
not adequate to verify protection from lightning surges.
The model does not include any provisions for delaying the response to abnormal
voltage or current conditions. These provisions can be added by the user.
Only one SVC model can be included in the system. If more than one SVC needs to be
modelled at the same time, the modules must be modified to give unique names to all
of the TACS variables that are not flagged as dummy variables.
20.3 Computer Aspects

20.3.1
Time Step
The maximum time step recommended for use with the detailed SVC EMTP model is 50 (s or
roughly 1o at 60 Hz. The time step can be reduced for higher frequency transients.
20.3.2
Dimensioning
The default list sizes should be adequate.
20.4 Structure
20.4.1
Overview
The detailed SVC model can be divided into three subsystems, namely:
1.
The external AC equipment such as the step down transformer and the filter
banks.
2.
The thyristor switches and the switched inductors and capacitors.
3.
The control system.

20 - 2

There are essentially no restrictions on the layout of the AC system connected to the SVC. The
model does create several internal node names which are discussed below.
Figure 20 -1 presents an overview of the control structure of the SVC model.
EMTP Data File
SVCx
Filters
Transformer
TCR
Controls
TSC
Figure 20.1: Overview of EMTP SVC Model

The EMTP data file contains all of the cards needed to define the system to which the SVC is
connected. It also calls the SVCA module and passes the names of the high and low voltage buses,
the initial bus voltage, the initial susceptance setting for the SVC, and flags indicating the initial
status of the TSCs.
20.4.2
Internal EMTP Node Names
The internal node names used within the SVC model have been standardized using the six
character positions of the name as follows:
The first two positions are reserved for the module type.
CR for the TCR portion
SC for the TSC portion
SF for single tuned filters
DF for damped filters
XR for the internal nodes of the step down transformer
The third position of the name is a number identifying the particular unit
20 - 3
The last three digits identify the particular internal node. The exact method varies
depending on the module.
TCR and TSC:

Two types of representations are used within these two modules. The node names and TACS
variables associated with the thyristors use positions 4 and 5 to designate the connections to the
phases. The fourth position designates the nearest phase for the anode side of the thyristor and the
fifth position designates the nearest phase for the cathode. The sixth position is set to a D for the
anode side of the thyristor and to a K for the cathode side. The nodes in the TSC and TCR external
to the snubber and the branches representing the internal impedance of the thyristor use a V in the
fourth position, the phase (A, B, or C) in the sixth position, and a number (1, 2, or 3) designating
which the leg of the bridge in the fifth position.
Filters:
The fourth and fifth character in the filter names are M. and are used to designate the node
between the inductor and the capacitor. The sixth character is the phase of the branch (A, B, or C).
In addition, the neutral of the wye-connected filter is not grounded but is designated by the last
three characters ..N. If the filters are to be grounded, this node must be connected to ground in
the main EMTP data deck.
Transformer:
The only node names created for the transformers are for the internal nodes which uniquely
identify the transformers. The last three characters for these nodes are ..p where p represents the
phase (A, B, or C) of each winding.
20.4.3
Control System Signal Names
Named control signals are used for the interconnection of the low level control modules in the
model. Signals internal to a module are declared as dummy (DUM) and the actual names are
generated by the EMTP. The signal names which are not declared to be dummy variables are listed
in Section 20.9
20.4.4
Output and Plot
Output and batch-mode plotting cards would normally be inserted in the main data file, but they
can also be placed in any of the modules if desired. Some of the TACS output variables have been
pre-selected in the modules as shown in Section 20.9. Additionally, the thyristor voltage and
current signals have been selected by a 3 in column 80 of the TSC and TCR modules. The total TCR
currents are also pre-selected.
20 - 4

Internal TACS signals specified as dummy variables (DUM) within a module are not available as
output in the higher level modules. If an output other than those described above is desired, the
user must modify the appropriate module to allow for that specific output.
It should be noted that the number of output variables available are limited by the EMTP. The
number of output variables from the SVC model should be specified with care.
20.5 Main Data File

The main data file consists of standard EMTP cards, such as:
Miscellaneous data cards
TACS HYBRID card
TACS output requests
EMTP branch, switch, and source cards
Node voltage output requests
Batch mode plotting cards
BLANK ENDS cards separating the different data cards
In addition, the file must include a module request for either SVCA or SVCB to insert the SVC
model into the study file and must also include TACS source statements or assignment statements
that define the following TACS variables:
UREF
the target voltage for the voltage regulator of the SVC
FRZPL1
the PLL disable flag
FRZUMS
the voltage measurement disable flag
FRZVRG
the voltage regulator disable flag
FRZIOV
the overcurrent check disable flag
FRZUOV
the overvoltage check disable flag
DUREFR
user supplied voltage error modification
BLKTCR
flag to block TCR operation
USTBON
flag to choose filtered voltage measurement
20 - 5
20.6 SVC Module Files

20.6.1
Overview
The arguments passed to each module can fall into several classes. These classes can be described
as follows:
Node or signal name characters. - These will be in the form of character strings of
from 1 to 6 characters. The characters will either give the complete name of the node
or variable (generally the case when 6 characters are passed) or will be used as the
first characters to construct the final name. For instance, the first five characters of a
three phase node will be passed and the module will then add A, B, or C to the end.
Numeric variables. - These are TACS variable names whose value can be modified by
one of the modules and used in mathematical calculations by other TACS routines.
Numeric constants. - These are numeric values that are not modified during the EMTP
run. This would include such things as the equipment parameters, system initial conditions, etc.
Note that any TACS variables that are not marked as dummy in a module are available to any other
module. As a result, many modules use variables that are not specifically passed to them. Also, the
name that appears in a module as an argument may or may not be the same as the variable actually
used by the module. The variable used depends on the variable passed by the calling module.
20.6.2
Top Level Modules - SVCA or SVCB
Either of the modules SVCA or SVCB can be used to define the SVC to be studied. Both models have
an identical main circuit representation but differ in the control system representation. These
modules serve as the link to all of the other needed modules.
Arguments:
The $INCLUDE statement for the SVC modules are shown below. Note that SVCB has an additional
argument to directly measure the total current into the high voltage terminals of the transformer.
$INCLUDE
HTERM
- ; Node to which the high side of the SVC is connected
LTERM
- ; Node to which the low side of the SVC is connected
INITVOLTPU
- ; Initial voltage on the high voltage bus in per unit
BSTARTING
- ; Initial susceptance request (capacitive is positive)
SVCA
20 - 6
$INCLUDE
HTERM
- ; Node to which the high side of the SVC is connected.

- ; This node is connected to the high side of the
- ; transformer through a measuring switch.
ITERM
- ; Node to which the high side of the SVC is connected
LTERM
- ; Node to which the low side of the SVC is connected
INITVOLTPU
- ; Initial voltage on the high voltage bus in per unit
BSTARTING
- ; Initial susceptance request (capacitive is positive)
SVCB
The - in the $INCLUDE statement allows the statement to be continued on the next card. The ;
indicates that the remainder of the card is to be interpreted as a comment. The module requires the
arguments listed in the following table.
Argument
Type*
Width
Units
HTERM
Node
ITERM
Node
LTERM
Node
INITVOLTPU
NC
10
per unit
BSTARTING
NC
MVAr
* Node = Node or signal name characters

NC = Numeric constant
Table 1: SVCA and SVCB Arguments
20 - 7
PRMBS
XR1
SECBS
SC2
CR1
SC3
HARMONIC
FILTERS
SC1
NOTSCS
TSCPULS
URESPU
MEASUREA
UVSTATEGY
TSCPULS
PLLA
CLMPVR
UOVLIM
TSCPULS
ALLOCATOR
TSCPULS
TCRCONV
VREG
BREF
URFILTA
BTCR
TCROVCUR
UREF
Block Diagram of the SVCA Module
Figure 20.2:
PRMBS
XR1
SECBS
SC2
SC3
NOTSCS
TSCPULS
URESPU
MEASUREB
TSCPULS
PLLB
ALLOCATOR
URFILTB
UREF
VREG
BREF
TCRCONV
TSCPULS
TSCPULS
BTCR
Figure 20.3:
Block Diagram of the SVCB Module

20 - 8
CR1
HARMONIC
FILTERS
SC1
Module Constants:
The majority of the SVC constants that define the model are in the SVCA or SVCB modules. All of
the specifications in these modules are in the form of numeric constants passed to other modules.
The following table lists the modules called and the values assumed in the model.
Module
Called
Value in
Model
Description
DAMPFILT
0.06
2.41
61.2
25.1
Coil resistance in Ohms

Coil inductance in mH
Filter Capacitance in F
Damping resistance in Ohms
MEASUREA
345.0
550.0
0.01
Base voltage in RMS kV phase-to-phase

Base susceptance in MVAr
Slope of the SVC steady-state characteristic
MEASUREB
345.0
0.0
550.0
0.01
Base voltage in RMS kV phase-to-phase

Initial PLL A-phase angle
PLLA
18.0
0.0
60.0
Base voltage on low side of transformer in RMS kV phaseto-phase

System operating frequency
PLLB
18.0
0.0
60.0

RMSMETER
1.0
18.0
Initial system voltage in per unit

STFILT
0.06
4.89
61.2
Coil resistance in Ohms

Coil inductance in mH
Filter Capacitance in (F
TCRCONV
163.0
Susceptance of the TCR when fully conducting
TCROVCUR
3200.0
3200.0
1.3E-5
TCR current reference

Initial TCR current
Integrator gain
Table 2: Fixed Values used by SVCA and SVCB
20 - 9

UOVLIM
425
0.100
Initial value for the overvoltage

Sample interval in seconds
UVSTRATEGY
1.0
0.69
0.60
Initial state of module (1 = no blocking)

Blocking stopped when voltage above this level
Blocking started when voltage below this value
VREG
-125.0
425.0
Minimum susceptance allowed in MVAr

Maximum susceptance allowed in MVAr
XFORM2W
300.0
345.0
18.0
0.1
60.0
Base MVA of transformer

Primary side voltage in RMS kV phase-to-phase
Secondary side voltage in RMS kV phase-to-phase
Leakage reactance in pu of the transformer base MVA
Table 2: Fixed Values used by SVCA and SVCB
20.6.3
Distribution Unit, ALLOCATOR
The distribution unit converts the susceptance reference (BREF) signal from the voltage regulator
to logical orders (on/off signals) for the TSCs and arithmetic orders for the TCRs. The module
outputs two quantities:
1.
NOTSCS - the number of TSC units required to be on, and
2.
BTCR - the susceptance order for the TCR.
The output signals are piecewise functions of BREF. Since the range of BTCR is normally greater
than the susceptance of each TSC, there is a hysteresis between BREF and the two outputs. In the
hysteresis region, there are two output states producing the same MVAr supply from the SVC to the
network.
The susceptance quantities are related as follows:
BREF = - BTCR + (Total TSC Susceptance) + (Total filter Susceptance).
20 - 10
TSCMVr
_
BREF
Filter
MVAr
BTCR
OR
A
A
A>B
TCRMVAr
A<B
B
Figure 20.4:
S&H
S
NOTSCS
Block Diagram of the ALLOCATE Module
Arguments:
The following is a list of the arguments passed to the ALLOCATOR module in the order that they
are specified.
Argument Type*/
Name
Size
Value
in
Model
Description
NOTSCS
NV/6
Output of the ALLOCATOR module
BTCR
NV/6
Output of the ALLOCATOR module
BREF
NV/6
Calculated by the VREG module
MAXNUM
NC/6
The number of TSCs installed
Table 3: Arguments for ALLOCATOR
20 - 11

INITNUM
NC/7
Calculated in SVCA or SVCB based on the values of STEP1,

STEP2, and STEP3. This is the number of TSCs energized at
t = 0.
* NC = Numeric constant
NV = Numeric variable
Table 3: Arguments for ALLOCATOR
Module Constants:
There are several numeric constants in the ALLOCATOR module that would need to be changed if a
different SVC is to be studied. These are the boundaries of the piecewise representation of BREF
and the maximum inductive limits for BTCR. These values are summarized in the following table.
Number
BREF
of TSCs minimum
BREF
maximum
BTCR
maximum
-125.0
30.0
161.108
2.0
164.0
150.311
132.0
294.0
147.070
267.0
435.0
160.000
Table 4: Fixed Values internal to ALLOCATE
20.6.4
Flip-Flop, BCOUNTER
The BCOUNTER is a binary counter or a flip-flop circuit. The status of the flip-flop switches
between zero and one when the counter receives an input pulse. The module generates two
outputs:
1.
PPULSE indicating the one status, and
2.
NPULSE indicating the zero status.

PPULSE
Input
S&H
Figure 20.5:
NPULSE
Block Diagram of the BCOUNTER Module

20 - 12
Arguments:
The following is a list of the arguments passed to the BCOUNTER module in the order that they are
specified.
Argument
Name
Type* /
Size
Value
in
Model
Description
PPULSE
NV/6
1 status of the flip-flop
NPULSE
NV/6
0 status of the flip-flop
TRIGER
NV/6
Timing signal to switch flip-flop
ANGLPHASEA
NC/10
0.0
Initial phase angle for phase A
* NV = Numeric variable
Table 5: Arguments for BCOUNTER
20.6.5
Binary Function with Hysteresis, BFHYST
The binary function BFHYST has two possible states: zero and one. The functional value changes
from zero to one when the input increases above a set level (level-high) and changes from one to
zero when the input decreases below another level (level-low). The level-high is greater than the
level-low.
Input
Level Low
Input
Level High
A>B
&
T
OR
A>B
Output
Figure 20.6:
Block Diagram of the BFHYST Module
20 - 13
Arguments:
The following is a list of the arguments passed to the BFHYST module in the order that they are
specified.
Argument
Name
Type* /
Size
Value
in
Model
Description
OUTPUT
NV/6
Output of the module
VINPUT
NV/6
Input signal to be compared to the limits
INITVAL
NC/7
1.0
Initial status of the output signal
HIGHERLEVL
NC/10
0.69
Level for transition from 0 to 1
LOWERLEVEL
NC/10
0.60
Level for transition from 1 to 0
Table 6: Arguments for BFHYST
20.6.6
Damped Filter, DAMPFILT
This is one of the two types of harmonic filters included in the SVC model. The DAMPFILT consists
of three single-tuned filter branches in a wye-connection. Its neutral is floating, but it can be
grounded through an external impedance branch if desired. Electrical characteristics of damped
filters are discussed in Reference 20.
ACBUSA
L mH
ACBUSB
L mH
DF1M.A
CF
ACBUSC
DF1M.B
CF
L mH
DF1M.C
CF
DF1...N
Figure 20.7:
Circuit Diagram for the DAMPFILT Module
20 - 14
Arguments:
The following is a list of the arguments passed to the DAMPFILT module in the order that they are
specified.
Argument Type* /
Name
Size
Value
in
Model
Description
DF1
NODE/3
Component name
ACBUS
NODE/5
Connecting node name
RRRRRR
NC/6
0.06
Resistance of inductive branch - Ohms
LLLLLL
NC/6
2.41
Inductance - mH
CCCCCC
NC/6
61.2
Capacitance - F
DAMPER
NC/6
25.1
Damping resistance - Ohms

Table 7: Arguments for the DAMPFILT Module
20.6.7
Transformation from 3-Phase to (d,q)-components, DQTRNS
This module is used only if SVCB is selected as the main SVC module and is called from the
MEASUREB module.
The (a,b,c)-phase components of a balanced 3-phase quantity can be transformed to (d,q)
components relative to a reference phasor. This is achieved by applying the (,)-transformation
and the symmetrical component transformation. The (,)-transformation is defined by
v
v
va
1 2 1 1
= --vb
3 0
3 3
vc
where v,v,v are the phase components of a three-phase quantity and v,v are the (a,b)- components.
The symmetrical component transformation is defined by
where (=t) is the angle of the reference phasor and v,v are the (d,q)-components.
20 - 15
vd
vq
cos ( ) sin ( ) v
sin ( ) cos ( ) v
Arguments:
The following is a list of the arguments passed to the DQTRNS module in the order that they are
specified.
Argument Type* /
Name
Size
Value
in
Model
Description
ISVC1D
NV/6
D-axis component, output by DQTRNS
ISVC1Q
NV/6
Q-axis component, output by DQTRNS
ISVC.
Node/5
First 5 characters of the signal name
PL1
Node/3
Name of the PLL used
NV/6
Calculated by MEASUREB
SIGNAL
Table 8: Arguments for the DQTRNS Module
DQTRNS contains a numeric value for omega calculated for 60 Hz operation. If the frequency is
different than 60 Hz, this constant must be changed and the module recompiled.
20.6.8
TCR Current Measurement, ITCRMEAS
Measurement of currents in TCR branches is provided for the TCR overcurrent control. The average
current is calculated over a half cycle of the fundamental frequency (T):
t+T
t
1
ITCR av = --- i ( t ) dt i ( t ) dt
T
0
20 - 16
PL1PLS
TCR
current
S&H
S
ITCR
Figure 20.8:
Block Diagram of the ITCRMEAS Module
Arguments:
The following is a list of the arguments passed to the ITCRMEAS module in the order that they are
specified.
Argument
Name
Type* /
Size
Value
in
Model
IRMS1A
NV/6
Measured RMS current - output by module
PL1
Node/3
Name of the PLL device
CURTCR
Node/6
Name of the current signal
3200
Initial value of the measured current
INITAMVALU NC/10
Description
Table 9: Arguments for the ITCRMEAS Module
20.6.9
Measurement Circuit Type-A, MEASUREA
This module is called only if SVCA module is used. The MEASUREA module generates two signals
to be used by the voltage regulator:
1.
URESP
2.
URESP
20 - 17

The URESP signal is a measurement of the power system voltage response in per unit. To obtain
this, the RMS voltage is measured with the module RMSMETER and then divided by the nominal
voltage.
The URESP signal represents the signal required to achieve a desired slope in the steady-state UI
characteristics. The slope is typically between 1 % to 10 %. The signal is calculated by
URESP
BREF
- ---------------------------------------------- slope
URESP = ------------------------Base KV S uscep tan ceBase
PL1SMP
Three-phase
Voltage
RMSMETER
S&H
URESP
Base kV
URESP
BREF
Base Susceptance
Slope
Figure 20.9: Block Diagram of the MEASUREA Module
Arguments:
The following is a list of the arguments passed to the MEASUREA module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
URESPU
NV/6
Measured RMS voltage in per unit of BASERMSKV
DURESP
NV/6
Modification to voltage response due to the slope
Description
Table 10: Arguments for the MEASUREA Module
20 - 18

SOURC
Node/5
Name of node
INITPUVALU
NC/10
Initial value of voltage. Specified in call to SVCA or SVCB
BASERMSKV
NC/9
PL1
Node/3
FREEZE
NV/6
345.0
Base voltage on high side of SVC in kV phase-to-phase

Signal to freeze voltage measurement. Must be set in main
EMTP deck.
BREF
NV/6
Calculated by VREG
BMVARBASE NC/9
550.0
NC/7
0.01
SLOPEPU
Table 10: Arguments for the MEASUREA Module
20.6.10 Measurement Circuit Type-B, MEASUREB

This module is only called if SVCB is used. The MEASUREB module generates the two signals
URESP and URESP. The purpose of the MEASUREB module is identical to that of the MEASUREA
module. However, this module utilizes the transformation method of the DQTRNS module. First,
the transformed voltages (v,v) and currents (i,i) are calculated by measuring the input voltages
(v,v,v) and currents (i,i,i) relative to a (d,q)-reference frame synchronous to the PLL output.
The URESP signal is obtained by URESP =
2
2
v d + v q . The reactive power is obtained by
QMVAr = vq id - vd iq.
The URESP signal is calculated by

BaseKV
QMV Ar
URESP = ---------------------- ------------------------------ slope
URESP BaseMVAr
20 - 19
BCOUNTER
Slope
Ia
abc
Ib
Vq * Id - Vd * Iq
Ic
dq
PL1ARM
Va
Base MVAr
URESP
abc
Vb
SQRT(Vd * Vd + Vq * Vq
Vc
dq
PL1ARM
Base kV
Figure 20.10:
Block Diagram of the MEASUREB Module
Arguments:
The following is a list of the arguments passed to the MEASUREB module in the order that they are
specified.
Value
in
Model
Argument
Name
Type* /
Size
URESPU
NV/6
Measured RMS voltage in per unit of BASERMSKV
DURESP
NV/6
Modification to voltage response due to the slope
VSVC.
Node/5
Name of node where voltage is measured
ISVC.
Node/5
Name of node where current is measured
PL1
Node/3
FREEZE
NV/6
Signal to freeze voltage measurement. Must be set in

main EMTP deck.
INITPUVALU
NC/10
Description
Initial value of voltage. Specified in call to SVCA or

SVCB
Table 11: Arguments for the MEASUREB Module

20 - 20

BASERMSKV
NC/9
ANGLPHASEA NC/10
345.0
Base voltage on high side of SVC in kV phase-to-phase
0.0
Calculated by VREG
BMVARBASE
NC/9
550.0
SLOPEPU
NC/7
0.01
Table 11: Arguments for the MEASUREB Module
20.6.11 Phase Comparator Type-A, PHCOMPA

This module is called only if SVCA is used. The PHCOMPA module calculates the phase error
between a single-phase voltage v(t) and a PLL1 output f(t). The f(t) is a square-wave of the PLL
frequency. The error is given as a function of the phase difference. The phase error signal e() is
calculated as below:
t0 + T
e ( ) = ------2T
f ( t )v ( t ) dt
t = t0
where tis a time of zero-crossing for the f(t), and T the period of the PLL. If the input signal is a
pure sine wave, e() = sin , where is the phase error in radians.
20 - 21
PL1ADT
&
PL1ANC
SOURCA
K
S
PL1APC
PL1APC
PL1ADT
PL1ATG
S&H
ERRPHS
&
PL1APC
SOURCA
K
S
PL1ANC
Figure 20.11: Block Diagram for the PHCOMPA Module
Arguments:
The following is a list of the arguments passed to the PHCOMPA module in the order that they are
specified.
Argument Type*/
Name
Size
Value
in
Model
Description
ERRPHS
NV/6
The phase error. Output of module.
SOURCA
NV/6
Single phase reference voltage.
PL1
Node/3
Name of PLL device.
Table 12: Arguments for the PHCOMPA Module
20.6.12 Phase Comparator Type-B, PHCOMPB

This module is only called if SVCB is used. The PHCOMPB module calculates the phase difference
between the PLL and the three-phase synchronizing voltages v,v,v. First, v and v are obtained by
using the (,)-transformation:
20 - 22
va
1 2 1 1
= --v
3 0 3 3 b
v
vc
The phase error signal is generated by

e ( ) = v sin ( ) v cos ( )
abc
UNORMA
UNORMB
UNORMC
PL1ARP
sin ()
PL1ARM
cos ()
Figure 20.12:
Block Diagram of the PHCOMPB Module
Arguments:
The following is a list of the arguments passed to the PHCOMPB module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
ERRPHS
NV/6
The phase error. Output of the module.
SOURCA
NV/6
A-phase reference voltage.
SOURCB
NV/6
B-phase reference voltage.
SOURCC
NV/6
C-phase reference voltage.
Description
Table 13: Arguments for the PHCOMPB Module

20 - 23

PL1
Node/3
ANGLPHASEA NC/10
Name of the PLL device.
0.0
Initial phase angle for phase A.
Table 13: Arguments for the PHCOMPB Module
20.6.13 Proportional and Integral Regulator, PIREG

The PIREG module is used for the phase-locked-loop (network synchronization) and for the voltage
Ki
control. This regulator has the transfer function G(s) = K+ ----- . The parameter Kp determines the
S
proportional response and Ki is the gain of the integral part. The regulator output is subjected to
upper and lower dynamic (non-windup) limits.
Upper Limit
KP
Input
Freeze
KI
Output
1
S
Lower Limit
Figure 20.13:
Block Diagram of the PIREG Module
Arguments:
The following is a list of the arguments passed to the PIREG module in the order that they are
specified.
Argument
Name
Type*/
Size
TOTOUT
NV/6
Value
in
Model
-
Description
Output of module.
Table 14: Arguments for the PIREG Module

20 - 24

ERRINP
NV/6
Input of module, supplied by calling module.
FREEZE
NV/6
Signal to suspend integration.
INITIALPER
NC/10
Initial value of the integrator.
KAYPEEVALU NC/10
Proportional gain constant.
KAYAYEVALU NC/10
Integral gain constant.
MINLIM
NV/6
##
Lower limit of the output.
MAXLIM
NV/6
##
Upper limit of the output.
# The value depends on which module calls PIREG
## These could either be a fixed number or a variable depending on the calling module.
Table 14: Arguments for the PIREG Module
20.6.14 Phase Locked Loop for One Phase, PLL1

This module is used in the PLLA module (type-A PLL circuit) to accomplish the phase-locking for
an individual phase. The PLL1 module consists of the following module blocks:
1.
flip-flop (BCOUNTER)
2.
phase comparator (PHCOMPA)
3.
proportional and integral regulator (PIREG)
4.
voltage controlled oscillator (VCO).
The PLL1 produces an output pulse train at the rate of its internal frequency (f). The PLL1
accomplishes this by using a VCO which runs at the rate twice the internal frequency and a flipflop which divides the frequency by half. The phase comparator (PHCOMPA) calculates the
control-loop error as a function of the phase difference between the input and output signals. The
error signal is fed to a PI regulator which controls the frequency of the VCO. The error is reduced
by the negative feed-back control with proper loop gains. When the error is zero, the PLL1 is
locked to the input signal. When the PLL1 is locked, the synchronizing pulses will come exactly at
the positive and negative peaks of the input AC voltage.
The PLL1 generates the following signals:
1.
PLnpRM - a ramp function which is proportional to the PLL angle,
2.
PLnpTG - the pulse train indicating the moment of voltage peaks,
20 - 25

3.
PLnpSM - the pulse train advancing the PL1ATG signal by 1.5 ms,
4.
PLnpPC, PLnpNC - the outputs of the flip-flop.
The PLL function is deactivated when the input signal FRZPL1 becomes active (i.e. FRZPL1=1).
The integral part of the PIREG maintains a constant output until the signal becomes inactive.
FRZPL1
BASE
FREQUENCY
(Freeze)
Max f deviation
UNORMA
_
PHCOMPA
PL1APC
PL1ANC
PIREG2
VCO
PL1ATG
Min f deviation
BCOUNTER
Figure 20.14:
Block Diagram of the PLL1 Module
Arguments:
The following is a list of the arguments passed to the PLL1 module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
PL1A
NODE/4
Name and phase of the PLL.
SOURCA
NV/6
Per unit phase voltage from PLLA.
NV/6
Flag to stop PLL action (FRZPL1). Must be set in the main

EMTP deck.
LOWERLIMIT
NC/10
-15.0
Lower limit for frequency deviation.
UPPERLIMIT
NC/10
15.0
Upper limit for frequency deviation.
ANGLPHASEA NC/10
0.0
Initial phase angle for the input phase voltage.
FREEZE
Description
Table 15: Arguments for the PLL1 Module
20 - 26

BASFREQNCY NC/10
60.0**
System base frequency from PLLA.
** This is the default value in PLLA but the call to PLLA from SVCA can override the value.
Table 15: Arguments for the PLL1 Module
20.6.15 Phase Locked Loop for a 3-Phase System Type-A, PLLA

This module is only called if SVCA is used. The PLLA module represents a typical PLL circuit. The
module consists of three PLL1 modules to achieve a six-pulse SVC operation. Two PLLA modules
are required for a twelve-pulse SVC configuration.
The module generates the following signals:
1.
PL1ARM, PL1BRM, PL1CRM - ramp function for the three phases,
2.
PL1ATG, PL1BTG, PL1CTG - pulse trains indicating the voltage peaks,
3.
PL1APC, PL1ANC, PL1BPC, PL1BNC, PL1CPC, PL1CNC - the output of the flip-
flops,
4.
PL1PLS - summation of the signals PL1ATG, PL1BTG, PL1CTG,
5.
PL1SMP - summation of the signals PL1ASM, PL1BSM, PL1CSM.
The PL1SMP signal is used to update the content of control signals.
UPHASA
Voltage Rating
UPHASB
Voltage Rating
UPHASC
PL1ATG
PLL1
OR
PL1PLS
OR
PLLSMP
PL1ASM
PL1BTG
PLL1
PL1BSM
PL1CTG
PLL1
PL1CSM
Voltage Rating
Figure 20.15: Block Diagram of the PLLA Module

20 - 27
Arguments:
The following is a list of the arguments passed to the PLLA module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
PL1A
NODE/4
UPHAS
NODE/5
Name of node where voltage is measured.
NV/6

EMTP deck.
FREEZE
REFSYSTEMKV NC/11
18.0
ANGLPHASEA
NC/10
0.0
BASFREQNCY
NC/10
60.0**
Description
Phase-to-phase voltage in kV RMS.

Table 16: Arguments for the PLLA Module
20.6.16 Phase Locked Loop for a 3-Phase System Type-B, PLLB

This module is called only when SVCB is used. The PLLB module represents another type of PLL
circuit. Operation of this module is similar to that of the PLLA module except that synchronization
is carried out simultaneously for the three phases. This PLL circuit calculates the control loop error
using the PHCOMPB module.
Limits
FRZPLL
va
vb
vc
(Freeze)
BASE
FREQUENCY
BCOUNTER
PHCOMPA
PIREG2
VCO
BCOUNTER
PL1ARM
VCO
BCOUNTER
VCO
Figure 20.16:
PL1APC
PL1ANC
Block Diagram of the PLLB Module

20 - 28
PL1BPC
PL1BNC
PL1PLS
OR
PL1CPC
PL1CNC
Arguments:
The following is a list of the arguments passed to the PLLB module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
PL1A
NODE/4
UPHAS
NODE/5
Name of node where voltage is measured.
NV/6

EMTP deck.
REFSYSTEMKV
NC/11
18.0
Phase-to-phase voltage in kV RMS.
LOWERLIMIT
NC/10
-15.0
Lower limit for frequency deviation.
UPPERLIMIT
NC/10
15.0
Upper limit for frequency deviation.
ANGLPHASEA
NC/10
0.0
BASFREQNCY
NC/10
60.0**
FREEZE
Description
Table 17: Arguments for the PLLB Module
20.6.17 Potential Transformer, PT

This module represents an idealized transformer that can be used for voltage measurement
between any two points in the power system.
VBUSK
N1 : N2
VOUT
VBUSM
Figure 20.17:
Circuit Diagram of the PT Module

20 - 29
Arguments:
The following is a list of the arguments passed to the PT module in the order that they are specified.
Argument Type*/
Name
Size
Value
in
Model
Description
VBKM.A
NV/6
Name of secondary side of transformer - output of module.
BUSK.A
NODE/6
Name of first primary node.
BUSM.A
NODE/6
Name of second primary node.
PRIME.KV
NC/8
1.0
Primary side voltage.
SECND.KV
NC/8
1.0
Secondary side voltage.
Table 18: Arguments for the PT Module
20.6.18 Measurement of a 3-Phase Voltage or Current, RMSMETER

This module is called only when SVCA is used. The RMSMETER measures the magnitude of threephase voltages or currents. The measurement involves rectification and active filtering. The threephase voltage signals are individually rectified and added to obtain a six-pulse rectified response.
The rectified signal is then filtered in a 360 degree running average filter to eliminate the
fundamental frequency and its harmonics.
PHASEA
PHASEB
K
S
OUTPUT
PHASEC
T
Figure 20.18:
Block Diagram of the RMSMETER Module
20 - 30
Arguments:
The following is a list of the arguments passed to the RMSMETER module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
QNTRMS
NV/6
Per unit RMS output value.
SIGNL
NODE/5
Three phase quantity to be measured.
INITPUVALU
BASERMSKV
Description
NC/10
1.0
Initial value of output which is maintained for the first

cycle.
NC/9
18.0
Base voltage in phase-to-phase kV RMS.
Table 19: Arguments for the RMSMETER Module
20.6.19 Single Tuned Filter, STFILT

This is one of the two types of harmonic filters included in the SVC model. The STFILT consists of
three single-tuned filter branches in a wye-connection. The neutral is floating, but it can be
grounded through an external impedance branch if desired. Electrical characteristics of singletuned filters can be found in Reference 20.
ACBUSA
ACBUSB
SF1M.A
SF1M.B
ACBUSC
SF1M.C
SF1...N
Figure 20.19:
Circuit diagram for the STFILT Module

20 - 31
Arguments:
The following is a list of the arguments passed to the STFILT module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
SF1
NODE/3
Component name.
ACBUS
NODE/5
Connecting node name.
RRRRRR
NC/6
0.06
Resistance of inductive branch - Ohms.
LLLLLL
NC/6
4.89
Inductance - mH
CCCCCC
NC/6
61.2
Capacitance - F
Description

Table 20: Arguments for the STFILT Module
20.6.20 Six-Pulse Delta-Connected TCR Unit, TCR6

A three-phase TCR unit is formed by three delta-connected single-phase branches. Each branch has
a thyristor valve unit connected in series between two reactors. The valve unit is modelled as a pair
of back-to-back thyristors in parallel with a simple snubber circuit. Each thyristor has been
modelled by the type-11 TACS controlled switch of the EMTP. The thyristors in the three-phase TCR
unit are controlled as a six-pulse group by gate signals coming from a TCRPULS module. A TCR is
normally connected to the AC system through a step-down transformer (module XFORM2W).
A twelve-pulse TCR configuration consists of two TCR units and a three-phase three-winding
transformer with wye-connected and delta-connected secondary windings (module XFORM3W).
The twelve-pulse TCR requires two six-pulse control systems.
20 - 32
SVCBSA
SVCBSB
CR1V1A
CR1ABD
CR1V2B
CR1BCD
CR1BAK
CR1ABG
CR1V3C
CR1CBG
CR1BCK
CR1BAD
CR1V1B
CR1CAD
CR1CBK
CR1BCG
CR1BAG
CR1ABK
SVCBSC
CR1CAG
CR1CBD
CR1ACG
CR1CAK
CR1V2C
CR1IAB
CR1ACK
CR1IBC
CR1ACD
CR1V3A
CR1ICA
o
o
o
o
o
o
SVCBSB
SVCBSC
SVCBSA
Figure 20.20:
Circuit Diagram for the TCR6 Module
Arguments:
The following is a list of the arguments passed to the TCR6 module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
CR1
NODE/3
Component name.
ACBUS
NODE/5
Description

Table 21: Arguments for the TCR6 Module
The reactance value for each phase-to-phase leg of the TCR is set in this module to 7.915 mH
which corresponds to 163 MVAr at 18 kV phase-to-phase. If a different rated TCR is to be used,
this value must be changed and the module re-compiled.
20 - 33
20.6.21 Six-Pulse Delta-Connected TSC Unit, TSC6

A three-phase TSC consists of three single-phase branches in delta connection. Each phase of a TSC
branch has a thyristor valve unit connected in series between a capacitor and a tuning reactor or a
current limiting reactor. The thyristor valves in each phase switch capacitor banks either fully on or
fully off. Each TSC unit is controlled by signals coming from a TSCPULS module. Identical TSC
units are often connected in parallel so that the valve-control might provide multiple-step changes
in capacitive MVAr output from the SVC.
SVCBSA
SVCBSB
CR1V1A
CR1ABD
CR1V2B
CR1BAK
CR1ABG
CR1BCD
CR1BAG
CR1ABK
SVCBSC
CR1V3C
CR1CBK
CR1BCG
CR1BAD
CR1CBG
CR1BCK
CR1V1B
CR1CBD
CR1CAD
CR1CAG
CR1ACG
CR1CAK
CR1V2C
SVCBSB
CR1ACK
SVCBSC
CR1ACD
CR1V3A
SVCBSA
Figure 20.21: Circuit Diagram for the TSC6 Module
Arguments:
The following is a list of the arguments passed to the TSC6 module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
CR1
NODE/3
Component name.
ACBUS
NODE/5
Description

Table 22: Arguments for the TSC6 Module
20 - 34

The capacitance value for each phase-to-phase leg of the TSC is set in this module to 315.0 F
which corresponds to 115 MVAr at 18 kV phase-to-phase. If a different rated TSC is to be used, this
value must be changed and the module re-compiled.
20.6.22 TCR Firing Angle Calculation, TCRCONV

The TCRCONV module converts the susceptance order (BTCR) from the ALLOCATOR to a firing
angle order which is measured from the time of the last voltage peak. To maintain the same
control response over the entire SVC operating range, the angle is determined as a non-linear
function of the BTCR. This function is given as a table. The table is derived from the following
formula:
sin ( )
1 X L B ( ) = + -------------------
where X is the TCR reactance at the fundamental frequency, B() the susceptance of the TCR
fired at , and the angle in per unit of 90 degrees.
The module generates two output signals derived from the :
1.
ALPHON - the angle given in per unit of 90 degrees,
2.
ALPHOF - the angle of thyristor turn-off.
The difference between ALPHON and ALPHOF is the duration of the firing pulses which is
approximately equal to the thyristor conduction angle ():
ALPHOF - ALPHON = .
1.0
BTCR
PL1HLD
S&H
Base
MVAr
_
S&H
T/2 sec
(1/2 cycle)
Figure 20.22:
Block Diagram for the TCRCONV Module

20 - 35
ALPHON
ALPHOF
Arguments:
The following is a list of the arguments passed to the TCRCONV module in the order that they are
specified.
Value
in
Model
Argument
Name
Type*/
Size
ALPH
NODE/4
Firing angle name - output of module.
PL1
NODE/3
Name of PLL circuit.
BTCR
NV/6
Susceptance order for TCRs from ALLOCATOR.
BMAXOFTCR
NC/9
163.0
Description
Susceptance of fully conducting TCR.
Table 23: Arguments for the TCRCONV Module
20.6.23 TCR Overcurrent Control, TCROVCUR

The TCR overcurrent control prevents excessive current in the TCR during conditions of high
voltage by imposing a delay of the trigger pulses to the TCR. This added delay reduces the TCR
current to a preset level. The module consists of three control loops operating independently for
individual phases. These loops are slower than the voltage regulator loop. The module also
includes ITCRMEAS modules for the TCR current measurement.
20 - 36
PH1HLD
CR1IAB
ITCRMEAS
ITCR reference
1
S
S&H
ALDLAB
Lower Limit = 0
PH1HLD
CR1IBC
ITCRMEAS
ITCR reference
1
S
S&H
ALDLBC
Lower Limit = 0
PH1HLD
CR1ICA
ITCRMEAS
ITCR reference
1
S
S&H
ALDLCA
Lower Limit = 0
Figure 20.23:
Block Diagram for the TCROVCUR Module
Arguments:
The following is a list of the arguments passed to the TCROVCUR module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
Description
NODE/4
Name of modification to the firing angle name - output of

module.
PL1
NODE/3
CR1
NODE/3
Name of TCR circuit.
FRZIOV
NV/6
Signal to deactivate overcurrent control.
IREFERENCE
NC/10
3200.0
ALDL
Maximum continuous current for TCR.
Table 24: Arguments for the TCROVCUR Module
20 - 37

INITAMVALU
NC/10
3200.0
Initial value of TCR current.
INTEGRGAIN
NC/10
1.3E-5
Gain for the TCROVCUR integrator.
Table 24: Arguments for the TCROVCUR Module
20.6.24 Firing Pulse Generator for a Six-Pulse TCR Unit, TCRPULS

The module TCRPULS supplies firing pulses for a three-phase TCR unit. Inputs to the module are:
1.
signals from the PLL circuit ,
2.
ALPHON, ALPHOF from the TCRCONV module,
3.
modifications to the firing angle from TCROVCUR.
4.
BLKTCR signal set to one will block all three phases of the TCR units.
A firing pulse is formed when the elapsed time since the occurrence of the last peak voltage is
greater than the current value of the ALPHON signal but less than that of the ALPHOF. The pulse is
directed to one of the back-to-back thyristor pairs using the PLL flip-flop outputs.
SAMPLE
BLKTCR
oS&H
TCRDBK
PL1APC
&
ALPHOF
ALDLAB
CR1ABG
A
A>B
B
&
PL1ARM
A
ALPHON
ALDLAB
A>B
B
&
PL1ANC
Figure 20.24:
CR1BAG
Block Diagram of the TCRPULS Module
20 - 38
Arguments:
The following is a list of the arguments passed to the TCRPULS module in the order that they are
specified.
Argument Type*/
Name
Size
Value
in
Model
Description
CR1
NODE/3
Name of TCR circuit.
PL1
NODE/3
ALPH
NODE/4
Firing angle name.
NODE/4
Name of the modification to the firing angle form

TCRPULS.
NV/6
Signal to block the TCR.
ALDL
BLKTCR
Table 25: Arguments for the TCRPULS Module
20.6.25 Firing Pulse Generator for a Six-Pulse TSC Unit, TSCPULS

The TSCPULS module generates firing pulses for a three-phase TSC unit. Each TSC unit is assigned
a priority number by the user and the TSC is activated when the NOTSCS signal from the
ALLOCATOR is greater than or equal to this number. In this way, TSC units are turned on
sequentially from the lowest priority number to the highest. The BLKTSC signal from the
UVSTRATEGY module blocks all three phases of the TSC unit when BLKTSC = 1.
20 - 39
PL1APC
&
SC1ABG
&
SC1BAG
&
SC1BCG
PL1BNC
&
SC1CBG
PL1CPC
&
SC1CAG
PL1CNC
&
SC1ACG
PL1ANC
BLKTSC
PL1HLD
PL1BPC
NOTSCS
A
A>B
&
S&H
TSC Unit
Number
Figure 20.25:
Block Diagram of the TSCPULS Module
Arguments:
The following is a list of the arguments passed to the TSCPULS module in the order that they are
specified.
Argument Type*/
Name
Size
Value
in
Model
Description
SC1
NODE/3
Name of TSC circuit.
PL1
NODE/3
NOTSCS
NV/6
Number of TSCs required as determined by ALLOCATOR.
BLKTSC
NV/6
Signal to block the TSC.
ORDER
NV/5
The order in which the TSCs are turned on.
STDYST
NV/6
Initial status of TSC supplied in call to SVCA or SVCB.
Table 26: Arguments for the TCRPULS Module
20 - 40
20.6.26 Secondary Overvoltage Limiter, UOVLIM

The secondary overvoltage limiter prevents the possibility of excessive voltage on the secondary
side of the SVC transformer. The secondary voltage response is compared to the secondary voltage
limit. If the response exceeds the limit, the limiter output acts on the upper limit for the voltage
regulator output, forming an additional control loop. This loop is much slower than the primary
voltage regulator loop.
CLOCK
100 ms
V limmax
Secondary
voltage limit
Secondary
voltage response
1
S
S&H
DBMAX
V limmin
Figure 20.26:
Block Diagram of the UOVLIM Module
Arguments:
The following is a list of the arguments passed to the UOVLIM module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
Description
NV/6
Amount of reduction in the maximum susceptance limit

- output of module.
BREF
NV/6
Susceptance request by the voltage regulator.
V2RESP
NV/6
Voltage on low side of transformer in per unit.
FREEZE
NV/6
Flag to deactivate overvoltage protection.
INITVALUE
NC/8
425.0
DLTBMX
Initial output in MVAr.
Table 27: Arguments for the UOVLIM Module

20 - 41

SAMPLPERIOD
NC/11
0.10
Sample period in seconds.
Table 27: Arguments for the UOVLIM Module
Several parameters are set within the UOVLIM module. These are:
1.
The point where overvoltage protection starts - 1.22778
2.
The maximum amount of reduction - 550
3.
The integrator gain - 0.300
If the model is changed, it may be necessary to change one or more of these values and recompile
the module.
20.6.27 Voltage Response Filter Type-A, URFILTA

The voltage response is often filtered before fed to the voltage regulator to improve the control
system response. The URFILTA module represents a lead-lag circuit.
Arguments:
The following is a list of the arguments passed to the URFILTA module in the order that they are
specified.
Argument Type*/
Name
Size
URESPF
URESPU
Value
in
Model
Description
NV/6
Filtered output of the module.
NV/6
The measured voltage response from the voltage

measurement module.
Table 28: Arguments for the URFILTA Module
The module contains fixed values for the transfer constants in the form of:
1,0 + 0,013s
-----------------------------1,0 + 0,032s
20 - 42
20.6.28 Voltage Response Filter Type-B, URFILTB

The URFILTB module shows another type whose main purpose is to eliminate certain harmonics.
The module consists of three filters, a 60-Hz notch filter, a 120-Hz notch filter and a low-pass
filter.
Arguments:
The following is a list of the arguments passed to the URFILTB module in the order that they are
specified.
Argument Type*/
Name
Size
URESPF
URESPU
Value
in
Model
Description
NV/6
Filtered output of the module.
NV/6
The measured voltage response from the voltage

measurement module.
Table 29: Arguments for the URFILTB Module
The module contains fixed values for the transfer constants for each of the three filters in the form
of:
URESPU
s2 + 2
s2 + 2
s2 + 2
s2 + BS + 2
s2 + BS + 2
s2 + BS + 2
60 Hz notch filter
120 Hz notch filter
low pass filter
= 2f, f = 60
= 2f, f = 120
= 2f, f = 85
Q = = 3.2
B
Q = = 3.2
B
Q = = 0.8
B
Figure 20.27:
URESPF
Block Diagram of the URFILTB Module
20.6.29 Undervoltage Strategy, UVSTRATEGY

An SVC control normally has a logic to apply predetermined control actions during conditions of
abnormally low AC voltages. The UVSTRATEGY module implements a typical strategy. The
20 - 43

module takes the following actions when the voltage response signal from the measurement circuit
drops below a preset limit (typically 0.6 pu):
1.
clamp the voltage regulator output (BREF) to zero,
2.
block all TSCs.
The module outputs BLKTSC and CLMPVR are activated when the undervoltage condition occurs.
The BLKTSC signal controls the module TSCPULS and the clmpvr the module VREG.
The module allows the TSCs to deblock and remove the clamp on the voltage regulator output
after the voltage recovers.
URESP
IN
Level High
&
Level Low
BLKTSC
B
CLMPVR
Figure 20.28:
Block Diagram of the UVSTRATEGY Module
Arguments:
The following is a list of the arguments passed to the UVSTRATEGY module in the order that they
are specified.
Argument Type*/
Name
Size
Value
in
Model
Description
NV/6
Signal to block operation of any TSCs - output by module.
NV/6
Signal to set output of voltage regulator to zero - output by

module.
URESPU
NV/6
Per unit value of measured system voltage.
INITVALU
NC/7
1.0
Initial status. 1 no blocking.
LEVELHIGH
NC/8
0.69
Level to restore control action when voltage recovers.
LEVELLOW
NC/8
0.60
Level to suspend control action when voltage falls.
BLKTSC
CLMPVR
Table 30: Arguments for the UVSTRATEGY Module
20 - 44
20.6.30 Voltage Controlled Oscillator, VCO

A VCO is required for each phase of a PLL circuit. The VCO accepts a level input (PL1AVF) and
generates four output signals with a frequency which is twice the input signal:
1.
a train of pulses (PL1ATG) which occur exactly at the peaks of the voltage wave
when the PLL is locked,
2.
a ramp function (PL1ARM) whose value indicates the elapsed time in seconds
since the last voltage peak,
3.
a pulse train (PL1ADT) which lags the PL1ATG by one time step (T),
4.
a pulse train (PL1ASM) which leads the PL1ATG by 1.5 ms.
A ramp generator and a level comparator are the essential components of the VCO. The resolution
of the VCO is limited by the step size and error may accumulate over time. Therefore, a
stabilization circuit is required.
The stabilization circuit employs an error register and a delay unit. The stabilization circuit
prevents error accumulation by correcting the ramp height in the ramp generator with the error (BA) in the level comparator. The error in the level comparator is due to the fixed time-step used in
the EMTP/TACS.
The VCO module includes a calculation section which initializes the ramp generator. The initial
value of the ramp generator output (VRAMP) is determined by:
AMOD ( + 360,0, 180,0 )

A
VRAMP = -------------------------------------------------------------------
360,0 f o
where Aand frespectively are the initial phase angle and the frequency of the A-phase PLL circuit.
The function AMOD returns the remainder when A+360.0 is divided by 180.0.
20 - 45
SETRMO
R
Error
Register
T
T
RAMPFN
Ramp R
Generator
A
B
PULSE
AB
VOFREQ
1.5 ms
Figure 20.29:
AB
PPULS4
Block Diagram of the VCO Module
Arguments:
The following is a list of the arguments passed to the VCO module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
Description
RAMPFN
NV/6
Synchronization ramp - output by module.
PULSE
NV/6
Synchronization reference pulse - output by module.
Table 31: Arguments for the VCO Module
20 - 46
NV/6
Synchronization reference pulse delayed by (t - output

by module.
PPULS4
NV/6
Sampling pulse - output by module.
VOFREQ
NV/6
VCO Frequency set by PLL1.
ANGLPHASEA
NC/10
0.0
Initial angle of phase A.
TYMESHIFT
NC/9
1.5E-3
NC/8
60.0
SETRMO
FUNDFREQ
Time shift to calculate PL1ASM.

Default value. Can be changed by passing frequency
from SVCA or SVCB.
Table 31: Arguments for the VCO Module
20.6.31 Voltage Regulator, VREG

The voltage regulator performs the closed-loop voltage control. The VREG module is of
proportional and integrating type and uses the PIREG module. The output of the voltage regulator
is the susceptance reference (BREF). The control-loop error (UERROR) is the summation of the four
inputs:
1.
UREF - the voltage reference set by the user in per unit,
2.
UREF - input for special function blocks added by the user,
3.
URESP - voltage response signal from the measurement circuit,
4.
URESP - a signal from the measurement circuit.
The VREG module accepts the following additional input signals:

1.
FREEZE - this user supplied signal disables the regulator when it is active
(FREEZE =1),
2.
CLMPVR - an output of the module UVSTRATEGY which clamps the regulator

output to zero if CLMPVR =1,
3.
DBMAX - an output of the module UOVLIM which affects the upper limit of the
regulator.
20 - 47
DBMAX
B max
CLMPVR
UREF
UREF
URESP
URESP
FREEZE
PIREG
BREF
B min
CLMPVR
Figure 20.30:
Block Diagram of the VREG Module
Arguments:
The following is a list of the arguments passed to the VREG module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
Description
NV/6
Susceptance reference in MVAr - output by module.
NV/6
Voltage reference set point - input by user in main EMTP

deck.
DUREFR
NV/6
Input for special function blocks provided by user.
URESPU
NV/6
The measured voltage.
NV/6
Modification to voltage signal to achieve the slope

characteristic.
DBMAX
NV/6
Maximum allowable MVAr from UOVLIM.
FREEZE
NV/6
Flag to stop the integration of the error signal.
CLMPVR
NV/6
Flag to set output of voltage regulator to zero.
NC/10
Initial MVAr setting - input by user in call to SVCA or

SVCB.
BREF
UREF
DURESP
INITIALMVA
Table 32: Arguments for the VREG Module
20 - 48

NC/9
-125.0
Minimum susceptance range in MVAr.
BMVARMAXM NC/9
425.0
Maximum susceptance range in MVAr.
BMVARMINM
Table 32: Arguments for the VREG Module
20.6.32 Three-Phase Two-Winding Transformer, XFORM2W

Three-phase two-winding transformers are used for six-pulse SVC arrangements. This transformer
is modelled as three single-phase transformers connected wye on the power network side and delta
on the SVC side. The primary side is solidly grounded.
PRMBS
XF1
SECBS
Figure 20.31:
Circuit Diagram of the XFORM2W Module
Arguments:
The following is a list of the arguments passed to the XFORM2W module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
Description
XF1
NODE/3
Name of transformer.
PRMBS
NODE/5
Name of node on primary side of transformer.
Table 33: Arguments for the XFORM2W Module
20 - 49

SECBS
NODE/5
BASE.MVA
NC/8
300.0
NC/8
345.0
NC/8
18.0
NC/8
0.10
NC/8
60.0
PRIME.KV
SECND.KV
XXXXX.PU
BASEFREQ
Name of node on secondary side of transformer.

MVA of transformer. Set by call from SVCA or SVCB.
Primary voltage rating of transformer. Set by call from

SVCA or SVCB.
Secondary voltage rating of transformer. Set by call

from SVCA or SVCB.
Per unit impedance of transformer. Set by call from
SVCA or SVCB.
Base frequency. Defaults to 60 Hz but can be set by

passing from SVCA or SVCB.

20.6.33 Three-Phase Three-Winding Transformer, XFORM3W

Three-phase three-winding transformers are required for twelve-pulse SVC arrangements. This
transformer is modelled as three single-phase transformers with a wye-winding on the power
network side, and delta- and wye-windings on the SVC side. The primary side is solidly grounded.
SCBSY
PRMBS
XF1
SECBS
Figure 20.32:
Circuit Diagram of the XFORM3W Module

20 - 50
Arguments:
The following is a list of the arguments passed to the XFORM3W module in the order that they are
specified.
Argument
Name
Type*/
Size
Value
in
Model
Description
XF1
NODE/3
Name of transformer.
PRMBS
NODE/5
Name of node on primary side of transformer.
SCBSD
NODE/5
Name of node on secondary side of transformer (delta).
SCBSY
NODE/5
Name of node on secondary side of transformer (wye).
BASE.MVA
NODE/5
MVA of transformer. Set by call from SVCA or SVCB.
NC/8
NC/8
Secondary voltage rating of transformer. Set by call

from SVCA or SVCB.
NC/8
Per unit impedance primary-delta of transformer. Set by

call from SVCA or SVCB.
NC/8
Per unit impedance primary-wye of transformer. Set by

call from SVCA or SVCB.
NC/8
Per unit impedance delta-wye of transformer. Set by call

from SVCA or SVCB.
NC/8
60.0
PRIME.KV
SECND.KV
XXXPD.PU
XXXPY.PU
XXXDY.PU
BASEFREQ
Primary voltage rating of transformer. Set by call from

SVCA or SVCB.
Base frequency. Defaults to 60 Hz but can be set by

passing from SVCA or SVCB.

20.7 Initialization
The SVC model is initialized by using the EMTP steady-state initialization. All of the electrical
power system except the TCRs are activated during the initial calculation process. The calculation
mode determines how many of the TSCs are initially in service based on the value passed to the
20 - 51

SVCx module by the main data deck. If the initial specification is in the overlap area, the SVC will
be initialized such that the minimum number of TSCs are energized. During the steady-state
initialization phase of the EMTP run, the TCR is assumed to be in a non-conducting mode.
Additionally, some but not all of the control signals are initialized in the calculation sections of the
modules. As a result, the control system is not allowed to operate in the full closed loop mode
immediately after t=0. The control system is set up to produce equally spaced firing pulses for a
specified initialization period. The SVC model automatically sets the length of this period to 20 ms.
The user can lengthen this period by setting FRZPL1 and FRZVRG to one (1) during the desired
initialization period.
A recommended procedure for the simulation of a power system including SVCs is to first do a
steady state solution to determine the voltage magnitude and phase angle at t=0. then got back into
the model and input these values in the SVCA or SVCB modules before starting the actual runs. This
will help to minimize the startup transients that may occur. It is also a good idea to do the first run
without any other switching taking place to verify that the system has reached steady-state before
proceeding.
20.8 Some Comments About Modules

The module handling features of the EMTP are described in detail in Section 18 (EMTP data
modules). The features that are most relevant to the use and modification of the SVC model are
similar to the ones relevant to the hvdc model. Refer to Section 19.15 for more details.
20.9 TACS Signal Names

The TACS external signal names are listed below, together with an explanation of their use. These
names should be considered reserved. As written here, n represents the SVC number and p
represents the phase to which the signal refers.
Signal Name
Description
Plot
ALDLpp
Modifies ALPHON & ALPHOF due to overcurrent in TCR
ALPHOF
Firing angle value to turn TCR off
ALPHON
Firing angle value to turn TCR on
BLKTCR
Block the TCR if equal to 1
BLKTSC
Block the TSC if equal to 1
BREF
Desired susceptance signal as set by the voltage regulator
BTCR
Susceptance order for the TCRs
20 - 52

CLMPVR
Set output of voltage regulator to zero if equal to 1
CRnIpp
TCR current signal
CRnppG
TCR gate signal
DBMAX
Reduction in the maximum BREF allowed due to secondary

overvoltage
DUREFR
Provision for user supplied input to modify the voltage error

signal
DURESP
Change in the voltage error signal due to slope
FRZIOV
Disable overcurrent checking if equal to 1
FRZPL1
Disable PLL controller if equal to 1
FRZUMS
Disable voltage measurement if equal to 1 (the voltage output is

constant)
FRZUOV
Disable overvoltage checking if equal to 1
FRZVRG
Disable voltage regulator if equal to 1 (stops the error integration)
NOTSCS
Number of TSCs requested by the allocator
PLnHLD
Pulses synchronized to the low bus phase-to-phase voltage peaks
PLnpDT
PLnHLD delayed by one time step
PLnPLS
Pulses synchronized to the low bus phase-to-phase voltage zeros
PLnpNC
Equal to 1 while the voltage is decreasing from positive peak to

negative peak
PLnpNG
Error correction signal for the PLL
PLnpPC
Equal to 1 while the voltage is increasing from negative peak to

positive peak
PLnpRM
Sawtooth timing signal with a frequency of twice nominal
PLnpRP
Phase error signal
PLnpSM
Pulses synchronized to the phase-to-phase zero crossings
PLnpTG
Pulses for each phase synchronized to the phase-to-phase voltage

crests
PLnpVF
Corrected frequency signal used by the PLL
SCnppG
TSC gate signal
U2RESP
Per unit RMS measured low side voltage (average of three phases)
20 - 53

UREF
Voltage set point for high voltage bus in per unit
URESPF
Filtered value of URESPU
URESPU
Per unit RMS measured high side voltage (average of three phases) X
URSPFL
Voltage signal passed to the voltage regulator
USTBON
Flag to choose between filtered or unfiltered value of URESPU (1

=> filtered)
VOLT1p
Phase-to-phase low side voltage
20 - 54
Section 21
21.1
21.2
General Rules ...................................................................................... 21-2
21.3
Input Requirements and Formats for Load Flow Components ........... 21-5
21.4
21.5
21.6
21.3.1
Single Phase PQ Load ..................................................... 21-5
21.3.2
Voltage Source ............................................................... 21-6
21.3.3
Current Source ............................................................... 21-6
21.3.4
Three-phase Static Load ................................................. 21-7
21.3.5
Slack (Vq) Generator ..................................................... 21-8
21.3.6
PQ Generator .................................................................. 21-10
21.3.7
PV Generator .................................................................. 21-12
Miscellaneous Data ............................................................................. 21-14

21.4.1
End of Network Load Flow Data (Set #1) ..................... 21-14
21.4.2
Convergence Criteria (Set #2) ........................................ 21-14
21.4.3
Case Heading (Set #3) .................................................... 21-16
Examples of Generator Connections ................................................... 21-17

21.5.1
WYE Solidly Grounded .................................................. 21-17
21.5.2
WYE Grounded through an Impedance* ........................ 21-18
21.5.3
WYE Ungrounded ........................................................... 21-18
21.5.4
DELTA ............................................................................. 21-19
Sample Input/Output* ......................................................................... 21-20

21.6.1
Sample Input File ........................................................... 21-21
21.6.2
Sample Output: Format I ................................................ 21-22
21.6.3
Sample Output: Format II .............................................. 21-25
Section 21
21.1 General Description
Starting in version 3, a three-phase load flow (unbalanced power flow) option/program
complements the earlier load flow options FIX SOURCE 1 and FIX SOURCE 2 (which are intended
primarily for Synchronous machine initialization) This load flow has been implemented as a
stand-alone program "LDFL". It uses the Newton-Raphson method to iteratively solve a threephase network until all nodal constraints are within a specified tolerance. The network
components (loads, generators, etc.) are represented using branch equations, as opposed to the
conventional nodal equations. This leads to a greater flexibility when defining connections of the
network components (i.e., delta-connected generators, phase-to-phase voltage sources, etc.) .
LDFL supports a variety of network models which include:
(A)
Single-phase loads with specified constant active and reactive power.
(B)
Voltage and current sources.
(C)
Balanced/unbalanced three-phase loads.
(D)
Three-phase slack (V) generators, which correspond to the conventional

slack bus in positive sequence load flows.
(E)
Three-phase PQ generators, where the active and reactive power output are
specified.
(F)
Three-phase PV generators, where the active power output and the regulated
voltage magnitude are specified. If this generator regulates a remote bus, the
remote bus voltage magnitude is specified.
The three-phase load flow can be executed without defined EMTP sources. An EMTP source is any
voltage source that is defined in the EMTP source data section (see Section 10). However, all
nodes represented in the load flow must be previously defined in the EMTP branch data section.
Transient simulations (TMAX > 0) do not, at present, follow the three-phase load flow solution
automatically. A subsequent transient simulation based on the three-phase load flow can be
generated manually. However, nodal constraints enforced during the load flow solution will not
be contained in the transient simulation. The load flow solution only provides a starting point
(voltage magnitudes and angles), and has no other effects in transient simulations.
21 - 1
21.2 General Rules

Rule 1:
In order to execute the load flow, one or both of the following are needed:
(A)
A slack generator
(B)
A voltage source to ground
Rule 2:
All data must be in consistent units, e.g. V, A, W, VAr or kV, kA, MW, MVAr, or P.U.
(See Section 21.4.2).
Rule 3:
Voltage and current magnitudes are entered in peak quantities, and angles are entered
in degrees. Powers are specified as average powers.
Rule 4:
Isolated delta connections and isolated systems (from switches) are not allowed.
Rule 5:
Comments are allowed in the load flow data deck using the same format as that in
EMTP (I.E., "C ". However, comment cards are not allowed within subsets defining
loads or generators. (See example below.)
C
345678901234567890123456789012345678901234567890123456789012345678901234567890
C
1
2
3
4
5
6
7
8
6EGAVGA
7GAV A
972.0
626.5562
-920.0
435.0
EGAVGB
.00000001
.1149
EGAVGC
.00000001
.2231
C ------------THIS IS A LEGAL COMMENT CARD
------------17SYS A
631.4
0.0
17SYS B
634.6
-120.0
17SYS C
625.2
120.0
3EGAV AEGAV B
63.0
47.0
C -----------THIS IS AN ILLEGAL COMMENT CARD
----------EGAV BEGAV C
EGAV CEGAV A
Rule 6:
Load flow data begins after a "BLANK" card ending the EMTP sources.
Rule 7:
EMTP source nodes cannot be defined in the Load Flow Data Section (i.e., if "BUSA" is
defined as a type-14 voltage source, it cannot be redefined in the load flow data section).
Rule 8:
"LOAD FLOW" card should follow the "BEGIN NEW DATA CASE" card. (i.e., "LOAD
FLOW" card instead of "FIX SOURCE1" or "FIX SOURCE2" cards.)
Rule 9:
Three case-heading cards are required to end each load flow case data.
21 - 2

Rule 10:
Bus Naming Convention (for load flow output Format II only (see Section 21.4.2).
Bus names (six-character long) must follow the naming convention stated below so
that the program will interpret per unit bases correctly.
1.
2.
The first character of the bus name should identify bus voltage class (see Table
1) with the following two exceptions.
Exception I:
The first character of a neutral node name should be "N".
Exception II:
The first character of the transformer magnetizing branch

node name should be "X".
The last character of the bus name (sixth character) designates the phase. It
should be labeled as A, B, or C.
Table 1: Voltage Class IDs

Voltage Class
ID
Voltage Class
765
kV
500
kV
345
kV
230
kV
161
kv
138
kV
69
kV
46
kV
34.5
kV
26
kV
24
kV
23
kV
22
kV
17.25
kV
15.5
kV
15
kV
21 - 3
Table 1: Voltage Class IDs

Voltage Class
ID
Notes:
Voltage Class
14.4
kV
13.8
kV
13.2
kV
12.47
kV
12
kV
is reserved for labeling neutral nodes. Do not use this character as a voltage
class ID.
is reserved for labeling the grounding node (labeled as 'TERRA' by EMTP) in the
input file. Do not use this character as a voltage class ID.
is reserved for labelling the internal node of the transformer magnetizing

branch. Do not use this character as a voltage class ID.
Naming Convention Examples:

(A)
(GAV
VA
VB
VC
=
=
=
"EGAV A", "EGAV B", "EGAV C" will be interpreted as phases A, B, and C of
the Gavin 26 kV bus. In this example, the first character of the bus name (E)
identifies bus base voltage (26 kV), and the last character of the bus name (A,
B, C) identifies the phase. In the output, this bus name will appear as (GAV 26
kV). A sample of voltage output at (GAV 26 kV) bus is shown below.
26 kV)
V (PU)
ANGLE
0.9745
0.9876
0.9803
-23.52
-143.57
95.79
VPOS
VNEG
VZER
=
=
=
V (PU)
ANGLE
0.9808
0.0076
0.0000
-23.77
122.45
-90.19
VAB
VBC
VCA
=
=
=
V (PU)
ANGLE
0.9813
0.9871
0.9740
6.67
-114.01
126.03
(B)
"NGAV1" represents a neutral node and contains neither base voltage nor phase
ID. In the output, this node will appear as (NGAV1 ).
(C)
"XGAV1A", "XGAV1B", "XGAV1C" will be interpreted as magnetizing branch

node of phases A, B, and C of Gavin transformer #1. In the output, this node
will appear as (XGAV1 ). A sample of voltage output at (XGAV1 ) is shown
below.
21 - 4

(XGAV1
V (RMS VOLTS)
VA
VB
VC
=
=
=
15403.7
15644.6
15417.4
ANGLE
-26.88
-146.26
92.78
21.3 Input Requirements and Formats for Load Flow

Components
21.3.1
Single Phase PQ Load

3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
ITYPE
1
2
12 345678 901234 567890
KBUS
MBUS
I2
A6
A6
E16.0
E16.0
ITYPE
Enter 0 (zero) in Column 2, or leave blank.
(2)
KBUS
Enter the name of the sending node.
(3-8)
MBUS
(9-14)
Enter the name of the receiving node. This allows loads to be connected
from phase to ground or phase to phase. Leave blank for ground.
P
(21-36)
Enter the average active power of the single phase load.

P > 0; Power flows from KBUS to MBUS.
P < 0; Power flows from MBUS to KBUS.
Q
(37-52)
Enter the average reactive power of the single phase load.

Q > 0; Power flows from KBUS to MBUS.
Q < 0; Power flows from MBUS to KBUS.
21 - 5
21.3.2
Voltage Source
ITYPE
1
2
12 345678 901234 567890
KBUS
MBUS
I2
A6
A6
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
VMAG
THETA
E16.0
E16.0
Enter 1 in Column 2.
ITYPE
(2)
KBUS
(3-8)
Enter the name of the receiving node. This allows voltage sources to be
connected from phase to ground or phase to phase. Leave blank for ground.
MBUS
(9-14)
(21-36)
Enter the magnitude of the voltage (in peak values) as measured from KBUS
to MBUS.
THETA
Enter the angle (in degrees) of the voltage as measured from KBUS to MBUS.
VMAG
(37-52)
21.3.3
Current Source
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
ITYPE
1
2
12 345678 901234 567890
KBUS
MBUS
IMAG
THETA
I2
A6
A6
E16.0
E16.0
ITYPE
(2)
KBUS
(3-8)
MBUS
(9-14)
Enter the name of the receiving node. This allows current sources to be
connected from phase to ground or phase to phase. Leave blank for ground.
21 - 6
(21-36)
Enter the magnitude of the current (in peak values) flowing from KBUS to
MBUS.
THETA
Enter the angle (in degrees) of the current flowing from KBUS to MBUS.
IMAG
(37-52)
21.3.4
Three-phase Static Load
Three cards are required.
Card #1:
3
4
5
6
7
8
1234567890123456 7890123456789012 34567890 12345678 901234567890
ITYPE
1
2
12 345678 901234 567890
KBUS
MBUS
P-3PH
Q-3PH
I2
A6
A6
E16.0
E16.0
ITYPE
RE(RATIO) IM(RATIO)
E8.0
E8.0
(2)
KBUS
Enter the name for Phase "A" of the sending node.
(3-8)
MBUS
Enter the name for Phase "A" of the receiving node.
(9-14)
P-3PH
(21-36)
Q-3PH
(37-52)
RE(RATIO)
(53-60)
IM(RATIO)
(61-68)
Enter the average active power of the three-phase load.

P > 0; Power flows from KBUS to MBUS.
P < 0; Power flows from MBUS to KBUS.
Enter the average reactive power of the three-phase load.
Q > 0; Power flows from KBUS to MBUS.
Q < 0; Power flows from MBUS to KBUS.
Enter the real part of the positive to zero sequence impedance ratio for the
load.
Real(Zpos / Zzero)
Enter the imaginary part of the positive to zero sequence impedance ratio
for the load.
Im(Zpos / Zzero)
21 - 7
Card #2:
1
2
3
4
5
6
7
8
12 345678 901234 567890123456789012345678901234567890123456789012345678901234567890
KBUS
MBUS
A6
A6
Enter the name for Phase "B" of the sending node.
KBUS
(3-8)
Enter the name for Phase "B" of the receiving node.
MBUS
(9-14)
Card #3:
1
2
3
4
5
6
7
8
12 345678 901234 567890123456789012345678901234567890123456789012345678901234567890
KBUS
MBUS
A6
A6
Enter the name for Phase "C" of the sending node.
KBUS
(3-8)
Enter the name for Phase "C" of the receiving node.
MBUS
(9-14)
21.3.5
Slack (V) Generator
Card #1:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 7890123456789012 3456789012345678 9012345678901234567890
KBUS
MBUS
I2
A6
A6
VMAG
THETA
E16.0
E16.0
21 - 8
ITYPE
(2)
Enter the name for Phase "A" of the generator terminal. Do not leave this
field blank.
KBUS
(3-8)
Enter the name for Phase "A" of the generator internal node which is used to
determine the connectivity of the generator. (wye-grounded, delta etc.).
Leave blank for ground. (See examples of generator connections in Section
21.5).
MBUS
(9-14)
Enter the magnitude of the positive sequence voltage (in peak value) at the
generator terminals.
VMAG
(27-42)
Enter the angle (in degrees) of the positive sequence voltage

at the generator terminals.
THETA
(43-58)
Card #2:
1
2
12 345678 901234 567890
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
KBUS
MBUS
RE(ZZERO)
IM(ZZERO)
A6
A6
E16.0
E16.0
Enter the name for Phase "B" of the generator terminal. Do not leave this
field blank.
KBUS
(3-8)
Enter the name for Phase "B" of the generator internal node. Leave blank
for ground.
MBUS
(9-14)
RE(ZZERO)
(21-36)
IM(ZZERO)
(37-52)
Enter the real part of the generator's zero sequence impedance.

Re(Zzero)
Enter the imaginary part of the generator's zero sequence impedance.
Im(Zzero)
Card #3:
1
2
12 345678 901234 567890
KBUS
MBUS
A6
A6
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
RE(ZNEG)
IM(ZNEG)
E16.0
E16.0
21 - 9
Enter the name for Phase "C" of the generator terminal. Do not leave this
field blank.
KBUS
(3-8)
(9-14)
Enter the name for Phase "C" of the generator internal node. Leave blank
for ground.
RE(ZNEG)
(21-36)
Enter the real part of the generator's negative sequence impedance.

Real(Zneg)
IM(ZNEG)
Enter the imaginary part of the generator's negative sequence impedance.

Im(Zneg)
MBUS
(37-52)
Note:
In order to produce a symmetric 3X3 phase impedance matrix, it is assumed that

ZPOS = ZNEG.
21.3.6
PQ Generator
Card #1:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 7890123456789012 3456789012345678 9012345678901234567890
KBUS
MBUS
P-3PH
Q-3PH
I2
A6
A6
E16.0
E16.0
ITYPE
(2)
KBUS
(3-8)
MBUS
(9-14)
P-3PH
(27-42)
Q-3PH
(43-58)
field blank.
Enter the name for Phase "A" of the generator internal node. Leave blank for
ground. (See examples of generator connections in Section 21.5).
Enter the average active three-phase power of the generator.
P > 0; Power flows from generator terminals to the system.
P < 0; Power flows from system to the generator terminals (absorbing).
Enter the average reactive three-phase power of the generator.
Q > 0; Power flows from generator terminals to the system.
Q < 0; Power flows from system to the generator terminals (absorbing).
21 - 10
Card #2:
1
2
12 345678 901234 567890
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
KBUS
MBUS
RE(ZZERO)
IM(ZZERO)
A6
A6
E16.0
E16.0
field blank.
KBUS
(3-8)
for ground.
MBUS
(9-14)
RE(ZZERO)
(21-36)
IM(ZZERO)
(37-52)

Re(Zzero)
Im(Zzero)
Card #3:
1
2
12 345678 901234 567890
KBUS
MBUS
A6
A6
KBUS
(3-8)
MBUS
(9-14)
RE(ZNEG)
(21-36)
IM(ZNEG)
(37-52)
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
RE(ZNEG)
IM(ZNEG)
E16.0
E16.0
field blank.
for ground.
Real(Zneg)
Im(Zneg)
21 - 11
21.3.7
PV Generator
Three cards required are required.
Card #1:
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 7890123456789012 3456789012345678 90123456789 01234567890
KBUS
MBUS
KREG
MREG
P-3PH
VREG
I2
A6
A6
A6
A6
E16.0
E16.0
ITYPE
VARMIN
E11.0
VARMAX
E11.0
(2)
KBUS
(3-8)
MBUS
(9-14)
KREG
(15-20)
field blank.
Enter the name for Phase "A" of the generator internal node. Leave blank
for ground. (See examples of generator connections in Section 12.5).
Enter the node name whose voltage will be regulated by this generator. The
regulated voltage can be phase to ground, phase to phase (either at generator
terminal or at remote nodes), or Vpos at the generator terminals. If KREG is
left blank, the magnitude of the positive sequence voltage of the generator
terminals will be regulated.*
* The magnitude of the positive sequence voltage is defined as:
1/3 [1 a a2] ([Vk] - [Vm])
Where:
a = 1 /120
= 1 /240
a2
Care should be taken when entering the regulated voltage for deltaconnected generators.
MREG
(21-26)
P-3PH
(27-42)
Enter the node name whose voltage will be regulated by this generator.
Needed only if phase to phase voltage will be regulated.
Enter the average active three-phase power of the generator.
P > 0; Power flows from generator terminals to the system.
P < 0; Power flows from system to the generator terminals (absorbing).
21 - 12
(43-58)
Enter the regulated voltage magnitude (in peak value). This is either the
magnitude of the positive sequence terminal voltage (KREG = BLANK), or
the magnitude of V(KREG) - V(MREG).
VARMIN
Enter the generator's three-phase minimum reactive power limit.
VREG
(59-69)
Enter the generator's three-phase maximum reactive power limit.
VARMAX
(70-80)
Card #2:
1
2
12 345678 901234 567890
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
KBUS
MBUS
RE(ZZERO)
IM(ZZERO)
A6
A6
E16.0
E16.0
field blank.
KBUS
(3-8)
for ground.
MBUS
(9-14)
RE(ZZERO)
(21-36)
IM(ZZERO)
(37-52)

Re(Zzero)
Im(Zzero)
Card #3:
1
2
12 345678 901234 567890
KBUS
MBUS
A6
A6
KBUS
(3-8)
MBUS
(9-14)
3
4
5
6
7
8
1234567890123456 7890123456789012 3456789012345678901234567890
RE(ZNEG)
IM(ZNEG)
E16.0
E16.0
field blank.
for ground.
21 - 13

RE(ZNEG)
(21-36)

Real(Zneg)
IM(ZNEG)

Im(Zneg)
(37-52)
21.4 Miscellaneous Data

Three sets of miscellaneous data cards which are required to follow the network model data are
described in this section. These sets of cards signify the end of three-phase load flow data.
21.4.1
End of Network Load Flow Data (Set #1)
This set consists of only one data card.
ITYPE
1
2
3
4
5
6
7
8
12 345678901234567890123456789012345678901234567890123456789012345678901234567890
I2
Enter 99 to signify the end of the network load flow data.
ITYPE
(1-2)
21.4.2
Convergence Criteria (Set #2)
This set consists of only one data card.
NITERA
I8
NPUNIT
I8
XPQCHK
E8.0
MREG
IVUSER
E8.0
I8
AEPOUT
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345 67890123456789012345678901234567890
I5
21 - 14
NITERA
Enter the maximum number of iterations allowed. Defaults to 7.
(1-8)
NPUNIT
Enter units for data.
(9-16)
NPUNIT = 0 (Default);
NPUNIT = 1;
NPUNIT = 2;
XPQCHK
(17-24)
Data is in MW, MVAr, kV, kA and ohm.

Data is in P.U.
Data is in W, VAr, V, A, and ohm.
Enter power mismatch for convergence checking. (Default = .01) The

power convergence tolerance (CT) is calculated in the following manner:
NPUNIT = 1;
CT = XPQCHK
NPUNIT 1;
CT = XPQCHK Max {Pmax, Qmax} specified in
load flow input.

Example:
XVCHK
(25-32)
IVUSER
(33-40)
Max {Pmax, Qmax} = 1000 W or VAr

XPQCHK = .001
Maximum allowable mismatch is equal to 1 W or
VAr. (1000 .001)
Enter the voltage mismatch for convergence checking. (Default = .02) The
voltage convergence tolerance (VCT) is calculated in the following manner:
NPUNIT = 1;
VCT = XVCHK
NPUNIT 1;
VCT = XVCHK Max (V) specified in load flow
input.
Enter whether the load flow run is to be started from a user specified set of
voltages, or a "flat start".
IVUSER = 0 (Default);
Flat start.
If the program is run in the VM environment, a file "STVOLT.DATA"

containing bus voltages will be (1) created/updated on the User's "A" disk if
the job is run in interactive mode, or (2) sent to the User's reader if the job is
run in batch mode.
IVUSER = 1;
User specified starting voltages in file

"STVOLT.DATA" will be used.
User can either (1) create a new "STVOLT.DATA"
file in the format shown below; or (2) use an
existing "STVOLT.DATA" file created in the
previous run with or without modifications.
21 - 15

If the job is run in interactive mode in VM
environment, the file "STVOLT.DATA" will be
created/updated with the latest calculated
voltages. If the job is run in batch mode in VM
environment, a file "STVOLT.DATA" will be sent to
the User's reader.
Format for "STVOLT.DATA" file:

(whether created by the user, or created/updated by the program)
1
2
3
4
5
6
7
8
12345678901234 567890 1234 567890123 45678 901234567 890123456789012345678901234567890
BUS
NAME
VMAG
A6
Notes:
AEPOUT
(41-45)
21.4.3
THETA
E9.0
E9.0
1.
In the VM environment, the "STVOLT.DATA" file sent to the

User's reader after a batch job execution may be discarded or
received to replace the old "STVOLT.DATA" file.
2.
In the VM environment, "STVOLT.DATA" file will not be created/

updated if the solution does not converge within the specified
number of iterations.
Enter the desired output format (See samples in Section 21.6.)

Format I.
Format II.
AEPOUT = 0 (default);
AEPOUT = 1;
Case Heading (Set #3)

This set consists of three data cards. Enter the case heading/description which will be printed on
each page of the output.
1
2
3
4
5
6
7
8
1 2345678901234567890123456789012345678901234567890123456789012345678901234567890
DESCR
A79
21 - 16
Card #1:
Line 1 of case description. Leave blank if line 1 of case heading is not
desired.
DESCR
(2-80)
Card #2:
desired.
DESCR
(2-80)
Card #3:
desired.
DESCR
(2-80)
21.5 Examples of Generator Connections

This section illustrates how to enter KBUS and MBUS in order to model various generator
connections.
21.5.1
WYE Solidly Grounded
A
TERRA
B
C
Zzero
Zneg
Zpos = Zneg
Input:
KBUS (3-8)
MBUS (9-14)
EGAV A
EGAV B
EGAV C
(BLANK)
(BLANK)
(BLANK)
21 - 17
EGAVA
EGAVB
EGAVC
21.5.2
WYE Grounded through an Impedance*
A
NGAV
Zzero
Zneg
Zpos = Zneg
EGAVA
EGAVB
EGAVC
Input:
KBUS (3-8)
EGAV A
EGAV B
EGAV C
MBUS (9-14)
NGAV
NGAV
NGAV
The connection from NGAV through impedance Z to ground must be defined in the
EMTP branch data section.
21.5.3
WYE Ungrounded
This connection is treated the same way as the grounded neutral connection of section 21.5.2,
except that a large value for Z (grounding impedance) must be used in the branch data section.
A
NGAV
Zzero
Zneg
Zpos = Zneg
Input:
KBUS (3-8)
EGAV A
EGAV B
EGAV C
MBUS (9-14)
NGAV
NGAV
NGAV
21 - 18
EGAVA
EGAVB
EGAVC
21.5.4
The connection from NGAV through impedance Z to ground must be defined in the
EMTP branch data section.
DELTA
A
B
C
Zzero
Zneg
Zpos = Zneg
Input:
KBUS (3-8)
ECOOK A
ECOOK B
ECOOK C
MBUS (9-14)
ECOOK B
ECOOK C
ECOOK A
21 - 19
ECOOKA
ECOOKB
ECOOKC
21.6 Sample Input/Output*
26/765 Kv
GENERATOR
Gavin 25 Kv
Gavin 765 Kv
GSU TRANSFORMER
Saux
AUXILIARY LOAD
AUXILIARY LOAD
EGAVGA
EGAVGB
PHASE2
EGAVGC
PHASE3
EGAVA
Gavin 26 kV generator 1
X1 = X2 = 0.2231 ohms, Xo = 0.1149 ohms
PHASE1
GSU TRANSFORMER
XGV1A
24.63 Kv
7GVA
726.8/ 3 Kv
Saux
EGAVB
XGV1B
25.00 Kv
7GVB
726.8/ 3 Kv
Saux
Saux
EGAVC
XGV1C
24.63 kV
7GVC
726.8/ 3 kV
Z0
Figure 21.1:
Gavin Unit 1 Unbalance Study
21 - 20
EQUIVALENT
7SYSA
765 kV subsystem equivalent

X1 = X2 = 16.89 ohms, Xo = 19.31 ohms
GENERATOR
PHASE1
7SYSB
PHASE2
7SYSC
PHASE3
21.6.1
Sample Input File
BEGIN NEW DATA CASE

C * MISC DATA CARD REQUESTING 3@LF SOLUTION & MISC SOLUTION CARDS
.0001
0.0
60
60
LOAD FLOW
0
0
0
1
C
C GAVIN 1 GENERATOR (X0 AND X2 IN OHMS)
51EGAVGAEGAV A
.1149
52EGAVGBEGAV B
.2231
53EGAVGCEGAV C
C
C SYSTEM EQUIVALENT (X0 AND X2 IN OHMS)
C
517GAV A7SYS A
19.31
527GAV B7SYS B
16.89
537GAV C7SYS C
C
C GAVIN #1 GSU
C
TRANSFORMER
XGAV1A
9999
1EGAV AEGAV B
.00144.0928 24.63
27GAV ANGAV1
.4180 26.94 419.62
TRANSFORMER
XGAV1B
9999
1EGAV BEGAV C
.00096.0856 25.00
27GAV BNGAV1
.2714 24.10 419.62
TRANSFORMER
XGAV1C
9999
1EGAV CEGAV A
.00138.0928 24.63
27GAV CNGAV1
.4003 26.94 419.62
21 - 21
C
C
Three-Phase Load Flow

Data Section
GAVIN #1 GSU NEUTRAL NODE

NGAV1
.00001
BLANK CARD ENDING EMTP SOURCE DATA
C LDFL DATA SECTION BEGINS

C
C AUX LOAD
(MW)
(MVAR)
0EGAV AEGAV B
21.0
15.7
0EGAV BEGAV C
21.0
15.7
0EGAV CEGAV A
21.0
15.7
C
C GAVIN #1 (MODELED AS VOLTAGE SOURCES)
C
1EGAVGA
20.81
-4.27
1EGAVGB
20.81
-124.27
1EGAVGC
20.81
115.73
C
C 765 KV SYSTEM EQUIVALENT (MODELED AS VOLTAGE SOURCES)
C
17SYS A
631.4
0.0
17SYS B
634.6
-120.0
17SYS C
625.2
120.0
C
C
END OF LOAD FLOW; MISC SOLUTION DATA & CASE
HEADINGS
C
99
50
.0000001
1
EMTP THREE-PHASE LOAD FLOW EXAMPLE
GAVIN UNIT 1 UNBALANCE STUDY
BLANK CARD ENDING NODE NAMES FOR VOLTAGE OUTPUT
BLANK CARD TERMINATING PLOT SPEC. CARDS
BLANK TERMINATION-OF-RUN CARD
BEGIN NEW DATA CASE
BLANK
21.6.2
Sample Output: Format I

THREE PHASE LOADFLOW
STEADY STATE SOLUTION
SUMMARY OF NODE VOLTAGES
BUS #
2
3
4
5
6
7
NODENAME
EGAVGA
EGAV A
EGAVGB
EGAV B
EGAVGC
EGAV C
VOLT(KV)
20.8100
20.6875
20.8103
20.9657
20.8097
20.8105
ANG(DEG)
-4.2712
-23.5225
-124.2694
-143.5742
115.7304
95.7884
21 - 22

8
9
10
11
12
13
14
15
16
17
7GAV A
7SYS A
7GAV B
7SYS B
7GAV C
7SYS C
XGAV1A
NGAV1
XGAV1B
XGAV1C
626.5293
631.4000
627.0842
634.6000
620.8917
625.2000
21.7841
0.0000
22.1248
21.8035
1.5657
0.0000
-118.5004
-120.0000
121.3964
120.0000
-26.8834
-117.0025
-146.2600
92.7816

SUMMARY OF STATION LOADS
FROM
EGAV
EGAV
EGAV
BUS
A
B
C
TO BUS
EGAV B
EGAV C
EGAV A
P (MW)
21.0000
21.0000
21.0000
Q (MVAR)
15.7000
15.7000
15.7000
S (MVA)
26.2200
26.2200
26.2200
ANG (DEG)
36.7825
36.7825
36.7825

SUMMARY OF LINE FLOWS
TERRA
15
EGAVGC
318.115
P (MW)
EGAVGA
XGAV1A
XGAV1C
-318.115
140.485
156.834
P (MW)
EGAV B
EGAV B
4
14
16
EGAV A
EGAVGB
5
0.000
P (MW)
EGAV A
2
14
17
NGAV1
EGAVGA
3
P (MW)
323.260
P (MW)
EGAVGB
XGAV1A
XGAV1B
-323.260
176.777
125.391
P (MW)
Q (MVAR)
0.000
S (MVA)
0.000
Q (MVAR)
S (MVA)
59.668
323.663
Q (MVAR)
S (MVA)
48.269
-120.242
56.284
321.757
184.917
166.627
Q (MVAR)
S (MVA)
47.732
326.765
Q (MVAR)
S (MVA)
62.281
62.211
-140.371
Q (MVAR)
PF
329.205
187.404
188.220
S (MVA)
21 - 23
PLOSS (MW)
0.000
PF
0.983
PF
-0.989
0.760
0.941
PF
0.989
PF
-0.982
0.943
0.666
PF
PLOSS (MW)
0.000
PLOSS (MW)
0.000
0.230
283.259
PLOSS (MW)
0.000
PLOSS (MW)
0.000
317.032
0.155
PLOSS (MW)
QLOSS (MVAR)
0.000
QLOSS (MVAR)
107.937
QLOSS (MVAR)
107.937
14.829
-65.608
QLOSS (MVAR)
110.013
QLOSS (MVAR)
110.013
-72.861
13.798
QLOSS (MVAR)
EGAV C
6
16
17
10
7SYS A
NGAV1
7GAV A
316.802
-316.802
-316.731
P (MW)
7SYS B
NGAV1
308.390
-308.390
336.095
0.985
Q (MVAR)
S (MVA)
PF
-308.487
336.109
186.827
167.618
Q (MVAR)
S (MVA)
-87.692
87.692
328.715
328.715
Q (MVAR)
S (MVA)
97.064
331.271
Q (MVAR)
S (MVA)
-131.874
131.874
P (MW)
7GAV B
58.156
58.236
36.085
-109.851
P (MW)
7SYS B
10
-331.025
183.310
126.604
P (MW)
7GAV B
11
15
11
EGAVGC
XGAV1B
XGAV1C
7SYS A
8
331.025
P (MW)
7GAV A
9
15
EGAV C
335.403
335.403
Q (MVAR)
S (MVA)
141.576
339.423
-0.985
0.981
0.755
PF
0.964
-0.964
PF
-0.956
PF
0.919
-0.919
PF
-0.909
0.000
116.392
PLOSS (MW)
QLOSS (MVAR)
0.000
308.546
0.179
116.392
-118.085
12.041
PLOSS (MW)
QLOSS (MVAR)
0.071
-316.802
PLOSS (MW)
0.071
PLOSS (MW)
-0.097
-308.390
PLOSS (MW)
-0.097
9.372
87.692
QLOSS (MVAR)
9.372
QLOSS (MVAR)
9.702
131.874
QLOSS (MVAR)
9.702

SUMMARY OF LINE FLOWS
12
7GAV C
13
15
13
16
EGAV A
EGAV B
-140.255
140.255
P (MW)
TERRA
7GAV A
7GAV B
7GAV C
0.000
0.000
0.000
0.000
P (MW)
EGAV B
EGAV C
XGAV1C
3
7
-283.054
P (MW)
XGAV1B
5
7
17
7GAV C
NGAV1
1
8
10
12
283.080
-283.080
P (MW)
XGAV1A
3
5
15
7SYS C
NGAV1
7SYS C
12
14
P (MW)
-125.236
125.236
P (MW)
EGAV A
EGAV C
126.425
-126.425
Q (MVAR)
S (MVA)
-77.651
77.651
293.537
293.537
Q (MVAR)
S (MVA)
85.113
295.574
Q (MVAR)
S (MVA)
135.072
-135.072
Q (MVAR)
0.000
0.000
0.000
0.000
Q (MVAR)
154.169
-154.169
Q (MVAR)
-121.892
121.892
194.720
194.720
S (MVA)
0.000
0.000
0.000
0.000
S (MVA)
198.626
198.626
S (MVA)
175.616
175.616
21 - 24
PF
0.964
-0.964
PF
-0.958
PF
-0.720
0.720
PF
0.000
-0.695
0.929
-0.280
PF
-0.631
0.631
PF
0.720
-0.720
PLOSS (MW)
0.026
-283.080
PLOSS (MW)
0.026
PLOSS (MW)
0.230
317.032
PLOSS (MW)
0.000
-316.802
-308.390
-283.080
QLOSS (MVAR)
7.462
77.651
QLOSS (MVAR)
7.462
QLOSS (MVAR)
14.829
-72.861
QLOSS (MVAR)
0.000
87.692
131.874
77.651
PLOSS (MW)
QLOSS (MVAR)
0.155
308.546
13.798
-118.085
PLOSS (MW)
QLOSS (MVAR)
283.259
0.179
-65.608
12.041
21.6.3
Sample Output: Format II
11/17/95
16:57:50
PAGE
INPUT FILE NAME : GAVIN
TABLE OF VOLTAGE CLASS ID's

VOLTAGE
CLASS
ID
-7
5
3
2
A
1
B
C
D
E
F
G
H
I
J
K
L
M
O
P
Q
VOLTAGE
CLASS
---------765 kV
500 kV
345 kV
230 kV
161 kV
138 kV
69 kV
46 kV
34.5 kV
26 kV
24 kV
23 kV
22 kV
17.25 kV
15.5 kV
15 kV
14.4 kV
13.8 kV
13.2 kV
12.47 kV
12 kV
CASE
HEADINGS
NOTES:
N
IS RESERVED FOR LABELING NEUTRAL NODE.
IS RESERVED FOR LABELING THE GROUNDING NODE

(LABELED AS 'TERRA' BY EMTP) IN THE INPUT FILE.
IS RESERVED FOR LABELING THE INTERNAL NODE

OF THE TRANSFORMER MAGNETIZING BRANCH.
NEMA UNBALANCE = (MAXIMUM DEVIATION FROM AVERAGE) / AVERAGE
21 - 25
1
DAT

11/17/95
16:57:50
PAGE
2
DAT
THREE-PHASE LOAD FLOW

BUS VOLTAGES
(GAVG
26 KV)
V (PU)
VA =
VB =
VC =
NEMA
0.9803
0.9803
0.9803
= 0.00 %
(GAV
ANGLE
V (PU)
ANGLE
-4.27
-124.27
115.73
VPOS = 0.9803
VNEG = 0.0000
VZER = 0.0000
-4.27
-102.87
-90.38
ANGLE
V (PU)
ANGLE
-23.52
-143.57
95.79
VPOS = 0.9808
VNEG = 0.0076
VZER = 0.0000
-23.77
122.45
-90.19
ANGLE
V (PU)
ANGLE
1.57
-118.50
121.40
VPOS = 1.0003
VNEG = 0.0036
VZER = 0.0029
1.49
78.18
-47.80
V (PU)
VAB = 0.9803
VBC = 0.9803
VCA = 0.9803
NEMA = 0.00 %
ANGLE
25.73
-94.27
145.73
26 KV)
V (PU)
VA =
VB =
VC =
NEMA
0.9745
0.9876
0.9803
= 0.69 %
(GAV
V (PU)
VAB = 0.9813
VBC = 0.9871
VCA = 0.9740
NEMA = 0.69 %
ANGLE
6.67
-114.01
126.03
765 KV)
V (PU)
VA =
VB =
VC =
NEMA
1.0031
1.0039
0.9940
= 0.63 %
(SYS
VAB = 1.0038
VBC = 0.9995
VCA = 0.9977
NEMA = 0.35 %
1.0109
1.0160
1.0009
= 0.82 %
ANGLE
V (PU)
ANGLE
0.00
-120.00
120.00
VPOS = 1.0093
VNEG = 0.0044
VZER = 0.0044
0.00
79.56
-79.56
V (PU)
VAB = 1.0134
VBC = 1.0085
VCA = 1.0059
NEMA = 0.41 %
Sequence
Voltages
Phase-to-Neutral
Voltages
31.55
-88.72
151.63
30.08
-90.25
150.16
11/17/95
16:57:50
PAGE

NEUTRAL NODE VOLTAGES
V (RMS VOLTS)
)
ANGLE
Phase-to-Phase
Voltages

(NGAV1
ANGLE
765 KV)
V (PU)
VA =
VB =
VC =
NEMA
V (PU)
V =
0.0
21 - 26
ANGLE
-117.00
3
DAT

11/17/95
16:57:50
PAGE
4
DAT

MAGNETIZING NODE VOLTAGES
(XGAV1
V (RMS VOLTS)
VA =
VB =
VC =
15403.7
15644.6
15417.4
ANGLE
-26.88
-146.26
92.78

11/17/95
16:57:50
PAGE
5
DAT

VOLTAGE SOURCES
V (PU)
ANGLE
EGAVGA
EGAVGB
EGAVGC
TO
TO
TO
TERRA
TERRA
TERRA
0.9803
0.9803
0.9803
-4.27
-124.27
115.73
7SYS A
7SYS B
7SYS C
TO
TO
TO
TERRA
TERRA
TERRA
1.0109
1.0160
1.0009
0.00
-120.00
120.00

11/17/95
16:57:50
PAGE

SINGLE PHASE LOADS
I (RMS AMP)
EGAV A
EGAV B
EGAV C
TO
TO
TO
EGAV B
EGAV C
EGAV A
1027.67
1021.65
1035.40
ANGLE
P (MW)
-30.11
-150.80
89.25
21.00
21.00
21.00
21 - 27
Q (MVAR)
15.70
15.70
15.70
S (MVA)
PF (%)
26.22
26.22
26.22
80.09
80.09
80.09
6
DAT

11/17/95
16:57:50
PAGE
7
DAT

BRANCH FLOWS
(GAVG
26 KV)
TO
I (RMS AMP)
IA =
IB =
IC =
(GAV
21995.6
22206.1
22840.8
26 KV)
TO
I (RMS AMP)
IA =
IB =
IC =
(GAV
21995.6
22206.1
22840.8
765 KV)
TO
I (RMS AMP)
IA =
IB =
IC =
(SYS
742.0
756.4
668.6
765 KV)
TO
I (RMS AMP)
IA =
IB =
IC =
742.0
756.4
668.6
(GAV
ANGLE
I (RMS AMP)
-14.89
-132.67
105.77
(GAVG
IPOS =
INEG =
IZER =
I (RMS AMP)
165.11
47.33
-74.23
IPOS =
INEG =
IZER =
-162.96
84.65
-43.26
Branch Currents
22344.5
510.3
0.0
I (RMS AMP)
17.04
-95.35
136.74
ANGLE
P (MW)
Q (MVAR)
S (MVA)
PF (%)
-13.93
-147.56
45.00
SA =
318.1
SB =
323.3
SC =
331.0
------TOTAL
972.4
59.7
47.7
58.2
------165.6
323.7
326.8
336.1
-----986.4
98.29
98.93
98.49
-----98.58
P (MW)
Q (MVAR)
S (MVA)
PF (%)
SA = -318.1
SB = -323.3
SC = -331.0
-------TOTAL -972.4
48.3
62.3
58.2
------168.8
321.8
329.2
336.1
------986.9
98.87
98.19
98.49
-----98.53
P (MW)
Q (MVAR)
S (MVA)
PF (%)
SA =
316.8
SB =
308.4
SC =------283.1
TOTAL
908.3
-87.7
-131.9
--------77.7
-297.2
P (MW)
Q (MVAR)
S (MVA)
-316.7
-308.5
-283.1
-------TOTAL
-908.3
97.1
141.6
85.1
-------323.8
331.3
339.4
295.6
------964.2
ANGLE
166.07
32.44
-135.00
765 KV)
ANGLE
(GAV
22344.5
510.3
0.0
ANGLE
To-From
26 KV)
ANGLE
(SYS
From-To
26 KV)
IPOS =
INEG =
IZER =
720.8
20.4
56.9
ANGLE
19.60
176.00
-27.00
328.7
335.4
------293.5
955.7
96.38
91.95
-----96.44
95.04
765 KV)
I (RMS AMP)
IPOS =
INEG =
IZER =
720.8
20.4
56.9
ANGLE
-160.40
-4.00
153.00
Sequence Currents
21 - 28
SA =
SB =
SC =
Power Flows
PF (%)
95.61
90.89
95.76
-----94.19

11/17/95
16:57:50
PAGE
8
DAT

TRANSFORMER FLOWS
(GAV
26 KV)
TO
I (RMS AMP)
IA =
IB =
IC =
DELTA
Delta-Connection
ANGLE
20789.9
21053.1
21649.1
-11.38
-129.07
109.18
I1
I2
I3
I (RMS AMP)
ANGLE
IPOS = 21160.9
INEG =
510.2
IZER =
0.0
-10.43
-147.55
45.00
SA =
SB =
SC =
TOTAL
P (MW)
Q (MVAR)
S (MVA)
PF (%)
297.3
302.2
309.9
-------909.4
-64.0
-78.2
-73.8
--------215.9
304.1
312.1
318.6
------934.7
97.76
96.81
97.28
-----97.30
P (MW)
Q (MVAR)
S (MVA)
PF (%)
317.3
308.7
283.4
-------909.4
-58.0
-104.3
-53.6
--------215.9
322.5
325.8
288.5
------934.7
98.37
94.74
98.26
-----97.30
P (MW)
Q (MVAR)
S (MVA)
PF (%)
-316.8
-308.4
-283.1
--------908.3
87.7
131.9
77.7
-------297.2
328.7
335.4
293.5
------955.7
96.38
91.95
96.44
-----95.04
P (MW)
Q (MVAR)
S (MVA)
PF (%)
0.0
0.0
0.0
-------0.0
0.0
0.0
0.0
-------0.0
0.0
0.0
0.0
------0.0
69.52
92.94
28.00
-----0.00
DELTA WINDING CURRENT FLOWS

I (RMS AMP)
IAB =
IBC =
ICA =
ANGLE
12641.1
12696.2
11390.8
17.04
-95.35
136.74
I4
I5
I6
SAB =
SBC =
SCA =
TOTAL
(GAV
765 KV)
TO
I (RMS AMP)
IA
IB
IC
IN
=
=
=
=
(NGAV1
I (RMS AMP)
IA
IB
IC
IN
=
=
=
=
742.0
756.4
668.6
170.7
-162.96
84.65
-43.26
153.00
TO
(GAV
ANGLE
17.04
-95.35
136.74
-27.00
Wye-Connection
ANGLE
742.0
756.4
668.6
170.7
)
(NGAV1
I (RMS AMP)
I7 IPOS
I8 INEG
I9 IZER
I10
=
=
=
720.8
20.4
56.9
ANGLE
-160.40
-4.00
153.00
SA =
SB =
SC =
TOTAL
765 KV)
I (RMS AMP)
IPOS =
INEG =
IZER =
720.8
20.4
56.9
ANGLE
19.60
176.00
-27.00
SA =
SB =
SC =
TOTAL
21 - 29

11/17/95
16:57:50
PAGE

NEUTRAL FLOWS
I (RMS AMP)
( NGAV1
TO
( TERRA )
I =
21 - 30
170.65
ANGLE
153.00
9
DAT
Section 22
Network Reduction
22.1
Introduction To Network Reduction ................................................... 22-1
22.2
General Structure Of Input File For Network Reduction .................... 22-1
22.3
Cards Defining Network Reduction .................................................... 22-2

22.3.1
$NETRED ........................................................................ 22-2
22.3.2
Parameter Card ............................................................... 22-2
22.3.3
Node Names of Retained Nodes (for ITYPE=1 or 2) ...... 22-3
22.3.4
Network Reduction Branch Data

(R in Ohms, L in Henrys) .............................................. 22-3
22.4
22.5
22.3.5
Network Reduction Capacitance Data (C in Farads) ..... 22-4
22.3.6
Transformer Reduction (ITYPE = 3), Special Case ........ 22-4
Transformer Input File Formats .......................................................... 22-6

22.4.1
Capacitance File ............................................................. 22-6
22.4.2
Resistance File ............................................................... 22-6
22.4.3
Inductance File ............................................................... 22-7
22.4.4
Connection File .............................................................. 22-7
Example ............................................................................................... 22-8
Section 22
Network Reduction
22.1 Introduction To Network Reduction
The network reduction feature of the EMTP allows the simulation of large networks which consist
of linear RLC branches, which would not be possible to model using standard coupled the RLC
branches described in Section 5 (primarily due to input format limitations).
Network reduction can be applied to any linear RLC branches, and in particular, to a circuit of such
branches representing a high frequency transformer model based on construction data and
specified as separate data files for the resistances, inductances, capacitances and connections.
22.2 General Structure Of Input File For Network Reduction

(A)
Create an EMTP file as usual
(B)
Alter the "BRANCH DATA" section in the following manner to invoke the
reduction utility.
EMTP
branch data
NETRED
capacitance data
NETRED branch data
node names retained nodes
NETRED parameter card
$NETRED request card
EMTP
branch data
(C)
Up to 9 network reductions can be performed per run.

22 - 1
Network Reduction
22.3 Cards Defining Network Reduction

22.3.1
$NETRED
Network reduction request card
22.3.2
Parameter Card
User data containing the followin

<N_nodes>< N_branches>< N_retained_nodes>< N_capa>< N_connection>< ITYPE>< IREF>
FORMAT- 7I4:
N_nodes
Number of nodes in the system to be reduced, max=150/

reduction.
N_branches
Number of branches in the system to be reduced , max=150/

reduction.
N_retained_nodes
Number of retained nodes of the reduced model, max=30/

reduction.
N_capa
Number of capacitances in the system to be reduced,

max=150/reduction.
N_connection
Number of nonzero elements in the connection matrix. only

used for transformer reductions (ITYPE=3) max = 500/
reduction.
ITYPE
Specifies how the branch data is entered.

ITYPE = 1 Mutually coupled RL branches using type 51,
52, ... mutually coupled RL branches (see Section 5.3.2).
ITYPE = 2 Lumped uncoupled RL branches (see Section
5.1.2).
ITYPE = 3 Transformer Reduction (this is a special case).
Note. Capacitance data is required for all values of ITYPE.

7
IREF
If the system to be reduced is identical to a previous system,

enter the number of that network reduction (this eliminates
the need to repeat RL and C branch data ).
22 - 2
Network Reduction
22.3.3
Node Names of Retained Nodes (for ITYPE=1 or 2)
This is the six character EMTP node name This card is repeated up to the number of retained nodes
(ex if there are to be 4 retained nodes you need to specify 4 node names)
FORMAT (2x A6)
Note: If this reduction is based on a previous one, IREF > 0, it is required that the node names
arc entered in the same order as the corresponding node names in the referenced reduction (ex: if
the referenced nodes are entered in as Hl Xl Yl Y2 for a transformer, make sure that new node
names are entered such that they correspond to Hl Xl Y1 Y2).
22.3.4
Network Reduction Branch Data (R in Ohms, L in Henrys)

1.
For ITYPE = 1 (Mutually coupled) the same rules as Section 5.3.2 apply, namely
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234 567890 123456789012 345678 901234567890
ITYPE
Node Names
BUS1
BUS2
I2
A6
A6
elements (k, m)
elements (k, m+2)
elements (k, m+2)
E6.2
E12.2
E6.2
E12.2
E6.2
E12.2
where
1
2
3
4
5
6
7
8
12 345678901234 567890123456 789012 345678901234 567890 123456789012 345678 901234567890
51
Node Names of
1st Branch
R11
L11
52
Node Names of
2nd Branch
R21
L21
R22
L22
53
Node Names of
3rd Branch
R31
L31
R32
L32
R33
L33
54
Node Names of
4th Branch
R41
L41
R42
L42
R43
L43
R44
L44
2.
For ITYPE = 2 (Uncoupled) the same rules as Section 5.1.2 apply, namely
22 - 3
Network Reduction
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 7890123456789012 3456789012345678 9012345678901234 567890
I2
BUS1
BUS2
L (L)
A6
A6
E16.0
E16.0
3.
22.3.5
For ITYPE = 3 (TRANSFORMER REDUCTION) follow the rules of Section 22.3.6

below
Network Reduction Capacitance Data (C in Farads)
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890123456 7890123456789012 3456789012345678 9012345678901234 567890
I2
BUS1
BUS2
C (C)
A6
A6
E16.0
22.3.6
Transformer Reduction (ITYPE = 3), Special Case
If itype = 3, this type of reduction will read in R,L,C data files that are supplied from the
manufacturer (they must be in the specific format described below).
1.
Names of transformer files
There are four cards after the "parameter data card" representing the names the files
which contain transformer data
1
2
3
4
5
6
7
8
12345678901234567890 123456789012345678901234567890123456789012345678901234567890
RESISTANCE FILENAME
A20
1
2
3
4
5
6
7
8
12345678901234567890 123456789012345678901234567890123456789012345678901234567890
INDUCTANCE FILENAME
A20
22 - 4
Network Reduction
1
2
12345678901234567890
3
4
5
6
7
8
123456789012345678901234567890123456789012345678901234567890
CAPACITANCE FILENAME
A20
1
2
12345678901234567890
3
4
5
6
7
8
123456789012345678901234567890123456789012345678901234567890
CONNECTION MATRIX
FILENAME
A20
(Filename- Filetype- Filemode)

FORMAT = (A20)
2.
(A)
Resistance data file name
(B)
Inductance data file name
(C)
Capacitance data file name
(D)
Connection Matrix data file name
Retained nodes
Retained internal node numbers of the transformer, and EMTP node names
This is the internal transformer node number, and the corresponding six character
EMTP node name. This card is repeated up to the number of retained nodes (ex: if there
are to be 4 retained nodes, you need to specify this card 4 times)
NODE
NUMBER
1
2
3
4
5
6
7
8
1234 56 789012 34567890123456789012345678901234567890123456789012345678901234567890
I4
NODE
NAME
A6
22 - 5
Network Reduction
Note: If this reduction is based on a previous one, IREF > O. it is required that the
node names are entered in the same order as the corresponding node names in the
referenced reduction (ex if the referenced nodes are entered in as Hl,Xl,Yl,Y2 for a
transformer, make sure that new node names arc entered such that they correspond to
Hl,Xl,Yl,Y2)
22.4 Transformer Input File Formats

This step reduces the detailed transformer representation into a reduced model to be used in EMTP.
Several files must be specified and they contain the data as described below.
22.4.1
Capacitance File
1
2
3
4
5
6
7
8
12345 678901234567890 12345 6789 01234 5678 901234567890123456789012345678901234567890
CAPSIZE
E15.8
SB
RB
I4
I4
CAPSIZE
Capacitance in Farads between nodes SB and RB of

the transformer.
SR,SB
Internal node numbers of the transformer.
It is a requirement that ground be the last node number. For example, if the transformer has 136
nodes, then ground is node 137.
22.4.2
Resistance File
1
2
3
4
5
6
7
8
12345 678901234567890 12345 6789 012345678901234567890123456789012345678901234567890
RESISTANCE
E15.8
SEG#
I4
22 - 6
Network Reduction
22.4.3
RESISTANCE
Resistance in ohms for corresponding transformer

segment.
SEG#
Transformer internal segment number.
Inductance File
1
2
3
4
5
6
7
8
12345 678901234567890 123456789012345 678901234567890 123456789012345 678901234567890
LB(1)
LB(2)
LB(3)
LB(4)
LB(5)
E15.8
E15.8
E15.8
E15.8
E15.8
Contains KK elements, 5/row where KK = nsegt * (nsegt+1)/2. This corresponds to the upper
triangle of the inductance matrix including diagonals.
22.4.4
NSEGT
Number of winding segments of the transformer.
LB(I)
Inductance element, in Henrys, per segment.
Connection File
1
2
3
4
5
6
7
8
1234 56789012 34567890 12345678 90123456 78901234 56789012 34567890 12345678 90123456 7890
IPOS(1)
IPOS(2)
IPOS(3)
IPOS(4)
IPOS(5)
IPOS(6)
IPOS(7)
IPOS(8)
IPOS(9)
I8
I8
I8
I8
I8
I8
I8
I8
I8
Contains MMI elements, 9/row where MMI is the number of non-zero elements in the Connection
matrix. This file defines the connection of the transformer as well as current conventions, negative
out of the node. The matrix dimensions are (number_nodes,number_segments). The matrix is
used to evaluate the inverse nodal inductance matrix given by
INV = T LB Tt
where T is the connection matrix, and LB the inductance matrix.

22 - 7
Network Reduction
22.5 Example
An example for transformer reduction is given below:
BEGIN NEW DATA CASE

.2E-6 20.E-6
60.0
60.0
0
1
0
1
C
C START OF NETRED TRANSFORMER DATA
C
$NETRED
136 139
4 492 270
3
aeptstr.inc
aeptsti.inc
aeptstc.inc
aeptstt.inc
56
BUS56
1
BUS1
136 BUS136
16
BUS4
C
C END OF NETRED TRANSFORMER DATA
C
$VINTAGE,1
BUS4
30.0
BUS1
.0001
BUS56
285.00
BUS136
28500.0
$VINTAGE,0
C
C
BLANK LINE ENDING BRANCH DATA
BLANK LINE ENDING SWITCH DATA
14BUS136
593.4
60.0
-90.0
BLANK LINE ENDING SOURCE DATA
BUS1 BUS4 BUS56BUS136
BLANK
BLANK CARD ENDING BASE-CASE PLOT CARDS
BEGIN NEW DATA CASE
BLANK
22 - 8
Section 23
Protection Systems
23.1
Introduction ......................................................................................... 23-1
23.2
CT Model ............................................................................................ 23-2
23.3
23.4
23.5
23.6
23.7
23.2.1
Introduction .................................................................... 23-2
23.2.2
A Discussion of the CT Model ....................................... 23-3
23.2.3
Data Format .................................................................... 23-8
23.2.4
Output ............................................................................. 23-11
CVT Model .......................................................................................... 23-11
23.3.1
Introduction .................................................................... 23-11
23.3.2
A Discussion of the CVT Model .................................... 23-12
23.3.3
Data Format .................................................................... 23-14
23.3.4
Output ............................................................................. 23-17
Instructions For Using CEY51A Relay ............................................... 23-19

23.4.1
Description of CEY51A .................................................. 23-19
23.4.2
Data Format .................................................................... 23-20
23.4.3
Output of the CEY51A Relay .......................................... 23-22
Instructions For Using D202 Relay ..................................................... 23-22

23.5.1
Description of D202 ....................................................... 23-22
23.5.2
Data Format .................................................................... 23-23
23.5.3
Output of the D202 Relay .............................................. 23-25
Instructions For Using Sly Relay ........................................................ 23-25

23.6.1
Description of the Sly Relay .......................................... 23-25
23.6.2
Data Format .................................................................... 23-26
23.6.3
Output Of Sly Relay ....................................................... 23-29
Instructions For Using BDD15B Relay ................................................ 23-30

23.7.1
Description of BDD Relay .............................................. 23-30
23.7.2
Data Format .................................................................... 23-31
23.7.3
Output of BDD15B Relay ............................................... 23-32
Section 23
23.8
23.9
Protection Systems
Instructions For Using CDR_101 Relay .............................................. 23-33

23.8.1
Description of CDR_101 Relay ...................................... 23-33
23.8.2
Data Format ................................................................... 23-35
23.8.3
Output of CDR_101 Relay ............................................. 23-35
Instructions For User_defined FORTRAN Relay ................................. 23-36

23.9.1
Introduction ................................................................... 23-36
23.9.2
Development Of User_Defined Relay Model ............... 23-36
Section 23
Protection Systems
This chapter of the rule book covers instructions for using models of Current and Capacitorcoupled Voltage transformers, instructions for using relay models for line and transformer
differential protection in EMTP, and instructions for incorporating Fortran-relays in EMTP .
23.1 Introduction
A protection system consists of instrument transformers, relays, and circuit breakers. Protection
systems are critical power system components, and their behavior often determines the response
of a power system to a transient event.
Power system response to faults and other sudden disturbances includes "transient" and "steadystate" components. For low speed protection systems, the transient component is generally
ignored; only the steady-state component is used for analysis. For high-speed protection systems,
the transient component must be considered as well, since it creates a serious risk to protection
system security and dependability. [ Security refers to the fact that the relay does not operate
erroneously, e.g. for faults outside its zone of protection. Dependability refers to the fact that the
relay operates for a fault its zone.]
Designers of protection equipment have used miniature system models to determine relay
response under selected system conditions. As with other studies using miniature system models,
it is not possible to include substantial portions of the power system in the model. Also, it is not
possible to easily vary parameters, such as the remanent flux in the current transformer, or the
magnetization curve of a the current transformer, or the inertia of the machines in the system. The
dynamics of the interaction between the power system and the protection system cannot be
studied. The time and cost of such studies may also be prohibitive.
There are three main motivating reasons for developing relay system representation in EMTP.
In relay system simulation problems, ideally it is not acceptable to study the protection
system response in an off-line mode. Relay system models must respond to transient
events and participate in influencing their course by taking appropriate sequences of
control actions. Instances when this is crucial are numerous. For example: sequential
phase opening when single phase relaying is used; studies of evolving faults and relay
responses to them. It is often necessary to examine a complex transient event in its
natural order, and this order is partially determined by the protection system.
23 - 1
Protection Systems
Utility engineers (as well as relay designers) need a tool which can be used to analyze
relay responses under specific system conditions. There is no tool which can be used
by such a broad group of potential users.
The advances in digital computer relaying have made it highly desirable that the
simulations be in digital form. Since EMTP results are already in digital forms it can
readily interface with digital relay system representations.
This chapter of the rule book covers instructions for using models of Current and Capacitorcoupled Voltage transformers, instructions for using relay models for line and transformer
differential protection in EMTP, and instructions for incorporating Fortran-relays in EMTP .
Note:
It is extremely important to connect resistances from the current transformer secondary

terminals to the relay terminals. The resistances serve to isolate switches which would,
under certain conditions, be connected between the same nodes. TACS sources of Type
91 pass the current flowing through the first switch, if two switches are connected
between the same nodes. A suggested value of the resistance is 0.1 ohm.
23.2 CT Model
23.2.1
Introduction
The Current Transformer (CT) model is shown in Figure 23.1. The reader is refered to Section 18
for an explanation of the data module concept. The following files are required in the users
directory or the EMTP directory to use the CT model :
CTBL.INC
This is the Burden and Lead module. It is included by EMTP when the
user inputs IBURFG=O1 and ILEDFG=O1. The values of the burden
resistance, burden inductance, and burden capacitance are given by
the user and connected within the CT model on the ct secondary side.
An internal node for the lead impedance is created, and the values for
the lead resistance, lead inductance, and lead capacitance are given by
the user and connected in the ct model.
CTNBL.INC
This is the No Burden, Lead module. It is included by EMTP when the

user inputs IBURFG=OO and ILEDFG=01. The burden values must be
connected by the user on the ct secondary side. An internal node for
the lead impedance is created, and the values for the lead resistance,
lead inductance, and lead capacitance are given by the user and connected in the CT model.
23 - 2
Protection Systems
CTBNL.INC
This is the Burden, No Lead module. It is included by EMTP when the

user inputs IBURFG=OI and ILEDFG=OO. The values of the burden
resistance, burden inductance, and burden capacitance are given by
the user and connected within the CT model on the CT secondary side.
No provision is made for the lead impedance.
CTNBNL.INC
This is the No Burden, No Lead module.It is included by EMTP when

the user inputs IBURFG=OO and ILEDFG=OO. The burden values must
be connected by the user on the ct secondary side. No provision is
made for the lead impedance.
One feature of the CT model is the Type 96 Non-linear hysteretic inductor. The Type 96 element
allows the modelling of the hysteresis loops and the inclusion of the remanent flux in the CT. The
type 96 element requires the (peak)-i(peak) characteristic of the core material. Typically, the
user has access to a V(rms)-l(rms) curve. The CVT model has an inbuilt facility to convert the
V(rms)-l(rms) curve of the ARMCO M4 oriented silicon steel to the (peak)-i(peak) curve required
by the type 96 element. This conversion process can be invoked by setting IHYSFG = 00. If, the
user already has knowledge of the (peak)-i(peak) curve of the core, then the user must enter
IHYSFG= 01.
Note:
22 points of (peak)-i(peak) curve are required. The user must also set both
IBURFG=01 and ILEDFG=01.
23.2.2
A Discussion of the CT Model
EMTP requirements:
A good representation of a current transformer equivalent circuit for modeling purposes in EMTP
is shown in Figure 23.2. In this representation the secondary winding resistance and the secondary
winding leakage inductance are included with the lead impedance and the burden impedance.
In this context, it is necessary to distinguish between a low reactance current transformer, and a
high reactance current transformer. A uniformly wound toroidal core CT with a centrally placed
primary is considered a low reactance CT, with negligible secondary leakage reactance. This
implies that the CTs excitation characteristic can be used to predict its performance, particularly at
high levels of flux density, corresponding to high overcurrents in the primary circuit. A wound
primary CT is considered a high reactance CT, where tests have to be done to determine its
performance. In high reactance current transformers, the primary exciting current is influenced by
the leakage flux set up by the primary current and is not equal to the secondary exciting current
multiplied by the turns ratio. The secondary excitation characteristic cannot be used to
predetermine the current transformers performance during primary fault conditions. Thus the
model shown in Figure 23.2 is strictly valid for low reactance current transformers only.
23 - 3
Protection Systems
The EMTP representation of a current transformer is shown in Figure 23.3. In this representation
the secondary winding resistance and the secondary winding leakage inductance are included with
the lead impedance and the burden impedance. In this model Zm is placed after ZEMTP, in parallel
with (Zsec+ Zlead+ Zburden). This representation is required because finite values must be given
for Zpri and ZEMTP. More precisely, LEMTP must be non-zero, and REMTP can be zero; and either
Rpri or Lpri can be zero, but not both. This data requirement of the EMTP transformer model can be
satisfied by separating Zsec into two parts ZEMTP and Zsec . ZEMTP should be small, however,
very small values of ZEMTP give rise to very large values of the admittance 1/ZEMTP in the
matrices [Y] and [G]. This can "swamp out" the effects of other elements connected to ZEMTP,
leading to inaccurate results. It is not possible to give an absolute lower bound on the magnitude of
ZEMTP, because it depends on the relative ratio of Zpri and ZEMTP- The lower bound on ZEMTP
also depends on the magnitude of the other impedances in the network: this is due to the different
multipliers used in the forward reduction of the matrices [G] and [Y], which affect the diagonal
element corresponding to the node being eliminated. TOLMAT is a near-zero tolerance, which is
used to check the singularity of the steady-state complet matrix [Y]. If ZEMTP is very small, then
the EMTP will give the disconnected subnetwork error message. EPSILN is a near-zero tolerance,
which is used to check the singularity of the real coefficient matrix [G] within the time-step loop,
where
[G] Vnode ( t) = Inode(t)
If ZEMTP is very small, then the EMTP will give the disconnected subnetwork error message in the
steady-state solution, or the floating network error message in the time-step solution. With the
approximation of placing Zm as in Figure 23.3, the secondary side of the current transformer can
be represented by a current source in series with ZEMTP. Because of the current source behavior of
the CT, the value of ZEMTP does not affect the secondary voltage drop or the onset of saturation .
Open Circuit Magnetization Characteristics:

The determination of Lsec from secondary open circuit tests is difficult. The waveforms of
currents and voltages are required, because for impressed sinusoidal voltages, the currents are nonsinusoidal. Thus the Irms is only a measure of the effective magnitude of the current. The
secondary open circuit test is used to plot the magnetization characteristic: the (Vrms- Irms) curve.
This magnetization characteristic is the data that is available most of the time. This characteristic is
measured from the secondary terminals and hence measures (Zsec + Zm), as shown in Figure 23.2.
In the unsaturated region (Zsec < < Zm)~ and in the saturated region Zsec is of the same order of
magnitude as Zm.
One procedure of separating Zsec from the measurement of (Zsec + Zm) is as follows:
1.
Subroutine CONVERT of AUX is used to convert the ( Vrms - Irms) to the (peakIpeak). Subroutine CONVERT assumes that there is no voltage drop across the
secondary winding resistance and the secondary winding leakage reactance. Rsec
can be measured easily. A value of Lsec is assumed. The assumed voltage drop
across Zsec can be calculated using Irms. This assumed voltage drop is
subtracted from Vrms, to give a new ( Vrms~ -Irms ) curve. The new ( VrmsIrms) curve is converted to the (peak-Ipeak) curve.
23 - 4
Protection Systems
2.
From the test current oscillogram read the value of Ipeak.
3.
Corresponding to Ipeak in Step (2) read the value of peak from the (peakIpeak) curve generated in Step (1).
4.
Convert l_peak in Step (3) to Vpeak. This Vpeak represents the voltage drop
across Zm.
5.
The value of Vpeak in Step (4) is subtracted from the value of Vpeak which is
impressed across the secondary terminals to give the voltage drop across Zsec.
6.
The voltage drop across Zsec in Step (5) is compared to the assumed voltage
drop in Step (I) converted to its peak value. If there is a significant difference
between the two voltage drops, then the procedure is repeated from Step (1), till
there is no significant difference between the assumed and calculated voltage
drops.
7.
When the assumed and calculated voltage drops are equal, then the value of Lsec
can be calculated. This procedure can be repeated for other points on the (Vrms Irms) curve and the corresponding oscillograms.
The magnetization characteristic should ideally be used with the value of Zm separated from the
measurement of (Zsec + Zm). This separated value of Zm should be used to compute the
hysteresis curve and then included in the current transformer model. When this is done the value
of Zsec should be included with the secondary burden. The model used does not preclude this
exact procedure from being followed.
CT Model for EMTP:

For low reactance current transformers, generally (Zsec < (Zlead + Zburden)), and it can be
included as shown in Figure 23.1. When there is no fault on the primary circuit of the current
transformer, the voltage drop across Zsec is not significant. The small voltage drop does not
influence the exciting current and the pseudononlinear hysteretic inductor significantly. When
there is a fault on the primary circuit of the current transformer, the voltage drop across Zsec can
be significant. The large voltage drop across the secondary terminals ( including Zsec + Zlead +
Zburden) demands a large flux in the core. This large flux causes a sharp increase in the exciting
current and a reduction in the value of the pseudononlinear hysteretic inductor. The addition of
Zsec to (Zleadd + Zburden) does hasten the onset of saturation, depending on the relative
magnitude of Zsec and (Zlead + Zburden) .
In summary, the model in Figure 23.1 is used because of the following reasons: the data
requirements for the EMTP transformer model, the difficulty of separating Zm from the
measurement of (Zsec + Zm), and the low magnitude of Zsec relative to (Zlead + Zburden) for low
reactance current transformers. It is possible to use the model in Figure 23.2 when the following
is done: Zm is separated from the secondary open circuit measurement of (Zsec + Zm), and the
data requirement of the EMTP transformer model can be satisfied by separating Zsec into two parts
23 - 5
Protection Systems
ZEMTP and Zsec~ Because of the difficulty of separating Zm from the open circuit measurement,
it may in practice be possible to only achieve the data requirement of the transformer model.
The modules represent the basic structure of the CT model. The details needed by this module are
the following :
the names of the primary nodes to which the CT is connected
the names of the secondary nodes of the CT
the parameters of the transformer, such as the primary and secondary resistance and
leakage inductance, and number of turns on each winding.
the details of the burden and lead impedances.
the remanent flux, the frequency, and the coordinates of the saturation point.
the magnetization curve, either V(rms)-I(rms) or (peak)-i(peak) hysteresis curve.
flags for selecting default ct values, the burden connection, the lead impedance, and
the magnetization curve options.
EMTP processes the input data case and computes the hysteresis curve for ARMCO M4 oriented
silicon steel required by the Type 96 nonlinear hysteretic element. It then inserts the appropriate CT
module with the required inputs including the hysteresis curve in the data case. This module is then
sorted and included in the data case.
R1
L1
R2
N1
L2
Lead
Impedance
Switch
N2
Type 96
Nonlinear
Hysteretic
Inductor
R1, L1 : Transformer primary resistance and leakage inductance

R2, L2 : Transformer secondary resistance and leakage inductance
Switch closes at T = 0.0
Figure 23.1: Current Transformer Model

23 - 6
Protection Systems
Zsec + Lead
Impedance
Zpri
Switch
N2
N1
Type 96
Nonlinear
Hysteretic
Inductor
Zm
Zpri : Transformer primary resistance and leakage inductance

Zsec : Transformer secondary resistance and leakage inductance
Zm : Transformer magnetizing impedance
Figure 23.2: Good Representation of a CT Model in EMTP
Zpri
N1
Zemtp
N2
Zsec + Lead
Impedance
Switch
Type 96
Nonlinear
Hysteretic
Inductor
Zm

Zemtp: Transformer impedance required by EMTP
Zsec = Zsec - Zemtp
Figure 23.3: EMTP Requirements of a CT Model

23 - 7
Protection Systems
23.2.3
Data Format
The data format of the CT model is shown below. Many numerical quantities are read in as
characters. On top of the parameters controlling the type of models, the following information is
needed. Note these must be entered as comment lines in the data file, i.e. with C in columns 1-2.
Keyword:
1
2
3
4
5
6
7
8
12 3456789 01234567890123456789012345678901234567890123456789012345678901234567890
C
CTMODEL
CTMODEL
3-9
CT Node Names:
I2
CTPRI1 CTPRI2 CTOUT1 CTOUT2

A6
A6
A6
A6
CTRIDN
IDEFLG
1
2
3
4
5
6
7
8
12 34 567890 123456 789012 345678 901 2345678901234567890123456789012345678901234567890
A3
IDEFLG
3-4
00 user supplies all data

01 all default values used by EMTP
CTPRI1
5-10
CT primary terminal node name
CTPRI2
11-16
CT primary terminal node name
CTOUT1
17-22
CT output (secondary) terminal node name
CTOUT2
23-28
CT output (secondary) terminal node name [usually
grounded, enter ######]

CTRIDN
29-31
a three character, unique identifier of a particular CT
23 - 8
Protection Systems
CT Transformer Parameters:
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
C
RNPRIM RNSEC PRIRES PRILEK SECRES SECLEK
A6
A6
A6
A6
A6
A6
RNPRIM
3-8
number of primary turns of CT
RNSEC
9-14
number of secondary turns of CT
PRIRES
15-20
resistance of primary winding of CT (ohms)
PRILEK
21-25
leakage inductance of primary winding of CT. (in mH )
SECRES
27-32
secondary winding resistance of CT
SECLEK
33-38
leakage inductance of secondary winding of CT. (in mH)
CT Burden and Lead Parameters :
I2
BURRES BURIND BURCAP
A6
A6
A6
ILEDFG
IBURFG
1
2
3
4
5
6
7
8
12 34 567890 123456 789012 34 567890 123456 789012 34567890123456789012345678901234567890
I2
RLDRES RLDIND RLDCAP
A6
A6
A6
IBURFG
3-4
00 user will connect the burden across the output terminals

01 burden values supplied by user and connected by EMTP
BURRES
5-10
resistance of the CT burden
BURIND
11-16
inductance of the CT burden (in mH if XOPT =0)
BURCAP
17-22
capacitance of the CT burden (in F if COPT =0)
ILEDFG
23-24
00 no lead impedance
01 lead impedance values supplied by user and connected
by EMTP
RLDRES
25-30
resistance of lead, if required, in ohms
RLDIND
31-36
inductance of lead, if required, in mH ( if XOPT =0)
RLDCAP
37-42
capacitance of lead, if required, in ,F ( if COPT =0)
23 - 9
Protection Systems
CT Magnetization Parameters:
IHYSFG
1
2
3
4
5
6
7
8
12 34 567890 12345678 90123456 78901234 12345678901234567890123456789012345678901234567890
I2
REMFLX
FREQ
RISAT
VSAT
A6
E8.0
E8.0
E8.0
IHYSFG
3-4
00 user supplies V(rms)-I(rms) curve, REMFLX, FREQ,

RISAT, VSAT
01 user supplies (peak)-I(peak) curve, and REMFLX
REMFLX
5-10
remanent flux in the core (in Wbt)
FREQ
11-18
frequency of the power system
RISAT
19-26
the current at the saturation point on the V(rms)-I(rms) curve
VSAT
27-34
the voltage at the saturation point on the V(rms)-I(rms) curve
CT Magnetization Curve :
1
2
3
4
5
6
7
8
12 3456789012345678 9012345678901234 5678901234567890123456789012345678901234567890
C
RIRMS
VRMS
E16.0
E16.0
RIRMS
3-18
the I(rms) or the I(peak), depending on IHYSFG
VRMS
19-34
the V(rms) or the (peak), depending on IHYSFG
Termination:
1
2
3
4
5
6
7
8
12 3456789012345678 90123456789012345678901234567890123456789012345678901234567890
C
9999.
9999.
3-18
the number 9999. signals the end of the variable lenrth

V(rms)-I(rms) curve. If the user inputs the (peak)-i(peak)
curve, then 22 points are required to define the curve, and
then the number 9999. should be entered.
23 - 10
Protection Systems
23.2.4
Output
The outputs of the CT model are the current through the burden and the voltage across the burden.
These outputs are available in the cases where the user has selected the option IBURFG=01.
23.3 CVT Model

23.3.1
Introduction
The Capacitor Voltage Transformer (CVT) model is shown in Figure 23.4. The user is refered to
Section 18 for an explanation of the data module concept. The following files are required in the
users directory or the EMTP directory to use the CVT model:
For cases including the Type 96 nonlinear Hysteretic inductor:
CVTBLH.INC
This is the burden, lead and hysteresis module. It is included when

IBURFG = 01, ILEDFG = 01, and IHYSFG = 00 or 01. The hystheresis
curve is either computed from the V(rms)-I(rms) curve or 22 points of
the (peak)-I(peak) curve.
CVTNBLH.INC
This is the no burden, lead and hysteresis module. It is included when

IBURFG = 00, ILEDFG = 01, and IHYSFG = 00. The hystheresis curve is
computed from the V(rms)-I(rms) curve. The burden values must be
connected by the user on the CVT secondary side.
CVTBNLH.INC
This is the burden, no lead and hysteresis module. It is included when

IBURFG = 01, ILEDFG = 00, and IHYSFG = 00. The hystheresis curve is
computed from the V(rms)-I(rms) curve.
CVTNBNLH.INC
This is the no burden, no lead and hysteresis module. It is included

when IBURFG = 01, ILEDFG = 00, and IHYSFG = 00. The hystheresis
curve is computed from the V(rms)-I(rms) curve. The burden values
must be connected by the user on the CVT secondary side.
For cases with no Hysteresis:

CVTBLNH.INC
This is the burden, lead and no hysteresis module. It is included when

IBURFG = 01, ILEDFG = 01, and IHYSFG = 02. No hystheresis curve is
required.
23 - 11
Protection Systems
CVTNBLNH.INC
This is the no burden, lead and no hysteresis module. It is included

when IBURFG = 00, ILEDFG = 01, and IHYSFG = 02. No hystheresis
curve is required. The burden values must be connected by the user on
the CVT secondary side.
CVTBNLNH.INC
This is the burden, no lead and no hysteresis module. It is included

curve is required.
CVTNBNLNH.INC
This is the no burden, no lead and no hysteresis module. It is included

curve is required.
One feature of the CVT model is the Type 96 hysteresis which allows the inclusion of the remanent
flux in the CVT. The type 96 element requires the (peak)-i(peak) characteristic of the core
material. Typically, the user has access to a V(rms)-l(rms) curve. The CVT model has an inbuilt
facility to convert the V(rms)-l(rms) curve of the ARMCO M4 oriented silicon steel to the (peak)i(peak) curve required by the type 96 element. This conversion process can be invoked by setting
IHYSFG = 00. If, the user already has knowledge of the (peak)-i(peak) curve of the core, then the
user must enter IHYSFG= 01.
Note:
22 points of (peak)-i(peak) curve are required. The user must also set both
IBURFG=01 and ILEDFG=01].
The default values of the CVT model are for a 110kV CVT.
23.3.2
A Discussion of the CVT Model
EMTP requirements:
A good representation of a capacitor voltage transformer equivalent circuit for modeling purposes
in EMTP is shown in Figure 23.5. In this representation the secondary winding resistance and the
secondary winding leakage inductance are included with the lead impedance and the burden
impedance. The EMTP representation is shown in Figure 23.6. In this model Zm is placed after
ZEMTP, in parallel with (Zsec+ Zlead+ Zburden). This representation is required because finite
values must be given for Zpri and ZEMTP. More precisely, LEMTP must be non-zero, and REMTP
can be zero; and either Rpri or Lpri can be zero, but not both. This data requirement of the EMTP
transformer model can be satisfied by separating Zsec into two parts ZEMTP and Zsec. ZEMTP
should be small, however, very small values of ZEMTP give rise to very large values of the
admittance 1/ZEMTP in the matrices [Y] and [G]. This can "swamp out" the effects of other
elements connected to ZEMTP, leading to inaccurate results. It is not possible to give an absolute
lower bound on the magnitude of ZEMTP, because it depends on the relative ratio of Zpri and
ZEMTP- The lower bound on ZEMTP also depends on the magnitude of the other impedances in the
network: this is due to the different multipliers used in the forward reduction of the matrices [G]
and [Y], which affect the diagonal element corresponding to the node being eliminated. TOLMAT is
23 - 12
Protection Systems
a near-zero tolerance, which is used to check the singularity of the steady-state complet matrix
[Y]. If ZEMTP is very small, then the EMTP will give the disconnected subnetwork error message.
EPSILN is a near-zero tolerance, which is used to check the singularity of the real coefficient
matrix [G] within the time-step loop, where
[G] Vnode ( t) = Inode(t)
If ZEMTP is very small, then the EMTP will give the disconnected subnetwork error message in the
steady-state solution, or the floating network error message in the time-step solution. With the
approximation of placing Zm as in Figure 23.6, there is no appreciable change in the voltage seen
by the relays. The magnitude of Zm can be as high as six megohms (see Reference 33). The
magnitude of Zburden for which CVTs are tested varies from 40 ( corresponding to 100 VA) to
160 ( corresponding to 25 VA ) for a 63.5 V(rms) secondary rating. There is a lower bound on
the magnitude of Zburdens, but there is no upper limit on the magnitude of the connected
Zburden~ The order of magnitude Zsec is 1 . The magnitude of Zsec is much smaller than both
Zburden and Zm. Thus the position of Zsec does not affect the relay voltage in any significant
manner.
Open Circuit Magnetization Characteristics:

A similar discussion as in the case of current transformers applies here for the exact procedure for
the determination of Lsec, and the separation of Zm from the measured value of (Zsec + Zm).
However, in the case of the capacitor voltage transformer it is not necessary to do this because the
difference in the subsidence voltage transient with and without the Zm is negligible in most cases,
and is extremely small (around three percent ) in those cases where it exists.
CVT model for EMTP:

For capacitor voltage transformers, generally Zsec << (Zlead+ Zburden), and it can be included
as shown in Figure 23.4. This does not change the voltage seen by the relays both in the steadystate and during the subsidence transient. This is because the magnitude of Zm is much larger
than Zburden and has a negligible influence both, during steady-state operation, and during the
subsidence transient regime. Zm has been included in the CVT model because the nonlinear
magnetization characteristic is essential to simulate ferroresonance.
In summary, the model in Figure 23.4 is used because of the following reasons: the data
requirements for the EMTP transformer model, the difficulty of separating Zm from the
measurement of (Zsec + Zm)} and the low magnitude of Zsec relative to Zm and (Zlead +
Zburden) for capacitor voltage transformers. It is possible to use the model in Figure 23.6 when
the following is done: Zm is separated from the secondary open circuit measurement of (Zsec +
Zm), and the data requirement of the EMTP transformer model can be satisfied by separating Zsec
into two parts ZEMTP and Zsec. Because of the difficulty of separating Zm from the open circuit
measurement, it may in practice be possible to only achieve the data requirement of the
transformer model.
23 - 13
Protection Systems
In most cases, the CVT will be used to simulate the subsidence transient and its effect on connected
relays. For such an application the user need not include Zm in the CVT model, with negligible
difference in the voltage at the relay terminals. The user can include the different CVT models
without the nonlinear hysteretic inductor option.
23.3.3
Data Format
The data format of the CVT model is shown below. Many numerical quantities are read in as
characters. On top of the parameters controlling the type of models, the following information is
needed:
Keyword:
1
2
3
4
5
6
7
8
12 34567890 1234567890123456789012345678901234567890123456789012345678901234567890
C
CVTMODEL
CVTMODEL
3-10
CVT Node Names:
I2
CVTPR1 CVTPR2 CVTOT1 CVTOT2
A6
A6
A6
A6
CVTIDN
IDEFLG
1
2
3
4
5
6
7
8
12 34 567890 123456 789012 345678 901 2345678901234567890123456789012345678901234567890
A3
IDEFLG
3-4
00 user supplies all data

01 all default values used by EMTP
CVTPR1
5-10
CVT primary terminal node name
CVTPR2
11-16
CVT primary terminal node name (usually ground: ######)
CVTOT1
17-22
CVT secondary terminal node name
CVTOT2
23-28
CVT secondary terminal node name (usually ground: ######)
CVTIDN
29-31
a three character, unique identifier for a particular CVT
23 - 14
Protection Systems
Capacitor Divider and Tuning Inductor Values:

1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234567890123456789012345678901234567890
C
CPDIV1 CPDIV2 SLFIND RESIND CAPIND
A6
A6
A6
A6
A6
CPDIV1
3-8
value of capacitor of capacitor C1 voltage divider
CPDIV2
9-14
value of capacitor of capacitor C2 voltage divider
SLFIND
15-20
value of inductance of tuning inducto, mH
RESIND
21-26
value of resistance of tuning inductor,
CAPIND
27-32
value of the capacitance of the tunig inductor (in F if

COPT=0)
CVT Transformer Parameters:

1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
C
RNPRIM RNSEC PRIRES PRILEK SECRES SECLEK
A6
A6
A6
A6
A6
A6
RNPRIM
3-8
number of primary turns of CVT
RNSEC
9-14
number of secondary turns of CVT
PRIRES
15-20
resistance of primary winding of CVT,
PRILEK
21-26
leakage inductance of primary winding of CVT, mH
SECRES
27-32
secondary winding resistance of CVT
SECLEK
33-38
leakage inductance of secondary winding of CVT (in mH if

XOPT = 0)
23 - 15
Protection Systems
CVT Burden and Lead Parameters:
I2
BURRES BURIND BURCAP
A6
A6
A6
ILEDFG
IBURFG
1
2
3
4
5
6
7
8
12 34 567890 123456 789012 34 567890 123456 789012 34567890123456789012345678901234567890
I2
RLDRES RLDIND RLDCAP
A6
A6
A6
IBURFG
3-4
00 user will connect the burden across the output terminals

01 burden values supplied by user and connected by EMTP
BURRES
5-10
resistance of the CVT burden
BURIND
11-16
inductance of the CVT burden (in mH if XOPT = 0)
BURCAP
17-22
capacitance of the CVT burden (in F if COPT=0)
ILEDFG
23-34
00 no lead impedance
01 lead impedance values supplied by user and connected
by EMTP
RLDRES
25-30
resistance of lead, if needed,
RLDIND
31-36
inductance of lead, if needed (in mH if XOPT = 0)
RLDCAP
37-42
capacitance of lead, if needed (in F if COPT=0)
CVT Magnetization Parameters:
IHYSFG
1
2
3
4
5
6
7
8
12 34 567890 12345678 90123456 78901234 12345678901234567890123456789012345678901234567890
I2
REMFLX
FREQ
RISAT
VSAT
A6
E8.0
E8.0
E8.0
IHYSFG
3-4
00 user supplies V(rms)-I(rms) curve, REMFLX, FREQ,

RISAT, VSAT
01 user supplies (peak)-I(peak) curve, and REMFLX
02 type-96 element not included in model
REMFLX
5-10
remanent flux in the core (Wbt)
FREQ
11-18
frequency of the power system
RISAT
19-26
current at the saturation point on the V(rms)-I(rms) curve
VSAT
27-34
voltage at the saturation point on the V(rms)-I(rms) curve

23 - 16
Protection Systems
CVT Magnetization Curve:

1
2
3
4
5
6
7
8
12 3456789012345678 9012345678901234 5678901234567890123456789012345678901234567890
C
RIRMS
VRMS
E16.0
E16.0
RIRMS
3-18
the I(rms) or the I(peak), depending on IHYSFG
VRMS
19-34
the V(rms) or the (peak), depending on IHYSFG
Termination:
1
2
3
4
5
6
7
8
12 3456789012345678 90123456789012345678901234567890123456789012345678901234567890
C
9999.
23.3.4
9999.
3-18
the number 9999. signals the end of the variable lenrth

V(rms)-I(rms) curve. If the user inputs the (peak)-i(peak)
curve, then 22 points are required to define the curve, and
then the number 9999. should be entered.
Output
The outputs of the CVT model are the current through the burden and the voltage across the
burden. These outputs are available in the cases where the user has selected the option IBURFG =
01.
23 - 17
Protection Systems
C1
L0
R1
L1
C2
R2
Lead
Impedance
L2
Switch
N1
N2
Type 96
Nonlinear
Hysteretic
Inductor
C1, C2 : Capacitor Divider

L0
: Tuning Inductor
R1, L1 : Transformer primary resistance and leakage inductance

R2, L2 : Transformer secondary resistance and leakage inductance
Figure 23.4: Capacitor Voltage Transformer Model
C1
L0
C2
Zsec + Lead
Impedance
Zpri
N1
Switch
N2
Type 96
Nonlinear
Hysteretic
Inductor
Zm

Figure 23.5: Good Representation of a CVT Model in EMTP

23 - 18
Protection Systems
C1
L0
C2
Zpri
N1
Zemtp
N2
Zsec + Lead
Impedance
Switch
Type 96
Nonlinear
Hysteretic
Inductor
Zm

Zemtp: Transformer impedance required by EMTP
Zsec = Zsec - Zemtp
Figure 23.6: EMTP Requirements of a CVT Model
23.4 Instructions For Using CEY51A Relay

The CEY51A is a three-phase, single-zone, mho directional distance relay manufactured by
General Electric Company of USA. Details of the CEY51A relay such as calculation of settings is
given in Reference 22. The CEY relay model developed is a single phase model. To simulate the
protection for a three phase line, three CEY51A models will have to be used.
23.4.1
Description of CEY51A
The connections to the CEY51A relay model are the secondaries of the current transformer, and the
secondaries of the voltage transformer. The settings of the relay required are:
the basic minimum ohmic tap required (allowed values are 0.75, 1.5, or 3.0).
SETOHM
the tap value of the autotransformer (allowed values -10 -100 ). SETTAP
23 - 19
Protection Systems
the maximum torque angle (allowed range 60-75 degrees). TORANG
the ohmic reach adjustment (allowed variation is 0-1.0 p.u.). TRESTR
a variable time delay (TDELAY).
The CEY51A model has been incorporated in EMTP using the module CEY51.INC. The module
represents the basic structure of the CEY51A relay. The details needed by it are the following:
the node names of the voltage input to the relay
a three character unique identifier for the particular relay
the number of turns on the primary and secondary of the transactor
the value of the inductance required to get the desired angle of maximum torque
the value of the adjustable restraint resistance
the tap setting on the potential transformer input
the variable time delay for the relay trip output
EMTP processes the input data case and computes quantities required by the module. The input
data case can have either the relay settings, or the protected line parameters and instrument
transformer data. In both the cases the quantities required by the module are computed. The
module CEY5.INC with the required input is inserted in the data case. The module is then sorted and
included in the data case. The outputs of the module are the torque developed by the relay, the
angle of the moving contact, and the final trip output of the relay.
23.4.2
Data Format
The data format for the CEY5IA relay is given below. An explanation of the various terms is now
given.
Keyword:
1
2
3
4
5
6
7
8
12 345678 901234567890123456789012345678901234567890123456789012345678901234567890
C
CEY51A
CEY51A
3-8
23 - 20
Protection Systems
Input To Relay:
CTPRI1 CTPRI2 VOLTR1 VOLTR2
I2
A6
A6
A6
A6
CEYIDN
ISETFG
1
2
3
4
5
6
7
8
12 34 567890 123456 789012 345678 901 2345678901234567890123456789012345678901234567890
A3
ISETFG
3-4
00 : relay settings read from card 3

01: line/transformer data read from card 3, EMTP calculates
relay settings
CTPRI1
5-10
First node name of the current input
CTPRI2
11-16
Second node name of the current input
VOLTR1
17-22
First node name of the voltage input
VOLTR2
23-28
Second node name of the voltage input
CEYIDN
29-31
Three-character unique identifier for a relay
Option 1:
ISETFG = 0
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234567890123456789012345678901234567890
C
SETOHM TORANG SETTAP TRESTR TDELAY
F6.2
F6.2
F6.2
F6.2
F6.2
SETOHM
3-8
Basic minimum ohmic tap required (either 0.75, 1.5 or 3.0)
TORANG
9-14
Maximum torque angle (60deg<TORANG<75deg)
SETTAP
15-20
Tap value of autotransformer (10.0<SETTAP<100.0)
TRESTR
21-26
Ohmic reach adjustment (0<TRESTR<1.0, suggest:0.57)
TDELAY
27-32
Variable time delay for the trip output
23 - 21
Protection Systems
Option 2:
ISETFG = 01
1
2
3
4
5
6
7
8
12 34567890 12345678 9012345678 9012345678 901234 567890 123456789012345678901234567890
C
PROTEZ
F8.2
ANGLEZ
F8.2
PTRATO
F10.2
CTRATO
F10.2
TORANG TDELAY
F6.2
F6.2
PROTEZ
3-10
Magnitude of line impedance to be protected,
ANGLEZ
11-18
Angle of line impedance to be protected, degrees
PTRATO
19-28
Potential transformer ratio (prim. volt/sec. volt)
CTRATO
29-38
Current transformer ratio (prim. amp/sec. amp)
TORANG
39-44
Maximum torque angle (60deg<TORANG<75deg)
TDELAY
45-50
Variable time delay for the trip output
The program will calculate the relay settings from the given data.
23.4.3
Output of the CEY51A Relay
The outputs of the relay model are TORAAA, ANGAAA, TRPAAA where AAA is the threecharacter unique identifier of this particular CEY51A relay. TORAAA is the torque produced on the
relay. ANGAAA is the angle of the moving contact of the relay. TRPAAA is the final trip output of
the relay which includes the time delay.
23.5 Instructions For Using D202 Relay

23.5.1
Description of D202
The D202 is a harmonic restrained percentage differential relay designed for the protection of two
winding transformers and unit-connected transformers and generators. The D202 relay is
manufactured by the Brown Boveri Corporation (BBC), now Asea Brown Boveri (ABB). Details of
the D202 relay are given in Reference 6.
The typical connections to the D202 relay are the secondaries of the current transformers on either
side of the protected devices. There are two settings required for the D202 relay. They are
23 - 22
Protection Systems
1.
g = I/IN
g = the pickup value corresponds to the basic setting g
I = Difference between the two secondary currents (I1-I2)
IN = Nominal current rating of D202 relay (5A [rms]).
The values of g allowed are 0.2, 0.3, 0.4, and 0.5.

2.
V = I/IH = (I1-I2) / 0.5(I1+I2)

V = pickup ratio or bias [the slope of the differential characteristic]
IH = multiple of the rated current IN
The allowed values of V are 1.2, 1.3, 1.4. and 1.5.

The D202 model has been incorporated in the module D202.INC. The module should be present in
the EMTP directory and the user should have access to the module. The module represents the
basic structure of the D202 relay. The details needed by the module are the following:
the node names of the current input from one side of protected device
the node names of the current input from the other side of the protected devlce
a three character unique identifier of the particular D202 relay
the current through the zener diode, at the breakdown voltage
the value of resistance across the operating coil
EMTP processes the input data cases and computes the quantities required by the module. It then
inserts the D202 module with the required quantities in the data case. This module is then sorted
and included in the data case. The output of the D202 model is the trip signal.
23.5.2
Data Format
The format for the data format cards is given below. An explanation of the various terms is riven
below:
23 - 23
Protection Systems
Keyword:
1
2
3
4
5
6
7
8
12 3456 78901234567890123456789012345678901234567890123456789012345678901234567890
C
D202
D202
3-6
Inputs to Relay:
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234567890123456789012345678901234567890
C
NODCT1 NODCT2 NODDT1 NODDT2 D22IDN
A6
A6
A6
NODCT 1
3-8
NODCT2
9-14
NODDT1
15-20
NODDT2
21-26
D22IDN
27-29
A6
A6
Node names of current input from one side of the protected

device
Node names of current input from the other side of the protected device
Three character unique identifier of a particular relay
Relay Settings:
1
2
3
4
5
6
7
8
12 345678 901234 567890123456789012345678901234567890123456789012345678901234567890
C
F6.2
F6.2
3-8
The pickup value corresponds to the basic setting g (F6.2).
9-14
Pickup ratio or bias (the slope of the differential characteristic)
23 - 24
Protection Systems
23.5.3
Output of the D202 Relay
The outputs of the D202 relay model are the following :

1.
the trip signal TR6IDT
2.
the current through the trip coil, identified as OR7IDT - OR5IDT.
3.
the current through the primary of the operating ( the differential) current
transformer, identified as CP1IDT - CTOIDT.
IDT refers to the three character unique identifier of a particular D202 relay.
23.6 Instructions For Using Sly Relay

23.6.1
Description of the Sly Relay
The SLY relay is a static mho phase distance relay manufactured by General Electric Company of
USA. The type SLY designation covers a family of static mho phase distance relays intended for
the protection of transmission lines. The specific SLY relay modeled is the SLY12C. The SLY12C is
specifically designed to provide high-speed multi-phase fault protection in step distance and/or
directional comparison blocking schemes. Details of the SLY and SLY12C are given in Reference
29 and Reference 30. Some knowledge of the material in these references is necessary for
determining the relay settings.
Typical connections to the SLY12C are the secondaries of the current transformers for phase A,
phase B. and phase C. Also connected to the SLY12C are the secondaries of the voltage
transformers for phase A, phase B. and phase C. The required settings of the relays are
the forward ohmic setting of the M1 function (OHMFM1)
the reverse ohmic setting of the M1 function (OHMRMI)
the total ohmic reach of the M2 function (OHMTM2)
the forward offset ohmic reach of the M2 function (OHMOM2)
the restraint tap setting for the M1 function
the restraint tap setting for total ohmic reach of the M2 function (RSTRM2)
the restraint tap setting for forward offset ohmic reach of M2 function (RSTRM3)
angle of maximum reach in degrees (ANGMRE)
the pickup current for the supervisory function in A rms (IPCKUP)

23 - 25
Protection Systems
The relay has been modelled by creating the SLY12C.INC module which needs data such as the
following:
three pairs of current transformer secondary names
three voltage transformer secondary names
a two character unique identifier for the relay
three multipliers for the voltage tap settings of the relay
three gains for the transactors
a value for the pickup current to be compared with the I3maximum current
a value for the inductor to give a variable phase shift for the angle of maximum
reach.
Creation of the module SLY12C.INC required changes to subroutine SUBR40 of AUX. The
parameters MAXARG and MAXCHA must be changed from 99 and 200 to 1000 and 4000
respectively, because of the complexity and the large number of arguments required by this relay.
23.6.2
Data Format
To run an EMTP data case the user has to input data as shown in below. The relay settings given in
the data case are used to compute the following values :
gain for the M1 forward reach transactor
gain for the M1 reverse reach transactor
gain for the M2 total reach transactor
gain for the M2 forward offset reach transactor
multiplier for the restraint tap setting for the forward ohmic reach of the M1 function
multiplier for the restraint tap setting for the total ohmic reach of the M2 function
multiplier for the restraint tap setting for the forward offset ohmic reach for the M2
function
the value of the inductance (in mH) to give the desired phase shift to match the angle
of maximum reach
An explanation of the various data items follows:
23 - 26
Protection Systems
Keyword :
1
2
3
4
5
6
7
8
12 345678 901234567890123456789012345678901234567890123456789012345678901234567890
C
SLY12C
SLY12C
3-8
Inputs To Relay :
C CTTRA1 CTTRA2 CTTRB1 CTTRB2 CTTRC1 CTTRC2

A6
A6
A6
A6
A6
A6
VOLTR
VOLTRB VOLYRC
A
A6
A6
A6
SLYIDN
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 567890 123456 78 9012345678901234567890
A2
CTTRA1
CTTRA2
3-8
9-14
Node names of the current input to the relay of phase A

[Relay model shorts the two
nodes, and uses a TACS Type 91 current source - CTTRA1, as
current input]
CTTRB1
CTTRB2
15-20
21-26
Node names of the current input to the relay of phase B

nodes, and uses a TACS Type 91 current source - CTTRB1, as
current input
CTTRCI
CTTRC2
27-32
33-38
Node names of the current input to the relay of phase C

nodes, and uses a TACS Type 91 current source - CTTRC1, as
current input
VOLTRA
39-44
Node name of the voltage input to the relay of phase A.

[Relay model uses a TACS Type
90 voltage source VOLTRA, as voltage input]
VOLTRB
45-50
Node name of the voltage input to the relay of phase B.

90 voltage source VOLTRB, as voltage input]
VOLTRC
51-56
Node name of the voltage input to the relay of phase C.

90 voltage source VOLTRC, as voltage input]
SLYIDN
57-58
Two character unique identifier of a particular relay

23 - 27
Protection Systems
Relay Settings :
OHMFM2
OHMTM2
OHMOM2
RSTRM1
RSTRM2
RSTRM3
ANGMRE
IPCKUP
OHMFM1
1
2
3
4
5
6
7
8
12 34567 89012 34567 89012 34567 89012 34567 89012 34567 890123456789012345678901234567890
F5.1
F5.1
F5.1
F5.1
F5.1
F5.1
F5.1
F5.1
F5.1
OHMFM1
3-7
The forward ohmic setting of the M1 function - in ohms
OHMRM1
8-12
The reverse ohmic setting of the M1 function - in ohms.

Value is 0.0 or 0.5 ohms.
OHMTM2
13-17
The total ohmic reach of the M2 function - in ohms.
OHMOM2
18-22
The forward offset reach of the M2 function - in ohms .

OHMOM2 < OH MTM2)
RSTRM1
23-27
Restraint tap setting for the M1 function - In percentage,

value is between 0-100
RSTRM2
28-32
Restraint tap setting for the total ohmic reach of the M2

function - In percentage, value is between 0-100
RSTRM3
33-37
Restraint tap setting for forward offset ohmic reach of the

M2 function - In percentage, value is between 0-100
ANGMRE
38-42
angle of maximum reach in degrees - value is between 60.0 75.0
IPCKUP
43-47
pickup current for the supervisory function in A(rms) Value is between 1.0 - 8.0. Compare to I max.
In the data case the file SLY12C.INC is added along with the node names, the identifier and the
computed values. This is then processed by the EMTP. The values of the variables are taken and
substituted in the required places in the EMTP data case. The data case is then sorted to group all
elements in the required order, eg. suitches before sources etc.
Note:
The large number of substitutions required by the the module SLY12C.INC requires
LIMSUB and LIMARG of DATAIN to be dimensioned at least 999 and 3000
respectively.
23 - 28
Protection Systems
23.6.3
Output Of Sly Relay
The outputs of the SLY relay are the current inputs to the SLY relay for phase A and phase B. The
trip output for the ML function for Phase (A - B) faults TRX61D is available for plotting. The trip
output for the M2 function for Phase (A - B) faults TRU6ID is also available for plotting. Other trip
output is given in Figure 23.7, below. The trip outputs are derived from the outputs of the M1 and
M2 measurement functions for the Phase (A - B), Phase (B - C), and Phase (C - A) faults. The trip
signals are combined as shown in Figure 23.7. The signals correspond to Fig. 8 on Page 16 of
Reference 30. These trip signals can be used to develop further tripping logic schemes.
TRX6ID*
TRIP - M1 PHASE A-B
TRY6ID
TRIP - M1 PHASE B-C
TR73ID*
OR
TRZ6ID
TRIP - M1 PHASE C-A
TRU6ID*
TRIP - M2 PHASE A-B
TRV6ID
TRIP - M2 PHASE B-C
SUPDID
(I(max).GT.IPCKUP)
AND
TR72ID*
TR75ID*
OR
TRW6ID
TRIP - M2 PHASE C-A
SUPDID
(I(max).GT.IPCKUP)
AND
TR76ID*
SUPAID
SUPBID
OR
TR79ID*
SUPCID
TR72ID*
TR76ID*
OR
TR00ID*
DELAY
4 ms
TR77ID*
* TRIP OUTPUT
Figure 23.7: Final trip outputs of SLY relay
23 - 29
Protection Systems
23.7 Instructions For Using BDD15B Relay

23.7.1
Description of BDD Relay
The BDD type relays are differential type relays designed specifically for transformer protection.
The relays are provided with the features of percentage and harmonic restraint and use a polarized
main unit with an auxiliary relay as the operating element. An instantaneous unit picks up on
extremely large internal fault currents and completes the trip circuit. Percentage restraint permits
discrimination between internal and external faults at high fault currents. Harmonic restraint
enables the relay to distinguish, by the difference in waveform, between the differential current
caused by the internal fault, and that of transformer magnetizing inrush. The BDD15B relay is for
the protection of two winding transformers and has two through current restraint circuits and one
differential current circuit. Details of the BDD15B relay are given in Reference 38.
The module BDD15B contains the structure of the relay. It needs the following:
the node names of the secondaries of the current transformers on either side of the
protected transformer.
a two character unique identifier for the relay
ampere tap on winding 1
ampere tap on winding 2
resistance to control the pickup of the relay [Suggested value = 79.2 ohms, max.
value = 150.0 ohms].
resistance to control the harmonic restraint [Suggested value = 342.0 ohms, max.
value = 1000.0 ohms]
resistance to adjust the percentage slope (For 15%, 25%, and 40% the value is 5.7,
9.3, and 14.7 ohms respectively.)
EMTP processes the input data and the the relay settings given in the data case are used to compute
the following values.
number of turns in primary of winding 1 of through restraint current transformer
number of turns in primary of winding 2 of through restraint current transformer
number of turns in primary of winding 1 of differential current transformer
number of turns in primary between winding 1 and winding 2 of differential current

transformer
EMTP then inserts the module BDD15B.INC in the data case, along with the node names, the
identifier and required numerical values of quantities needed by the module. The values of the
variables are taken and substituted in the required place in the BDD15B module. The data case is
23 - 30
Protection Systems
then sorted to group all elements in the required order, for example, TACS before branches,
switches before sources, etc.
23.7.2
Data Format
Data formats are shown below.
Keyword:
1
2
3
4
5
6
7
8
12 345678 901234567890123456789012345678901234567890123456789012345678901234567890
C
BDD15B
BDD15B
3-8
Input to Relay:
CTSCB1 CTSCB2
A6
A6
BDDIDN
1
2
3
4
5
6
7
8
12 345678 901234 56 7890123456789012345678901234567890123456789012345678901234567890
A2
CTSCBI
3-8
Node name of current input from one side of the protected

device to the relay. [Ground is assumed to be the other current input]. This node is connected to the higher ampere tap
required for the relay settings.
CTSCB2
9-14
Node names of current input from other side of the protected

device to the relay. [Ground is assumed to be the other current input.
BDDIDN
15-16
Two character unique identifier of a particular relay.
23 - 31
Protection Systems
Relay Settings:
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678901234567890123456789012345678901234567890
C
WD1TAP WD2TAP R1OPCL R2HACL R3SLOP
F6.2
F6.2
WD1TAP
F6.2
3-8
F6.2
F6.2
This is the higher of the ampere tap rating on the winding.

The ampere tap on this winding corresponds to the node
CTSCBI
WD2TAP
9-14
The ampere tap rating of the winding corresponds to node

CTSCB2. The allowed values of both WD1TAP and WD2TAP
are 2.9, 3.2, 3.5, 3.8, 4.2, 4.6, 5.0, and 8.7.
RLOPCL
15-20
The resistor which is in parallel with the operating coil. This

resistor controls the pickup of the relay. Range of values
allowed is 1.0-150.0 ohms. Suggested value is 79.2 ohms.
R2HACL
21-26
The resistor is in parallel on the AC side of the harmonic

restraint rectifier. This resistor controls the amount of harmonic restraint applied to the relay. Range of values allowed
is 1.0 - 1000.0. Suggested value is 342.0 ohms.
R3SLOP
27-32
This resistor is in parallel with the restraining winding. This

resistor controls the slope of the differential characteristic.
The range of values allowed is 1.0-25.0 ohms.
For 15% slope, the suggested value = 5.7 ohms.
23.7.3
Output of BDD15B Relay
The output of the BDD relay is as follows:
trip output of main operating unit and auxiliary relay
trip output of instantaneous relay
trip output which is the result of 'OR'ing of the previous two trip signals.
23 - 32
Protection Systems
23.8 Instructions For Using CDR_101 Relay

23.8.1
Description of CDR_101 Relay
One of the earlier distance measurement algorithm suitable for computer relaying was introduced
by E.O. Schweitzer and A.J. Flechsig, Jr. in 1977. Details of the algorithm and its derivation and
implementation are given in Reference 27. Here a brief overview of the algorithm, its
implementation in EMTP, and details of how to use it in a data case are given. This acronym
CDR_101 stands for Computer Distance Relay.
The algorithm of the CDR_101 distance relay is the following. At the relaying point, the voltage
and current waveforms can be described by the following equations :
v(t) = V sin(t + )
i(t) = I sin(t + )
where V, I, _and_ are constants which change to new values at t = 0, the instant of
fault occurrence.
Consider two finite duration impulse response (FIR) filters, SAL(x) and CAL(x), defined as follows:
SAL(x)
= 1
= 1
= 0
CAL(x) = 1
= 1
= 0
0 x < 1/2
1/2 x < 1
for all other x
0 x < 1/2 or 3/4 x < 1
1/4 x< 3/4
for all other x
where x = t/2..
Let x(t) = Xsin(t +) be sampled N times per cycle at times t = kt, where t = 2/
N.
If the samples are denoted by x_k, then
x_k = Xsin( kt +) = Xsin( (2k_/ N_)__+_)
If N is a multiple of four, the results of time-discrete convolution of the input samples xk, with the
SAL and CAL finite-duration impulse responses are
Sx = AX cos(_+_/ N) and Cx = AX sin(_+_/ N)
23 - 33
Protection Systems
where A = 2 / sin (/ N). These are calculated from the samples x_k using
N/2
Sx = [x_k - x_(k+N/2)]
k= 1
N/4
Cx= ~[ x_k - { x_(k+N/4) + x_(k+ N/2) } + X_(k+3N/4) ]
k= 1
The N-th sample is the most recent and the N samples preceding the time of calculation are used in
the calculation of Sx and Cx. The filter outputs Sv, Cv, Si and Ci are found using the two equations
above. There are now two rules:
If Sv and Ci have opposite signs, and if |Sv| / |Ci| < Z protecred, then trip.
If Cv and Si have the same signs, and if |Cv| / |Si| < Z protecred , then trip.
Two successive trip signals must be calculated by the relay to produce a final trip output. This
feature increases the security of the relay.
In the data case the file CDR_101.INC is added along with the node names, the identifier and the
computed values. This is then processed by the EMTP. The values of the variables are taken and
substituted in the required places in the EMTP data case. The data case is then sorted to group all
elements in the required order, eg. suitches before sources etc.
Note:
The large number of substitutions required by the the module CDR_101.INC requires
LIMSUB and LIMARG of DATAIN to be dimensioned at leat 400 and 800 respectively.
The module CDR_IOI.INC contains the structure of the relay. It needs the following data:
two node names for the current input to the relay. The module shorts the two nodes
by a measuring switch and passes the current into TACS as a Type 91 source.
two node names for the voltage input to the relay. The first voltage input name is
passed into TACS as a Type 90 source.
a three character unique identifier for a particular relay.
the protected line impedance in ohms ( the allowed field width is six - F6.0 to F6.5).
the angle of the protected line in degrees ( the allowed field width is six).
the current transformer ratio ( Primary current / Secondary current rating). The
allowed field width is six.
the potential transformer ratio ( Primary voltage / Secondary voltage rating). The
allowed field width is six.
23 - 34
Protection Systems
23.8.2
Data Format
To run an EMTP data case using the CDR_101 relay, the user needs to type the following example
data in the data case.
$INCLUDE CDR_IOI.INC CTBECI, CTBEC2, VTBECI, VTBEC2, BBB, 0.1005,90.0,1.000,1.0
where:
CTBEC1
A6
example names for the Current Transformer secondary node

names
CTBEC2
A6
example names for the Current Transformer secondary node

names
VTBEC1
A6
example names for the Voltage Transformer secondary node

names
VTBEC2
A6
example names for the Voltage Transformer secondary node

names
BBB
A3
example of a three character name that uniquely identifies

this particular CDR_101 relay
0.1005
F6.0
example of the line impedance to be protected
90.0
F6.0
example of the angle of the line impedance to be protected
1.000
F6.0
example of the CT ratio
1.0
F6.0
example of the VT ratio
23.8.3
Output of CDR_101 Relay
The output of the relay model will be identified by the name TRPBBB in this case. More generally,
the name of the output will begin with TRP and the next three characters will be the unique
identifiers for this particular relay. A value of 1.0 indicates the relay decision to trip the line. A
value of 0.0 indicates no trip signal is issued.
23 - 35
Protection Systems
23.9 Instructions For User_defined Fortran Relay

23.9.1
Introduction
The capability to interface EMTP electrical network variables and TACS control signals to form a
hybrid EMTP-TACS interactive configuration, makes it possible to study the dynamic interaction
between the power system and the control system. Thus network variables such as voltage at a
node, current through a switch etc. are passed to TACS. TACS will in turn pass signals for switch
operation or converter valve firing. The signals from TACS are inevitably delayed by one time step.
This delay is generally acceptable.
23.9.2
Development Of User_Defined Relay Model
The user has the choice between interfacing his own subroutine modeling a relay through
1.
the general subroutine interface provided with TACS
2.
the specific Computer Relay Algorithm subroutine associated with relays
In both cases, the method consists of writing the relay model as a Fortran subroutine, which must
include all the necessary common blocks. This subroutine must be called by an appropriate
existing EMTP subroutine. This subroutine must be compiled, and saved as an object file. Then
both the existing EMTP object/library files and the relay subroutine object file must be linked to
create a run time version.
The first technique is described in Section 14. Here we described the second approach. Two
Fortran subroutines are written which allow the user to write his/her own Computer Relay
Algorithm. The Fortran Subroutines do the following:
1.
Pass the EMTP variables of interest to this Computer Relay Subroutine, i.e. the
values of currents, voltages, switch status and the internal variable of special
EMTP components required should be passed as needed.(Subroutine SOURCES)
2.
The outputs of this routine are moved into the TACS portion of EMTP.(Subroutine
OUTPUT)
These Fortran routines are called from the TACS portion of EMTP. The Type-94 TACS source allow
us to move variables from the Subroutine COMPREL (Computer Relay) to TACS. This Type 94 can
be used in a data case as any other TACS variable, for example, it can he used with a time delay to
turn on or turn off a TACS controlled switch.
Subroutine COMPREL must include a call to subroutine SOURCES to get the inputs such as TACS
sources Type 90 -93. These sources can be the current and voltage seen by the relay, or the status
23 - 36
Protection Systems
of some switch, or some synchronous machine variable. These inputs at a particular time step are
of course vital for any algorithm. Subroutine SOURCES is called as follows:
CALL SOURCES(NCRSRC,CRSRCS,SVALUE)
where
NCRSRC
Number of sources of Type 90 - 93.
CRSRCS(NCRSRC)
Array defining the names of the sources, format = A6.
SVALUE(NCRSRC)
Array of values of corresponding sources at time = T.
The inputs to this subroutine are NCRSRC and CRSRCS(NCRSRC). The value of the sources at time
= T is returned in array SVALUE.
After a call to SOURCES, the user has access to the values of the sources at time = T. The computer
relay algorithm can be run at this time, if T corresponds to a sampling instant. It is more efficient
to call SOURCES at a sampling instant. The variable T is the time the simulation has reached and is
contained in the common blocks. The user's Fortran code for the relay is then executed.
After the algorithm is executed, it is necessary to pass variables from the subroutine COMPREL to
TACS, the user should have defined a Type 94 source of any name in the users EMTP data case.
The user should use the same name to pass the value of this variable to TACS from COMPREL. The
user can call subroutine OUTPUT as
CALL OUTPUT(NOUTPT,CROTPT,CRVLOT)
where
NOUTPT
Number of outputs
CROTPT(NOUTPT)
Array defining the names of the outputs, format = A6
CRVLOT(NOUTPT)
Array defining the corresponding outputs at time = T
For example if, CROTPT(1) = 'NODEA1' and CRVLOT(1) = 5.3 , then, the value assigned to a Type
94 source NODEAL is 5.3, at time = T. The Type 94 source NODEAI must be previously defined in
the EMTP data case. This Type 94 source can be used in a data case as any other TACS variable,
e.g., it can be used with a time delay to turn on/off a TACS controlled switch. To plot or print this
variable the TACS Type 33 flag can be used.
23 - 37
Section 24
Air Gap Arc Model
24.1
Introduction ......................................................................................... 24-1
24.2
Determination of Parameters Through the Integral Law .................... 24-2
24.3
Elaboration of the Spark Gap Model in EMTP .................................... 24-4
24.4
Application .......................................................................................... 24-6
24.5
EMTP Routine of Air Gap Model ........................................................ 24-7
Section 24
Air Gap Arc Model
This section presents a spark gap model which follows the integral law. It is suited to all fast front
wave forms, whatever the polarity. For a given spark gap, the user must first determine the three
basic parameters of the integral law on the basis of results of the spark gap behavior against
standard waves, and the solution of a set of three integral equations. At the moment of arcing, the
model takes into account the progressive build-up of the arc current in a simplified manner, by
imposing an exponential-form decrease of the spark gap voltage.
24.1 Introduction
A spark gap in air protects station equipment by limiting overvoltages in the network. This
limitation takes place through creation of an arc and a short-circuit to earth. When the
overvoltage at the spark gap terminals exceeds the dielectric strength of air between its horns, the
air ionizes and starts to conduct. Ionization conditions depend on several factors, the major ones
of which are the form and polarity of the incident voltage wave and the shape of the spark gap.
The model we are proposing in this section is limited to representation of the behavior of spark
gaps in air against fast front waves.
Today's literature indicates several methods for evaluation of the moment and level of arcing of
spark gaps although we have mainly determined two methods suited to the very varied wave
forms which occur in networks after a lightning strike: the integration method and the leader
propagation method.
This section addresses the first of these methods, by describing how its parameters are
determined, and by proposing an EMTP module.
We start by reminding the equation which determines dielectric withstand with this method:
(Form 1)
Tb
n
( U ( t ) U0 ) dt = D
T0
24 - 1
Air Gap Arc Model

In this equation, parameters D, n and U0 characterize the air gap. Arcing takes place at moment
Tb, when the integral becomes equal to D. Integration starts at T0, when U(t) exceeds the constant
voltage U0. The arcing voltage is U(Tb).
The model implanted into EMTP consists in integrating the voltage at the terminals of the spark gap
from the moment when U(t)>U0 on, then in closing a switch when the integral becomes equal to
D.
Besides encoding of the model, the main problem is the determination of the characteristic
parameters of a given air gap.
24.2 Determination of Parameters Through the Integral Law

The integral law enables calculating the moment (ta) and the level (Uarc) of arcing of a spark gap,
whatever the shape of the wave, notably with standard 1.2/50 s waves. However, these
characteristics for these waves are available thanks to the numerous laboratory tests carried out.
Plotting the results of these tests on a Uarc-ta graph leads to a nebula of points of roughly
hyperbolic envelopes. This characteristic is called "voltage-time characteristic" of the spark gap.
Against a slight approximation, use of three pairs Uarc-ta for a standard wave enables to determine
coefficients U0, n and D of the spark gap, through numerical solving of a set of three integral
equations.
We have used this procedure on the basis of tests carried out by EDF in its laboratories [4] on
insulator strings fitted with spark gaps set between 15 and 80 cm. These tests led to determination
of a formula for the voltage-time characteristic of the spark gap:
(Form 2)
5
5 ( 1,36 d 0,15 )
Va = 4,8 10 d + 1,04 10 + ---------------------------------------t
d = spacing of the spark gaps, in m
t = arcing time, in s
This formula is valid only for arcing times in excess of 0.5 s and for waves with negative polarity.
One then assumes that the insulator string gap and the spark gap support a line which conducts an
overvoltage of fast front standard form, so that the part of the wave used for determination of the
striking voltage and moment can be deemed as being linear. Uarc-ta has been determined for three
incident waves meeting these conditions, after which the MATHEMATICA software package has
24 - 2
Air Gap Arc Model

been used to solve numerically the set of three integral equations in order to determine U0, n and
D. The calculation results are tabulated below:
d, in m
U0, in kV
0.8
0.71
0.61
0.4
0.35
0.28
0.92
0.92
0.93
0.93
0.93
0.92
343
311
276
205
188
167
1.40E-01
1.30E-01
1.10E-01
6.90E-02
5.60E-02
3.70E-02
Table 1: Value of parameters of the integral

law according to the length of the air gap D
The method used for determination of U0, n and D is wholly general and can be extended to any
gap value and wave polarity. It suffices to know the voltage-time characteristic for the
configuration to be investigated. Moreover, one notes that the results obtained from the triplet of
parameters are all the more accurate as the incident wave is closer to the waves used for
theoretical calculations.
We indicate as examples two other formulas for the voltage-time curves.
There are two formulas for voltage-time curves:
(A)
the first one is that of MM. Darveniza, Popolansky, Whitehead (see Reference
5 and Reference 26) and is expressed as follows:
(Form 3)
0,71 d
V = 0,4 d + ------------------0,75
t
V is the arcing voltage, in MV
d is the spacing of the air gap, in m
t is time, in s
This formula has been determined in large spark gaps, with a length in excess of 5.6 m.
(B)
the second one is that of the Insulation Coordination Work Station of EDF's
ERMEL Department and has been proposed by Prof. S. SADOVIC (see
Reference 39).
24 - 3
Air Gap Arc Model

(Form 4)
0,9
V = U50 0,87 + -------
t
V is the arcing voltage, in kV
U50 is in kV
d is the spacing of the air gap, in m
t is time, in s
These relationships are based on tests, the results of which are uncertain for arcing delays in excess
of one s, although this type of situation occurs, for instance, in case of direct lightning strikes.
Then, in the absence of a more accurate modelization, one extrapolates the expression of the
voltage-time function below the s level.
24.3 Elaboration of the Spark Gap Model in EMTP

The model of spark gap which has been created is mainly comprised of a switch, driven by a
variable TACS. Knowing that current at the moment of arcing does not build-up in the same
manner in a switch and in a spark gap, we have arranged the open-closed transition so as not to
generate non-physical fast fronts.
The model consists of a main arm with switch K and an electromotive force which smoothes the
open-closed transition of the switch, and a sub-network which imposes the form of smoothing.
The moment of closing is determined through direct summing:
DETECT =
TENSI1 M t
TENSI1 > UO
TENSI1 is the voltage at the spark gap terminals
obtained by means of an "accumulator", module TACS No. 65. K closes when DETECT becomes
higher than D. The value of DETECT is brought back to zero at each time that the absolute value of
the spark gap voltage falls below U0.
As long as DETECT<D in the sub-network, TENSI1 charges C via K1 which is closed, and U=0. U
is the electromotive force which is injected permanently into the main network via R and the two
current sources. At the moment when K closes, K1 opens and K2 closes, which causes C to
discharge into R1 and appearance, in the main circuit, of a voltage which decreases exponentially
down to zero and which opposes sharp build-up of current. This decrease takes place with a time
constant R1C=10 ns, so that the current source becomes ineffective after 30 ns.
24 - 4
Air Gap Arc Model
TENSI1
K
GAUCHE
MILIE1
MILIE2
MILIE3
DROIT
1m
1m
%FERME
I=U/R
NOT(%FERME)
TENSI1
%FERME
1m
AAAAAA
K1
BBBBBB
%%%%E0
TENSI1
K2
R1
OUTPUT
COMPTEUR
MODULE TACS 65
DETECT
(ABS(TENSIO1)**N)DELTAT INPUT
RAZ
HOLD
-1
ABS(TENSIO1)<U0
Figure 24.1:
Spark gap model makeup diagram
24 - 5
%FERME=
(DETECTD)
Air Gap Arc Model

The spark gap model has been designed as a sub-routine. It is called-up at the start of programs
through prefix $INCLUDE. Five parameters are entered in the order given: name of the nodes
which connect the spark gap to the network, D, n, U0.
The EMTP file of the module is given in Section 24.5.
24.4 Application
Our purpose is to verify operation of the model against a simple example, that of a large-length,
long-span, 90 kV single-phase line, suspended from a pylon via a the insulator string gap and a 710
mm spark gap. A 20 kA lightning strike hits the conductor at a distance of 1 km from the pylon.
Rt
Figure 24.2:
Diagram of application to a 90 kV single-phase line of infinite length,

supported at one point by a pylon fitted with a spark gap. The active
wire is hit by a 20 kA lightning strike in the vicinity of the pylon.
24 - 6
Air Gap Arc Model

The plot below shows the voltage at the spark gap terminals:
Figure 24.3:
Voltage at the spark gap terminals
Arcing takes place over 0.5 s, at a voltage of 2.12 MV. The voltage-time curve (form 1) applied
to this case indicates an arcing level of 2.08 MV, thus verifying proper behavior of the model.
This curve also verifies the progressive decrease of voltage at the spark gap terminals, associated
with progressive build-up of the arc current after dielectric bursting.
24.5 EMTP Routine of Air Gap Model

BEGIN NEW DATA CASE
C
C
C
C
MODULE
ECLA.INC
ARG GAUCHE,DROITE,%DDDE%,%KKKK%,%UUU0%
NUM %DDDE%,%KKKK%,%UUU0%
DUM TENSIO,TENSI1,%DEBUT,SURFAC,XXXRAZ,XXHOLD,%%INCR
DUM DETECT,%FERME,%mar1%,MFERME,
DUM OUVRIR,
DUM AAAAAA,BBBBBB,%%%%E0,
DUM MILIE1,MILIE2,MILIE3,MILIE4
C
24 - 7
Air Gap Arc Model

C
C
C
C
/BRANCH
C <-B1-><-B2-><- B3><-B4 ><R- ><-L -><-C- ><long>LPP
S
GAUCHEMILIE1
1.E-3
AAAAAABBBBBB
1.E-3
MILIE2MILIE3GAUCHEMILIE1
MILIE1MILIE2
1.
2
C
/SWITCH
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <-B1-><-B2-><--VVVV--><---III-->< TDion >
<FERM>
<grid><open>
s
12MILIE3DROITE
%FERME
C
C
/SOURCE
C <-B1->VI<--AMPL--><-- f --><--phase-><---A1---><---T1---><--TDEB--><--TFIN-->
60MILIE1-1
60MILIE2-1
C
/TACS DATA
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <-OUT> <----A---><----B---><----C--->
<-T-START><-T-STOP->
90GAUCHE
0.
1.
90DROITE
0.
1.
C
C <-SOR> =<------------------EXPRESSION FORTRAN----------------------------99TENSIO =
ABS(GAUCHE-DROITE)
99%DEBUT =
(TENSIO.GT.%UUU0%)
99SURFAC =
(ABS(TENSIO-%UUU0%)**%KKKK%)
C <nom-> S< IN > S< IN > S< IN > S< IN > S< IN > <-A- ><-B--><-C--><-D--><-E-->
MFERME +%FERME
1.
C <OUT >CD <IN1 > <IN2 > <IN3 > <IN4 > <IN5 > < A >< B >< C >< D >< E >
99DETECT65+%%INCR
1.
0.
0.XXXRAZXXHOLD
99%%INCR = SURFAC*DELTAT
99XXHOLD = -1
99XXXRAZ = NOT(%DEBUT)
C
99%FERME = (DETECT.GT.%DDDE%) .OR. MFERME
C
99OUVRIR = NOT(%FERME)
C
C
C
C
Amortissement du transitoire d'amorcage
C
Sous reseau R-C pour generer un FEM decroissante
C
C
/BRANCH
C <-B1-><-B2-><- B3><-B4 ><R- ><-L -><-C- ><long>LPP
S
BBBBBB
1.E-3
%%%%E0
10.
/SWITCH
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C <-B1-><-B2-><--VVVV--><---III-->< TDion >
<FERM>
<grid><open>
s
12TENSI1AAAAAA
OUVRIR
12BBBBBB%%%%E0
%FERME
/SOURCE
24 - 8
Air Gap Arc Model

C <-B1->VI<--AMPL--><-- f --><--phase-><---A1---><---T1---><--TDEB--><--TFIN-->
60TENSI1
/TACS DATA
C
90BBBBBB
99TENSI1 = GAUCHE-DROITE
C
99MILIE1 = BBBBBB
99MILIE2 = -MILIE1
C
C
33DETECTBBBBBBTENSIO
C 345678911234567892123456789312345678941234567895123456789612345678971234567898
C <nom-> < valeur>
77%DEBUT
0
77MDEBUT
0
END OF DATA
BLANK
BEGIN NEW DATA CASE
BLANK
24 - 9
Section 25
1.
P.M. Anderson and A.A. Fouad, Power System Control and Stability. Ames, Iowa
(USA): The Iowa State University Press, 1977.
2.
L.O. Barthold and G.K. Carter, "Digital Travelling-Wave Solutions", AIEE Trans., Part
III, Vol. 80, p. 812, 1961.
3.
S. Carneiro and J. Marti, "Evaluation of Corona and Line Models in Electromagnetic

Transients Simulations". IEEE Trans. on Power Delivery, Vol. 6, No.1, January 1991
4.
CIGRE Working Group 13.1, "Practical Applications of Arc Physics in Circuit Breakers:
Survey of Calculation Methods and Application Guide", Electra 1988.
5.
P. Darveniza, F. Popolansky and E.R. Whitehead, "Lightning Protection of UHV

Transmission Lines", Electra no. 41, July 1975.
6.
Differential Relays Type D200, CH-ES 61-65.03 E, Edition November 1980, Brown
Boveri Corporation.
7.
H.W. Dommel, "Digital Computer Solution of Electromagnetic Transients in Single and

Multiphase Networks", IEEE Trans., vol.PAS-88, pp. 388-399, April 1969.
8.
H.W. Dommel, Editor, "Electromagnetic Transients Program Reference Manual (EMTP

Theory Book)", Bonneville Press Administration, August 1986.
9.
H. W. Dommel. The EMTP Theory Book. Second Edition. The University of British
Columbia, Vancouver, BC, 1994.
10.
Ekstrm, G Liss "A Refined HVDC Control System", IEEE Transaction on PAS, Vol 89,
May/June 1970.
11.
C. Gary, G. Dragan and D. Critescu, " Attenuation of Travelling Waves Caused By

Corona". CIGRE Report 33-13, 1978
12.
C. Gary et M. Moreau, "Leffet Couronne en Tension Alternative. Pertes et

Perturbations Radiolectriques Engendres par les Lignes de Transport dnergie
lectrique". Collection de la Direction des Etudes et Recherches dElectricit de France.
EYROLLES 1976.
25 - 1

13.
C. Gary, A. Timotin and D. Critescu, "Prediction of Surge Propagation Influenced By

Corona and Skin Effect". Proc. IEE, 130-A, pp. 264-272, July 1983
14.
H.K. Lauw, W.S. Meyer, "Universal Machine Modeling for the Representation of
Rotating Electric Machinery in an Electromagnetic Transients Program", IEEE Trans.,
Vol. PAS-101, pp. 1342-1351, June 1982.
15.
Jiming Lin and J.R. Marti, "Implementation of the CDA procedure in the EMTP", IEEE
Trans. Power Systems, Vol. 5, pp.394-402, May 1990.
16.
J. Mahseredjian, Support Routine for the EMTP Dynamic Corona Model. Rapport (IREQ4534, dcembre 1989
17.
J.R. Mart, L. Mart, and H.W. Dommel, "Transmission Line Models for Steady-State
and Transients Analysis", Invited paper. Proceedings IEEE Joint International Power
Conference, Athens Power Tech APT'93, Athens, Greece, Sept. 5-8, 1993, pp. 744-750.
18.
J.R. Marti and Jiming Lin, "Suppression of Numerical Oscillations in the EMTP", IEEE
Trans. Power Systems, Vol. 4, pp. 739-747, May 1989.
19.
P. S. Maruvada, H. Menemenlis and R. Malewski, "Corona Characteristics of Conductor

Bundles Under Impulse Voltages". IEEE Trans. on Power Apparatus and Systems,
Vol. 96, No. 1, January 1977, pp. 102-115
20.
R.W. Menzies and G.B. Mazur, "Advances in the Determination of Control Parameters
for Static Compensators", IEEE Transactions on Power Delivery, Vol. 4, No. 4, October
1989.
21.
W.S. Meyer, "Transients Program Memoranda." Approximately 150 pages per volume,
this concerned the EMTP development memoranda which were regularly issued by
Bonneville Power Administration (BPA). Individual memos were bound when there has
been sufficient accumulation to form a volume. As of February 1984, there were 14
volumes. As of April 1984, Memoranda writing has been suspended indefinitely with
the 74-page contribution dated 19 January 1984 (finished on 25 March 1984) being the
final one.
22.
Mho Distance Relay, Type CEYSIA, ~ GEK-1275F, Power Systems Management

Department, General Electric, Philadelphia, PA.
23.
A.S. Morched and V. Brandwajn, "Transmission Network Equivalents for

Electromagnetic Transients Studies", IEEE Trans., vol. PAS-102, pp. 2984-2994, Sept.
1983.
24.
A.S. Morched, L. Mart and J.H. Ottevangers, "A High Frequency Transformer Model
for the EMTP", Paper 92 SM 359-0 PWRD, Presented at the PES-IEEE 1992 Summer
Meeting, Seattle, May 1992.
25 - 2

25.
J. Mahseredjian, M. Landry and B. Khodabakhchian, "The New EMTP Breaker Arc

Model", IPST97-International Conference on Power System Transients, Seattle, June
22-26, 1997, pp. 245-249
26.
M. Rioual, Short and Long Air Gaps (Insulator Strings and Spark Gaps) Modelling for
Lighting Studies with EMTP Program (EPRI-DCG version 2.0), Research project - Final
Report, March 88.
27.
E. O. Schweitzer, and A. J. Flechsig, An Efficient Directional Distance Algorithm For

Digital Computer Relaying, IEEE PES Summer Meeting, Mexico City, Mexico, July 1722, 1977.
28.
A. Semlyen and H. Wei-Gang, "Corona Modeling For Calculation Of Transients On

Transmission Lines", IEEE Trans. on Power Delivery, Vol. 1, No. 3, July 1986, pp.
228-239
29.
Static Phase Distance Relays, Type SLY, GEK-26487B, Power Systems Management
Department, General Electric, Philadelphia, PA.
30.
Static Phase Distance Relays, Type SLY12C, GEK-27948, Insert Booklet-GEK26487, Power Systems Management Department, General Electric, Philadelphia, PA.
31.
G. St-Jean and R. F. Wang, "EQuivalence Between Direct And Synthetic Interruption

Tests On High Voltage Circuit Breakers". IEEE Trans. PAS, July/August 1983, pp. 22162223.
32.
M.M. Suliciu and I. Suliciu, "A Rate Type Constitutive Equation for the Description of
the Corona Effect", IEEE Trans., Vol. PAS-100, No. 8, pp. 3681-3685, August 1981.
33.
Andrew Sweetana, Transient Response Characteristics of Capacitive Potential

Devices, IEEE Trans. Power Apparatus and Systems, vol. PAS-89, pp. 1989-1997,
Nov./Dec. 1970.
34.
E.J. Tarasiewicz, A.S. Morched, A. Narang, E.P. Dick, "Frequency Dependent Eddy
Current Models for Nonlinear Iron Cores", IEEE Transactions on Power Systems, Vol. 8,
No. 2, pp. 588-597, May 1993.
35.
H.B. Thoren and K.L. Carlsson, "A Digital Computer Program for the Calculation of
Switching and Lightning Surges on Power Systems" IEEE Trans., Vol. PAS-89, pp. 212218, February 1970.
36.
E. Thuries, P. Van Doan, J. Dayet and B. joyeux-Bouillon, "Synthetic Testing Method

For Generator Circuit Breakers". IEEE Trans. on Power Delivery, Jan. 1986, pp.179184.
25 - 3

37.
W.F. Tinney and J.W. Walker, "Direct Solutions of Sparse Network Equations by
Optimally Ordered Triangular Factorization", Proc. IEEE, vol. 55, pp. 1801-1809,
November 1967. Also available in 1967 IEEE PES PICA Conference Record.
38.
Transformer Differential Relay with Percentage and Harmonic Restraint, GEH-2057

E, Power Systems Management Department, General Electric, Philadelphia, PA.
39.
D. Vidal, S. Sadovic, R. Soler, "Dossier de spcification d'une maquette de logiciel pour

des tudes statistiques de foudre dans les postes (INCA)", HM-77/94/080.
40.
White and Woodson, Electromechanical Energy Conversion, John Wiley.
41.
P. Lacasse and J. Mahseredjian, "Implementation of a Multiphase Corona Line Model in

the EMTP Users Manual". IREQ Report 94-326, October 94
25 - 4
Bibliography
I
A. Ametani, "A Highly Efficient Method for Calculating Transmission Line Transients",
IEEE Trans., vol. PAS-95, pp. 1545-1551, Sept/Oct 1976.
II
V. Brandwajn, "Synchronous Generator Models for the Simulation of Electromagnetic

Transients", Ph.D. thesis written at the University of British Columbia (Vancouver,
B.C., Canada), April 1977.
III
V. Brandwajn and H.W. Dommel, "A New Method for Interfacing Generator Models
with an Electromagnetic Transients Program", IEEE PES PICA Conference Record, Vol.
10, pp. 260-265, 1977.
IV
Determination of Torsional Fatigue Life of Large Turbine Generator Shafts, Final

Report on Project 1531-1, EL-3083, April 1984.
Direct Current Transmission - Volume I, Kimbark
VI
H.W. Dommel, Editor, "EMTP Newsletter". Published approximately quarterly in

Vancouver, issue number one appeared during July of 1979.
VII
H.W. Dommel, "Nonlinear and Time Varying Elements in Digital Simulation of

Electromagnetic Transients", IEEE Trans., vol. PAS-90, pp. 2561-2657, Nov/Dec 1971.
VIII
L. Dube and H.W. Dommel, "Simulation of Control Systems in an Electromagnetic

Transients Program with TACS," IEEE PES PICA Conference Record, vol. 10, pp. 266271, 1977.
IX
J.G. Frame, N. Mohan, T.H. Liu, "Hysteresis Modeling in an Electromagnetic Transients

Program", IEEE Trans., vol.PAS-101, pp. 3403-3412, September 1982.
IEEE Committee, "Conventions for Block Diagram Representation", IEEE Tran., Vol.
PWRS-1, No. 3, pp. 95-100, August 1986.
XI
K.C. Lee and H.W. Dommel, "Addition of Modal Analysis to the U.B.C. Line Constants
Program", research report to B.C. Hydro and Power Authority, Vancouver, Canada,
January 1980, published by the Electrical Engineering Department of the University of
British Columbia.
XII
J.R. Marti, "Accurate Modelling of Frequency-Dependent Transmission Lines in

Electromagnetic Transients Simulations", Proc. IEEE Power Industry Computer
Applications (PICA) Conference, Philadelphia, PA, 9 pages, May 1981.
XIII
L, Marti, "Simulation of Transients in Underground Cables with Frequency-Dependent

Modal Transformation Matrices", presented at the IEEE/PES Winter Power Meeting,
New Orleans, LA, February 1-5, 1987.
25 - 5

XIV
W.S. Meyer and H.W. Dommel, "Numerical Modelling of Frequency Dependent

Transmission Lines with Ground Return Using Recursive Convolutions", IEEE Trans.,
vol. PAS-93, pp. 1401-1409, Sep/Oct 1974.
XV
A. Narang, R.H. Brierley, "Topology-Based Magnetic Model for Steady-State and and
Transient Studies for Three-Phase Core-Type Transformers". T-PWRS, pp. 1337-1349,
August 1994.
XVI
A.G. Phadke, Course Organizer, "Digital Simulation of Electrical Transient

Phenomena", IEEE Tutorial Course No. 81 EH0173-5-PWR, last given at the 1982 IEEE
PES Winter Meeting in New York City.
XVII
A. Semlyen, "Contributions to the Theory of Calculation of Electromagnetic Transients

on Transmission Lines with Frquency Dependent Parameters", IEEE Trans., vol. PAS100, pp. 848-856, Feb. 1981.
XVIII
A. Semlyen, A. Dabuleanu, "Fast and Accurate Switching Transient Calculations on

Transmission Lines with Ground Return Using Recursive Convolutions", IEEE Trans.,
vol. PAS-94, pp. 561-571, 1975.
25 - 6
EMTP Rulebook 2
Table of Contents
Section 1
Introduction
Section 2
Computer Aspects of the EMTP Support Progrms
Section 3
Structure of AUX Input Data
Section 4
Section 5
Transformer Parameter Calculation
Section 6
FDBFIT: High Frequency Transformer Fitter
Section 7
Saturation, Hysteresis and ZnO Routines
Section 8
Eddy Currents Calculations
Section 9
Line Constants
Section 10
Cable Constants
Section 11
FDNE
Section 12
References
Section 1
Introduction
1.1
A Brief Overview .................................................................................... 1-1
1.2
Applications of the AUX ........................................................................ 1-1
1.3
Applications of the FDNE ...................................................................... 1-2
1.4
Applications of the FDBFIT ................................................................... 1-2
1.5
Input and Output ..................................................................................... 1-2
1.6
About the Rulebook 2 ............................................................................. 1-3
Section 1
Introduction to the EMTP Support Routines
1.1 A Brief Overview
Most of the Electromagnetic Transients Program's (EMTP) input data requirements are different
from, and more extensive than those of other programs such as load flow, short-circuit, and
stability. This is because the EMTP is multiphase, can simulate nonlinear elements, and, in
general, uses more detailed models than the other programs. These features are needed to
accurately simulate high frequency transients which occur during short time periods.
models such as frequency dependent transmission lines, require rather complicated input
parameters generally derived from physical data such as tower and conductor characteristics.
Whenever a model requires an amount of pre-processing that does not have to be done every time
the EMTP is run, such pre-processing is done in one of the auxilliary programs such as AUX, FDNE,
or FDBFIT.
EMTP
For example, in the case of frequency dependent line models, EMTP model input is the partial
fraction expansion of the approximation of the modal characteristic impedance and propagation
functions as a function of frequency (lots of numbers with little obvious relationship to a physical
line). In AUX, the line configuration and tower configuration is used as input data, and the EMTP
model input data is generated (generally in the form of a "punch" file.
1.2 Applications of the AUX

The AUX covers a wide range of support routines:
(A)
- calculation of transmission line parameters and EMTP line

models from their geometry. It is separated into two modules: "lineparameters" for the calculation of line impedance and admittance matrices,
and "line-model" for the creation of distributed parameter models.
(B)
XFORMER - calculation of transformer models from their ratings and shortcircuit test data. Three major types of models are available: BCTRAN, TRELEG
and TOPMAG.
(C)
SATURATION - derivation of magnetic saturation data = f(i) required by

EMTP nonlinear reactance models. Also included is the derivation of an
LINE CONSTANTS
1-1
the

exponential model of the nonlinear R = f(i) data for the representation of the
protective value characteristic of ZnO and SiC arresters.
(D)
CABLE CONSTANTS - calculation of cable parameters and EMTP cable models

from their geometry. It is separated into two modules: "cable-parameters" for
the calculation of cable impedance and admittance matrices, and "cable-model"
for the creation of distributed parameter models
1.3 Applications of the FDNE

The FDNE (Frequency Dependent Network Equivalent) is a stand-alone support program designed
to derive frequency-dependent network equivalents for a reduced representation of large networks.
The derivation of the equivalents is based upon the topology of the network, transmission line
geometry and other system data.
1.4 Applications of the FDBFIT

FDBFIT is a stand-alone support program for the generation of the HFT (High Frequency
Transformer) model. It takes as input the frequency response at the terminals of a transformer, and
it produces a multi-phase FDB (Frequency Dependent Branch) model that reproduces the frequency
characteristics of the measured response.
1.5 Input and Output

The input for the EMTP support routines consists of data often available from manufacturers test
data. AUX, FDNE and FDBFIT normally create an output file (*.out) which provides additional
information about the model being processed or data being requested. When a model is generated,
as in the case of a transmission line model, a "punch" file suitable for inclusion into the EMTP data
deck is generated.
In some cases, the "punch" can be merged into the EMTP data case "as is" using $INCLUDE filename
card, but sometimes the contents of the "punch" may have to be edited before merging into an
EMTP data case, to include appropriate node names, etc.
1-2
1.6 About the Rulebook 2

The Rule Book 2 is organized in 10 sections. First are the introductory section and this section.
Sections 3 through 10 are devoted to various aspects of the EMTP support modules:
Section 2:
Computer aspects of the AUX.
Section 3:
Structure of AUX input data.
Section 4:
Cards to begin a new data case.
Section 5:
Transformer parameter calculation.
Section 6:
FDBFIT
Section 7:
Magnetic saturation routines.
Section 8:
EDDYC
Section 9:
Line constants routines.
Section 10:
Cable constants supporting routines.
Section 11:
Frequency dependent network equivalent (FDNE).
Section 12:
References
- High Frequency Transformer model generator
- Eddy Currents support module
RuleBooks 1 and 2 have been designed to be viewed on-line with FrameViewer or FrameReader,
and are supported on Windows as well as some UNIX systems such as Sun/Solaris.
At the top of each section there are navigation buttons (in addition to those provided by the
viewing programs) to provide hypertext links to various parts of the manual. Also at the top of
each section, there is a "printer" icon. Clicking on this icon links to the corresponding "print" file.
These "print" files are re-formatted versions of the screen files suitable for printing from within
FrameReader or FrameViewer. Trying to print directly from the screen files generally produces
poor results as the contents are truncated to fit a 8.5" x 11" page.
Credit for the design, porting, formatting and artwork in RuleBooks 1 and 2 belongs to Chris
Marti, under contract from the IREQ (Canada), on behalf of the EMTP Development Coordination
Group (DCG). Technical contents of the RuleBooks has been provided by the various model
developers over a period of many years and several generations of EMTP. Technical editing of this
version of the RuleBooks has been provided by Luis Marti.
1-3
Section 2

Support Programs
2.1
Availability on Different Computer Systems ...................................... 2-1
2.2
Installation-Dependent Considerations ............................................... 2-1
2.3
2.2.1
Calendar Date Format .................................................... 2-2
2.2.2
Input/Output Unit Numbers ........................................... 2-2
User-Controlled Variable Dimensioning of the AUX ........................ 2-3
Section 2
Computer Aspects of the EMTP Support Programs
2.1 Availability on Different Computer Systems
The computer industry is continuously changing. Operating systems and hardware platforms do
not last very long in the market. EMTP96 is supported on all the platforms used by the members of
DCG, which probably reflects a substantial cross-section of the hardware/software market. The
platforms supported by DCG in 1996 are:
IBM RS6000/AIX
HP9000/UX
Sun/Solaris
This section explains several common computer-dependent aspects of AUX, the main EMTP
Support Program.
2.2 Installation-Dependent Considerations

has put considerable effort into providing a uniform interface between the AUX and the user
on all supported platforms. However, a number of aspects are inevitably dependent on the
hardware and operating system. For example, the installation procedure, how to execute a
program, assigning input/output files, plotting, etc. All these system-dependent aspects are
covered in the installation guide that accompanies the package.
DCG
2-1
2.2.1
Calendar Date Format
The date format convention is the one used most often in the United States; that is, "MM/DD/YY",
where:
"MM"
"DD"
"YY"
two decimal digits for the month

(e.g., "03" for March);
two decimal digits for the day within
the month;
two decimal digits for the year
(e.g., "95" for 1995).
At present, the user has no control over date format.
2.2.2
Input/Output Unit Numbers
This information is only relevant in operating systems such as VAX/VMS, where the assignment of
input, output, and plot files is done through assignments to specific logical units. Note that in order
to present a common interface between all platforms, the logical unit assignment in the distribution
version of VAX/VMS is done through a command file.
Logical unit 5
Input data file for EMTP input data cards
Logical unit 6
EMTP
printed output
Logical unit 7
EMTP
punched-card output
The following logical unit assignments are internal to the program. If these are explicitly assigned
by the user in a VMS environment, data loss may occur. In other operating systems, logical unit
assignments should only be apparent in the case of a system crash or some other unusual form of
execution termination.
Logical unit 1
scratch file
Logical unit 2
scratch file to store memory contents for each new energization

of a "STATISTICS" or "SYSTEMATIC" data case.
Logical unit 3
scratch file
Logical units 9-15
scratch files
2-2
2.3 User-Controlled Variable Dimensioning of the AUX

stores most data in tables or arrays, which can be sized by the user by recompiling module
"newmods" and then re-linking it to the rest of the EMTP libraries. This means that only users with
the appropriate compiler and access to the appropriate object libraries can re-dimension AUX. Redimensioning was available to users in version 3.0. Starting with version 3.1 re-dimensioning of
AUX was discontinued (it appears that nobody had the need to do so in fifteen or so years).
AUX
2-3
Section 3
3.1
Structure of a Data Deck .....................................................................3-1
3.2
Cards to Begin a New Data Case ........................................................3-2
3.3
$-Cards For Special File Operations ...................................................3-3
3.4
FORTRAN
Data Format Specifications .................................................3-3
Section 3
Any number of cases can be run with one data file. Simply stack the data decks for the different
cases as shown below:
special end-of-job termination card,

if required by users system
blank card for termination of cases
data deck for the nth case
data deck for the 1st case
cards to begin a new data case and extensions
3.1 Structure of a Data Deck

Each AUX problem must be fully described in a data deck which consists of the following parts (in
this order):
1.
Cards to begin a new data case. Included are a "BEGIN NEW DATA CASE" card
and special request cards of various types (if any).
2.
Specially-requested extensions to the miscellaneous data cards.
3.
Specific data case card.
blank card
Data deck for first case
$ cards
Cards to begin a new data case
begin a new data case card
3-1

The blank card is extensively used as a terminator card for various classes of data. Any blank card
can be replaced by a card with the characters "BLANK " in the first 6 columns and any comments in
the remaining columns. For example, "BLANK CARD ENDING TRELEG CASES". The AUX
recognizes all such cards and blank them out internally before processing them (but after listing on
the printer; they appear as entered by the user).
Comment lines may be added to the data deck at any point. Comments must have "C " in columns
1-2 (Upper case "c" and a blank character). Such comment cards will appear as part of the input
data listing, but will be otherwise disregarded by the AUX.
Also, whole blocks of data can be so ignored by the AUX if recourse is made to $DISABLE and
$ENABLE cards.
3.2 Cards to Begin a New Data Case

Each new case begins with a number (possibly zero) of special-request and miscellaneous cards.
These cards are described briefly in the following. More information can be found in Section 4.
BEGIN NEW DATA CASE
REDEFINE TOL. EPSILN
redefine EPSILN
KILL CODES
XFORMER
derive impedance matrix for transformers
SATURATION
access B-H curves, hysteresis, ZnO data
EDDY CURRENTS
calculate eddy currents model in iron cores
LINE CONSTANTS
calculate overhead line constants
CABLE CONSTANTS
calculate cable system constants
DIAGNOSTIC
OUTPUT WIDTH 80
OUTPUT WIDTH 132
CONVERT ZNO
convert old data of ZnO to new formats
END LAST DATA CASE
optional card to end the AUX calculations
3-2
3.3 $-Cards For Special File Operations

A number of $-cards similar to the ones implemented in the EMTP (see Section 3.6, RuleBook 1)
have also been implemented in AUX. These are:
1.
$DISABLE/$ENABLE. A $DISABLE card tells the AUX to treat all following data
cards as if they were comment cards, until the subsequent appearance of a
$ENABLE card. This feature can be used for big blocks of comments (without
having to add "C " on columns 1-2
2.
$LISTOFF/$LISTON. A $LISTOFF card tells the AUX not to print and interpret
data cards which follow. This continues until a $LISTON card is encountered.
The object is to minimize the size of the output file by omitting the listing and
interpretation of blocks of data which have been tested and used before.
3.
$INCLUDE, [filename]. When $INCLUDE is encountered, the contents of

[filename] are included into the data file.
3.4 FORTRAN Data Format Specifications

Throughout this manual, data formats are indicated in fixed-format FORTRAN notation, with
symbols like F10.3, I8, A6, or E10.3. Note that FORTRAN syntax and rules are implicitly
followed in all data card descriptions. The following examples should illustrate most instances of
FORTRAN fixed-format usage:
F10.3:
A decimal number is to be inserted in this space (field) of width 10 columns. In

general, a decimal point should be entered. Otherwise, it is assumed that digits in the
last 3 of the 10 positions are to the right of the decimal point. As long as the decimal
point is entered, the ".3" is overridden, and positioning within the field is immaterial.
Examples follow:
1 2 3 4 5 6 7 8 9
gives 39.5
gives 132.745
gives 3.14159
3-3
I8:
Integer data (no decimal point allowed) is to be inserted in the space (field) of width 8
columns. Integers should be "right-adjusted," as far to the right in the field as possible,
since blanks are interpreted as zeros. Examples follow:
1 2 3 4 5 6 7 8 9
A6:
gives 39
gives 1024
gives 99000
An arbitrary collection of characters (letters; numbers; punctuation) is to be inserted in

this field of width 6. Blanks are a special unique character, remember, so "BUS A" and
"BUS A" are not the same. Examples follow:
BUS-1A
1.3+A5
GROUND
E10.3:
A decimal number is to be inserted in the field of width 10 columns. Use of scientific

notation, with a power of 10 following the symbol "E", is added onto the previous rules
for F-formats. Omitting the E and its following exponent is interpreted as a unity
multiplier, so the E10.3 format then behaves just like F10.3. Right adjust the number
if an E is entered, since just as with the I-format, any trailing blanks are interpreted as
zeroes. Examples follow:
1 2 3 4 5 6 7 8 9
1
E 3
E 3
E + 4
gives 1609
gives 1.609 10300
gives 3.1415926
gives 5.1
3-4
gives 0.00052
gives 52000
Section 4
4.1

Special Request Cards .........................................................................4-1
4.1.1
Optional Case-Separation Cards ...................................... 4-2
4.1.2
Blank Card for the Termination of AUX Execution ...... 4-3
4.1.3
Special Request Cards to Redefine Built-In

Program Parameters ......................................................... 4-3
4.1.4
"ABORT": Card to Disable an

Entire AUX Data Case ..................................................... 4-4
4.1.5
"KILL CODES": Request for the

Printing of AUX Kill Codes (Error Messages) .............. 4-4
4.1.6
"XFORMER": Request for the

Transformer Routine ........................................................ 4-4
4.1.7
"SATURATION": Request for Magnetic

Saturation and ZnO Routines ......................................... 4-5
4.1.8
"EDDY CURRENTS": Request for

Eddy Currents Model ....................................................... 4-5
4.1.9
"LINE CONSTANTS": Request for the

Line Constants Routine ................................................... 4-6
4.1.10
"CABLE CONSTANTS": Request for the

Cable Constants Routines ................................................ 4-6
4.1.11
"DIAGNOSTIC": Request for Selective

Diagnostic Printout ........................................................ 4-6
4.1.12
"OUTPUT WIDTH 80": Request for

80-Column Output ......................................................... 4-7
4.1.13
"OUTPUT WIDTH 132": Request for

132-Column Output ....................................................... 4-8
Section 4
Each new data case to be solved begins with one or more of the data types described in this
section. There can be any number of special-request input cards described in Section 4.1. The
special request cards with the exception of "BEGIN NEW DATA CASE", which must come first
can be in any order. The $-cards of Sections 3.3 come after the special request cards.
4.1 Special Request Cards

In the order that they are explained in subsequent sections, the following key words are
recognized at this point of AUX data input:
BEGIN NEW DATA CASE
redefine EPSILN
ABORT DATA CASE
abort current data case
KILL CODES
XFORMER
derive impedance matrix for transformers
SATURATION
access B-H curves, hysteresis, ZnO data
EDDY CURRENTS
eddy currents in iron cores
LINE CONSTANTS
calculate overhead line constants
CABLE CONSTANTS
calculate cable system constants
DIAGNOSTIC
request for diagnostics printout
OUTPUT WIDTH 80
OUTPUT WIDTH 132
4-1
4.1.1
Optional Case-Separation Cards
While not mandatory, it is a good practice to make the very first card of each data case a special
case-separation marker:
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
For runs which terminate normally, or single-data-case runs, this card serves no operational
purpose. On the other hand, if AUX terminates a particular data-case solution prematurely, then a
case-separation marker does serve a purpose, since the program automatically steps over
remaining input cards, discarding them one at a time until it finds a "BEGIN NEW DATA CASE"
record. At this point, the new data-case is solved. This is referred to as AUX error-recovery
capability, during multiple-data-case runs.
Another situation where a "BEGIN NEW DATA CASE" record is useful is when during solution of
the final data case of the run that a fatal AUX error termination occurs. In this case there would be
no following data case, so further solutions would not be possible. To distinguish this situation,
the final data case should always be followed by a "BEGIN NEW DATA CASE" card and a blank card
which will then end execution. The blank card is, in this situation, read as the start of a new
data case, and lack of data in this data case is recognized as the end of all data cases.
Note that "BEGIN NEW DATA CASE" cards must be properly positioned in the data deck; they do
not change the need for blank termination cards. The identity and function of these special cards is
recognized by AUX in only two places:
1.
Following an AUX fatal error, such cards are watched out for, as other records of
the input data are discarded.
2.
Following normal completion of the solution for any particular data case, the
very first non-comment record of the following data case will be checked for
"BEGIN NEW DATA CASE".
Any other placement of the "BEGIN NEW DATA CASE" key cards will generally cause an error.
4-2
4.1.2
Blank Card for the Termination of AUX Execution
To indicate the termination of all data cases add a "BEGIN NEW DATA CASE" record followed by a
blank.
4.1.3
Special Request Cards to Redefine Built-In Program Parameters
There are two parameters which can be redefined by the user, if need be, as follows:
Variable "EPSILN": Small Floating Point Number

In AUX, a very small number is defined as 10-8. This default value can be over-ridden using this
special-request card.
1
2
3
4
12345678901234567890123456789012 34567890
A32
5
6
7
8
1234567890123456789012345678901234567890
EPSILN
E8.0
Variable "STATFR": Synchronous Frequency

Variable "STATFR" is the default synchronous power system frequency, in Hertz. The default is
set to 60Hz.
1
2
3
4
5
6
7
8
123456789012345 67890123456789012 34567890 1234567890123456789012345678901234567890
POWER FREQUENCY
A15
STATFR
E8.0
4-3
4.1.4
"ABORT": Card to Disable an Entire AUX Data Case
When this card is used, the EMTP will skip input records until the next "BEGIN NEW DATA CASE",
is encountered. This card can be useful when more than one data case is stacked in the same input
file.
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
ABORT DATA CASE
4.1.5
"KILL CODES": Request for the Printing of AUX Kill Codes

(Error Messages)
This option will generate the text of any particular fatal EMTP error message (KILL CODE) without
actually having to set up a data case that will produce such an error condition.
1
2
3
4
5
6
7
8
12345678901234567890123456789012 34567890 12345678 90123456789012345678901234567890
KILL CODES
A32
KILL1
KILL2
I8
I8
The key-word "KILL CODES" is entered in column 1-10, and KILL1 and KILL2 are the beginning
and ending KILL numbers of the consecutive KILL codes which are to be printed. Note that KILL2
must not be less than KILL1.
4.1.6
"XFORMER": Request for the Transformer Routines
AUX supports three different transformer model generation routines: BCTRAN, TRELEG and
TOPMAG. These are selected by using the keyword "XFORMER" in the request card shown below
1
2
3
4
1234567 890123456789012345678901234567 890
XFORMER
FLAG
A7
F3.0
4-4
5
6
7
8
1234567890123456789012345678901234567890

where:
MODEL
4.1.7
FLAG
TRELEG
33.
BCTRAN
44.
TOPMAG
55.
"SATURATION": Request for Magnetic Saturation and ZnO

Routines
The special request keyword "SATURATION" in columns 1-10 provides access to the supporting
routines for the derivation of the EMTP saturation. Also provided is the access to the generation of
ZnO (SiC) arrester models
.
1
2
3
4
5
6
7
8
1234567890 1234567890123456789012345678901234567890123456789012345678901234567890
SATURATION
A10
The details of the available models and their data requirements are described in Section 7.
4.1.8
"EDDY CURRENTS": Request for Eddy Currents Model
The special request keyword "EDDY CURRENTS" in columns 1-13 provides access to the
supporting routines for modelling of eddy current losses in laminated iron cores.
.
1
2
3
4
5
6
7
8
1234567890123 4567890123456789012345678901234567890123456789012345678901234567890
EDDY CURRENTS
A13
4-5
4.1.9
LINE CONSTANTS": Request for the Line Constants Routines
The special request word "LINE CONSTANTS" in columns 1-14, provides access to a set of modules
for the calculation of overhead transmission line parameters and EMTP line models.
1
2
3
4
5
6
7
8
12345678901234 567890123456789012345678901234567890123456789012345678901234567890
LINE CONSTANTS
A14
4.1.10
"CABLE CONSTANTS": Request for the Cable Constants Routines
The special request word CABLE CONSTANTS in columns 1-15, provides access to a set of modules
for the calculation of underground cable parameters and EMTP cable models.
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
CABLE CONSTANTS
A15
4.1.11
DIAGNOSTIC": Request for Selective Diagnostic Printout
The diagnostic of AUX can be controlled overlay by overlay. A special-request card with the key
word "DIAGNOSTIC" in columns 1-10 is used in this case. The following format is applicable for
this card which must precede the miscellaneous data cards:
A20
(IPRSOV(I) , I=1, 30)
I2 format
4-6
21
22
23
24
25
26
27
28
29
30
19
20
16
17
18
6
7
8
9
10
11
12
13
14
15
4
5
DIAGNOSTIC
1
2
3
1
2
3
4
5
6
7
8
12345678901234567890 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90 12 34 56 78 90
IPRSOV(K)
diagnostic printout control parameter as restricted to overlay number "K".

Within overlay "K", IPRSOV(K) has the same meaning as "IPRSUP" as
described in RuleBook1 Section 4.2.2.
The following overlay numbers should be used for the purpose of the
diagnostic printout.
OVERLAY NO
4.1.12
MODULE
10
MODULE
11
XFORMER
12
SATURATION
14
LINE CONSTANTS
17
CABLE CONSTANTS
OUTPUT WIDTH 80": Request for 80-column Output
As in the case of the EMTP, it is possible in AUX to switch to 80column output (rather than the
more common 132-column output) by means of an "OUTPUT WIDTH 80" request. In the absence
of such a request, AUX printout will normally be 132 columns.
Use of 80-column output has some serious limitations which should be understood. First, the
interpreted tabulation of input data cards, will be truncated at column 80, and all output to the right
of this point will simply be lost. Second, not all displays any longer convey full information. If in
doubt, try the feature on a small problem, and see if the output is satisfactory. Note that the request
"OUTPUT WIDTH 80" is a single-time request which then would normally remain in effect for the
entire AUX solution.
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
OUTPUT WIDTH 80
4-7
4.1.13
"OUTPUT WIDTH 132": Request for 132-column Output
This request card forces output width output to 132 columns in those systems where 80-column
output is used as default.).
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
OUTPUT WIDTH 132
4-8
Section 5
Transformer Parameter Calculations
5.1
Introduction ......................................................................................... 5-1
5.2
Module "TRELEG" ............................................................................. 5-2
5.3
5.2.1
Data-Deck Structure for "TRELEG" ............................. 5-2
5.2.2
Sample "TRELEG" Data Case ....................................... 5-7
5.2.3
Interpretation of TRELEG's Output ............................... 5-8
Module "BCTRAN" ............................................................................ 5-9

5.3.1
Linear Magnetizing Impedance and

Excitation Test Data ....................................................... 5-9
5.4
5.3.2
Description of the Data Deck ......................................... 5-11
5.3.3
Exciting Current in Zero Sequence Excitation Test ....... 5-20
5.3.4
Error Messages ............................................................... 5-22
5.3.5
Description of the Output ............................................... 5-23
5.3.6
Test Example .................................................................. 5-24
Subroutine "TOPMAG" ...................................................................... 5-25

5.4.1
Introduction .................................................................... 5-25
5.4.2
Description of the Data Deck ......................................... 5-27
5.4.3
Class 1: Excitation Data ................................................ 5-29
5.4.4
Class 2: Winding Data .................................................. 5-31
5.4.5
Class 3: Data for Duality Winding (and Optionally,

Limb Dimensions) .......................................................... 5-32
5.5
5.4.6
Class 4: Short-Circuit Test Data ................................... 5-33
5.4.7
Sample Data File ........................................................... 5-36
Modeling Hints for all Transformer Models ....................................... 5-37

5.5.1
Saturation Effects ........................................................... 5-37
5.5.2
Floating Delta Connection ............................................. 5-37
5.5.3
High-Frequency Effects ................................................. 5-38
5.5.4
Autotransformers ........................................................... 5-38
Section 5
5.1 Introduction
AUX supports three different modules for the calculation of transformer parameters at power
frequency:
(A)
Section 5.2: Module TRELEG for the calculation of [R], [L] or [R], [L]-1
parameters of single and three-phase transformers;
(B)
Section 5.3: Module BCTRAN for the calculation of [R], [L]-1 parameters of
single and three-phase transformers.
(C)
Section 5.4: Module TOPMAG for the calculation of [R], [L] or [R], [L]-1
parameters of three-phase core-type transformers, taking into account the
topology of the magnetic circuit formed by the core and the windings.
All these models are designed to match the standard open and short circuit tests of a transformer at
power frequency. They represent transformers as a coupled impedance matrix (represented either
as [R], [L] or as [R], [L]-1). This class of models is valid as long as the capacitances between
windings, from windings to the tank and core, and between layers or windings can be ignored.
The range of validity of these models is usually between power frequency and 6 kHz to 10 kHz,
depending on the type of transformer. Adding capacitances to the terminals of the model can
approximate the asymptotic behaviour of the frequency response of the windings at very high
frequencies, but cannot model the dynamics of the transformer in the mid-frequencies range. For
a more accurate transformer model with a wide frequency response, the HFT (High Frequency
Transformer) model should be used. The HFT model is generated using the support program
FDBFIT described in Section 6.
All three models described in this Section reproduce the behaviour of a transformer at power
frequency. BCTRAN allows an infinite magnetizing impedance, whereas TRELEG assumes that the
magnetizing impedance is finite (for large values of the magnetizing impedance TRELEG may
produce an ill-conditioned model). Also TRELEG degenerates to a coupled impedance matrix at
dc (which may or may not be stable), whereas BCTRAN becomes an uncoupled resistance matrix
at dc which represents winding resistances. In general, BCTRAN is better behaved that TRELEG.
Although conceptually different, from a numerical point of view, TOPMAG is an extension of
BCTRAN that allows the reproduction of specific zero sequence power frequency tests that
BCTRAN cannot reproduce. However, TOPMAG requires additional test data that may not be
available for standard factory data sheets.
5-1
5.2 Module "TRELEG"

The effects of a three-leg core in a transformer are apparent in the different values determined for
short-circuit impedances in positive and zero sequence. These values are used by the program to
calculate a 3N X 3N impedance matrix model, where N = number of windings on any leg (N < 5). It
is also possible to use this program to calculate an N X N matrix to represent a single-leg core, a
shell type, or a 5-leg core transformer. In the latter cases, the values of the impedances (shortcircuit and magnetizing) will be equal in both zero and positive sequence.
The program accepts data for short-circuit tests performed with up to two of the windings
connected in delta. This is the standard form in which a manufacturer will provide test data.
TRELEG requires that winding data be entered so that delta-connected windings appear last. Also
the windings are assumed to be concentrically-located on the core, and entered (into the data file)
in the order from outer to inner winding. When windings are not concentric, or when the deltaconnected windings are not innermost on the core, the program provides an option for the user to
retain delta-connected windings as the last windings, and to provide the magnetizing impedance of
each winding. In the absence of any test data, the positive sequence magnetizing impedance of
windings in per unit can be assumed to increase with increasing diameter, while the zero sequence
magnetizing impedance decreases. The variation from one winding to another will be
approximately equal to the positive sequence short-circuit reactance between them. It does not
appear that this approximation of magnetizing impedances has any significant effect on the
resulting model.
5.2.1
Data-Deck Structure for "TRELEG"
The structure of the data deck for a TRELEG transformer model is as follows:
1.
First comes a "BEGIN NEW DATA CASE" card (see Section 4.1.1).
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
2.
Next comes an "XFORMER" special-request card, with a value of "33." in

columns 3840 (see Section 4.1.6).
1
2
3
4
1234567 890123456789012345678901234567 890
XFORMER
A7
33.
F3.0
5-2
5
6
7
8
1234567890123456789012345678901234567890

3.
Next come data cards which give the electrical parameters of the transformer.
Blank card
TRELEG data cards
4.
Parameters for more than one transformer can be provided by repeating the data
of Point 3 as many times as desired. Each such grouping is a separate data case
within "TRELEG", corresponding to a different transformer. A blank card
terminates these.
5.
To indicate the end of all AUX data cases add a "BEGIN NEW DATA CASE" card
at this point, followed by a blank card, as described in Section 4.1.1.
Blank card
Begin new data case card
"TRELEG" Data cards:

Class I
Card 1
1
2
3
4
5
6
7
8
1 23 45 678901234567 890123456789 01 12345678901234567890123456789012345678901234567890
N
ND
I2
I2
E12.0
SBVA
IS
E12.0
I2
number of windings
(2-3)
ND
number of delta windings (<2)
(4-5)
F
frequency in Hz (60 in North America)
(6-17)
SBVA
base MVA (3-phase rating)
(18-29)
5-3

IS
(30-31)
flag indicating the presence of a 3-leg core transformer

= 0 3-leg core transformer
= 1 single-phase transformer
Class II: (present only if NDELTA = 2 for transformers with N=3 windings.
Card 2 & 3 (exactly 2 cards)
1
2
3
4
5
6
7
8
1 234567890123 456789012345 6789012345678901234567890123456789012345678901234567890
TPKMR
E12.0
TPKMR
(2-13)
TPKMX
(14-25)
TPKMX
E12.0
real and imaginary parts of the positive sequence test between the wye
and two delta-connected windings. This data can be left blank, if
unknown. In such case, the program will internally simulate the test and
generate the requested data.
1
2
3
4
5
6
7
8
1 234567890123 456789012345 6789012345678901234567890123456789012345678901234567890
TZKMR
E12.0
TZKMR
(2-13)
TZKMX
E12.0
real and imaginary parts of the zero sequence test between wye and two
delta windings.
TZKMX
(14-25)
Class III
Exactly (N-1) N/2 cards containing the short-circuit test data between different windings. If
NDELTA=2, the zero sequence data can be left blank since the program will generate appropriate
numbers based upon the data from cards 2 & 3.
5-4
1
2
3
4
5
6
7
8
1 23 45 678901234567 890123456789 012345678901 234567890123 456789012345678901234567890
J
TPR
TPX
TZR
TZX
I2 I2
E12.0
E12.0
E12.0
E12.0
I,
(2-3)
numbers of windings between which the test has been conducted
(4-5)
real and imaginary parts of the positive sequence test in p.u.
TPR
(6-17)
TPX
(18-29)
real and imaginary parts of the zero sequence test in p.u.
TZR
(30-41)
TZX
(42-53)
Terminate the Class III data cards with a blank card.
Class IV
One card specifying the flag (KZOUT) to determine whether the output impedance matrix is to be
in p.u., ohms, mH, 1/ohms or 1/mH.
KZOUT
1
2
3
4
5
6
7
8
1 23 45678901234567890123456789012345678901234567890123456789012345678901234567890
I2
KZOUT
(2-3)
= 0
output of [R] and [X] in p.u.
= 1
output of [R] and [X] in ohms
= 2
output of [R] in ohms and [L] in mH
= 1
output of [R] in ohms and [X]-1 in mho (S)
= 2
output of [R] in ohms and [L]-1 in 1/mH
5-5
Class V
Exactly N cards containing winding number, rated voltage of the winding, indicator of delta
windings and the node names to be used by the program in punching the branch impedance cards.
I2
I1 IND
1
2
3
4
5
6
7
8
1 23 4 5 6 789012345678 901234567890 123456 789012 345678 901234 567890 123456 78901234567890
VR
E12.0
(2-3)
R
E12.0
NA(I)
NB(I)
NA(I1)
NB(I1)
NA(I2)
NB(I2)
A6
A6
A6
A6
A6
A6
winding number (delta windings should always have the highest

numbers!)
(5)
winding connection flag

= 0 for wye-connected windings
= 1 for delta-connected windings
VR(J)
rated voltage of winding J, in kV rms
IND
(7-18)
R(J)
dc resistance of winding J, in ohms
(19-30)
NA(I),
node names for phase A of winding J
(31-36)
NB(I)
(37-42)
NA(I1),
node names for phase B of winding J
(43-48)
NB(I1)
(49-54)
NA(I2),
node names for phase C of winding J
(55-60)
NB(I2)
(61-66)
Class V data cards have to be terminated with a blank card.
Card VI
1
2
3
4
5
6
7
8
1 23 45678901234567890123456789012345678901234567890123456789012345678901234567890
NT
I2
5-6
NT
(2-3)
Flag indicating how the magnetizing impedances (XPOZ, XZERO)

are specified.
=1
magnetizing impedances are known for each winding, (N

data cards follow).
it is assumed that XPOZ and XZERO is known only for the

first winding.
Class VII
One or N cards containing values for magnetizing impedances in positive and zero sequence for
the first (when NT=1) or all the windings (NT 1).
1
2
3
4
5
6
7
8
1 234567890123 456789012345 6789012345678901234567890123456789012345678901234567890
XPOZ
E12.0
XPOZ
XZERO
E12.0
Positive and zero sequence magnetizing impedance in p.u.
(2-13),
XZERO
(14-25)
Note that when XPOZ and XZERO are known for each winding they should be entered in order
corresponding to the assigned number starting from the lowest numbered (see Class V data cards).
For single-phase transformers, set XPOZ = XZERO. Class VII data has to be terminated with a
blank card in this case. For a more complete discussion of XZERO for a three leg-core
transformers see Section 5.3.3.
5.2.2
Sample "TRELEG" Data Case
Consider a case of 3-phase, 3-leg, core-type transformer. A listing of this data file follows:
BEGIN NEW DATA CASE

C
BENCHMARK DC-36A
C
TEST OF "TRELEG" SUPPORTING PROGRAM
C
3-PHASE, THREE-LIMB CORE TRANSFORMER
C
YIELDS (9 X 9) MATRICES [R] AND [WL] AS OUTPUT.
XFORMER
33.
C 3456789012345678901234567890123456789012345678901234567890
5-7

3 1
60.
750.0
1 2 .0017
.13
.0057
.115
1 3 .0042
.35
.0096
.268
2 3 .0044
.2
.0143
.136
BLANK CARD ENDING MEASUREMENTS.
1
C 1
1 0 288.6751346 .473
HIGHA
HIGHB
HIGHC
2 0 138.5640646 .029875
LOWA
LOWB
LOWC
3 1
28.
.01128
TERTA TERTB TERTB TERTC TERTC TERTA
BLANK CARD ENDING WINDINGS.
1
100.0
1.
99.87
1.13
99.67
1.33
BLANK CARD END MAGNETIZING IMPEDANCES.
BLANK CARD ENDING "TRELEG" DATA CASES.
BEGIN NEW DATA CASE
BLANK
An excerpt from the corresponding output file follows:

**********
80-COLUMN CARD-IMAGE LISTING OF UNIT-7 PUNCHED CARDS
*********
-------------------------------------------------------------------------------1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
-------------------------------------------------------------------------------51,HIGHA ,
,,,
0.473000000000E+00, 0.223333333341E+05 ,,,,,
52,LOWA ,
,,,
-0.646793103332E-01, 0.107059927425E+05 $
0.298750000000E-01, 0.514227199959E+04 ,,,,,
53,TERTA ,TERTB ,,,
-0.362556943699E-01, 0.215874652794E+04 $
-0.357184214476E-01, 0.103714560936E+04 $
0.112800000000E-01, 0.209767040000E+03 ,,,,,
54,HIGHB ,
,,,
0.000000000000E+00, -0.110000000004E+05 $
.
.
.
59,TERTC ,TERTA ,,,
-0.494356580077E-01, -0.106342221589E+04 $
-0.325054568924E-01, -0.509650587464E+03 $
0.000000000000E+00, -0.102798080000E+03 ,,,,,
-0.494356580077E-01, -0.106342221589E+04 $
-0.325054568924E-01, -0.509650587464E+03 $
0.000000000000E+00, -0.102798080000E+03 ,,,,,
-0.362556943699E-01, 0.215874652794E+04 $
-0.357184214476E-01, 0.103714560936E+04 $
0.112800000000E-01, 0.209767040000E+03 ,,,,,
--------------------------------------------------------------------------------
5.2.3
Interpretation of TRELEG's Output
The impedance matrix produced by the program is symmetrical and is represented by its lower
triangular part only. As an example, the impedance matrix for a three-winding, three-phase
transformer would have the following general form:
5-8
Z11
Z12
Z13
Z22
Z23
Z33
M11
M12
M13
M12
M22
M23
M13
M23
M33
Z11
Z12
Z13
Z22
Z23
Z33
M11
M12
M13
M12
M22
M23
M13
M23
M33
M11
M12
M13
M12
M22
M23
M13
M23
M33
leg I
leg I
leg II
leg II
Z11
Z12
Z13
Z22
Z23
leg III
Z33
leg III
Zik = coupling between windings on one leg (including self impedance Zii).
Mik = coupling between windings on different legs.
5.3 Module "BCTRAN"

BCTRAN accurately reproduces the behaviour of a three-phase, multi-winding transformer at
power frequency. From a numerical point of view, it has a number of advantages over TRELEG.
For instance, BCTRAN allows an infinite magnetizing impedance, whereas TRELEG assumes that
the magnetizing impedance is finite (for large values of the magnetizing impedance TRELEG may
produce an ill-conditioned model). Also TRELEG degenerates to a coupled resistance matrix at dc
(which may or may not be stable), whereas BCTRAN becomes an uncoupled positive resistance
matrix at dc which represents winding resistances. In general, BCTRAN is better behaved that
TRELEG.
5.3.1
Linear Magnetizing Impedance and Excitation Test Data
The exciting current in 5-limb three-phase transformers and in single-phase transformers can often
be ignored. If the current is ignored, the [L]-1-matrix generated by BCTRAN is singular and cannot
be inverted to a [Z]-matrix. However, since the EMTP is based on a nodal admittance matrix
formulation, the fact that [Z] is not invertible is not a problem.
For three-phase transformers with three-limb core construction, the exciting current in the zero
sequence test is fairly high (e.g., 100%) and should not be ignored. The shunt admittance of the
5-9

magnetizing branch (see reference [2]), which is added to the [R], [L]-1-model to represent the
exciting current and excitation losses, is calculated from the excitation test data. There are two sets
of values: one for the positive and the other for the zero sequence test. These two values are
converted to a 3x3 submatrix with diagonal elements Ys and off-diagonal elements Ym,
Ys = (Yo + 2Y1)/3
(5.1)
Ym = (Yo Y1)/3
(5.2)
This shunt admittance matrix is then either connected across one winding, or (1/N)-th of the p.u.
values is connected across all N windings.
If shunt admittances are connected across all windings, as shown in Figure 5.1, then no correction
is made by the program to account for the influence of the short-circuit input impedances. As a
consequence, the exciting current in the model of Figure 5.1 will be slightly larger than the
specified value, and the short-circuit input impedance will be slightly smaller than the specified
value. For an exciting current of 0.01 p.u. (or Zm = 100 p.u.) and a short-circuit input impedance
of 0.10 p.u., these differences are approximately 0.1%. Note that the imaginary parts of (Ys, Ym)
become part of the [L]-1-matrix, as indicated by the large, dotted box in Figure 5.1. This
modification makes the [L]-1-matrix nonsingular and invertible. The resistances Rm, which
approximate hysteresis and eddy current losses, must be added as additional branches at the
"external nodes".
external
node
2Rm
(p.u.)
internal
node
internal
node
2Lm
(p.u.)
2Lm
(p.u.)
external
node
2Rm
(p.u.)
original [L]-1 matrix

(singular)
non-singular [L]-1 matrix after addition
of magnetizing inductances
Figure 5.1:
Addition of Magnetizing Branches Across all Windings

(shown for single-phase transformer with N=2)
If the shunt admittance matrix is connected across one winding only, then the program makes a
correction in the case where the excitation test is made across one winding "i" while the shunt
5 - 10

admittance matrix is connected across another winding "k". In that case, the short-circuit input
impedance between "i" and "k" is subtracted from the inverse of (Ys, Ym), and this modified shunt
admittance is then connected across "k". This way, the specified excitation data and the excitation
data obtainable from the model will be identical. If the user specifies zero excitation losses, they
will be raised to the value I2excitingRi in this case, because in reality these losses must be at least as
high as the I2R-losses in winding "i". As explained in reference [2], it may be best to connect the
shunt admittance across the winding closest to the core on transformers with cylindrical windings.
5.3.2
Description of the Data Deck
The structure of the data deck for a BCTRAN transformer model is as follows:
1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
2.

columns 3840 (see Section 4.1.6).
1
2
3
4
5
6
7
8
1234567 890123456789012345678901234567 890 1234567890123456789012345678901234567890
44.
XFORMER
A7
F3.0
3.
Next come data cards which give electrical parameters of the transformer.
These data cards consist of:
One card for excitation data
Exactly N cards for winding data, one for each transformer winding.
Exactly (N(N-1))/2 cards for short-circuit test data, one for each short-circuit test between a pair of windings. Terminate the short-circuit test data
with a blank card.
5 - 11
blank card
BCTRAN data cards
4.
within "BCTRAN", corresponding to a different transformer. A blank card
terminates these.
5.
To indicate the end of all AUX data cases add a "BEGIN NEW DATA CASE" card at
this point, followed by a blank card, as described in Section 4.1.1.
blank card
begin new data case card
Note that there will be two blank cards at the end of the last transformer data deck: one to
terminate the short-circuit test data and one to terminate "XFORMER" data cases.
Excitation Data
One card with the format shown below.
I2
E10.2
E10.2
pos
LOSS excit
zero
Iexcit
zero
Srating
zero
LOSSexcit
(MVA)
(kW)
(percent)
(MVA)
(kW)
E10.2
E10.2
E10.2
E10.2
E10.2
I2 I2 I2 I2
N
(12)
Number of windings per core leg (link). Present limit: N<10.

Example: A 230/500 kV three-phase transformer without a tertiary
winding has N=2; if a tertiary winding is added, then N=3.
f
(3-12)
Rated frequency in Hz (needed to convert reactances into inductances).
5 - 12
IPRINT
(percent)
pos
S rating
IPUT
(Hz)
pos
I excit
ITEST
f
N
NPHASE
1
2
3
4
5
6
7
8
12 3456789012 3456789012 3456789012 3456789012 3456789012 3456789012 3456789012 34 56 78 90

pos
pos
I excit
Exciting current in percent, based on three-phase power rating S rating and
(1322)
rated voltages, in the positive sequence excitation test.
pos
S rating
(23-32)
pos
LOSS excit
(3342)
zero
pos
Three-phase power rating in MVA, on which exciting current I excit of the

positive sequence test is based.
Excitation loss in kW in the positive sequence excitation test. See
Section 5.3.3 for possible modifications of this value by the program.
pos
I excit
(4352)
rated voltages, in the zero sequence excitation test.
zero
pos
S rating
(5362)
rated voltages, in the zero sequence excitation test
zero
LOSS excit
(6372)
Excitation loss in kW in the zero sequence excitation test. See

Section 5.3.3 for possible modifications of this value by the program
If the transformer has delta-connected windings, then the zero sequence excitation
test really becomes a short-circuit test, since a closed delta acts as a short-circuit
for zero-sequence currents. It is, therefore, assumed that delta connections are
open (Figure 5.2) in the zero sequence excitation test
C2
C1
Figure 5.2:
Open Delta Connection

zero
zero
On transformers with closed deltas, any reasonable value of I excit , S rating ,

zero
and LOSS excit can be used because the influence of these values will be
overridden by the short-circuit test data to the closed deltas. On transformers
with no delta-connected windings or open deltas, the zero sequence current
determines how much voltage will be induced in the two other phases of a
winding if one phase is energized.
5 - 13

For three-phase transformer banks consisting of single-phase units, input the
single-phase data as positive sequence parameters and leave the fields for the
zero sequence parameters blank.
NPHASE
=1
For three-phase transformer banks consisting of singlephase transformers.
= 0 or blank
For three-phase transformers.
(7374)
Number of the winding from which the excitation tests were made.
ITEST
(7576)
Number of the winding across which the magnetizing branch is to be
placed. If ITEST and IPUT are both zero or blank, then the program
connects magnetizing branches across all windings. If ITEST is
specified (>0), then IPUT must also be specified (>0). IPUT = ITEST is
permitted. For more details see Section 5.3.3.
IPUT
(7778)
IPRINT
= 0 or blank
Matrices [R] and [L]-1 will be printed and saved on file.
>0
Matrices [R] and [L] will be printed and saved on file.
<0
Matrices [R] and [L]-1 as well as matrices [R] and [L]

will be printed and saved on file.
(7980)
Winding Data
Exactly N cards, one for each transformer winding. The N cards can be read in arbitrary order.
The format is shown below.
1
2
3
4
5
6
7
8
123 4567890123 4567890123 4 567890 123456 789012 345678 901234 567890 12345678901234567890
NAME 1
k
Vrating-k (kV)
I3
E10.2
k
(13)
Rk ()
E10.2
NAME 2
winding k
phase 1
A6
A6
NAME 3
NAME 4
winding k
phase 2
A6
A6
NAME 5
NAME 6
winding k
phase 3
A6
A6
Winding number. Number windings consecutively 1, 2, 3, ..., N

(where N<10). A wye-wye-connected 230/500 kV three-phase
transformer with a delta connected tertiary of 30 kV would have
3 windings (i.e., 1 = high voltage 500 kV, 2 = low voltage 230
kV, 3 = tertiary voltage 30 kV).
5 - 14

Rated voltage in kV; line-to-ground for wye-connected
windings, and line-to-line for delta connected windings.
Vratingk
(413)
In the example above: V1=500/

V3= 30 kV.
Rk
(1423)
3 kV, V2 = 230/
3 kV,
Winding resistance in ohms of one phase (if the values differ in

the three phases, use the average value). If the winding
resistances are not known, they can be calculated from the load
losses supplied with the short-circuit data if N=2 or 3.
Strictly speaking, the load losses are not only I2R-losses, but
contain stray losses as well; however, this is ignored. In the
calculation of winding resistances from load losses, it is assumed
that R1 p.u. = R2 p.u. for two winding transformers. For threewinding transformers, there are three equations in three
unknowns R1 p.u., R2 p.u., R3 p.u. For transformers with four or
more windings (per phase), there is no easy way to find winding
resistances from the load losses. Therefore, winding resistance
must be specified as input data for N>4.
NAME 1
Columns 2530; 3136; 3742; 4348; 4954; 5560.
Node names. The terminals of the winding in each one of three

phases have to be assigned node names to produce output data in
the form of branch cards which can be used directly as input by
the EMTP. Exactly six node names are required per winding (one
pair for each one of the three phases). If a terminal is connected
to ground (e.g., the neutral in wye connection), then use a blank
field as the name for 'ground'.
.
.
NAME 6
Short-Circuit Test Data

Exactly N(N-1)/2 cards, one card for each short-circuit test between a pair of windings, terminated
by a blank card. The cards can be read in arbitrary order. The card format is shown below.
I2 I2
ik
(kW)
E10.2
pos
ik
(%)
E10.2
pos
rating
(MVA)
zero
ik
(%)
E10.2
E10.2
5 - 15
zero
rating
(MVA)
E10.2
I2 I2
ILOSS
IDELTA
1
2
3
4
5
6
7
8
12 34 5678901234 5678901234 5678901234 5678901234 5678901234 56 78 9012345678901234567890
i
(12)
k
(34)
Numbers of the pair of windings between which the short-circuit test is made.
Pik
(514)
Load losses in kW in the positive sequence test. If Pik>0, then this value is
used in Equation 5.3 below to find the positive sequence reactance:
Xik p.u. = (Zik p.u.)2 (Ri p.u. + Rk p.u.)
with
Zik p.u. =
(5.3)
p.u. short-circuit impedance in test between i and k,
and
Ri p.u. + Rk p.u. = p.u. load losses on the same MVA basis as Zik p.u. if load
losses are nonzero; if load losses are not given, these are
the specified p.u. winding resistances on the same MVA
basis.
Pik can also be used to calculate winding resistances for N<3, provided
Pik > 0 for all short-circuit test (see parameter ILOSS below). Read-in
winding resistances are then ignored.
Equation 5.3 is used for positive sequence values. It is also used for zero
sequence values if the zero sequence test does not involve a third (deltaconnected) winding. In the latter case, the following procedure is used. Let
us assume that the high-voltage and low-voltage windings are wye-connected
with their neutrals grounded. In this case, the zero-sequence short-circuit test
between the high- and low-voltage windings will not only have the lowvoltage winding shorted but the tertiary winding as well if the delta is closed
(which is usually the case). This special situation is handled by modifying
the data for an open delta so that the earlier approach can again be used. With
the well-known equivalent star circuit of Figure 5.3, the three test values
supplied by the user are:
closed
X 12
X13
X2 X 3
= X 1 + -------------------X2 + X 3
(5.4)
= X1 + X3
(5.5)
5 - 16
X23
= X2 + X3
(5.6)
which can be solved for X1, X2, X3:

X 1 = X 13 X X X X closed
23 13
23 12
(5.7)
X2 = X23 X13 + X1
(5.8)
X3 = X13 - X1
(5.9)
2 (low)
X2
1(high)
X3
X1
3 (tertiary)
Figure 5.3:
Equivalent star circuit for zero sequence short-circuit tests of a three-winding

transformer (all reactances must be in p.u.)
After this modification, the program works with short-circuit reactances

X1 + X2, X1+ X3 and X2 + X3, which implies that winding 3 is no longer
shorted in the test between 1 and 2. The modification scheme used in the
program is more complicated because the resistances are also included in
Equation 5.4, which becomes:
closed
Z 12
( R 2 + jX 2 ) ( R 3 + jx 3 )
= R 1 + j X 1 + ---------------------------------------------------------(R + R ) + j(X + X )
2
5 - 17
(5.10)

closed
with Z 12
being the values supplied by the user, the R1, R2, R3 being the
winding resistances which were either directly supplied by the user or which
were calculated from the load losses, as explained in the description of the
Winding Data Card
pos
Z ik
(1524)
Short-circuit input impedance in percent in the positive sequence test between

pos
windings i and k, based on S rating (three phase) and on the rated voltages of
both windings. In North-American standards, the short-circuit input
impedance is called "impedance voltage"; in some European standards it is
called "short-circuit voltage."
pos
S rating
pos
Three-phase power rating in MVA, on which Z ik is based.
(2534)
zero
Z ik
(3544)
zero
S rating
(4554)
Same as preceding two parameters, respectively, except for zero sequence

test. If IDELTA = 0, then Pik from the positive sequence test is also used to
zero
calculate the zero sequence reactance with Equation 5.3. If S rating = 0 or

pos
blank, it is assumed to be equal to S rating .

If IDELTA>0, then the winding resistances are used to obtain reactances from
impedances. In Ydd connections, the zero sequence test cannot be performed
zero
between the closed delta connected windings. Set Z ik
= 0 or blank in such
a case.
The program will then automatically calculate a reasonable value from
zero
zero
pos Z y 2d
X d d = X d d --------------pos
Zy 2d
(5.11)
Here "2d" in the subscript indicates that both deltas are shorted in parallel.
IDELTA
(5556)
=0
or blank
The zero sequence short-circuit test involves only windings i and

k, as in transformers where all windings are wye-connected with
grounded neutrals. If a transformer has a delta-connected winding
and if the winding is not k, then the delta must be open in the test
between i and k if IDELTA=0.
>0
Number of additional winding which is short-circuited in addition

to winding k in zero sequence test between i and k, as described
earlier. This additional winding will normally be delta-connected
(in the case of a closed delta). For the most important case of
three-winding transformers, the program can presently handle
Yyd-connections and Ydd-connections.
5 - 18
In the Yyd connection, "d" would be the additional shorted winding in the
zero sequence test between "Y" and "y". In the Ydd-connection (1=Y, 2=d,
3=d), 3 would be the additional winding in test between 1 and 2, and 2 would
be the additional winding in test between 1 and 3, but both tests would
produce identical impedances (this is recognized by the program, which prints
the message
"Input value of zero sequence short-circuit impedance from 'i'
to 'idelta' is ignored and set equal to value from 'i' To 'k'
because both impedances must be equal if there are closed
deltas in 'k' and 'idelta'".
The program cannot handle Ddd-connections with IDELTA>0.

For three-phase transformer banks consisting of single-phase transformers,
input the single-phase data as positive sequence parameters and leave the
fields for the zero sequence input parameters blank, including IDELTA.
ILOSS
Specify ILOSS on the first short-circuit test data card.
(5758)
=0
or blank
Read-in winding resistances will be used.
>0
Winding resistances will be calculated from load losses Pik,

provided N<3 and Pik>0 for all short-circuit tests. Read-in
5 - 19
5.3.3
Exciting Current in Zero Sequence Excitation Test
If the transformer has delta-connected windings, it will be assumed that the delta connections are
opened for the zero sequence excitation test. Otherwise, the test is not really an excitation test, but
a short-circuit test between the excited winding and the delta-connected windings, since closed
delta connections provide a short-circuit path for zero sequence currents.
= 1--- A
2
1
= --- A
2
VB
VC
VA
Figure 5.4:
Three-Limb Core-Type Design
Often, the zero sequence exciting current is not given by the manufacturer. In such cases, a
reasonable value can be found as follows: Excite one phase of a winding (A in Figure 5.4), and
estimate how much voltage will be induced in the other two phases (B and C). For the three-limb
core design of Figure 5.4, almost one half of flux A returns through phases B and C, which means
that VB and VC will be close to 0.5 VA (with reversed polarity). If we use k for this factor, then
zero
I excit
1+k
------------ = -------------pos
1 2k
I excit
(5.12)
Equation 5.12 is derived from:

VA = ZsIA
VB = ZmIA
VC = ZmIA
(5.13)
with Zs = self impedance of phase A in excitation test, and Zm = mutual impedance to phases B and
C. Then
5 - 20
Z m
Z pos Z z ero
------ V = k V
V B = V C = Z V A = --------------------------------A
A
s
2Z pos + Z zero
(5.14)
Since the exciting current is proportional to 1/Zpos in positive sequence, and to 1/Zzero in zero
sequence, the relationship for k in Equation 5.14 can be transformed into Equation 5.12.
Obviously, k cannot be exactly 0.5, because this would lead to an infinite zero sequence exciting
zero
pos
current. A reasonable value for I excit of a three-limb core-type design might be 100%. If I excit =
0.5%, this would produce k = 199/401 = 0.496, which comes close to the theoretical limit of 0.5
mentioned above.
Besides the three-limb core-type design, there are also five-limb core-type designs, (Figure 5.5)
and shell-type designs (Figure 5.6). In the five-limb core-type design, maybe two thirds of
approximately (1/2) A returns through legs B and C. In that case, k would be one third, or
zero
I excit
------------ = 4
pos
I excit
= 1--- A
2
1
--3
1
= ---
2 A
B
2
--3
Figure 5.5:
Five-Limb Core-Type
Design
Figure 5.6:
Shell-Type Design
The excitation loss in the zero sequence test is higher than in the positive sequence test, because
the fluxes A, B, C in the three cores are now equal, and in the case of a three-limb core-type
design, they must return through air and the tank, with additional eddy-current losses in the tank.
Neither the values for the zero sequence excitation current nor the value for the zero sequence
excitation loss are critical if the transformer has delta-connected windings because excitation tests
really become short-circuit tests in such cases.
5 - 21
5.3.4
Error Messages
The following messages indicate fatal errors in the input data. In each case program execution will
be terminated and no more input cards will be read from the data deck of this case or of any
following cases.
1.
"NUMBER OF WINDINGS = 'n'"

Number of windings is either 1 or greater than 10.
2.
"EITHER ITEST = 'i' OR IPUT = 'k' NOT PERMITTED"

A winding has been specified from which the excitation test has been made but
no winding has been specified across which the magnetizing branch should be
connected, or vice versa.
3.
"'i'k' WRONG WINDING NUMBERS"

Message refers to pairs of windings between which the short-circuit test was
made, in the following cases:
4.
(A)
Both windings have the same number.
(B)
Either one or both of the winding numbers are larger than N, the specified
number of transformer windings.
(C)
Data for this pair of windings has already been read in a preceding card.
"LOAD LOSSES OR WINDING RESISTANCES TOO LARGE 'i'k"

If argument of square root in Equation 5.3, Section 5.2.2 is negative.
5.
"ONLY 'n' SHORT-CIRCUIT TESTS SPECIFIED, BUT 'm' ARE NEEDED"

Not enough short-circuit test data have been read in.
6.
7.
"IDELTA='idelta' WRONG IN SHORT-CIRCUIT TEST BETWEEN 'i' AND 'k'"

(A)
IDELTA is either i or k (see Section 5.3.2, Short-circuit Tests Data Card).
(B)
IDELTA > N, with N being the specified number of transformer windings.
"MODIFICATION OF ZERO SEQUENCE SHORT-CIRCUIT TEST BETWEEN 'i' AND

'k' NOT POSSIBLE. ERROR CODE = 'm'"
This particular case cannot be handled by the present version of the program.
5 - 22

8.
"DIAGONAL ELEMENT IN ROW 'i' CLOSE TO ZERO"

Can happen during internal calculations to convert input data into [L-1] form
(i.e., in the inversion process with Equation 11 of reference [2] -very unlikely-,
or with the inversion of Equation 20b of reference [2] - possible if exciting
current is very small).
9.
5.3.5
"P.U. EXCITATION LOSS LARGER THAN P.U. EXCITING CURRENT (EITHER

IN POS. OR ZERO SEQUENCE)"
Description of the Output
Section 5.3.6 shows a sample output. The results consist of two parts:
Shunt Resistances for Representation of Excitation Losses

Depending upon parameters IPUT and ITEST on the excitation data card (Section 5.3.2), the
program will provide one of the following results with short explanations:
"SHUNT RESISTANCES FOR REPRESENTATION OF EXCITATION LOSSES:"
(A)
"PLACE SHUNT RESISTANCE MATRIX ACROSS WINDING 'IPUT' WITH

R(SELF/OHM)= '_____' AND R (MUTUAL/OHM) = '_____'"
(B)
"PLACE SHUNT RESISTANCE MATRIX ACROSS ALL TERMINALS WITH THE

FOLLOWING VALUES:"
"WINDING NO.
(C)
R(SELF/OHM)
R(MUTUAL/OHM)"
'_____'
'_____'
'_____'
"LEAVE OFF, BECAUSE SERIES RESISTANCES ALREADY PRODUCE LOSSES

WHICH ARE GREATER THAN INPUT VALUES OF EXCITATION LOSSES."
Resistance and Reactance (or Inverse Inductances) Matrix

These matrices are printed, as well as written on file for direct input into the EMTP (see parameter
IPRINT,in the Excitation Data card of Section 5.3.2). As an example, the impedance matrix for a
three-winding, three-phase transformer would have the following general form (only lower
triangular part of the symmetric matrix provided):
5 - 23
Z11
Z12
Z22
Leg I
Z13
Z23
Z33
M11
M12
M13
Z11
M12
M22
M23
Z12
Z22
M13
M12
M33
Z13
Z23
Z33
M11
M12
M13
M11
M12
M13
Z11
M12
M22
M23
M12
M22
M23
Z12
Z22
M13
M12
M33
M13
M12
M33
Z13
Z23
Leg I
Leg II
Leg II
Leg III
Z33
Leg III
Zik = coupling between windings on one leg (including self impedance Zii).
Mik = coupling between windings on different legs.
5.3.6
Test Example
To illustrate the AUX-request for "BCTRAN", a partial listing of benchmark DCNEW-8 is shown
below:
BEGIN NEW DATA CASE
C
BENCHMARK DCNEW-8A
C
TEST OF "BCTRAN" 3-PHASE TRANSFORMER ROUTINE OF EMTP.
THIS
C
PARTICULAR TEST CASE IS FROM HERMANN'S ORIGINAL UBC WRITEUP.
C 45678901234567890123456789012345678901234567890123456789012345678901234567890
XFORMER
44.
360.
.428
300.
135.73
.428
300.
135.73
1 3 0
1132.79056 .2054666
H-1
H-2
H-3
263.393059 .0742333
L-1
L-2
L-3
350.
.0822
T-1
T-2
T-2
T-1
1 20.
8.74
300.
7.3431941 300.
3 1
1 30.
8.68
76.
26.258183 300.
2 30.
5.31
76.
18.552824 300.
BLANK CARD TO TERMINATE THE SHORT-CIRCUIT TEST DATA
BLANK CARD TO TERMINATE "XFORMER" DATA CASES
BEGIN NEW DATA CASE
BLANK
5 - 24

The corresponding punch file of the AUX simulation contains the EMTP branch cards of a coupled
RL branch with the high-precision format. Note that the "USE RB" option is enabled since BCTRAN
generates a [R] and [L-1] matrices. A partial listing for the above example is shown here:
$VINTAGE, 1
USE RB
1H-1
2L-1
3T-1
0.2054666000E+000.2651269237E-01
0.0000000000E+00-.5957848438E-01
0.7423330000E-010.1808547434E+00
0.0000000000E+000.5124542161E-02
0.0000000000E+00-.7106950227E-01
0.8220000000E-010.7656071131E-01
0.0000000000E+000.1317410104E-02
0.0000000000E+00-.1044760157E-02
0.0000000000E+00-.2174181664E-02
0.2054666000E+000.2651269237E-01
T-2
4H-2
.
.
.
9
T-1
0.0000000000E+00-.2174181664E-02
0.0000000000E+000.2647586814E-02
0.0000000000E+000.2417436248E-02
0.0000000000E+00-.2174181664E-02
0.0000000000E+000.2647586814E-02
0.0000000000E+000.2417436248E-02
0.0000000000E+000.5124542161E-02
0.0000000000E+00-.7106950227E-01
0.8220000000E-010.7656071131E-01
5.4 Subroutine "TOPMAG"

5.4.1
Introduction
Conceptually, this model takes into account the topology of the magnetic circuit formed by the
core and the windings to assemble an equivalent electric circuit representation for the transformer
(Figure 5.7 below). For three-limbed and five-limbed units, each core-limb is modelled
individually, and interfaced to an admittance matrix reproducing the correct magnetic coupling
among windings. An additional 3-phase winding (termed fictitious winding) is needed to establish
this interface, since core limbs are electrically isolated from the windings. Node names for
fictitious windings are assigned by the user, and ought to be unique for each transformer.
(A)
The output from the model consists of the following card images for direct
insertion in an EMTP runstream: (1) a symmetric admittance matrix of order up
to 3(N+1), depending on the core type and the specified modelling options, and
5 - 25

(B)
a network of parallel RL branches modelling magnetizing currents due to the

wound limbs (Zb), the horizontal yokes (Zk), and the zero-sequence return path
through air for 3-limbed transformers, or through the outer limbs in 5-limbed
transformers (Zo).
The program requires that N(N-1)/2 positive-sequence short circuit test values be specified for an
N-winding 5-limbed or single-phase transformers. For 3-limbed transformers, at least one
additional parameter is needed to characterize the zero-sequence performance, as follows:
(A)
(i) zero-sequence excitation current, or
(B)
(ii) one zero-sequence short-circuit test impedance, performed with excitation

on winding 1 and short-circuit on any other winding "m", provided that no
other winding is connected in delta during the test.
For more accuracy, a full complement of N(N-1)/2 zero-sequence short-circuit test values may be
specified, if they are available, provided that no more than one of the windings is connected in
delta during tests.
Windings should be numbered from 1 to N, such that the winding positioned outermost on the
wound limb is "1". This is normally the winding carrying the highest voltage rating. Winding "N"
ought to be innermost on the core, and is normally the one with the lowest voltage rating.
Windings are assumed to be concentric, fully covering the wound limb, (which may not always be
the case in practice). For example a tapped HV winding may comprise two separate winding
sections, one outermost on the core and a second (tapped) section innermost on the core. For
higher accuracy, and perhaps for some studies, each such winding section ought to be modelled as
a separate winding (requiring additional short-circuit test data, which is not normally available).
However this is unlikely to be needed for most studies.
5 - 26
Figure 5.7:
5.4.2
Zk
Zo
Winding Leakage
Zo
Core Topology
Zk
Zo
Ls,Lm
Ls,Lm
Zb
Zb
Zb
Phase A
Phase B
Phase C
Schematic representation of model for 3-phase, 3-winding core type transformer
Description of the Data Deck
The structure of the data deck for a TOPMAG transformer model is as follows:
1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
2.
3.
Columns 3840 (see Section 4.1.6).
5 - 27
1
2
3
4
1234567 890123456789012345678901234567 890
5
6
7
8
1234567890123456789012345678901234567890
55.
XFORMER
A7
F3.0
Next come the transformer data specification cards (similar to the input requirements for BCTRAN,
with slight variations as described next.
Class 1:
One card for excitation data.
Class 2:
Exactly N cards for winding data, one for each transformer winding.
Class 3:
One card for the fictitious winding which provides the interface to the circuit modelling core magnetics. Relative limb dimensions may be specified optionally.
Class 4:
Exactly N(N-1)/2 cards for short-circuit test data, one for each short-circuit test
between a pair of windings. Terminate the short-circuit data with a blank card.
blank card
TOPMAG data cards
4.
within "TOPMAG", corresponding to a different transformer. A blank card
terminates these.
5.
To indicate the end of all AUX data cases add a "BEGIN NEW DATA CASE" card at
this point, followed by a blank card, as described in Section 4.1.1.
blank card
Note that there will be two blank cards at the end of the last transformer data deck: one to
terminate the short-circuit test data and one to terminate "XFORMER" data cases.
5 - 28
5.4.3
Class 1: Excitation Data
One card with the format shown below.
I2
(percent)
E10.2
E10.2
N
(12)
pos
LOSS excit
zero
I excit
zero
S rating
zero
LOSS excit
(MVA)
(kW)
(percent)
(MVA)
(kW)
E10.2
E10.2
E10.2
E10.2
E10.2
I2 I2 I2 I2
Number of windings per core leg. Present limit: N10.

Example: A 230/500 kV three-phase transformer without a tertiary winding
has N=2; if a tertiary winding is added, then N=3.
f
(3-12)
Rated frequency in Hz (needed to convert reactances into inductances).
pos
pos
I excit
(1322)
rated voltages, in the positive sequence excitation test.
pos
pos
S rating
Three-phase power rating in MVA, on which exciting current I excit of the
(23-32)
positive sequence test is based.
pos
LOSS excit
(3342)
zero
I excit
(4352)
zero
S rating
(5362)
zero
LOSS excit
(6372)
Excitation loss in kW in the positive sequence excitation test. This value

may be changed by the program based on specification for "ILOSS" under
short-circuit test data.
Same as preceding three parameters, respectively, except for zero-sequence
excitation test. If the transformer has delta-connected windings, then the
excitation test really becomes a short-circuit test since a closed delta acts as
a short-circuit for zero-sequence currents. It is, therefore, assumed that
delta connections are open in the zero sequence excitation test.
For 3-phase transformer banks consisting of single-phase or 5-limbed units,
input the data as positive-sequence parameters and leave the fields for the
zero-sequence parameters blank.
5 - 29
IPRINT
(Hz)
pos
S rating
ISOLKG
pos
I excit
ITEST
f
N
NTYPE
1
2
3
4
5
6
7
8
12 3456789012 3456789012 3456789012 3456789012 3456789012 3456789012 3456789012 34 56 78 90

zero
For 3-limb units, I excit will be estimated by default if this entry is left
blank. This is the recommended option, provided that at least one
zero-sequence short-circuit test is specified with excitation on winding 1,
and without any other winding connected in delta during the test
zero
(See Class 4 data). If a value for I excit is specified, it is used directly (also
see ISOLKG under Class 1 cards).
NTYPE
=1
For three-phase transformer banks consisting of single-phase

units.
=5
For three-phase 5-limbed core.
1 or 5
For three-phase 3-limbed core.
(73-74)
If NTYPE is -ve, the debug facility will be activated and the

magnitude of NTYPE will have the same meaning as indicated
above.
(75-76)
Number of the winding from which the zero-sequence excitation test was
made. (DEFAULT = 1)
ISOLKG
This is the recommended option for modelling 3-limbed

transformers, provided a full complement of reliable N(N-1)/2
zero-sequence short-circuit impedances is available. The
resulting coupling matrix is of order 3(N+1), matching the
specified positive- and zero-sequence impedances. For singlephase or 5-limbed units, ISOLKG is irrelevant.
<0
In this event, even if zero-sequence short-circuit impedances are

specified for 3-limbed units, they are ignored except for
ITEST
(77-78)
zero
estimating I excit if it has not already been specified. This option

is recommended if a full complement of zero-sequence test data is
not available. Zero-sequence short-circuit impedances are
therefore established by default on a physical basis.
IPRINT
(79-80)
=0
or
blank
5 - 30
5.4.4
Class 2: Winding Data
Exactly N cards, one for each transformer winding. The cards can be read in arbitrary order,
however the outermost winding on the core (usually the one with the highest voltage rating)
should be designated as winding #1, and any winding remaining closed during short-circuit tests
(if one exists) must be winding number "N". The card format is shown below.
1
2
3
4
5
6
7
8
123 4567890123 4567890123 4 567890 123456 789012 345678 901234 567890 12345678901234567890
NAME 1
k
Vrating-k (kV)
I3
E10.2
Rk ()
E10.2
NAME 2
winding k
phase 1
A6
A6
NAME 3
NAME 4
winding k
phase 2
A6
A6
NAME 5
NAME 6
winding k
phase 3
A6
A6
k
(1-3)
Winding number. Number windings consecutively 1, 2, 3..., N (N10). A

wye-wye connected 230/500 kV three-phase transformer with a delta connected
tertiary of 30 kV would have 3 windings (e.g., 1 = high voltage 500 kV, 2 = low
voltage 230 kV, 3 = tertiary voltage 30kV).
Vrating-k
(4-13)
Rated voltage in kV; line-to-ground for wye-connected winding, line-to-line for

delta connected winding.
In the above example: V1 = 500/ 3 kV, V2 = 230/ 3 kV, V3 = 30 kV.
Rk
(14-23)
Winding resistance of one phase (in ohms). If the values differ among phases,
use the average value. If the winding resistances are not known, they can be
calculated from the load losses supplied with the short-circuit data if N=2 or 3.
Strictly speaking, the load losses are not only I2R-losses, but contain stray
losses as well; however, this is ignored. In the calculation of winding
resistances from load losses, it is assumed that R1 p.u. = R2 p.u. for two
winding transformers. For three-winding transformers, there are three
equations in three unknowns R1 p.u., R2 p.u., R3 p.u. For transformers with
four or more windings (per phase), there is no easy way to find winding
resistances from the load losses. Therefore, winding resistance must be
specified as input data for N4.
NAME 1
Columns 25-30; 31-36; 37-42; 43-48; 49-54; 55-60.
.
.
.
NAME 6
Node names. The terminals of the winding in each one of three phases have to
be assigned node names to produce output data in the form of branch cards
which can be used directly as input by the EMTP. Exactly six node names are
required per winding (one pair for each one of the three phases). If a terminal is
connected to ground (e.g. the neutral in wye connection), then use a blank field
as the name for 'ground'.
5 - 31
5.4.5
Class 3: Data for Duality Winding (and Optionally, Limb

Dimensions)
Exactly one card, specifying the fictitious winding which implements the duality based magnetic
model for the core. The card format is shown below.
The program produces a complete linear topological model for 3-limbed and 5-limbed core-type
transformers. This includes generation of the coupling matrix, and linear magnetizing branches
(including shunt conductance branches) for the wound limbs, the yokes and the zero-sequence
magnetic path.
1
2
3
4
5
6
7
8
123 4567890123 45678901234 567890 123456 789012 345678 901234 567890 12345 67890 12345 67890
NAME 1 NAME 2 NAME 3 NAME 4 NAME 5 NAME 6
CUPCO
E10.2
CUPCO
(4-13)
NAME 1
.
.
.
NAME 6
winding k
phase 1
A6
A6
winding k
phase 2
A6
A6
winding k
phase 3
A6
A6
LENGTH
AREA
YkLm
OtLm
YkLm
OtLm
E5.0
E5.0
E5.0
E5.0
This is normally left blank since a value is assigned by default. An option is

provided here to override the default value in case numerical ill-conditioning
problems are encountered with the model. This may be signalled by an error
message during time domain solution warning about floating subnetworks.
CUPCO defines the coupling between winding 1 and the fictitious winding. Its
value essentially represents the reciprocal of the p.u. leakage impedance
between windings 1 and 2, and should be large (eg. 103 to 104 range). If a
value of less than 10 is specified, it is ignored.
Columns 25-30; 31-36; 37-42; 43-48; 49-54; 55-60.
Node names, as for Class 2 Cards (unique for each transformer, otherwise an
EMTP simulation could end up with two transformers with identical internal
node names). The terminals of the winding in each one of three phases have to
be assigned node names to produce output data in the form of branch cards
which can be used directly as input by the EMTP. Exactly six node names are
required (one pair for each one of the three phases). If a terminal is connected
to ground (e.g. the neutral in wye connection), then use a blank field as the
name for 'ground'.
5 - 32

LENGTHS:
Ratio of limb lengths may be specified if desired, otherwise a default value is

assigned. This determines the unbalance in excitation current among phases
for 3-limbed and 5-limbed units. The computed model reproduces the
specified excitation current in the centre phase. For 3-limbed units, the
established excitation current is a little higher in the outer phases (eg. about
39% higher for YkLm=0.5).
YkLm
(61-65)
OtLm
(66-70)
AREA:
Ratio of yoke length to limb length;
OtLm:
Ratio of outer limb length (including horizontal section) to

vertical (wound) limb length.
Ratio of effective limb cross-sectional areas may be specified, if desired, for

5-limbed units. If no input is specified, a default value is assigned. For
3-limbed units, a value of 1 (unity) is always assumed.
YkLm
(71-75)
OtLm
(76-80)
5.4.6
YkLm:
YkLm:
Ratio of yoke area to limb area;
OtLm:
Ratio of outer limb area (including horizontal section) to vertical

(wound) limb area.
Class 4: Short-Circuit Test Data
Exactly N(N-1)/2 cards, one card for each short-circuit test between a pair of windings, terminated
by a blank card. The N cards can be read in arbitrary order. The format is shown below.
I2 I2
P ik
(kW)
E10.2
i,k
(1-2)
(3-4)
pos
ik
(percent)
pos
rating
(MVA)
E10.2
E10.2
zero
rating
(MVA)
E10.2
E10.2
I2 I2
zero
ik
(percent)
ILOSS
IDELTA
1
2
3
4
5
6
7
8
12 34 5678901234 5678901234 5678901234 5678901234 5678901234 56 78 9012345678901234567890
Numbers of the pair of windings between which the short-circuit test is

made. Excitation is applied to winding "i" with "k" short-circuited. The
numbering order is crucial for zero-sequence tests since the tests are not
necessarily reciprocal. For positive-sequence short-circuit tests, the order is
immaterial.
5 - 33

Pik
(5-14)
Load losses in kW in the positive-sequence test. If Pik > 0, then this value is
used in Equation 5.15 to find the positive-sequence reactance:
X ik pu =
with Zik pu =
( Z pu ) ( R pu + R pu )
ik
i
k
(5.15)
p.u. short-circuit impedance in test between i and k and
Ri pu + Rk pu = p.u. load losses on the same MVA base as Zik p.u. if load
losses are nonzero, or specified p.u. winding resistances on
the same MVA base if load losses are not given.
Pik can also be used to calculate winding resistances for N<3, provided
Pik0 for all short-circuit test (see parameter ILOSS). Read-in winding
resistances are then ignored.
pos
Z ik
(15-24)
Short-circuit input impedance in percent in the positive-sequence test

pos
between windings i and k, based on S rating (three phase) and on the rated
voltages of both windings. In North-American standards, the short-circuit
input impedance is called "impedance voltage" in some European standards,
it is called "short-circuit voltage".
pos
S rating
pos
Three-phase power rating in MVA, on which Z ik
(25-30)
5 - 34
is based.

zero
Z ik
(35-44)
zero
S rating
(45-54)
Same as the preceding two parameters, respectively, except that they

correspond to zero-sequence test. For single-phase and 5-limbed units,
these can be left blank.
For 3-limb units, a facility is provided for the model to be computed based
zero
on physical considerations, such that it is not necessary to specify all Z ik
values. This option is recommended only if a full complement of measured

test values is not available. In this event, ISOLKG must be set to a negative
zero
value (ie. <0). If ISOLKG < 0 and the specified I excit = 0 (or it is left blank),
zero
then I exci t is estimated using one zero-sequence test value and all others are
ignored. In these cases, the model is computed using positive-sequence
short-circuit impedance, based on physical considerations. All zerozero
sequence impedances are established by default. (See I excit and ISOLKG

under Class 1 cards).
If IDELTA = 0, then Pik from the positive-sequence test is also used to
zero
calculate the zero-sequence reactance with Equation 5.15. If S rating = 0 or

pos
blank, it is assumed to equal S rating

If IDELTA > 0, then the specified winding resistances are used directly to
determine reactances based on specified impedances.
IDELTA
(55-56)
=0
or blank
The zero-sequence short-circuit test involves only windings i

and k, as in transformers where all windings are wye-connected
with grounded neutrals. If a transformer has a delta-connected
winding and if the winding is not k, then the delta must be open
in the test between i and k if IDELTA=0.
>0
Number of the additional winding which is short-circuited in

addition to winding k in zero-sequence test between i and k, as
described earlier. This additional winding will normally be
delta-connected (closed delta). For the most important case of
three-winding transformers, the program can presently handle
Yyd-connections only.
In the Yyd connection, "d" would be the additional shorted
winding in the zero-sequence test between "Y" and "y".
5 - 35

The program cannot handle Ydd or Ddd-connections with
IDELTA>0.
For three-phase transformer banks consisting of single-phase
transformers, input the single-phase data as positive-sequence
parameters and leave the fields for the zero-sequence input
parameters blank, including IDELTA.
Specify ILOSS on the first short-circuit test data card.
ILOSS
(57-58)
5.4.7
= 0,
or blank
Specified winding resistances will be used directly.
>0
Winding resistances will be calculated from load losses Pik,

provided N3 and Pik0 for all short-circuit tests. Read-in
Sample Data File
C TOPMAG DATAFILE
C BEGIN NEW DATA CASE
XFORMER
55.
C --------------------------- XFMR DATA ---------------------------------------C LONGWOOD TS AUTO NEI #44484 500/240/28-KV 750-MVA 3-PHASE
C EXCITATION DATA:
<N<I<I<I
C <---FRQ--><
pos ><
pos ><
pos ><
zer >< zer
><
zer ><T<T<S<P
C
< %I
>< S MVA ><Loss(kW)>< %I
>< S MVA ><Loss(kW)><Y<S<L<R
C
<-- exc >< rating><
exc ><
exc >< rating><
exc ><P<T<K<N
3
60.0
0.030
750.0
200.0
5 1
C
C W<-Vrating><- R ----> <----------- NODE NAMES ----------->
C G<(kV-LG Y><
dc
> <- PHASE A-><- PHASE B-><- PHASE C->
C #<(kV-LL D>< (ohms) > <FROM><-TO-><FROM><-TO-><FROM><-TO->
1 288.67513 1.000000 RH1
HN
WH1
HN
BH1
HN
2 138.56406
.000000 RX1
WX1
BX1
3
28.00
.000000 RY1
RY2
WY1
WY2
BY1
BY2
C
<-LENGTH-><--AREA-->
C <- CUPCO->
<------------- NAMES --------------><YKLM<OTLM<YKLM<OTLM
RZ1
RZ2
WZ1
WZ2
BZ1
BZ2
C
C SHORT CCT TEST DATA: FOR TAP 11 (NEUTRAL TAP)
C W<--------><--------><---pos--><--------><---zer--><I<L
C D< pos
>< pos
>< S
><
zer >< S
><D<O
C G< Loss
>< Z
><
rat >< Z
><
rat ><E<S
C #< (kW) >< (%)
>< (MVA) ><
(%) >< (MVA) ><L<S
1 2
1032.20
13.780
750.00
12.82
750.00 3 0
1 3
131.10
33.300
750.00
29.89
750.00 0
2 3
130.10
18.000
750.00
17.21
750.00 0
C ---------------------------------- END OF XFMR DATA --------------------------BLANK CARD TERMINATING DATA CASE
BLANK CARD TERMINATING TOPMAG
BEGIN NEW DATA CASE
5 - 36
5.5 Modeling Hints for all Transformer Models

5.5.1
Saturation Effects
Since the air-core inductance (which is the slope of the /i-curve in the fully saturated region) is
fairly low (typically twice the value of the short-circuit inductance), it may make a difference
where the nonlinear inductance is added. It is best to put the nonlinear inductance across the
terminals of the winding closest to the core, which is usually the tertiary winding in three-winding
transformers. Supporting evidence may be found in reference [6] and reference [8].
This nonlinear inductance will be in parallel with the unsaturated value of the magnetizing
inductance. Example: if the saturation curve is defined by three points at im = 0.03%, 0.06% and
0.12%, and if 0.03% was used as magnetizing current for finding the impedance matrix, then the
value of 0.03% must be subtracted in defining this nonlinear inductance (dashed line in Figure 5.8,
below).
For a more detailed discussion of the inclusion of saturation effects, please refer to reference [2].
L3
L2
L1
im
0.03
Figure 5.8:
5.5.2
0.06
0.12%
Definition of Nonlinear Inductance
Floating Delta Connection
If transformer windings are connected in delta and nothing else is connected to it, then the delta is
"floating". In a floating delta connection, the voltages to ground are not defined but only the
voltages across the windings. This leads to a singular matrix with a respective error message
termination. Therefore, either ground one terminal or add ground capacitance.
5 - 37
A
B
terminal C grounded
Figure 5.9:
5.5.3
ground capacitances added
Means of avoiding floating delta connections
High-Frequency Effects
The above described models represent the linear behaviour of the transformer with reasonable
accuracy from very low frequencies up to 6 kHz to 10 kHz or so. At higher frequencies,
capacitances would have to be added to model the asymptotic behaviour of the windings, e.g., as
described in reference [1].
A more accurate representation can be obtained using the HFT model described in Section 6,
RuleBook 1. The HFT model, however, requires measurements of the transformer impedances as a
function of frequency, and these are genrally not available from in standard factory tests.
5.5.4
Autotransformers
If the user treats an autotransformer the same way as a regular transformer (that is, if one only
looks at the outside terminals and ignores the fact that two windings have a common section
inside), reasonably accurate results will be obtained with the models produced by the described in
this Section. It is possible, however, to develop more accurate models by modifying the shortcircuit test data. In the case of Figure 5.10, the short-circuit test data between H-L, H-T, L-T
would have to be changed into short-circuit test data between I-II, I-III, II-III. The transformer
would then simply be represented as three coupled windings I, II, III with winding I going from
node 1 to 2 and winding II going from 2 to 3.
5 - 38
1
I
H
2
T
II
III
Figure 5.10: Autotransformer
5 - 39
Section 6
6.1
Background ......................................................................................... 6-1
6.2
Fitting Process ..................................................................................... 6-5

6.2.1
How FDBFIT Works ..................................................... 6-5
6.2.2
Non-Default Situations .................................................. 6-10
6.3
Running FDBFIT ................................................................................ 6-11
6.4
Parameter File ..................................................................................... 6-12
6.5
6.4.1
Control Card 1 ................................................................ 6-12
6.4.2
Control Card 2 ................................................................ 6-14
6.4.3
Fit Control Cards ............................................................ 6-16
6.4.4
Node Name Cards .......................................................... 6-17
6.4.5
File Name Cards ............................................................. 6-17
6.4.6
Power Frequency Transformer Cards ............................ 6-19
6.4.7
Linear Magnetizing Branch Cards ................................. 6-19
Example ............................................................................................... 6-20
Section 6
6.1 Background
FDBFIT is an EMTP or support program designed to generate High Frequency Transformer (HFT)
models. The HFT model is one module of the more comprehensive Frequency-dependent
Transformer Model, which consists of several components that can be combined to obtain the
desired simulation capabilities. These components or building blocks are:
1.
Power frequency module (i.e, BCTRAN, TRELEG, TOPMAG)
2.
3.
Eddy currents module
4.
High Frequency Transformer (HFT) module
Hysteresis Model
HFT Model
Power Frequency Model
Eddy Current Model
Figure 6.1: Frequency-Dependent Transformer Model
6-1

The HFT module reproduces the behaviour of a transformer over a wide frequency range. Raw
input data for the HFT model can be in the form of field measurements of [Y()]; that is, frequency
scan measurements of the nodal admittance matrix of the transformer. Alternatively, raw data can
also be supplied by some transformer parameter calculation program, possibly based on the
physical characteristics and construction details of the transformer. Whether the raw data is
synthesized or measured, it must be manipulated and approximated with rational functions, which
can later be represented as FDB branches in the EMTP. The FDBFIT support routine handles the data
manipulation and approximation process. On output, FDBFIT produces a punch file that contains a
multi-port, multi-phase pi-circuit. Each branch in the pi-circuit contains a multi-phase EMTP FDB
model (thus the name FDBFIT).
Yseries = Y12
Yshunt1 = Y11 + Y12
Figure 6.2:
Yshunt2 = Y22 + Y21
HFT Model for a Two-Winding Three-Phase Transformer
For example, a two-winding, three-phase transformer would be represented with a two-port threephase pi-circuit (see Figure 6.2, above).
The nodal admittance matrix [Y] for this transformer can be partitioned in 3x3 blocks
Y =
Y 11
Y 12
Y 21
Y 22
where [Yij] is an 3x3 block matrix of the form
6-2
y 11 y 12 y13
Y ij = y 21 y 22 y23
y 31 y 32 y33
If [Yij] is assumed to be balanced (or if all the diagonal elements are averaged to produce ys and
all the off-diagonal elements are averaged to produce ym) then,
ys ym ym
Y ij = y m y s y m
ym ym ys
Note that in the pi-circuit of Figure 6.2, we have that
Yshunt1 = [Y11] + [Y12]

Yshunt2 = [Y22] + [Y12]
Yseries = -[Y12]
Since the elements of the pi-circuit are described by balanced matrices, they can be modelled
using zero and positive sequence parameters. The FDB branches produced by FDBFIT represent
the approximations by rational functions of Yshunt and Yseries in zero and positive sequence.
When the frequency response of the transformer is obtained from field measurements, it is not
uncommon to find that due to instrumentation limitations, measured data that is accurate at high
6-3

frequencies (e.g., 1 kHz to 500 kHz) loses accuracy in the power frequency range, especially the
real part of the elements of [Y]. Therefore, it is good practice to rely on the power frequency
module for low frequency response and on the HFT module for high frequency response. This is
achieved by subtracting the frequency behaviour of the power frequency module from the
measured data, and then approximating the resulting difference function using FDB models. This is
default option assumed in FDBFIT. The punch file obtained after running FDBFIT contains the
EMTP data cards of both the power frequency module and the HFT module.
Y23
Y13
Y12
Y11 + Y12
Figure 6.3:
Y13 + Y23
Y22 + Y21
HFT Model for a Three-Winding Three-Phase Transformer
The elements of [Y] are approximated with rational functions which contain real as well as
complex conjugate poles and zeroes. The rational functions can then be realized with RLC
networks which can be combined to produce the parameters of the equivalent pi-circuit.
Complex Poles
Real Poles
RC Branch
Structure of an FDB branch of an HFT model
6-4

The general structure of these RLC networks reflects the known frequency characteristics of the
admittance functions of a transformer:
Inductive behaviour at low frequencies which includes frequency dependent effects due to
skin effect in the windings and iron core eddy current losses. These are simulated the power
frequency transformer model, if present, and by the RL branches.
Series and parallel resonances from mid to high frequencies caused by winding-to-winding
and winding-to-ground stray capacitances. These are reproduced by the RLC branches.
Predominantly capacitive behaviour at high frequencies, represented by the single RC branch.
6.2 Fitting Process

Although FDBFIT has been designed to require minimal user intervention, the nature of the
problem of fitting measured transformer admittance functions with closed-form rational functions
depends on a number of factors that cannot be fully accounted for in all cases, such as, the quality
of the measurements (e.g., proper grounding techniques, leads, bandwidth of the equipment, etc.),
the amount of noise present, and the self-consistency of the measurements. Self-consistency of
measurements is important to guarantee the numerical stability of the final HFT model.
6.2.1
How FDBFIT Works
This Section provides a simplified overview of the fitting process in order to give a better
understanding of the optional features of FDBFIT that may be required to obtain a successful HFT
model.
To create the FDB branches required to produce an HFT model, FDBFIT automatically goes through
the following steps:
1.
Read measured nodal admittance matrix of the transformer as a function of

frequency. One of the simplest ways to measure the nodal admittance matrix of
a transformer is to connect a variable-frequency voltage source to one of the
terminals, while all other terminals (to be retained in the model) are shorted. The
currents flowing into these terminals are proportional the elements of one
column of [Y].
6-5
I1
y 11 y 12 y 13 V
o
y 21 y 22 y 23 0 = I 2
y 31 y 32 y 33
I3
For example, If Vo = 1, then I1, I2, and I3 give y11, y21, and y31 directly. By
connecting the voltage source to terminals 2 and 3, the complete [Y] matrix as a
function of frequency can be obtained. Since [Y] is symmetric, only N(N+1)/2
currents have to be measured (where N is the product of the number of phases
times the number of windings to be modelled).
The preferred sampling scale is logarithmic. 50 points per decade is probably
sufficient, but more points allow a finer resolution of high frequency resonance
peaks. The lowest scan frequency is generally determined by the characteristics
of the instrumentation, but it should not be lower than the first resonant peak
(typically below 5 kHz for large HV transformers). The maximum frequency in
the scan is determined by the highest significant frequency response required in
the EMTP simulations. The higher fmax is, the larger the order of the fitted
functions. Therefore, there are computational advantages to keeping fmax
reasonably low (e.g., 200-300 kHz is probably sufficient for most transient
simulations). Note even if fmax in the measured data is not very high, the
asymptotic behaviour of the transformer at high frequencies is predominantly
capacitive, and it would still be properly reproduced by the HFT model at
frequencies beyond fmax. This will be discussed in more detail later on.
2.
Clean-up noisy data and possible scale mismatches. Although this type of
manipulation is best done before data is given to FDBFIT, the program provides
some simple tools for frequency scale synchronization and stray peak removal
3.
Build the raw-data nodal admittance matrix. By default, FDBFIT builds the
nodal admittance matrix assuming block-symmetry, in order to create zero and
positive sequence multi-port pi-representations of the transformer as seen at its
terminals. The current version of FDBFIT cannot describe a delta-connected
winding from its terminals unless the measurements are made on an open delta,
and the delta connection is made implicitly in the EMTP via node name
assignment. The modifications (in FDBFIT) necessary to produce HFT models for
delta-connected windings have already been developed and will be included in
the next release of the program.
4.
Read power frequency model and subtract low frequency response.

Experience in the measurement and fitting of transformer measured data suggests
that it is very difficult to obtain accurate measurements high frequencies
(hundreds of kHz) and at 50 or 60 Hz with the same instrumentation. For
6-6

example, at low frequencies, [Y] is almost purely imaginary and small errors in
the measurements can result in large errors in the discrimination of the real and
imaginary parts of [Y]. It is then advantageous to generate an HFT model that
"rides on top" of one of the standard low frequency models such as BCTRAN.
FDBFIT can read any of the EMTP power frequency transformer models that can
be represented as a series impedance matrix, such as TRELEG, or as [R] and
[Linverse] matrices, such as BCTRAN, and TOPMAG. FDBFIT will then fit the
frequency response of the transformer after the response of the power frequency
model has been subtracted (see Figures 6.4 and 6.5).
5.
Selecting the fitting method. There are three fitting options:
Figure 6.4:
Real part (absolute value) of Y11 zero sequence. TRELEG and

measured response.
6-7
Figure 6.5:
Imaginary part (absolute value) of Y11 zero sequence. TRELEG and

measured response.
(i)
Fit each element of [Y] (not supported in this release of the EMTP).
(ii)
Fit the branches of the equivalent multi-phase pi-circuit in zero and

positive sequence (can have accuracy problems)
(iii)
Fit the elements of [Y] and then combine the approximations to

produce the branches of the equivalent multi-phase pi-circuit in zero
and positive sequence (default).
Fitting by elements is the default option since it appears to produce better results
even if the raw measurements show some inconsistencies.
6.
Evaluate RC branch for high frequency asymptotic behaviour. This process

is fairly accurate and the optimization process seldom changes the initial estimate
significantly.
6-8
Figure 6.6:
Real part of Y11 zero sequence. Raw data and difference

function.
6-9
Figure 6.7:
7.
Imaginary part of Y11 zero sequence. Raw data and difference

function.
Determine fcut, assign complex conjugate poles and optimize. As illustrated

in Figure 6.6, the functions to be fitted are divided in two zones (from power
frequency to fcut, and from fcut to fmax). The transition frequency fcut is
normally determined automatically by FDBFIT. Typical values of fcut for large
HV transformers vary from 1 to 20 kHz. Ideally, fcut should be equal to the
power frequency (50 or 60 Hz). In practice there is always a certain amount of
disagreement between the frequency scan measurements and the power
frequency model obtained from nameplate data, and fcut takes place at a higher
frequency. Current thinking is that these discrepancies are mainly due to
instrumentation accuracy. In any event, this leaves a region between power
frequency and fcut where the accuracy of the fitted function is not very high after
the complex poles and zeros have been assigned and optimized.
The selection criterion for fcut is the frequency at which the first minimum in the
real part of y() takes place. In some functions it has been found that the
automatic selection criterion can choose a value of fcut that is too low because of
small measurement glitches. In such cases the user has the option to override the
automatic selection and to provide the correct fcut. Whether or not the region
between fmin and fcut should be fitted with extra (real) poles and zeros in order
to improve accuracy will be discussed next.
6 - 10
6.2.2
8.
Assignment of real poles at low frequencies. To improve the fitting between

power frequency and fmin, it is possible to add extra real poles and zeros in this
region. The addition of the extra poles and zeros is left as a user-defined option.
In some instances, the addition of real poles result in very high accuracy at both
high and low frequencies. On the other hand, there are instances where the
addition of real poles can actually degrade overall accuracy. If the option to add
real poles is selected, FDBFIT will automatically discard these real poles if they
degrade the accuracy of the final function. The number of real poles to be added
can also be controlled by the user. In any event, the addition of real poles
increases the order of the fitted functions and slows down the optimization
process. Note that when real poles are assigned, the entire function including
complex conjugate poles is optimized from fmin to fmax.
9.
Combine fitted functions and store in punch file. Also note that the punch
file will include the data cards for the power frequency transformer and the
additional magnetizing branches (if any), as well as the FDB branches generated
for the HFT model (see the example in Section 6.5).
Non-Default Situations
The default settings and options of FDBFIT have been selected to produce stable and accurate fits
of measured transformer data in most of the cases tested. It is possible, however, that a particular
transformer may require a certain amount of user intervention to produce a stable and accurate
model. The following are a number of suggested courses of action to be tried if something does
not work as expected:
1.
Check the validity and consistency of the raw data. If, for instance, the negative
of an admittance function is recorded, the resulting HFT model would be
inconsistent and unstable. Use a known set of admittance functions as reference
to give an idea of what these functions should look like.
2.
Use a reasonable value of fmax based on dominant resonance peaks and the
bandwidth required for the transient simulation. Past a certain frequency,
resonance peaks have very little influence in most transient simulations, and it is
the asymptotic behaviour of the transformer due to stray capacitances that
dominates the behaviour at very high frequencies. In some cases, non-dominant
high frequency resonance peaks beyond 200 kHz only result in a slower higher
order model without a significant improvement in accuracy. Note that FDBFIT
correctly models this asymptotic behaviour even if relatively low values of fmax
are used. This provides some leeway in the selection of fmax.
3.
Given a choice, request high accuracy measurements in the lower frequency

ranges near power frequency. Never try to splice measured functions from two
different instruments.
6 - 11

4.
Use the option to override fcut as a tool to improve the quality of the fit before
the addition of real poles.
5.
The fewer the accessible terminals to be modelled, the better. For example, if the
transformer neutrals are known to be solidly grounded during simulations,
perform the measurements with the transformer grounded, rather than increasing
the order of the model to be fitted and grounding the neutral through node name
assignment in the EMTP.
6.3 Running FDBFIT

To run FDBFIT from a DOS console, enter the command "fdbfit" (without quotation marks).
Measurement data and fitting control parameters are always specified with file "fdbfit.par". The
program will then prompt for output and punch file names. Responding with a carriage return to
these prompts tells the program to assume "fdbfit.out" and "fdbfit.pun" as default output and punch
file names, respectively. The punch file can be included directly into the EMTP data case with a
standard text editor, or by using $INCLUDE statements. Plotted output is directed to the Postscript
plot file "fdbfit.ps".
By default, FDBFIT will look for fdbfit.par and will create all files in the directory from which the
program is invoked. An optional command line argument [path] can be used to specify a nondefault path name for the location of these files (e.g., fdbfd d:\mydir\otherdir\).
In addition to the user-specified supplementary output files, the following files are also created
when running FDBFIT:
FDBFIT1.TMP
FDBFIT2.TMP
FDBFIT3.TMP
These files store intermediate data and can be used as a time saving device when the fitting process
requires some user interaction (e.g., due to imprecise or inconsistent measurement data, or if
adjustment of the fitting control parameters is desirable).
Measurement data and fitting control parameters are always specified with file "fdbfit.par".
Plotted output is directed to the Postscript plot file "fdbfit.ps".
6.4 Parameter File

The name of the control parameter file is "fdbfit.par". This file controls various aspects of the
generation of data for the High Frequency Transformer Model (HFT). Note that either "*" or "C "
in column 1 can be used to add comment lines.
6 - 12

Typical structure of this file follows:
6.4.1
(A)
Control card 1 (Section 6.4.1).
(B)
Control card 2 (Section 6.4.2).
(C)
Fit control cards (Section 6.4.3).
(D)
Node names cards (Section 6.4.4).
(E)
A total of M file names (where M=K*(K+1)/2 and K is the product of the

number of winding and the number of phases NWIND*NPHAS) (Section
6.4.5 File Name Cards).
(F)
Power frequency transformer cards (terminated by a blank card)

(Section 6.4.6).
(G)
Linear magnetizing branch cards (terminated by a blank card) (Section 6.4.7).
Control Card 1
Control card 1:
1
2
3
4
5
6
7
8
12345 67890 12345 67890 12345 67890 12345 67890 12345 67890 123456789012345678901234567890
NPHAS NWIND
I5
I5
OPT
I5
NPHAS
DEBUG PRINT
I5
I5
PLOT
I5
SYNC STOPS FILTR STOPF
I5
I5
I5
I5
Number of phases. Normally, NPHAS will be 1 or 3.
(1-5)
NWIND
(6-10)
OPT
Number of windings accessible to the EMTP. For example, a

wye-wye-delta transformer where the delta winding is buried would
be considered as having two accessible windings and NWIND would
be equal to 2. Note that NWIND must be greater than 1.
Flag to select the method to fit [Y].
(11-15)
=0
Fitting by elements (see Section 6.2.1). This is the default

setting.
=1
Fitting by branches (see Section 6.2.1).

6 - 13

=2
DEBUG
(16-20)
PRINT
(21-25)
IPLOT
Fit each element of yi,j of [Y] without assuming that [Y] is

block-balanced. Note that at this point in time, there is no
provision in the EMTP to simulate the model generated with
this option.
Set DEBUG = 0 for no debugging diagnostics. Set DEBUG = 1,2, or 3

for larger amounts of diagnostic messages. Use DEBUG = 3 only if
you have a lot of disk space (and if your system administrator has a
sense of humour).
This flag controls the amount of printed output to intermediate files
for examination of intermediate stages of the fitting process. Output
is directed to separate files, with the first column designated as
frequency, followed by 8 data columns (E11.3 format) per file, over
as many files as needed. Columns are arranged as real and
imaginary parts, zero and positive sequence, upper triangular
elements only. For example, Frequency, Yzero,real(1,1),
Yzero,imag(1,1), Ypos,real(1,1), Ypos,imag(1,1), Yzero,real(2,1),
Yzero,imag(2,1), Ypos,real(2,1), Ypos,imag(2,1), will be stored in
fort70.out; Frequency, Yzero,real(2,2), Yzero,imag(2,2), Ypos,real(2,2),
Ypos,imag(2,2), Yzero,real(3,1), Yzero,imag(3,1), Ypos,real(3,1),
Ypos,imag(3,1), will be stored in fort71.out; and so forth.
fort70.out, fort71.out,...
Measured function in modal quantities.
Response from power frequency model, in

modal quantities.
Difference function between the measured

data and the power frequency model in
modal quantities.
Controls plotting of the input and fitted functions.
(26-30)
<0
Generate hardcopy plotted output only to Postscript file

fdbfit.ps.
=0
Do not generate plotted output.
>0
In addition to hardcopy plots, also plot to the screen on those

systems where screen plotting is supported (DOS version
only). The larger the absolute value of IPLOT the more
intermediate plots will be displayed.
6 - 14

If SYNC = +1, the program will synchronize the frequency scale of
all the files containing data for the elements of [Y], using the file for
element y1,1 as reference. The program will generate a new
synchronized file with the character "s" added to the end of the
filename. For example, the synchronized version of datay11.dat is
datay11s.dat. SYNC = -1 assumes that all files contain a consistent
frequency scale. Default SYNC = -1.
SYNC
(31-35)
If STOPS = +1, the program will stop execution after synchronizing

data files if SYNC = +1. This gives the user the opportunity to edit
fdbfit.par with the new filenames generated after synchronization. If
STOPS = -1, the program will not stop if data filtering takes place.
Default STOPS = -1.
ISTOPS
(36-40)
If IFILTR = +1, the program does some "spike" filtering to remove

absurd data from the containing data for the elements of [Y]. This is
not a "true" digital filter. The program will generate a new filtered
file with the character "f" added to the end of the filename. For
example the filtered version of datay11.dat is datay11f.dat. IFILTR =
-1 will produce no data filtering. Default IFILTR = -1.\
IFILTR
(41-45)
If STOPF = +1, the program will stop execution after filtering data
files if IFILTR = +1. This gives the user the opportunity to edit
fdbfit.par with the new filenames generated after cleanup. If
STOPF = -1, the program will not stop if data filtering takes place.
Default STOPF = -1.
ISTOPF
(36-40)
6.4.2
Control Card 2
Control card 2:
1
2
3
4
5
6
7
8
12345 67890 12345 67890 12345 67890 1234567890 1234567890 1234567890 1234567890 1234567890
IREAD
I5
IUSER MAXIT NPDEC ILORE
I5
I5
I5
I5
ISCRN
I5
EPSLN
FMIN
XOPT
F60HZ
E10.0
E10.0
E10.0
E10.0
6 - 15
IREAD
(1-5)
IUSER
(6-10)
MAXIT
(11-15)
NPDEC
(16-20)
ILORE
(21-25)
ISCRN
(26-30)
EPSLN
(31-40)
FMIN
(41-50)
Time-saving option. If IREAD = -1, files containing the elements of

[Y] will be processed, and if ILORE = +1, the low frequency response
of the power frequency model will be subtracted from the raw data.
The processed information is stored in intermediate files
FDBFIT1.TMP, FDBFIT2.TMP, and FDBFIT3.TMP. If IREAD = +1, the
program bypasses the reading and processing of the raw data and
assumes that the processed data is already stored in FDBFIT*.TMP.
Note that IREAD must be set to zero at least once. Default
IREAD = -1.
Set IUSER = +1 to interactively fit every element of [Y]. If
IUSER = -1, the program fits all functions without prompting the
user. This option can be useful to adjust the user-definable
parameters during the early stages of the model generation process.
Default IUSER = -1.
MAXIT is proportional to the number of iterations allowed in the
optimization of the elements of [Y]. Default MAXIT = 10.
Number of points per decade used to create intermediate points for

the assignment of real poles and zeroes. Default NPDEC = 50.
Set ILORE = +1 to subtract the frequency response of the power
frequency model prior to fitting. Set ILORE = -1 to fit the raw data
function without first subtracting the response of the power
frequency model. Default ILORE = +1.
Set ISCRN = +1 to enable full-screen displays. In the present release
of FDBFIT, this feature is only available in the DOS version of the
program. Option ISCRN = +1, assumes that the program has been
compiled and linked with the appropriate DOS graphics libraries for
interactive screen plotting. Default ISCRN = -1.
This factor is proportional to the magnitude of disturbance applied to
the independent variables in the optimization process. Default
EPSLN = 0.01.
Minimum frequency at which raw data is assumed to be valid.
When raw data is obtained from measurements, the first few
measurements may be unreliable or inconsistent with the frequency
response of the power frequency model. This parameter gives the
option to skip these points. If Fmin 0, then all raw data is used in
the fitting process. Default FMIN = 0.
(51-60)
Same meaning as XOPT in the EMTP, and it only applies to the Z (or
Linverse matrices) and the magnetizing branches for the power
frequency model. Default XOPT = 0.
F60HZ
Power frequency in Hz (normally 50 or 60 Hz).
XOPT
(61-70)
6 - 16
6.4.3
Fit Control Cards
A total of N fit control cards follow the second control card, where N = NWIND*(NWIND+1)
Fit control card:

1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 12345 67890 12345 67890 12345678901234567890
FCUT
RJECT
FLAST
TOLTR
E10.0
E10.0
E10.0
E10.0
FCUT(k)
(1-10)
RJECT(k)
(11-20)
FLAST(k)
(21-30)
TOLTER(k)
(31-40)
NPOLR(k)
(41-45)
ADDRP(k)
(46-50)
NPR
I5
ADDPR
I5
OPTR CHOOS
I5
I5
This is the frequency that marks the beginning of the region to be

fitted with complex conjugate poles of the kth admittance function
to be approximated. It is generally chosen automatically by the
program, and it often coincides with the first as the first minimum
of the real part of yij. The user can override the automatic
calculation of fcut by entering a non-blank value in this field.
Tolerance used in the rejection criteria of measurement noise of the
kth function to be approximated. There are instances where RJECT
can be used to eliminate small spurious peaks that are probably not
legitimate resonance peaks. Default RJECT = 0.03.
Maximum frequency to be considered in the fitting process of the
kth admittance function to be approximated. As a default
(FLAST = 0 or blank) it is assumed to be the frequency of the last
measured data point.
Tolerance for one of the termination criteria of the optimization
process of the kth admittance function to be approximated.
TOLTER is the quotient between the least squares errors calculated
between consecutive iteration steps. A larger value of TOLTER
(e.g., 1.015) will speed up the fitting process substantially. Default
TOLTER = 1.001.
Number of real poles to be added to the kth fitted function. These
real poles are in addition to the real pole determined by the RC
branch. Default NPOLR = 4.
Flag to control the addition of real poles to the of the kth fitted
function. Set ADDRP -1 if no real poles are to be added. Set ADDPR
= +1 to add real poles in addition to the complex conjugate ones.
Default ADDRP = -1.
6 - 17

OPTR(k)
Set OPTR = -1 to avoid optimizing real poles of the kth fitted

function. Set OPTR = +1 to optimize the function including real
poles. Default OPTR = +1.
(51-55)
CHOOSE(k)
(56-60)
6.4.4
Valid only if ADDRP = +1. If CHOOSE = +1, FDBFIT will optimize

the kth fitted function with added real poles, but it will only retain
the solution with the lowest least squares error. The solution
retained may not even include real poles. If CHOOSE = -1 FDBFIT
will do exactly as the user asks with the different combinations of
control flags. Default CHOOSE = +1.
Node Name Cards
The node names for the terminals of the transformer are specified using one or more Node Name
Cards. Node names should be entered in the following sequence:
W1-PH1, W1-PH2, W1-PH3, W2-PH1, W2-PH2, W2-PH3, W3-PH1, W3-PH2, W3-PH3, etc.
where W* indicates the winding number (high, low, tertiary) and PH* indicates the phase number
(A, B, C).
Node Name Card:

1
2
3
4
5
6
7
8
123456 7890 123456 7890 123456 7890 123456 7890 123456 7890 123456 7890 123456 7890 123456 7890
W1-PH1
W1-PH2
W1-PH3
W2-PH1
W2-PH2
W2-PH3
W3-PH1
W3-PH2
A6
A6
A6
A6
A6
A6
A6
A6
6.4.5
File Name Cards
A total of M file names follow the node name cards, where M = NPHS*NWIND*(NWIND*NPHS+1).
Each file name corresponds to the measured data of one element of [Y]. The following rules
apply:
1.
Only enter the file name, not its extension. For example, if file "file1.dat"
contains the desired data, only enter "file1". File extension for measurement data
files is always assumed to be ".dat".
2.
File names must be entered in the same order as the elements of Y they represent,
by columns, and assuming that Y is upper triangular. For example, for a 4x4 Y
matrix
6 - 18

file-Y11
file-Y12
file-Y22
file-Y13
file-Y23
file-Y33
file-Y14
file-Y24
file-Y34
file-Y44
3.
The same file name may be used for more that one matrix element. For
example, the following file list would represent a 4x4 balanced matrix
file-YS
file-YM
file-YS
file-YM
file-YM
file-YS
file-YM
file-YM
file-YM
file-YS
4.
Comment cards are allowed inside the measurement data file.
5.
Data is read in free format. Expected data structure is the following:

Frequency
Real{Yi,j}
Imaginary{Yi,j}
For example, the file corresponding to element Y1,1 might look as shown below:
C FREQUENCY RESPONSE TEST ON TS 215.5/44KV T8 TRANSFORMER
C TEST ON TERMINAL H1H1
C
Frequency
Real{Yi,j}
Imaginary{Yi,j}
0.4000000000E+03
0.2278124593E-03
-0.1620969735E-02
0.8000000000E+03
0.7271661889E-04
-0.8228833321E-03
0.1200000000E+04
0.4186973456E-04
-0.5416742060E-03
0.1600000000E+04
0.3082307740E-04
-0.3978578316E-03
0.2000000000E+04
0.2523938929E-04
-0.3123318602E-03
0.2400000000E+04
0.2202727774E-04
-0.2522802097E-03
0.2800000000E+04
0.1981685637E-04
-0.2065615990E-03
6 - 19
6.4.6
Power Frequency Transformer Cards
When the HFT model is modelled as an add-on module of the power frequency transformer model
(default option), it is necessary to subtract the frequency response of the power frequency model
(e.g., BCTRAN, TRELEG, or TOPMAG) from the measured data.
The EMTP data cards that describe the low frequency model are included after the filename cards,
exactly as they would be included in a standard EMTP data case. Please note that these cards must
have the same format as the punch file of the AUX run used to generate the power frequency model
(without the vintage cards). The power frequency transformer cards must be terminated with a
blank card.
For example, in the case of TRELEG, the first few data cards would look as shown below:
51,H1-A
52,X1-A
,
,
,,,
,,,
53,T1-A
,T1-B
,,,
0.563200000000E+00,
0.164891784765E-01,
0.268000000000E-01,
0.207556476682E-02,
0.360838777035E+06
0.736626325226E+05
0.150410276706E+05
0.400149116732E+05
,,,,,
$
,,,,,
$
On the other hand, if a BCTRAN or TOPMAG model is used, the first few cards would require the
"USE RB" control card, as shown below
USE RB
51,H1-A
52,X1-A
,
,
,,,
,,,
53,T1-A
,T1-B
,,,
0.563200000000E+00,
0.164891784765E-01,
0.268000000000E-01,
0.207556476682E-02,
0.360838777035E+06
0.736626325226E+05
0.150410276706E+05
0.400149116732E+05
,,,,,
$
,,,,,
$
These "power frequency transformer" data cards will be copied directly to the punch file in
addition to the FDB branches generated for the HFT model.
6.4.7
Linear Magnetizing Branch Cards
These represent the optional magnetizing branch cards that are often added to the power frequency
transformer model. These cards have the same format as standard EMTP series RLC branches.
They must be terminated by a blank card. For example,
C Magnetizing branches follow
C
1
2
3
4
C 345678901234567890123456789012345678901234567890
T1-A
4595.5
T1-B
4295.5
T1-A T1-B
4295.5
BLANK card terminating magnetizing and other branches
6 - 20

Note that the data cards for the power frequency transformer model and for the magnetizing
branches are copied into the punch file along with the FDB branches of the fitted HFT model.
6.5 Example
The following "fdbfit.par" file represents an HV wye-wye connected three-phase transformer, with
buried (terminals not accessible) tertiary winding. The power frequency model used is a standard
BCTRAN transformer model. The magnetizing branch represents 50% of the magnetizing losses.
The other 50% is assumed to be represented using a hysteresis model (not shown).
*
* YY connected transformer with buried delta.
*
I
I
I
I
*
N
D
P
I
S
S
F
S
* N
W
I
E
R
P
I
T
I
T
* P
I
O
B
I
L
N
O
L
O
* H
N
P
U
N
O
C
P
T
P
*<-S><-D-><-T-><-G-><-T-><-T-><-H-><-S-><-R-><-F->
3
2
0
0
0
+3
0
0
0
0
*
* I
I
M
N
I
I
* R
U
A
P
L
S
* E
S
X
D
O
S
* A
E
I
E
R
C
*<-D><-R-><-T-><-C-><-E-><-R-><-EPSLN--><--FMIN--><--XOPT--><--F60HZ->
0
1
10
50
1
0
1.e-2
0000.0
60.
60.
*
*
c
*
n
a
o
h
*
p
d
p
o
*
o
d
t
o
*
l
r
r
s
*-FCUT(I)><RJECT(I)><FLAST(I)><-TOLTER-><-r-><-p-><-p-><-e->
0.0
0.0
0.0
1.015
4
2
1
1
y11,zero
0.0
0.0
0.0
1.015
4
+2
1
1
y11,pos
0.0
0.0
0.0
1.015
4
2
1
1
y12,zero
0.0
0.0
0.0
1.015
4
2
1
1
y12,pos
0.0
0.0
0.0
1.015
4
2
1
1
y22,zero
0.0
0.0
0.0
1.015
4
2
1
1
y22,pos
*
*
*<BUS>
<-BUS>
<-BUS>
<-BUS>
<-BUS>
<-BUS>
<-BUS>
<-BUS>
H1-A
H1-B
H1-C
X1-A
X1-B
X1-C
*
* Data files follow, stored as an upper triangular matrix
* (by columns; i.e., [1,1], [1,2], [2,2], [1,3], [2,3], ...)
*
cher04
cher12f
cher05
cher13
cher23s
cher06
6 - 21

cher15f
cher21f
cher17s
cher11s
cher16f
cher20f
cher16f
cher28
cher10
cher17s
cher16f
cher15f
cher29
cher32s
cher09
*
*
USE RB
51,H1-A
52,X1-A
,
,
,,,
,,,
53,T1-A
,T1-B
,,,
54,H1-B
,,,
55,X1-B
,,,
56,T1-B
,,,
57,H1-C
,,,
58,X1-C
,,,
59,
,T1-A
,,,
0.563200000000E+00,
0.164891784765E-01,
0.268000000000E-01,
0.207556476682E-02,
-0.646357321451E-02,
0.690000000000E-02,
0.000000000000E+00,
0.450729231370E-01,
0.166576311163E-01,
0.563200000000E+00,
0.450729231370E-01,
0.000000000000E+00,
-0.333695525588E-02,
0.164891784765E-01,
0.268000000000E-01,
0.166576311163E-01,
-0.333695525588E-02,
0.000000000000E+00,
0.207556476682E-02,
-0.646357321451E-02,
0.690000000000E-02,
0.000000000000E+00,
0.450729231370E-01,
0.166576311163E-01,
0.000000000000E+00,
0.450729231370E-01,
0.166576311163E-01,
0.563200000000E+00,
0.450729231370E-01,
0.000000000000E+00,
-0.333695525588E-02,
0.450729231370E-01,
0.000000000000E+00,
-0.333695525588E-02,
0.164891784765E-01,
0.268000000000E-01,
0.166576311163E-01,
-0.333695525588E-02,
0.000000000000E+00,
0.166576311163E-01,
-0.333695525588E-02,
0.000000000000E+00,
6 - 22
0.360838777035E+06
0.736626325226E+05
0.150410276706E+05
0.400149116732E+05
0.817057471861E+04
0.443869934400E+04
-0.180109786766E+06
-0.367680781658E+05
-0.199731463549E+05
0.360838777035E+06
-0.367680781658E+05
-0.750592977487E+04
-0.407734858919E+04
0.736626325226E+05
0.150410276706E+05
-0.199731463549E+05
-0.407734858919E+04
-0.221489337600E+04
0.400149116732E+05
0.817057471861E+04
0.443869934400E+04
-0.180109786766E+06
-0.367680781658E+05
-0.199731463549E+05
-0.180109786766E+06
-0.367680781658E+05
-0.199731463549E+05
0.360838777035E+06
-0.367680781658E+05
-0.750592977487E+04
-0.407734858919E+04
-0.367680781658E+05
-0.750592977487E+04
-0.407734858919E+04
0.736626325226E+05
0.150410276706E+05
-0.199731463549E+05
-0.407734858919E+04
-0.221489337600E+04
-0.199731463549E+05
-0.407734858919E+04
-0.221489337600E+04
,,,,,
$
,,,,,
$
$
,,,,,
$
$
,,,,,
,,,,,
$
$
,,,,,
$
,,,,,
$
$
,,,,,
$
$
,,,,,
$
$
,,,,,
$
$
,,,,,
,,,,,
$
$
,,,,,
$
$
,,,,,
$
,,,,,
$
$
,,,,,
$
$
,,,,,

0.207556476682E-02,
-0.646357321451E-02,
0.690000000000E-02,
*
* Comments are allowed...
blank card terminating transformer data
*
* Magnetizing branches follow
*
1
2
3
4
*2345678901234567890123456789012345678901234567890
T1-A
4595.5
T1-B
4295.5
T1-A T1-B
4295.5
blank card terminating magnetizing and other branches
6 - 23
0.400149116732E+05 $
0.817057471861E+04 $
0.443869934400E+04 ,,,,,
Section 7
7.1

Support Module CONVERT ............................................................... 7-1
7.1.1
Assumptions ................................................................... 7-1
7.1.2
Input Data Format for CONVERT ................................. 7-2
7.1.3
Alternate Input of Incremental Inductance vs.

Current Characteristic .................................................... 7-4
7.1.4
7.2
Module HYSDAT ............................................................................... 7-7

7.2.1
7.3
7.4
Example ......................................................................... 7-6

Input Data Format .......................................................... 7-7
Module FITSAT .................................................................................. 7-10

7.3.1
7.3.2
Control Parameter Card ................................................. 7-12
7.3.3
Node Names and Output Control ................................... 7-13
7.3.4
The Saturation Characteristic Input ............................... 7-14
7.3.5
Sample Data and Output ................................................ 7-14
Generation Of ZnO Arrester Parameters ............................................. 7-16

7.4.1
General Remarks About Module ARRDAT .................. 7-16
7.4.2
7.4.3
Examples ........................................................................ 7-22
Section 7
AUX
supports the following magnetic saturation/hysteresis routines:

(A)
Subroutine CONVERT which is designed to convert RMS v-i saturation curves

into peak -i curves with the hysteresis loop being ignored.
(B)
Subroutine HYSDAT which is designed to provide the hysteresis loop required

by the Type-96 pseudo-nonlinear hysteretic reactor.
(C)
Subroutine FITSAT which is designed to provide the hysteresis data required by

the Type-92 nonlinear hysteresis model.
7.1 Support Module CONVERT

Data specification for nonlinear reactors requires an instantaneous flux-current -i saturation
curve as described in Section 8 of RuleBook 1. However, typical test data consists of RMS voltage
and current readings. The support module CONVERT is designed to convert from an RMS v-i curve
to the -i curve needed by the EMTP.
7.1.1
Assumptions
The user inputs enters a v-i curve as a series of points (linear interpolation between these values is
assumed). On output, a piecewise-linear -i curve is produced with the same number of points as
the input curve. The resulting curve is exact, except for the following approximations:
1.
The use of linear interpolation between points.
2.
The use of a finite-difference approximation for the sinusoidal excitation. Onedegree step-size is used, along with trapezoidal-rule integration where needed.
3.
Hysteresis is ignored.
7-1
RMS
voltage v
RMS
7.1.2
current i
Input Data Format for CONVERT
The structure of the data deck for a CONVERT translation is as follows:

1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
2.
Next comes a "SATURATION" special-request card.
1
2
3
4
5
6
7
8
1234567890 1234567890123456789012345678901234567890123456789012345678901234567890
SATURATION
A10
3.
After the special request card "SATURATION" comes the following miscellaneousdata card, which has the following format:
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 1234567890123456789012345678901234567890
FREQ
VBASE
PBASE
E8.0
E8.0
E8.0
IPUNCH
I8
KTHIRD
I8
7-2
FREQ
Frequency of the sinusoidal excitation, in Hz.
(1-8)
VBASE
Base voltage in kV, on which the input points are based.
(9-16)
PBASE
Base power in MVA, on which the input points are based.
(17-24)
IPUNCH
(25-32)
KTHIRD
(33-40)
4.
Parameter controlling the punched-card output of the derived flux

current curve:
=0
Yes, the curve will be punched.
=1
No punched output.
Zero or blank field will produce only first-quadrant output;

KTHIRD = 1 will produce both the third and first quadrant points (the
full curve).
After the miscellaneous data card come the (IRMS, VRMS) break-points, followed
by a 9999-card.
9999 card
IRMS - VRMS data cards
The (IRMS, VRMS) points start with the point closest to the origin and then moving
continuously away. The curve must be single-valued, and not over 100 points.
No point (0, 0) for the origin should be entered. Values are in per unit on the
previously-specified base, where
Ibase = Pbase / Vbase
(7.1)
IRMS [p.u.] = IRMS [Amps] / Ibase [Amps]
(7.2)
VRMS [p.u.] = VRMS [kV] / Vbase [kV]
(7.3)
7-3
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
IRMS[p.u.]
vRMS[p.u.]
E16.0
E16.0
Terminate with a 9999-card (columns 13-16).
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
9999
5.
To indicate the end of all AUX data cases add a blank card, followed by a "BEGIN
NEW DATA CASE" card, followed by a blank card, as described in Section 4.1.1.
blank card
blank card
7.1.3
Alternate Input of Incremental Inductance vs. Current Characteristic
Instead of an RMS I-V characteristic, there is an option to enter the characteristic as a function of
current vs. incremental inductance. The trapezoidal rule of integration is then used, to convert the
inductance curve into the desired output curve of flux vs. current:
(i) =
0 L ( ) d
(7.4)
1 = 0
Lk + Lk 1
- ( ik ik 1 )
k = k 1 + -----------------------2
k = 2,3,
7-4
(7.5)
incremental inductance
L1 = L2
L3
L4
L5
i1
i2
i3
i4
i5
In order to use this special inductance-input option, the following modification of preceding rules
must be observed:
1.
Set the field FREQ = -1.0 in columns 1-8 of the miscellaneous data card. This is
the special request flag for inductance-input.
2.
Leave fields "VBASE" (columns 9-16) and "PBASE" (columns 17-24) blank or
zero if the supplied characteristic will be in units of amps and henries. Nonzero
values are taken to be scaling factors by which the breakpoint values are to be
multiplied in order to produce amps and henries:
3.
VBASE
= multiplying factor for input current coordinates;
PBASE
= multiplying factor for input inductance coordinates.
The current vs. incremental inductance characteristic is to be entered in columns

1-32 as per the following format, terminated by a 9999-card (columns 13-16):
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
ik
Lk
E16.0
E16.0
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
9999
7-5
7.1.4
4.
The current breakpoints ik must be monotonically increasing, starting with value

zero. Incremental inductance values d/di at current value i=ik must all be
positive. There also must be two or more points defining the characteristic.
5.
Variable "IPUNCH" controls punched-card output. Note that the (0.0, 0.0) point
of the derived -i curve is printed, but is not entered as input data. This is
because usage with Type-98 pseudo-nonlinear reactors and saturable
transformers requires that the origin-point be omitted.
Example
The data listing below shows an example of two conversion cases. Excerpts from AUX printed
output is also shown below.
BEGIN NEW DATA CASE
C
MODIFIED BENCHMARK DC-13
C
TEST OF MAGNETIC-SATURATION ROUTINE, THE SUPPORTING PROGRAM WHICH
C
CONVERTS AN RMS SATURATION CURVE OF V VS. I TO FLUX VS. CURRENT .
C
THE OPTION OF USING AN INPUT CHARACTERISTIC WHICH GIVES CURRENT
C
VS. INCREMENTAL INDUCTANCE IS ALSO TESTED.
SATURATION
-1.
10.
.001
1
0.0
5.0
2.0
5.0
3.0
3.5
4.0
2.0
5.0
1.0
10.0
1.0
9999
60.
.664
.01
1
1.0
100.
1.6
105.
2.75
110.
4.0
113.
6.0
116.
10.
119.
44.
131.
9999
BLANK CARD ENDING "SATURATION" CASES
BEGIN NEW DATA CASE
BLANK
DERIVED SATURATION CURVE GIVING PEAK CURRENT VS. FLUX
ROW
CURRENT (AMP)
FLUX (VOLT-SEC)
---------------------------------------------------------------------------------------------------REMEMBER.
THE JUST-COMPLETED CONVERSION BEGAN WITH A CURRENT VS. INCREMENTAL INDUCTANCE
CHARACTERISTIC, DUE TO MISCELLANEOUS DATA PARAMETER 'FREQ' OF COLUMNS 1-8 BEING PUNCHED WITH A VALUE OF
-1.0 .
TRAPEZOIDAL RULE INTEGRATION OF THE INDUCTANCE CURVE WAS USED, TO PRODUCE FLUX.
---------------------------------------------------------------------------------------------------1
0.0000000000
0.0000000000
2
20.0000000000
0.1000000000
3
30.0000000000
0.1425000000
4
40.0000000000
0.1700000000
5
50.0000000000
0.1850000000
6
100.0000000000
0.2350000000
9999
1
60.
.664
.01
1
7-6

+MISC. CONST.
0.600E+02 0.664E+00 0.100E-01

1
1
1.0
100.
+(I, V) POINT.
0.10000E+01 0.10000E+03
1
1.6
105.
+(I, V) POINT.
0.16000E+01 0.10500E+03
1
2.75
110.
+(I, V) POINT.
0.27500E+01 0.11000E+03
1
4.0
113.
+(I, V) POINT.
0.40000E+01 0.11300E+03
1
6.0
116.
+(I, V) POINT.
0.60000E+01 0.11600E+03
1
10.
119.
+(I, V) POINT.
0.10000E+02 0.11900E+03
1
44.
131.
+(I, V) POINT.
0.44000E+02 0.13100E+03
1
9999
+SPECIAL TERMINATION-OF-POINTS CARD.
DERIVED SATURATION CURVE GIVING PEAK CURRENT VS. FLUX
ROW
CURRENT (AMP)
FLUX (VOLT-SEC)
1
0.0000000000
0.0000000000
2
21.2983970237
249.0875141368
3
57.2428412031
261.5418898436
4
99.8053477329
273.9962655505
5
154.6164576772
281.4688909746
6
236.9020266047
288.9415163987
7
420.7672719235
296.4141418228
8
1601.6892070492
326.3046435192
9999
CHECK OF DERIVED CURVE BY INDEPENDENT REVERSE COMPUTATION. ASSUMING SINUSOIDAL VOLTAGE (FLUX) AT LEVEL OF EACH
POINT,
RMS CURRENT IS FOUND NUMERICALLY. THIS CURVE SHOULD BE EQUAL TO THE ORIGINAL I-V POINTS INPUTTED.
ROW
CURRENT IN P.U.
VOLTAGE IN P.U.
2
1.00000000
100.00000000
3
1.60000000
105.00000000
4
2.75000000
110.00000000
5
4.00000000
113.00000000
6
6.00000000
116.00000000
7
10.00000000
119.00000000
8
44.00000000
131.00000000
7.2 Module HYSDAT

This module generates data in a format suitable for the Type96 hysteretic inductor (see
Section 8.5 of RuleBook 1). Required input data includes the major loop characteristic specified
in terms of flux and current (i.e., instantaneous -i characteristic).
7.2.1
Input Data Format
The structure of the data deck for a HYSDAT request follows:

1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
7-7

2.
Next comes an "SATURATION" special-request card,
1
2
3
4
5
6
7
8
1234567890 1234567890123456789012345678901234567890123456789012345678901234567890
SATURATION
A10
3.
followed by the flag 88. in columns 6-8.
1
2
3
4
5
6
7
8
12345678 901234567890123456789012345678901234567890123456789012345678901234567890
88.
E8.0
4.
The next two cards then specify the necessary input data
1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012345678901234567890123456789012345678901234567890
ITYPE
LEVEL
IPUNCH
I8
I8
I8
ITYPE
This is used to specify the type of material as follows:
(1-8)
=1
LEVEL
(9-16)
ARMCO
M4 oriented silicon steel.
Level of accuracy desired (1, 2, 3, or 4). The level of accuracy is

specified by the number of points used to define the loop. The
specific number of points in each level may vary from material to
material but the following values are given as a rough idea:
Level
# of Points
1
2
3
4
4-5
10
15
20-25
Unless space or time constraints should prove to be important, level 3

or 4 is suggested.
7-8

Parameter controlling the punched-card output of the derived
fluxcurrent loop:
IPUNCH
(17-24)
=0
Yes, the curve will be punched
=1
No punched output.
The cards are punched one point per card, using a 2E16.7 format.
1
2
3
4
5
6
7
8
12345678 90123456 7890123456789012345678901234567890123456789012345678901234567890
CURSAT
FLXSAT
E8.0
E8.0
CURSAT
(1-8)
FLXSAT
(9-16)
The current coordinate of the positive saturation point for the actual
reactor being specified (i.e., the point in the first quadrant of the
flux-current plane where the hysteresis loop changes from being
multivalued to single-valued).
The flux coordinate of the positive saturation point for the actual
reactor being specified.
Recall that the shape of the hysteresis loop for an inductor depends primarily on
the material of the core, while the scaling of the hysteresis loop depends on
geometry, the number of turns, and other factors of the actual construction.
What is stored in this routine is essentially the shape of the loop for the material
specified. It is necessary for the user to provide the information for the actual
reactor being specified, which will allow the correct scaling to be performed.
This is the purpose of variables CURSAT and FLXSAT.
One suggested way of determining values for CURSAT and FLXSAT from normal
or DC magnetization curves (which seem to be more readily available than
hysteresis loop curves) is as follows: beginning at the right of the normal
magnetization curve, in its linear region, a straight-edge is used to extrapolate
this line back to the left; the point where this straight line and the actual curve
first being the diverge is then taken as the saturation point.
Finally a word of caution. When determining the saturation point, care should
be taken to determine it accurately since any error in CURSAT will cause a
corresponding error in the width of the loop. For example, if the value of
CURSAT was chosen to be a factor of ten too small (which might easily happen if
one failed to notice a change of scale, for instance) the width of the loop at
7-9

flux = 0 would also be a factor of ten too small. This is a fairly drastic error near
zero flux, but quite possibly a relatively small error near the saturation region.
5.
At this point, if it is desired to determine the hysteresis loop for a new hysteretic
reactor, the user should specify new values for ITYPE, LEVEL, IPUNCH, CURSAT,
and FLXSAT. A blank card will terminate hysteresis-curve requests. A second
blank card is necessary to terminate all SATURATION cases, followed by a "BEGIN
NEW DATA CASE" card, followed by a blank card, as described in Section 4.1.1
blank card
BEGIN NEW DATA CASE card
blank card to end SATURATION
blank card to end HYSDAT
CURSAT, FLXSAT card
ITYPE, ILEVEL, IPUNCH card
7.3 Module FITSAT

The module FITSAT is designed to generate a Type-92 hysteretic reactor (see Rulebook 1, Section
8.11) from the measured -i characteristic of the iron core of a transformer. It requires data
normally not available in standard test sheets, but the program provides automatic defaults when
some parameters are not available.
7.3.1
Input Data Format
The structure of the data deck for a FITSAT request follows:

1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
7 - 10

2.
1
2
3
4
5
6
7
8
1234567890 1234567890123456789012345678901234567890123456789012345678901234567890
SATURATION
A10
3.
1
2
3
4
5
6
7
8
12345678 901234567890123456789012345678901234567890123456789012345678901234567890
66.
E8.0
4.
Next is a control parameter card, followed by a node names card, followed by

the -i characteristic.
-i characteristic
node names & output card

parameter card
5.
Two blank cards indicate the end of FITSAT and saturation requests. A BEGIN
NEW DATA CASE followed by a blank cards signals the end of the AUX simulation
blank card
blank card to end FITSAT
7 - 11
7.3.2
Control Parameter Card
The subroutine is controlled by providing a parameter card as follows:

1
2
3
4
5
6
7
8
12345678 90123456 78901234 56789012 34567890 12345678 90123456 78901234 56789012 34567890
ESCALE
MAXIT
IPRNT
E8.0
E8.0
E8.0
ESCALE
(1-8)
MAXIT
(9-16)
IPRNT
(17-24)
TOL
(25-32)
TOL
E8.0
COER
E8.0
Yrev
Xrev
AMrev
Yloop
Xloop
E8.0
E8.0
E8.0
E8.0
E8.0
Controls the convergence rate of the solution during optimization.

Default = 10000.0.
Maximum number of iterations in the optimization or fitting
process. Default = 500.
Flag that controls the amount of output generated during the curve
fitting process. Default = zero or blank.
Tolerance for convergence. Defaults to 1% of the smallest current
value in the user-supplied data for the flux-current characteristic.
(33-40)
Coercive current (in A). If negative, it is set to the first current value
of the user-supplied flux-current characteristic (which must be nonzero). Default = 0.02.
Yrev
(41-48)
The Y-ordinate of the reversal point of a given loop (in Vs). Yrev
should not be zero. Default = 5.0
Xrev
(49-56)
The X-ordinate of the reversal point, normally zero (in A). If Xrex =
0 or blank, then Xrex is calculated by the program assuming that the
reversal point belongs to a major loop. Default = 0.0.
AMrev
(57-64)
Slope at the reversal point {Xrev, Yrev} (in Vs/A). Default = 0.5
Yloop
(65-72)
Y-ordinate of a second point in a trajectory where the reversal point

is defined by {Xrev, Yrev} and AMrev (in Vs). Default = 0.0.
Xloop
(73-80)
X-ordinate of a second point in a trajectory where the reversal point

is defined by {Xrev, Yrev} and AMrev (in A). If Xloop = 0 the
program calculates Xloop and Yloop assuming that the trajectory lies
on the major loop. Xloop is normally set to zero unless sufficient
data about minor loops is available. Default = 0.0.
COER
7 - 12
AMrev
{yloop,yloop}
2nd reversal point
i
{yrev,yrev}
1st reversal point
7.3.3
coercive current
Node Names and Output Control
This consists of one card containing input parameters as follows:

1
2
3
4
5
6
7
8
123456 789012 34567890 123456789012345678901234567890123456789012345678901234567890
NODE1 NODE2
A8
A8
NODE1
IPCTR
I8
The first node name.
(1-6)
NODE2
The second node name.
(7-12)
7 - 13

IPCTR
Control flag for output printing.
(13-20)
0 = request for minimum output.
2 = request for maximum output.
7.3.4
The Saturation Characteristic Input
The input format for the saturation characteristic is shown below:

1
2
3
4
5
6
7
8
1234567890123456 7890123456789012 345678901234567890123456789012345678901234567890
CURSAT
FLXSAT
E16.0
E16.0
(1-16)
The current of the saturation characteristic (in A). First quadrant

only.
FLXSAT
The flux of the saturation characteristic (in Vs).
CURSAT
(17-32)
Note that if COER < 0 in the control parameters card, the first pair CURSAT will be the coercive
current and FLXSAT will be zero. On the other hand, if COER = 0, the first point should be {0,0} and
the saturation curve will be shifted by COER along the x-axis.
7.3.5
Sample Data and Output
A sample input data deck for FITSAT is shown below:

BEGIN NEW DATA CASE
C
1
2
3
4
5
6
7
8
C 45678901234567890123456789012345678901234567890123456789012345678901234567890
SATURATION
66.
10000.
2000
-1
1.D-3
-0.02
5.0
0.0
0.5
0.0
0.0
NNNNNN
2.4462
0.0
16.5646
122.494
24.3791
128.670
35.1231
134.846
44.8207
137.861
61.8104
140.875
7 - 14
BLANK
BLANK
BLANK
BEGIN
97.7506
155.411
213.382
271.353
14734.2
TO END FITSAT
TO END SAT
TO END DATACASE
NEW DATA CASE
143.375
146.169
147.052
147.493
171.609
The output listing is shown as follows:

DONE READING DISK FILE INTO EMTP CACHE.
NUMCRD =
22 CARDS.
SUPPORTING AUXILIARY ROUTINES FOR EMTP - DCG/EPRI VERSION 3.0 UNIX TRANSLATION
DISTRIBUTED BY EPRI SOFTWARE CENTRE. RUN DATE (MM/DD/YY) AND TIME (HH.MM.SS.)= 9/12/94 13:42:11
FOR INFORMATION, CONSULT THE EMTP RULE BOOK.
PROGRAM VERSION="V3.0"
LENGTH OF /LABEL/ EQUALS 281040 INTEGER WORDS, LENGTH OF INPUT DATA FILE EQUALS 2000 CARDS.
0
1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
0
--------------------------------------------------+------------------------------------------------------------------------------1BEGIN NEW DATA CASE
+MARKER CARD PRECEDING NEW DATA CASE.
COMMENT CARD.
1C 45678901234567890123456789012345678901234567890123456789012345678901234567890
COMMENT CARD.
1C
1SATURATION
+REQUEST FOR MAGNETIC-SATURATION ROUTINE.
1
66.
+FREQ=66 REQUESTS FITSAT SUBROUTINE
110000. 2000
-1
1.D-3
-0.02 5.0
0.0
0.5
0.0
0.0
+CONTROL PARAMETERS
1NNNNNN
+NODE 1, NODE 2, PRINT CONTROL
12.4462
0.0
+
0.2446200E+01
0.0000000E+00
116.5646
122.494
+
0.1656460E+02
0.1224940E+03
124.3791
128.670
+
0.2437910E+02
0.1286700E+03
135.1231
134.846
+
0.3512310E+02
0.1348460E+03
144.8207
137.861
+
0.4482070E+02
0.1378610E+03
161.8104
140.875
+
0.6181040E+02
0.1408750E+03
197.7506
143.375
+
0.9775060E+02
0.1433750E+03
1155.411
146.169
+
0.1554110E+03
0.1461690E+03
1213.382
147.052
+
0.2133820E+03
0.1470520E+03
1271.353
147.493
+
0.2713530E+03
0.1474930E+03
114734.2
171.609
+
0.1473420E+05
0.1716090E+03
1BLANK TO END FITSAT
COMPARISON OF INPUT DATA AND RE-CALCULATED VALUES

INPUT CURRENT
INPUT FLUX
CALCULATED FLUX
0.0000000E+00
0.0000000E+00
0.0000000E+00
0.1411840E+02
0.1224940E+03
0.8337700E+02
0.2193290E+02
0.1286700E+03
0.1295290E+03
0.3267690E+02
0.1348460E+03
0.1468318E+03
0.4237450E+02
0.1378610E+03
0.1468471E+03
0.5936420E+02
0.1408750E+03
0.1468754E+03
0.9530440E+02
0.1433750E+03
0.1469357E+03
0.1529648E+03
0.1461690E+03
0.1470328E+03
0.2109358E+03
0.1470520E+03
0.1471305E+03
0.2689068E+03
0.1474930E+03
0.1472281E+03
0.1473175E+05
0.1716090E+03
0.1716014E+03
PUNCH FILE OUTPUT TO LU7. COPY FOLLOWS
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 1234567890
C
C
8888.
--->
--->
7 - 15

C ----Vref
--><------ Remflx ----><------ Eps1 ------><------ Eps2 ------>
92NNNNNN
8888.
8.67559443309
-2.75429672548
81.4277245567
2.44620000000
0.680703053079
0.194250455170E-03 -0.120326121546
146.774908703
500000.000000
1BLANK TO END SAT
+BLANK CARD TERMINATING ALL SATURATION CASES.
CORE STORAGE FIGURES FOR PRECEDING DATA CASE NOW COMPLETED. --------------------------------------A VALUE OF -9999 INDICATES DEFAULT WITH NO FIGURE AVAILABLE.
SIZE LIST 1.
TOTAL STORAGE SPACE ALLOCATED FOR THE EMTP AUXILIARY PROGRAMS
TIMING FIGURES (DECIMAL) CHARACTERIZING CASE SOLUTION SPEED. -------------------------------------
PRESENT
FIGURE
CP SEC
PROGRAM
LIMIT (NAME)
281040(LTLABL)
I/O SEC
SUM SEC
7.4 Generation Of ZnO Arrester Parameters

Module ARRDAT is used to generate the parameters needed for the Type-92 of ZnO from the v-i
characteristic of the arrester.
This module uses the following form of the exponential function describing the protective
characteristic of a ZnO arrester:
I = k( v / vref )m
(7.6)
where Vref is the reference voltage used to prevent the numerical overflow of the exponential terms
for large voltage values. The arrester v-i characteristic is then fitted by several exponential
functions, each valid in a particular range. The exponential segments are fitted on a log-log plane
(straight line segments) to avoid numerical ill-conditioning of the exponential fitting.
7.4.1
General Remarks About Module ARRDAT
This subroutine provides for a fit of exponential curves to a set of data points. The fitting is
performed in the log-log plane using the least squares approach. There are two options for the
determination of number of exponential segments to be fitted:
1.
The user specifies the number and the boundaries of the segments.
2.
The program determines the number of segments depending on the maximum

permissible relative error specified by the user.
is designed to be used with the multi-phase ZnO modelling of Section 8.9 of the
RuleBook 1. However, it can also be applied to SiC arresters (see Section 8.10 of RuleBook 1).
ARRDAT
7.4.2
Input Data Format
The structure of the data deck for an ARRDAT request follows:
7 - 16

1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
A19
2.
1
2
3
4
5
6
7
8
1234567890 1234567890123456789012345678901234567890123456789012345678901234567890
SATURATION
A10
3.
1
2
3
4
5
6
7
8
12345678 901234567890123456789012345678901234567890123456789012345678901234567890
77.
E8.0
4.
Next are a set of parameter cards followed by arrester characteristics, terminated

by a blank card.
Blank card
v-i characteristic
parameter cards
5.
Two blank cards indicate the end of ARRDAT and saturation requests. A BEGIN
NEW DATA CASE followed by a blank cards signals the end of the AUX simulation
7 - 17
blank card
blank card to end ARRDAT
Parameter Cards
Optional "BRANCH" Card
1
2
3
4
5
6
7
8
123456 78 901234 567890 123456 789012 345678 901234 567890 123456 789012 345678 901234 567890
PHASE 1
BRANCH
BUS 1
A6
A6
PHASE 2
BUS 2 BUS 1
A6
A6
PHASE 3
BUS 2
BUS 1
A6
A6
PHASE 4
BUS 2 BUS 1
A6
A6
PHASE 5
BUS 2 BUS 1
A6
A6
PHASE 6
BUS 2 BUS 1
A6
BUS 2
A6
A6
This data card contains the keyword "BRANCH" in columns 1-6. The card is optional (numerical
values generated within ARRDAT are unaffected by its presence or absence). The "BRANCH" card
provides the option to specify pairs of 6-character node names which will be reproduced in
columns 3-14 (fields "BUS1" and "BUS2", each A6 format) of the EMTP branch cards which will be
produced. Without such a "BRANCH" card, columns 3-14 of the branch cards which are to be
punched as output will be blank.
Parameter Card #1
1
2
3
4
5
6
7
8
123456789012 345678901234 567890123456 789012345678 901234567890 123456789012 34567890
NEXP
IPHASE
ERRLIM
IPRZNO
VREF
VFLASH
I 12
I 12
E12.0
I 12
E12.0
E12.0
NEXP
Flag indicating which of the two fitting options is requested.
(1-12)
<0
Request for automatic determination of the number of

segments.
7 - 18

=N
IPHASE
(13-24)
ERRLIM
(25-36)
Request for user-supplied number of segments as well as their

boundaries. The positive integer N defines the number of
segments to be fitted.
Number of phases of the multiphase compensation model to be used in

the EMTP. This number affects only the printed and punched output of
the program. If left unspecified, a default value of "1" will be assigned
automatically by the program.
Maximum relative error permitted during the automatic determination
of the number of segments (NEXP < 0). The default value of ERRLIM is
0.05, and it is defined as:
C1 C2
ERROR = ----------------------C1
(7.7)
Where C1 is value of the current as specified by the user, and C2 is

value of the current as calculated by the program using the fitted
characteristic.
IPRZNO
(37-48)
VREF
(4960)
VFLASH
(6172)
Control of diagnostic printout. Suggested values 0 to 3. Amount of

printout increases with the assigned values.
Reference voltage used to scale the voltage data points. The scaling
is used to prevent a possible overflow during the simulation of the
EMTP. The default value of VREF is taken to be twice of A2 on the
card #2 below.
Gap flashover (sparkover) voltage in Volts (crest). Leave this field
blank or punch with zero if the arrester is gapless.
Parameter Card #2
1
2
3
4
5
6
7
8
123456789012 345678901234 567890123456 789012345678 901234567890 123456789012 34567890
A1
A2
A3
A4
A5
AMIN
E12.0
E12.0
E12.0
E12.0
E12.0
E12.0
A1
(1-12)
Voltage rating of the arrester upon which the input data are based in
volts RMS.
7 - 19

A2
(13-24)
The desired voltage rating of the arrester in volts RMS. This variable
can be used to:
1)
convert the arrester input data (voltages) from p.u. to V

(volts);
2)
obtain the characteristic of an electrically similar arrester

with a different voltage rating. The properties of similar
arresters do not change with rating, providing a
proportional number of blocks is used to obtain the new
rating.
3)
if VREF is maintained proportional to A2, similar arresters

will have identical parameters at all ratings.
A3
(25-36)
Multiplier used in the additional scaling of voltage points. It can be,

for example, used to obtain a minimum characteristic (maximum
energy) from a maximum characteristic (maximum voltage). The
value to be normally used is 1.0.
A4
(37-48)
Current multiplier used in obtaining characteristics of arresters with

number of columns different from that of which the data points are
known. An even current distribution is assumed.
A5
(49-60)
Flag signalling the existence (or its lack) of additional data

describing the arrester after gap sparkover. This parameter permits
the representation of:
AMIN
(61-72)
1)
arresters with shunt passive gaps. The shunted portion of

the arrester can be of similar or dissimilar material;
2)
arresters with series passive gap. Before gap sparkover,

the arrester is inoperative. Therefore, this data specifies
the actual nonlinear material characteristic.
= 0.0
or 1.0
indicates a gapless arrester (no data follows);
< 0.0
indicates that additional data follows, to represent

dissimilar material for a shunt or a series gap;
> 0.0
but
= 1.0
indicates that there is no further data. Instead the original

input data are to be used for multiplication of the voltages
by A5.
Minimum value of current above which segment #1 of the arrester

characteristic begins. Even if fitting a straight line passing through
the origin, do not leave AMIN blank (zero), log (0.0) is not defined.
7 - 20
Cards Specifying Arrester Characteristic

These cards come in NEXP (see card #1) groups, each terminated by a blank card. For the option of
automatic determination of the number of segments (NEXP < 0), there is only one group of data.
1
2
3
4
5
6
7
8
123456789012 345678901234 56789012345678901234567890123456789012345678901234567890
A
E12.0
B
E12.0
A
(112)
Value of current (on the arrester characteristic)
B
(1324)
Value of the voltage (on the arrester characteristic)
The following two groups of data cards are present if and only if the parameter A5 on data
card #2 was less than zero (i.e., A5 < 0.), and VFLASH on card #1 was greater than zero.
Card #i
1
2
3
4
5
6
7
8
123456789012 34567890123456789012345678901234567890123456789012345678901234567890
NEXP
E12.0
NEXP
(1-12)
Number of exponential segments to be fitted. The user can, at this

point, either preserve or change the fitting option, i.e., it is possible
to fit the first part with NEXP < 0, and the second (after gap
sparkover) part with NEXP > 0 or vice versa.
All the other constants are assumed to be as those specified on cards #1 and #2. This card will be
followed by a set of cards to specify the additional arrester characteristic with the following
format:
7 - 21
1
2
3
4
5
6
7
8
123456789012 345678901234 56789012345678901234567890123456789012345678901234567890
A
E12.0
7.4.3
E12.0
A
(112)
Value of current (on the arrester characteristic)
B
(1324)
Value of the voltage (on the arrester characteristic)
Examples
Consider a single-column ZnO arrester. The voltage points are known for an arrester rated 1 kV
(crest), maximum voltage characteristic. The current points are given in A (crest). It is desired to
obtain parameters for a 192 kV (RMS), 3column arrester with a maximum energy characteristic.
The conversion factor to this characteristic is assumed to be .962. The arrester is equipped with a
shunt gap and the shunted part is electrically similar to the rest of the arrester. The shunt gap
which sparks over at 380000 V (crest) consists of 12% additional blocks, i.e., after gap sparkover
the arrester contains (1.0/1.12) = .89286 of the original blocks, and therefore A5 = 0.89286. A
reference voltage 412500 V was chosen for the expected operating range of the 3-phase arrester.
Double Exponent Fit (NEXP=2)

The data described above will be used to create two stacked ARRDAT calculations. The first one
with a 2-exponential fit with the following values for variables A1A5:
A1 = 1000./SQRT(2)
the original rating of the arrester in V (RMS),
A2 = 192000.
the required rating in V (RMS),
A3 = .962
scaling factor for obtaining the desired maximum

energy characteristic.
A4 = 3.0
the original data are given for a single-column arrester.

Factor of 3.0 is needed to obtain the characteristic of a
3column arrester.
A5 = 0.89286
the arrester, before sparkover, contains 12% additional

blocks
A standard choice of AMIN is .001 A, i.e., somewhere at the end of the leakage current region.
7 - 22

The characteristic described above is to be fitted with two exponentials. The border between the
two segments has been determined graphically on a log-log paper to lie between 100 A and 200 A.
Automatic Segment Determination (NEXP = -1)

In the second example, the same data will now be fitted with relative current error of 0.05, i.e.,
ERRLIM = 0.05.
Sample AUX Input Data, Punched, and Printed Options:

Input listing for the example described above is shown here:
BEGIN NEW DATA CASE
C
BENCHMARK DC-39
C
TEST OF SUPPORTING PROGRAM "ARRDAT" WHICH DERIVES ZINC OXIDE
C
ARRESTER CHARACTERISTIC CARDS BY LEAST MEAN SQUARE FITTING.
SATURATION
77.
BRANCH RECA
RECB
RECC
1
3
2
412500.
380000.
707.107
192000.
.962
3.0
.892857
.001
1.0
1164.8
2.0
1181.6
5.0
1198.4
10.
1209.6
20.
1232.0
50.
1260.0
100.
1288.0
200.
1323.84
500.
1388.8
1000.
1442.56
2000.
1512.0
3000.
1556.8
BLANK CARD ENDING CHARACTERISTIC
BRANCH RECA
RECB
RECC
2
3
1
412500.
380000.
707.107
192000.
.962
3.0
.892857
.001
1.0
1164.8
2.0
1181.6
5.0
1198.4
10.
1209.6
20.
1232.0
50.
1260.0
100.
1288.0
200.
1323.84
500.
1388.8
1000.
1442.56
2000.
1512.0
3000.
1556.8
C
SHOWN BELOW IS AN EXAMPLE OF AUTOMATIC SEGMENT SELCTION USING
C
AN UPGRADED FITTING PROCEDURE AVAILABLE AS OF "M38" VRESION.
BRANCH RECA
RECB
RECC
7 - 23

-1
3
.05
707.107
192000.
.962
1.0
1164.8
2.0
1181.6
5.0
1198.4
10.
1209.6
20.
1232.0
50.
1260.0
100.
1288.0
200.
1323.84
500.
1388.8
1000.
1442.56
2000.
1512.0
3000.
1556.8
BLANK CARD ENDING CHARACTERISTICLKLK
BLANK CARD ENDING "ARRDAT" CASES
BLANK CARD ENDING "SATURATION" CASES
BEGIN NEW DATA CASE
BLANK
1
3.0
412500.
.892857
380000.
.001
Excerpts of the corresponding output file are show below:

As for output, it begins with the familiar interpreted input data cards, through the blank card which
ends the second piece of the characteristic at 3000 amperes. Then internal printout continues, and
finally, there is printout of the card images which should be written to the punch (LUNIT7 file) -cards which are to be used in some subsequent simulation as data belonging to a multi-phase ZnO
arrester. After this, data input begins for the second example.
DONE READING DISK FILE INTO EMTP CACHE.
NUMCRD =
61 CARDS.
SUPPORTING AUXILIARY ROUTINES FOR EMTP - DCG/EPRI VERSION 3.0 DOS WINDOWS TRANSLATION
DISTRIBUTED BY EPRI SOFTWARE CENTRE. RUN DATE (MM/DD/YY) AND TIME (HH.MM.SS.)= 5/ 2/97 13. 7.51
PROGRAM VERSION="V3.0e"
0
1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
0
--------------------------------------------------+------------------------------------------------------------------------------1BEGIN NEW DATA CASE
COMMENT CARD.
COMMENT CARD.
COMMENT CARD.
1C
BENCHMARK DC-39
1C
TEST OF SUPPORTING PROGRAM "ARRDAT" WHICH DERIVES ZINC OXIDE
1C
ARRESTER CHARACTERISTIC CARDS BY LEAST MEAN SQUARE FITTING.
1SATURATION
+REQUEST FOR MAGNETIC-SATURATION ROUTINE.

1
+FREQ=77 REQUESTS
ZNO
77.
DATA GENERATOR.
1BRANCH
RECA
RECB
RECC
+BUS NAMES FOR EACH PHASE.

+ARRESTER.
+RATINGS.
0.71E+03
+(I,V) POINT.
0.412E+06
0.19E+06
0.100000E+01
0.96E+00
192000.
.962
1.0
1164.8
2.0
1181.6
5.0
1198.4
10.
1209.6
20.
1232.0
50.
1260.0
100.
1288.0
200.
1323.84
0.116480E+04
0.200000E+01
0.118160E+04
+(I,V) POINT.
0.500000E+01
0.119840E+04
+(I,V) POINT.
0.100000E+02
0.120960E+04
+(I,V) POINT.
0.200000E+02
0.123200E+04
+(I,V) POINT.
0.500000E+02
0.126000E+04
+(I,V) POINT.
0.100000E+03
0.128800E+04
0.200000E+03
1
707.107
0.30E+01
+(I,V) POINT.
+(I,V) POINT.
1
1
412500.
380000.
.892857
.001
0.380E+06
0.132384E+04
7 - 24
3.0
+(I,V) POINT.
0.500000E+03
0.138880E+04
+(I,V) POINT.
0.100000E+04
0.144256E+04
+(I,V) POINT.
0.200000E+04
0.151200E+04
+(I,V) POINT.
0.300000E+04
0.155680E+04
500.
1388.8
1000.
1442.56
2000.
1512.0
3000.
1556.8
1BLANK CARD ENDING CHARACTERISTIC

+BLANK CARD ENDS CHARACTERISTIC.
********** ERROR STATISTICS **********
SEGMENT
LOCAL ERROR
ACCUMULATED ERROR
1
0.9405627167E+00
0.9405627167E+00
SEGMENT
LOCAL ERROR
ACCUMULATED ERROR
1
0.9405627167E+00
0.9405627167E+00
**********
80-COLUMN CARD-IMAGE LISTING OF UNIT-7 PUNCHED CARDS.
*******
-------------------------------------------------------------------------------1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
-------------------------------------------------------------------------------C RATING =192000.00 V-MULT = 0.96200E+00 I-MULT = 0.30000E+01 GAPPED
92RECA
5555.
C
VREFERENCE
VFLASHOVER
0.412500000000000E+06
0.921212121212121E+00
C
MULTIPLIER
EXPONENT
VMIN
0.294795442961165E+05
0.265302624185338E+02
0.545050636122854E+00
9999
0.596059571777301E+06
0.265302624185338E+02
0.486652275816743E+00
9999
92RECB
RECA
5555.
92RECC
RECA
5555.
-------------------------------------------------------------------------------1BRANCH RECA
RECB
RECC
1
2
3
+ARRESTER.
2
3
1
0.412E+06
0.380E+06
1
707.107
192000.
.962
+RATINGS. 0.71E+03 0.19E+06 0.96E+00 0.30E+01
1
1.0
1164.8
+(I,V) POINT.
0.100000E+01
0.116480E+04
1
2.0
1181.6
+(I,V) POINT.
0.200000E+01
0.118160E+04
1
5.0
1198.4
+(I,V) POINT.
0.500000E+01
0.119840E+04
1
10.
1209.6
+(I,V) POINT.
0.100000E+02
0.120960E+04
1
20.
1232.0
+(I,V) POINT.
0.200000E+02
0.123200E+04
1
50.
1260.0
+(I,V) POINT.
0.500000E+02
0.126000E+04
1
100.
1288.0
+(I,V) POINT.
0.100000E+03
0.128800E+04
1
200.
1323.84
+(I,V) POINT.
0.200000E+03
0.132384E+04
1
500.
1388.8
+(I,V) POINT.
0.500000E+03
0.138880E+04
1
1000.
1442.56
+(I,V) POINT.
0.100000E+04
0.144256E+04
1
2000.
1512.0
+(I,V) POINT.
0.200000E+04
0.151200E+04
1
3000.
1556.8
+(I,V) POINT.
0.300000E+04
0.155680E+04
SEGMENT
LOCAL ERROR
ACCUMULATED ERROR
1
0.5838939740E-01
0.5838939740E-01
2
0.4318787687E-02
0.6270818508E-01
COMPARISON BETWEEN INPUT DATA AND THE RECALCULATED POINTS
INPUT VOLTAGE
INPUT CURRENT CALCULATED CURRENT &&&&&&&&
0.3042583643E+06
0.3000000000E+01
0.3697707231E+01
0.3086467061E+06
0.6000000000E+01
0.7188264878E+01
0.3130350479E+06
0.1500000000E+02
0.1384330666E+02
0.3159606091E+06
0.3000000000E+02
0.2131943045E+02
0.3218117315E+06
0.6000000000E+02
0.4996815251E+02
0.3291256345E+06
0.1500000000E+03
0.1418224270E+03
0.3364395374E+06
0.3000000000E+03
0.3934037950E+03
0.3458013333E+06
0.6000000000E+03
0.6480730609E+03
0.3627695882E+06
0.1500000000E+04
0.1440748041E+04
0.3768122819E+06
0.3000000000E+04
0.2714383329E+04
0.3949507613E+06
0.6000000000E+04
0.5945558306E+04
0.4066530061E+06
0.9000000000E+04
0.9675684425E+04
REPROCESS INPUT DATA. MULTIPLIER "A5" =
0.8928570000E+00
SEGMENT
LOCAL ERROR
ACCUMULATED ERROR
1
0.5838939740E-01
0.5838939740E-01
2
0.4318787687E-02
0.6270818508E-01
INPUT VOLTAGE
7 - 25
412500.
380000.
3.0
.892857
.001

0.2716592104E+06
0.3000000000E+01
0.3697707231E+01
0.2755773721E+06
0.6000000000E+01
0.7188264878E+01
0.2794955337E+06
0.1500000000E+02
0.1384330666E+02
0.2821076415E+06
0.3000000000E+02
0.2131943045E+02
0.2873318571E+06
0.6000000000E+02
0.4996815251E+02
0.2938621266E+06
0.1500000000E+03
0.1418224270E+03
0.3003923961E+06
0.3000000000E+03
0.3934037950E+03
0.3087511410E+06
0.6000000000E+03
0.6480730609E+03
0.3239013662E+06
0.1500000000E+04
0.1440748041E+04
0.3364394836E+06
0.3000000000E+04
0.2714383329E+04
0.3526345519E+06
0.6000000000E+04
0.5945558306E+04
0.3630829831E+06
0.9000000000E+04
0.9675684425E+04
**********
*******
-------------------------------------------------------------------------------1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
92RECA
5555.
C
VREFERENCE
VFLASHOVER
0.412500000000000E+06
0.921212121212121E+00
C
MULTIPLIER
EXPONENT
VMIN
0.505584788677157E+07
0.464199973324618E+02
0.632754084797273E+00
0.122767153039008E+05
0.166775903445227E+02
0.816748018907844E+00
9999
0.973862640531083E+09
0.464199973324618E+02
0.564958913889838E+00
0.812683352032397E+05
0.166775903445227E+02
0.729239185918001E+00
9999
92RECB
RECA
5555.
92RECC
RECA
5555.
-------------------------------------------------------------------------------COMMENT CARD.
1C
SHOWN BELOW IS AN EXAMPLE OF AUTOMATIC SEGMENT SELCTION USING
COMMENT CARD.
1C
AN UPGRADED FITTING PROCEDURE AVAILABLE AS OF "M38" VRESION.
1BRANCH RECA
RECB
RECC
1
-1
3
.05
1
412500.
380000.
+ARRESTER. -1
3
1
0.412E+06
0.380E+06
1
707.107
192000.
.962
3.0
.892857
.001
+RATINGS. 0.71E+03 0.19E+06 0.96E+00 0.30E+01
TOLERANCE "ERRLIM".
0.500000E-01
1
1.0
1164.8
+(I,V) POINT.
0.100000E+01
0.116480E+04
1
2.0
1181.6
+(I,V) POINT.
0.200000E+01
0.118160E+04
1
5.0
1198.4
+(I,V) POINT.
0.500000E+01
0.119840E+04
1
10.
1209.6
+(I,V) POINT.
0.100000E+02
0.120960E+04
1
20.
1232.0
+(I,V) POINT.
0.200000E+02
0.123200E+04
1
50.
1260.0
+(I,V) POINT.
0.500000E+02
0.126000E+04
1
100.
1288.0
+(I,V) POINT.
0.100000E+03
0.128800E+04
1
200.
1323.84
+(I,V) POINT.
0.200000E+03
0.132384E+04
1
500.
1388.8
+(I,V) POINT.
0.500000E+03
0.138880E+04
1
1000.
1442.56
+(I,V) POINT.
0.100000E+04
0.144256E+04
1
2000.
1512.0
+(I,V) POINT.
0.200000E+04
0.151200E+04
1
3000.
1556.8
+(I,V) POINT.
0.300000E+04
0.155680E+04
SEGMENT
LOCAL ERROR
ACCUMULATED ERROR
1
0.1565888897E-26
0.1565888897E-26
2
0.3227246021E-03
0.3227246021E-03
3
0.9177189649E-04
0.4144964986E-03
4
0.5854456312E-03
0.9999421299E-03
5
0.3761499771E-04
0.1037557128E-02
6
0.2661302274E-04
0.1064170150E-02
INPUT VOLTAGE
0.3042583643E+06
0.3000000000E+01
0.3000000000E+01
0.3086467061E+06
0.6000000000E+01
0.6000000000E+01
0.3130350479E+06
0.1500000000E+02
0.1554091710E+02
0.3159606091E+06
0.3000000000E+02
0.2936613917E+02
0.3218117315E+06
0.6000000000E+02
0.6121937705E+02
0.3291256345E+06
0.1500000000E+03
0.1486496042E+03
0.3364395374E+06
0.3000000000E+03
0.2851101814E+03
0.3458013333E+06
0.6000000000E+03
0.6137376433E+03
0.3627695882E+06
0.1500000000E+04
0.1481613818E+04
0.3768122819E+06
0.3000000000E+04
0.3020708694E+04
0.3949507613E+06
0.6000000000E+04
0.5939647269E+04
0.4066530061E+06
0.9000000000E+04
0.9056303281E+04
REPROCESS INPUT DATA. MULTIPLIER "A5" =
0.8928570000E+00
SEGMENT
LOCAL ERROR
ACCUMULATED ERROR
1
0.1517768382E-26
0.1517768382E-26
2
0.3227246021E-03
0.3227246021E-03
3
0.9177189649E-04
0.4144964986E-03
7 - 26

4
0.5854456312E-03
0.9999421299E-03
5
0.3761499771E-04
0.1037557128E-02
6
0.2661302274E-04
0.1064170150E-02
INPUT VOLTAGE
0.2716592104E+06
0.3000000000E+01
0.3000000000E+01
0.2755773721E+06
0.6000000000E+01
0.6000000000E+01
0.2794955337E+06
0.1500000000E+02
0.1554091710E+02
0.2821076415E+06
0.3000000000E+02
0.2936613917E+02
0.2873318571E+06
0.6000000000E+02
0.6121937705E+02
0.2938621266E+06
0.1500000000E+03
0.1486496042E+03
0.3003923961E+06
0.3000000000E+03
0.2851101814E+03
0.3087511410E+06
0.6000000000E+03
0.6137376433E+03
0.3239013662E+06
0.1500000000E+04
0.1481613818E+04
0.3364394836E+06
0.3000000000E+04
0.3020708694E+04
0.3526345519E+06
0.6000000000E+04
0.5939647269E+04
0.3630829831E+06
0.9000000000E+04
0.9056303281E+04
**********
*******
-------------------------------------------------------------------------------1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
92RECA
5555.
C
VREFERENCE
VFLASHOVER
0.412500000000000E+06
0.921212121212121E+00
C
MULTIPLIER
EXPONENT
VMIN
0.750282041336212E+07
0.484039509531973E+02
0.639499429775582E+00
0.244892766440758E+10
0.684088433742264E+02
0.748760917888158E+00
0.110464775993181E+07
0.394755404048778E+02
0.766237290359511E+00
0.846634377521169E+05
0.279345803478546E+02
0.800464072361844E+00
0.164899969440848E+05
0.187565472092776E+02
0.836738184231501E+00
0.111305268446438E+05
0.144460093305136E+02
0.912847535957407E+00
9999
0.180956701025696E+10
0.484039509531973E+02
0.570981542371137E+00
0.570070693112018E+13
0.684088433742264E+02
0.668536426862867E+00
0.968578003188489E+08
0.394755404048778E+02
0.684140328358522E+00
0.200716302920862E+07
0.279345803478546E+02
0.714699950256779E+00
0.138159910829870E+06
0.187565472092776E+02
0.747087544958385E+00
0.572164277990577E+05
0.144460093305136E+02
0.815042312412322E+00
9999
92RECB
RECA
5555.
92RECC
RECA
5555.
-------------------------------------------------------------------------------1BLANK CARD ENDING "ARRDAT" CASES
+BLANK CARD ENDS ARRESTER CASES.
1BLANK CARD ENDING "SATURATION" CASES
+BLANK CARD TERMINATING ALL SATURATION CASES.
CORE STORAGE FIGURES FOR PRECEDING DATA CASE NOW COMPLETED. --------------------------------------- PRESENT
PROGRAM
A VALUE OF -9999 INDICATES DEFAULT WITH NO FIGURE AVAILABLE.
FIGURE
LIMIT (NAME)
SIZE LIST 1.
TOTAL STORAGE SPACE ALLOCATED FOR THE EMTP AUXILIARY PROGRAMS
281040(LTLABL)
TIMING FIGURES (DECIMAL) CHARACTERIZING CASE SOLUTION SPEED. ------------------------------------CP SEC
I/O SEC
SUM SEC
TOTALS
0.170
0.000
0.170
SUPPORTING AUXILIARY ROUTINES FOR EMTP - DCG/EPRI VERSION 3.0 DOS WINDOWS TRANSLATION
DISTRIBUTED BY EPRI SOFTWARE CENTRE. RUN DATE (MM/DD/YY) AND TIME (HH.MM.SS.)= 5/ 2/97 13. 7.52
PROGRAM VERSION="V3.0e"
--------------------------------------------------+-------------------------------------------------------------------------------DESCRIPTIVE INTERPRETATION OF NEW-CASE INPUT DATA 1 INPUT DATA CARD IMAGES PRINTED BELOW, ALL 80 COLUMNS, CHARACTER BY CHARACTER.
0
1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
0
--------------------------------------------------+-------------------------------------------------------------------------------1BEGIN NEW DATA CASE
1BLANK
+BLANK TERMINATION-OF-RUN CARD.
___Normal end of aux run__
Finally, consider the way execution of this routine ends. Shown below is the very end of the
LUNIT7 listing for the final (second) example which was buried in the data case. After two blank
cards, the familiar summary statistics of data-case termination appear:
To conclude the illustrative documentation, a listing of the card images which were written on the
LUNIT7 file. Note that there really are two unrelated groupings, corresponding to the two subcases
that were buried in the AUX data case:
C RATING =192000.00 V-MULT = 0.96200E+00 I-MULT =
92RECA
5555.
C
VREFERENCE
VFLASHOVER
0.412500000000000E+06
0.921212121212121E+00
7 - 27
0.30000E+01
GAPPED

C
MULTIPLIER
EXPONENT
0.294795442961165E+05
0.265302624185338E+02
9999
0.596059571777301E+06
0.265302624185338E+02
9999
92RECB
RECA
5555.
92RECC
RECA
5555.
92RECA
5555.
C
VREFERENCE
VFLASHOVER
0.412500000000000E+06
0.921212121212121E+00
C
MULTIPLIER
EXPONENT
0.505584788677157E+07
0.464199973324618E+02
0.122767153039008E+05
0.166775903445227E+02
9999
0.973862640531083E+09
0.464199973324618E+02
0.812683352032397E+05
0.166775903445227E+02
9999
92RECB
RECA
5555.
92RECC
RECA
5555.
92RECA
5555.
C
VREFERENCE
VFLASHOVER
0.412500000000000E+06
0.921212121212121E+00
C
MULTIPLIER
EXPONENT
0.750282041336212E+07
0.484039509531973E+02
0.244892766440758E+10
0.684088433742264E+02
0.110464775993181E+07
0.394755404048778E+02
0.846634377521169E+05
0.279345803478546E+02
0.164899969440848E+05
0.187565472092776E+02
0.111305268446438E+05
0.144460093305136E+02
9999
0.180956701025696E+10
0.484039509531973E+02
0.570070693112018E+13
0.684088433742264E+02
0.968578003188489E+08
0.394755404048778E+02
0.200716302920862E+07
0.279345803478546E+02
0.138159910829870E+06
0.187565472092776E+02
0.572164277990577E+05
0.144460093305136E+02
9999
92RECB
RECA
5555.
92RECC
RECA
5555.
7 - 28
VMIN
0.545050636122854E+00
0.486652275816743E+00
0.30000E+01
GAPPED
VMIN
0.632754084797273E+00
0.816748018907844E+00
0.564958913889838E+00
0.729239185918001E+00
0.30000E+01
GAPPED
VMIN
0.639499429775582E+00
0.748760917888158E+00
0.766237290359511E+00
0.800464072361844E+00
0.836738184231501E+00
0.912847535957407E+00
0.570981542371137E+00
0.668536426862867E+00
0.684140328358522E+00
0.714699950256779E+00
0.747087544958385E+00
0.815042312412322E+00
Section 8
Eddy Currents Calculation
8.1
Introduction ......................................................................................... 8-1
8.2
Input Data Format ............................................................................... 8-2
8.3
Example ............................................................................................... 8-4
Section 8
is the AUX eddy currents calculation module. It generates a punch file that contains an RL
network that represents eddy currents effects in steel cores, based on physical information such as
size, number of turns, etc.
EDDYC
8.1 Introduction
Eddy currents are induced in transformer core laminations by the alternating flux in the core. As
frequency changes, flux distribution in the iron core laminations change. For high frequencies the
flux will be confined to a thin layer close to the lamination surface; the thickness of this layer
decreases as the frequency increases. Consequently, the inductances which represent the iron path
magnetization and the resistances which represent eddy current losses are frequency dependent.
There are a number of possible representations for eddy current effects, ranging from a single RL
branch to higher-order ladder networks. These representations are generally based on the physical
characteristics of the core, and core laminations (see Reference 7). The AUX module EDDYC,
produces a ladder network (see Figure 8.1 below) which provides a good compromise between
speed and accuracy. The resulting ladder network should be connected in parallel with the
nonlinear branch representing saturation and/or hysteresis (see Section 6.1).
R1
L1
L2
Figure 8.1:
R2
R3
L3
L4
R4
R5
L5
Eddy currents ladder network
The resistances and inductances of this "NTERM" ladder network are given by
Lo
L k = -------------4k 3
8-1
R k = R o ( 4k 1 )
where
N A
L o = --------------l
4N A
R o = -------------2
ld
Lo = Inductance in Henries
Ro =Resistance in ohms
= Conductivity of steel lamination in S/m
N =Number of turns
A = Total cross-section area in m2
d = Thickness of lamination in m
= oR, and o=410-7 H/m
l = Length of the core in m
8.2 Input Data Format

The structure of the data deck for an EDDYC request is as follows:
1.
1
2
3
4
5
6
7
8
1234567890123456789 0123456789012345678901234567890123456789012345678901234567890
8-2

BEGIN NEW DATA CASE
A19
2.
Next comes an "EDDY-CURRENTS" special-request card.
1
2
3
4
5
6
7
8
1234567890123 4567890123456789012345678901234567890123456789012345678901234567890
EDDY-CURRENTS
A13
3.
The EDDY CURRENTS request card is followed by one data card with the format
shown below:
BUSM
A6
A6
NTERM
BUSK
Seed
1
2
3
4
5
6
7
8
123456 789012 34 5678901234 5678901234 5678901234 5678901234 5678901234 56789 012 34567890
XOPT
A2
E10.0
E10.0
E10.0
E10.0
E10.0
I5
I3
E8.0
BUSK
The first node name.
(1-6)
BUSM
The second node name.
(7-12)
Seed
(13-14)
Alphanumeric seed used to create internal node names. If left blank,

then seed = "NN". The program automatically generates unique node
names using the pattern NN0001, NN0002, etc. Please note that
Seeds for different eddy current models within an EMTP run must be
unique.
r
(15-24)
Resistivity of the steel lamination in ohm.m (r=1/g)
A
(25-34)
Cross-section are in m2
d
(35-44)
Thickness of the laminations in m.
8-3

l
(45-54)
Length of core in m.
mr
(55-64)
Relative permeability
N
(65-69)
Number turns
NTERM
Number of terms in ladder network (default = 5)
(70-72)
Value of XOPT to be used in the EMTP
XOPT
(73-80)
4.
To indicate the end of all EDDYC data cases add a blank card, followed by a
"BEGIN NEW DATA CASE" card (optional), followed by a blank card.
blank card
blank card
8.3 Example
The following example is based on data from a 50 MVA 110/27.6 kV transformer, where,
A = 0.353 m2, d = 0.35 mm, l = 3.35 m, r = 2000, = 50 .cm, and N = 865. The request and
data cards for this example are shown below
C
1
2
3
4
5
6
7
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
EDDY CURRENTS
NODE1 NODE2 XY
50.E-8
0.353
0.35E-3
3.35
2000. 865 5 60.
BLANK
BEGIN NEW DATA CASE
BLANK
The resulting punch file for inclusion into the EMTP is shown below
C
EDDY CURRENTS
8-4

C NODE1 NODE2 XY
$VINTAGE, 1
NODE1 NODE2
NODE1 XY0001
XY0001NODE2
XY0001XY0002
XY0002NODE2
XY0002XY0003
XY0003NODE2
XY0003XY0004
XY0004NODE2
NODE2 XY0004
$VINTAGE, 0
50.E-8
0.353
0.35E-3
3.35
0.7470215E+05
0.3861689E+07
0.1494043E+05
0.9010607E+07
0.8300239E+04
0.1415953E+08
0.5746319E+04
0.1930844E+08
0.4394244E+04
0.2445736E+08
8-5
2000.
865
60.
Section 9
Line Constants
9.1
Introduction to Line Constants Routines ............................................. 9-1
9.2
Capabilities of the Line Model Module .............................................. 9-2

9.2.1
Models for Transients Analysis ..................................... 9-2
9.2.2
Models for Steady-State Analysis: PI-EXACT ............ 9-4
9.2.3
Dimensioning Limits ...................................................... 9-5
9.2.4
Data Entry Conventions ................................................. 9-5
9.2.5
Input/Output Units ......................................................... 9-6
9.3
General Structure of a Line Constants Data Deck .............................. 9-6
9.4
Line-Parameters Module ..................................................................... 9-9
9.5
9.6
9.4.1
Line-Parameters Control Card [2] .................................. 9-9
9.4.2
Units Control Card [3] ................................................... 9-9
9.4.3
Electric Field Strength Across Right-of-Way ................ 9-10
9.4.4
Conductor Data Cards [4] .............................................. 9-11
9.4.5
Alternative Self-Inductance Calculation ........................ 9-17
9.4.6
Frequency Cards [5] ....................................................... 9-18
9.4.7
Special Request .NODES Card ...................................... 9-23
9.4.8
Example of Data Deck for the Line-Parameters Module 9-23
9.4.9
Description of Output from the Line-Parameters Module 9-24
Special Calculation Options ................................................................ 9-26

9.5.1
Calculation of Electric Field Strength at Ground Level
9-26
9.5.2
Mutual Impedance with Communications Lines ........... 9-28
Line-Model Module ............................................................................ 9-29

9.6.1
Line-Model Control Card ............................................... 9-29
9.6.2
Units Control Card [3] ................................................... 9-39
9.6.3
Conductor Data Cards [4] .............................................. 9-41
9.6.4
Conductor Cards for the Line-Rebuild Option .............. 9-45
9.6.5
Conductor Cards for the Fixed-Parameters Option ........ 9-47
9.6.6
Frequency Card [5] ........................................................ 9-48
9.6.7
Optional Control Cards [6] ............................................ 9-49
Section 9
9.7
Line Constants
9.6.8
Transformation Matrices and Q-Error Indicators .......... 9-58
9.6.9
Examples of Data Decks for the Line-Model Module ... 9-60
FIT-S Module ..................................................................................... 9-62

9.7.1
FIT-S Control Card ........................................................ 9-62
9.7.2
Data Function File .......................................................... 9-63
9.7.3
Data Function Format .................................................... 9-64
9.7.4
Type Of Fit ..................................................................... 9-64
9.7.5
Optional Control Cards .................................................. 9-64
9.7.6
Examples of Data Deck for the FIT-S Module .............. 9-67
Section 9
Line Constants
9.1 Introduction to Line Constants Routines
The Line Constants routines evaluate the resistance, inductance, conductance, and capacitance of
an arbitrary arrangement of conductors of an overhead transmission line, where one or more
conductors can represent one phase (e.g., bundled phase conductors). The resulting impedance
and admittance matrices can then be reduced to find the phase equivalents. The phase equivalents
in turn can be transformed into modal components. With the calculated impedance and
admittance data the program can generate models to represent the line for transients analysis and
for steady-state solutions.
There are three main modules within this set of routines:
(A)
Line Parameters Module. Keyword: "LINE-PARAMETERS"
(B)
Line Model Module. Keyword: "LINE-MODEL"
(C)
External Data Fitting Module. Keyword: "FIT-S"
Line Parameters Module:

The Line Parameters module reads as input the configuration of the system of conductors and
produces on output the series impedance and shunt admittance matrices of the line.
The impedance and admittance matrices can be determined for the full system of physical
conductors, or for the reduced system of equivalent phase conductors. The equivalent phase
matrices can also be converted into symmetrical component impedance and admittance matrices
(zero, positive and negative sequence quantities) of the associated perfectly transposed line (after
averaging the self and mutual elements).
The Line Parameters module can also determine, at selected frequencies, the modal line
parameters and wave propagation functions of the untransposed line, using eigenvalue/
eigenvector routines.
The Line Parameters module can also produce a multiphase nominal pi-circuit for a given line
length. This nominal pi-circuit is an approximation that ignores the distributed nature and
frequency dependence of the line parameters and is valid only for short line sections. The cards
required to set up the data deck for this module are described in Section 9.7.
9-1
Line Constants
Line Model Module

The line model module reads as input the line configuration (in the same format as for the Line
Parameters module) and produces, on output, line models that can be used by the EMTP for
transients and steady state solutions.
The main difference between Line Parameters and Line Model Modules is that the Line Model
module generates EMTP models, whereas the Line Parameters module does not (with the exception
of the generation of nominal pi-circuits).
The data and control cards required to set up the data deck for this module are described in
Section 9.8.
External Data Fitting Module:

The keyword "FIT-S" can be used to fit an externally-supplied frequency-domain function. It
provides direct access to the rational function fitting routines associated with the frequency
dependent line models.
9.2 Capabilities of the Line Model Module

The Line Model module can produce line models suitable for EMTP transients simulations.
However, frequency domain models which are only suitable for steady-state analysis can also be
generated.
9.2.1
Models for Transients Analysis
The following line models can be generated:

FD-LINE
CP-LINE
LBUILD
CONSTANT
FD-LINE
The fd-line model (frequency-dependent line model), also known as the JMARTI line model,
provides an accurate representation of the distributed nature of all the line parameters: R, L, G, and
C, as well as the dependence of R and L with frequency.
9-2
Line Constants
The model is based on the approximation by rational functions of the line characteristic impedance
Zc and propagation function Ap, namely
Zc =
( ( R + jL ) ( G + jC ) )
Ap = e-l,
where,
=
( R + jL ) ( G + jC )
and "l" is the length of the line.

The main simplification used in this model is the validity of the assumption that a constant real
transformation matrix can be used to relate phase and modal quantities over an extended
frequency range. This allows the approximation of Zc and Ap as scalar quantities.
CP-LINE
The cp-line model (constant-parameter line model), also known as the Dommel line model,
assumes that the line parameters R, L, and C are constant, as calculated at the requested frequency.
The model considers L and C to be distributed ("ideal line") and R to be lumped at three places
(line ends and line middle). The conductance G is assumed to be zero.
The frequency dependence of the line parameters (as modelled by the fd-line model) is an
important factor for the accurate simulation of waveform and peak values. However, the cp-line
model is very robust and simple (about 30% to 50% faster than the fd-line model) and provides a
good alternative for a first approximation analysis and for the modelling of secondary lines.
LBUILD
This option provides an approximation to the line when the actual geometry is not known. The
input data are the 60-Hz (or any other particular frequency) zero and positive sequence series
impedance and shunt admittance of the line. From this data, an equivalent balanced-line geometry
is reconstructed. This geometry is then used to generate any requested line model (including
frequency dependent models) as in the normal case.
CONSTANT
This option uses as input the 60-Hz sequence impedances and admittances as in the 'lbuild' option.
However, the line geometry is not reconstructed and the parameters are assumed fixed at their 609-3
Line Constants
Hz values. Any line model can be requested but frequency dependence of the line parameters is
not considered. The complexity of the models generated with this option is basically the same as
the complexity of the models generated with the 'lbuild' option and, therefore, there is normally no
advantage in requesting this option instead of the 'lbuild' option.
9.2.2
Models for Steady-State Analysis: PI-EXACT
This model provides an exact single-frequency representation of the line in the form of a
multiphase pi-equivalent. The model can be read directly by the EMTP when frequency scans or
single-frequency steady-state solutions are requested.
On output, this option produces a "punch file" that contains [Zseries()] and [Yshunt()] at every
frequency of the EMTP frequency scan calculation (see Figure 9.1 below), Note that the number of
frequency points specified in the generation of the pi-exact model (in either a logarithmic or linear
frequency intervals) must match exactly the number of frequency points in the EMTP frequency
scan.
[Zseries()]
[Yshunt()]
Figure 9.1:
[Yshunt()]
Circuit Representation of a PI-EXACT model
Since the parameters of the exact pi-circuit change with frequency (even for constant line
parameters), the model is not valid for transients simulations.
9-4
Line Constants
9.2.3
Dimensioning Limits
Line-Model Module:
Total Number of Conductors (including subconductors and ground wires)
177
Total Number of Phases (or Modes)
21
Number of Transposition Sections
50
Number of Poles in Rational Function Approximations
100
Line-Parameters Module:
Total Number of Conductors (including subconductors and ground wires)
100
Total Number of Phases (or Modes)
9.2.4
50
Data Entry Conventions
The following terminology and conventions are used in the description of the input data file:
1.
The lines in the data file are called "cards".
2.
Character data entry is case insensitive (e.g., LiNe-mOdel is the same as LineModel).
3.
In the description of the card images, Fortran format convention is used to

specify the data types.
4.
If a data field is left blank, the default value is internally assigned by the
program (assuming there is a default value for that field).
5.
Control flags in I2-format fields normally have the following meaning:

1 = on; 1 = off; blank or 0 = default value.
6.
Control flags in I1-format fields normally have the following meaning:

1 = on; 0 = off.
9-5
Line Constants
9.2.5
Input/Output Units
Logical Unit
Purpose
Assignment
LU5
Input Data Deck
User
LU6
Output Listing
User
LU7
Punched Output
User
LU17
Scratch file (binary)
Internal
9.3 General Structure of a Line Constants Data Deck

The general structure of the input data file for running the line constants program is shown in Table
9.1 below. Comment lines are ignored by the program. They can be inserted anywhere in the data
deck. These lines are identified by "C " or "c " (letter C or c in the first column, followed by a
space in the second column). They can contain any alphanumeric characters in the rest of the line.
BLANK: End of Run

BEGIN NEW DATA CASE
BLANK End of LINE CONSTANTS
BLANK: End of MODULE
Optional "." Cards
Frequency Card
BLANK: End of Conductor Cards
CONDUCTOR CARDS
UNITS
MODULE
LINE CONSTANTS
BEGIN NEW DATA CASE
9-6
Line Constants
[1
"BEGIN NEW DATA CASE" and

"LINE CONSTANTS" request
card
Begin data case identifiers
[2]
MODULE
Line-Parameters, Line-Model, FIT-S
[3]]
UNITS
Metric or British
[4]
CONDUCTORS
Conductor Data Cards
...
...
BLANK card
BLANK card to indicate the end of conductor
data cards
[5]
FREQUENCY
Earth resistivity, frequency, etc.
[6]
OPTIONAL CONTROL CARDS
Additional control for Line-Model module
[7]
BLANK to end module
End of Data Case Identifiers (4 cards)
BLANK to end line constants

BEGIN NEW DATA CASE
BLANK to end run
Table 9.1: Structure of the Line Constants Program Data Deck

With reference to Table 9.1, the data deck consists of the following sections:
[1]
BEGIN DATA CASE Identifiers
The begin data case identifiers follow general EMTP/AUX usage. Two cards are
needed: a) the "BEGIN NEW DATA CASE" card, and b) the "LINE CONSTANTS"
request card
[2]
MODULE
This card identifies the desired type of calculation: LINE-MODEL, LINEPARAMETERS, or FIT-S.
[3]
UNITS
Selects the system of units in which the conductor data is specified, i.e., Metric
or English. It also controls the input format used in the CONDUCTOR data cards
9-7
Line Constants
[4]
CONDUCTORS
These cards specify the geometry of the system of conductors and the
characteristics of the individual conductors. A marker card is used to signal the
end of the group of conductor cards. The marker card can be a blank card, the
word "BLANK" (left-justified), or the characters "====" (at least four, leftjustified).
[5]
FREQUENCY
This card contains the frequency at which to evaluate the line parameters for the
LINE-PARAMETERS module, and general information about the transmission
system for the LINE-PARAMETERS and LINE-MODEL modules. This general
information includes the ground resistivity, the line length, and other
miscellaneous information. There is only one frequency card in the LINEMODEL module and any number of frequency cards in the LINE-PARAMETERS
module.
[6]
OPTIONAL CONTROL CARDS
These cards are characterized by a dot "." in the first column of the line, joined to
a keyword (e.g., ".ctlfit"). These lines are optional and allow the user to supply
additional control information for the processing of the line parameters and line
models, and to override internally preset default values.
The internal default values have been carefully chosen and most production
cases can be run without the optional control cards.
[7]
END DATA CASE Identifiers
These markers follow general EMTP/AUX usage. They include the following
cards: a) "BLANK", b) "BEGIN NEW DATA CASE", and c) next data case, or
"BLANK" to end the run.
Default Values
The following table gives the default values for some of the parameters and control variables used
by the program. Some of these values can be changed by indicating the desired new values in the
corresponding data cards. Other values are fixed internally and cannot be modified.
DELF
= 100 Hz
FDC
= 1.E-15 Hz
FINF
= 1.E+08 Hz
FMAX
= 5.0 kHz
FMIN
= 0.1 Hz
GPHASE
= 0.2E-09 S/km
IBUG1
=0
IBUGF
=0
IFITAL
=0
IFITZC
=0
ICOMPF
= -1
IMONIT
= -1
IPLOT
= +1
IPRAT
= +1
IPHASE
= +1
9-8
Line Constants
IROT
= +1
IWAVEF
= 1
IQUICK
= 1
IWD1
= 1
IIWQ
= 1
IWD
= 1
NDEC
=8
NORMAX
= 25
IXDYN
= +1
NPDEC
= 10
9.4 Line-Parameters Module

The Line-Parameters module can determine the resistance, inductance, and capacitance matrices
for a multiphase overhead transmission line consisting of an arbitrary configuration of conductors.
The data deck for this module follows the general structure shown in Section 9.3, Table 9.1. The
required cards and information are described next.
9.4.1
Line-Parameters Control Card [2]
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012345678901234567890
LINE-PARAMETERS
A15
Enter the keyword "LINE-PARAMETERS" in columns 1-15.
9.4.2
Units Control Card [3]
Specify S.I. Metric or English (British) system of units for line and conductor data.
UNITS
XMIN
XMAX
DELX
INOPT
1
2
3
4
12345 678 9012345678 9012345678 9012345678 9 0
A8
E10.0
E10.0
E10.0
I1
9-9
5
6
7
8
1234567890123456789012345678901234567890
Line Constants
I
UNITS
= "METRIC"
(1-5)
The S.I. system of units is used for conductor and line

data. Units of mm or cm are used for conductor diameter
and bundle spacing depending on the specified input
option INOPT described below:
Units in the CONDUCTOR cards:
Resistance
/km
Diameter
mm for INOPT = 1
cm for INOPT = 2
Height and Spacing
Bundle Spacing
m for INOPT = 1
cm for INOPT = 2
Units in the FREQUENCY card:

Line length
""ENGLISH"
or
"BRITISH"
km
The English system of units is used for conductor and line

data.
Units in the CONDUCTOR cards
Resistance
/mile
Diameter
inches
Height and Spacing
feet
Bundle spacing
inches
Units in the FREQUENCY card

Line length
9.4.3
miles
Electric Field Strength Across Right-of-Way
When fields 9 to 38 of the UNITS card are specified, the electric field strength E (in kV/m or kV/ft,
according to UNITS) at ground level (assuming flat ground) on a plane perpendicular to the line are
calculated. The values of E are determined at increments DELX, from XMIN to the left to XMAX to
the right.
Input option INOPT = 1 (columns 40) must be used for the CONDUCTOR data when electric field
calculations are desired. With this input option, fields 73 to 80 of the CONDUCTOR cards are used
9 - 10
Line Constants
to specify the voltages between conductors and ground, in magnitude and phase angle. See
Section 9.5 for further information on these calculations.
Maximum distance (in m or feet, depending on UNITS) to the left (from
the vertical reference axis used in the CONDUCTOR cards) at which E is
calculated. See the diagram of Figure 9.4 for distance measurements.
XMIN
(9-18)
XMAX
(19-28)
DELX
(28-38)
Increments from XMIN to XMAX (in the same units as XMIN and XMAX)
at which E is calculated. The output is printed in the sequence E(XMIN),
E(XMIN + DELX), E(XMIN + 2XDELX), ..., E(XMAX).
Input format for conductor cards (see image of CONDUCTOR data cards
in Section 9.4.4).
INOPT
(40)
9.4.4
Maximum distance (in m or feet, depending on UNITS) to the right

(from the vertical reference axis used in the CONDUCTOR cards) at which
E is calculated.
=2
This is the default option and it corresponds to the traditional

EMTP format for line constants (prior to version 2.0)
=1
This new format of the CONDUCTOR cards permits the addition

of fields VOL and PHA needed for the ELECTRIC FIELD
STRENGTH calculations described above.
Conductor Data Cards [4]
The conductor cards contain the geometrical data of the line and the characteristics of the
individual conductors.
As indicated in the UNITS card, two input formats for the conductor data are supported. INOPT = 2
in the UNITS card (default) corresponds to the traditional EMTP format. The new format INOPT = 1
allows for two additional fields that are used for electric field at ground level calculations (Section
9.4.2).
Format for INOPT = 2 (Default):
RESIS
I3
E5.0
E8.0
REACT
DIAM
HORIZ
VTOWER
VMID
SEPAR
ALPHA
NBUND
SKIN
IXTYPE
IPHASE
1
2
3
4
5
6
7
8
123 45678 90123456 78 90123456 78901234 56789012 34567890 12345678 90123456 789012 345678 90
I2
E8.0
E8.0
E8.0
E8.0
E8.0
E8.0
E6.0
I2
9 - 11
Line Constants
Format for INOPT = 1 (New Option)
RESIS
I3
E5.0
E8.0
REACT
DIAM
I2
E8.0
E8.0
HORIZ VTOWER
E7.0
E8.0
5
9012345
7
678 789012 3456
VMID
NBUND
SKIN
IXTYPE
IPHASE
1
2
3
4
123 45678 90123456 78 90123456 78901234 5678901 2345678
E7.0
I3
8
7890
ALPHA VOLT PHA

E6.0
E4.0
E4.0
There must be one conductor card for each physical conductor in the line. This includes each
subconductor in a bundle and also the ground wires. There is no conductor card associated with
the earth return path. Bundled conductors specified using the bundle input option SEPAR (59-66)
and ALPHA (67-72) require a single conductor card.
To facilitate data entry of identical conductors, if any or all of the first six fields (IPHASE, SKIN,
RESIST, IXTYPE, and REACT) is left blank, the value of the corresponding variable will be assumed
to be the same as in the previous card.
In the following description, N conductors are assumed (column numbers refer to INOPT=2 above):
IPHASE
(1-3)
The phase number to which the conductor belongs. If more than one
conductor is given the same phase number, this means that the conductors
are electrically connected in parallel. This is the case, for instance, of
individually-specified conductors in a bundle (fields SEPAR and ALPHA
left blank). It could also be used, for instance, to internally combine two
parallel lines when it is not desired to preserve their individual identity.
Phase numbers for conductors must follow the sequence 1, 2, 3, ..., N with
no missing phases.
Set IPHASE = 0 for a ground wire (ground is phase number zero, by
definition).
SKIN
(4-8)
Flag for skin effect correction defined as Thickness/Diameter (T/D) of an

equivalent tubular conductor (see Figure 9.2).
>0
Assume a tubular conductor with T/D = SKIN.
= 0.0 or
BLANK
Neglect the skin effect correction (see also the description

of RESIS below).
= 0.5
Assume a solid conductor.
= 1
Use Galloway-Wedepohl's formula for stranded

conductors.
9 - 12
Line Constants
RESIS
(9-16)
DC resistance of the conductor in units of /km or /mile, according to

UNITS.
If the skin-effect correction is to be bypassed (0.0 in field SKIN (columns

4-8), specify the conductor's AC (not DC) resistance. Recall that AC
resistance is equal to DC resistance plus a skin-effect contribution which
depends on frequency. Also, the internal inductance for IXTYPE = 4 will
not be corrected for skin effect when SKIN = 0. is specified.
For the Galloway-Wedepohl's formula ('SKIN' = -1.0), RESIS is the per
unit length resistance of only one strand: one of the outer strands.
IXTYPE
(17-18)
Determines the interpretation of the variable REACT below. Usually

KTYPE = 4. Other alternatives are explained in Section 9.4.5.
(19-26)
REACT Takes on different meanings depending on the value of IXTYPE.

For a conductor with IXTYPE = 4, REACT is the relative permeability and
DIAM
Outside diameter of the conductor in units of:
REACT
it is automatically defaulted to r = 1.0 if left blank. Section 9.4.5 for

other alternatives.
(27-34)
HORIZ
(35-42)
centimetres
if UNITS = METRIC and INOPT = 2
millimetres
inches
if UNITS = BRITISH
Horizontal distance of the conductor from the reference point x = 0 (see

diagram in Figure 9.2 below), in units of:
metres
if UNITS = METRIC
feet
if UNITS = BRITISH
The reference point x = 0 can be located anywhere since distances are

relative. Distances to the right of x = 0 are positive and distances to the
left are negative.
VTOWER
(43-50)
Vertical height of the conductor above the ground, at the tower in units
of:
metres
if UNITS = METRIC
feet
if UNITS = BRITISH
9 - 13
Line Constants
When both VTOWER and VMID are specified, an average height is
calculated by the program using the formula:
VTOWER VMID
2 VMID + VTOWER
VMID + -------------------------------------------------- = ---------------------------------------------------------3
3
If only VTOWER is specified, then this height will be assumed uniform

along the line.
Midspan height of the conductor above the ground, at the tower in units
of:
VMID
(51-58)
metres
if UNITS = METRIC
feet
if UNITS = BRITISH
Example of a conductor card:

C
1
2
3
4
5
6
7
8
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
1.3871 .03480 2
.7092
1.802
-.75
69.0
39.0
1.3871 .03480 2
.7092
1.802
.75
69.0
39.0
2.3871 .03480 2
.7092
1.802
40.0
69.0
39.0
18.0
0.0
2
0 0.5
3.10 1
.484
.495
20.0 133.0
Conductors
a, b, c, d:
earth wire
phase 1
a b
T/D = 0.3871
103 at midspan
133 at tower
39 at midspan
69 at tower
.75
reference
Figure 9.2:
GMR = 0.7092"
DIAMETER = 1.802"
phase 2
c d
Ground wire:
-.75
R = 0.0398 /mile
Example of Conductor Card (INOPT = 2)
9 - 14
R = 3.1 /mile
Reactance for 1
spacing at 60 Hz =
0.484
DIAMETER = 0.495"
T/D = 0.5 (solid)
Line Constants
Bundle Input Option:

Specified using parameters NBUND, SEPAR, ALPHA
Leave NBUND blank for single conductors. For a bundle with K conductors (see Figure 9.3 below)
there are three options available to enter the bundle data:
(A)
Enter a normal conductor card for each of the K conductors in the bundle, and
leave NBUND blank. This may be too time-consuming for a regular
symmetrical bundle, but it is the only option available for asymmetrical
bundles.
(B)
Convert the bundle into a single equivalent conductor using the distance
averaging formulas available for this purpose and enter the equivalent
conductor in a normal conductor card (leave NBUND blank). This alternative is
less accurate than options (A) and (C).
(C)
Specify a symmetrical bundle using the fields NBUND, SEPAR, and ALPHA.
There should be only one CONDUCTOR card per bundle.
The data fields for option (C) are specified as follows:

NBUND
Number of conductors in the bundle.
(79-80)
SEPAR
Spacing between adjacent conductors in the bundle, in units of:
(59-66)
ALPHA
(67-72)
centimetres
UNITS = "METRIC" and INOPT = 2
millimetres
inches
UNITS = "ENGLISH"
Angular position of the first conductor (or any conductor) of the bundle,
in units of degrees. Positive angles are measured counter-clockwise.
9 - 15
Line Constants
SEPAR
1
ALPHA
2
4
3
Figure 9.3:
Sample sketch of a bundle with NBUND = 6 conductors, with

angle ALPHA = 30
Electric Field Strength At Ground Level

This option applies only for INOPT = 1 in UNITS card
VOLT
(73-76)
PHA
(77-80)
These fields are used for electric field strength at ground level
calculations (see also UNITS card in Section 9.4.2 above). Leave blank if
these calculations are not desired.
VOL and PHA specify the magnitude and phase angle of the voltage
between the conductor and ground. The magnitude is the RMS value in
kV and the phase angle is expressed in degrees.

As an example, a 500 kV three-phase line with phases 1, 2, 3 could have
the following voltages for normal operation:
Phase
Amplitude (RMS)
Angle(degrees)
289.
0.
289.
240.
289.
120.
9 - 16
Line Constants
9.4.5
Alternative Self-Inductance Calculation
With the default option IXTYPE = 4, the conductor internal inductance is corrected for skin effect
assuming tubular conductor geometry. Alternative inductance options are possible that do not
correct the conductor's internal inductance. These options are requested using the flag IXTYPE
(columns 17-18). The data corresponding to the requested option is entered in field REACT
(columns 19-26).
IXTYPE
Flag controlling the method used to calculate conductor inductance
(17-18)
=0
Reactance for unit spacing, this reactance (not the inductance) is

assumed to remain constant regardless of the frequency that may
be specified in the FREQUENCY cards.
The unit spacing is:
1 meter if UNITS = "METRIC"
1 foot if UNITS = "ENGLISH"
The reactance REACT is in units of:
/km if UNITS = "METRIC"
/mile if UNITS = "ENGLISH"
=1
Reactance for unit spacing (as above) at 60 Hz. As opposed to

the case above where the reactance is assumed to remain
constant, it is now the inductance that is assumed to remain at its
60-Hz value. If frequencies other than 60 Hz are specified in the
FREQUENCY cards, the reactance will be changed
proportionately.
Note that the relationship between reactances for 1'-spacing and
GMR (geometric mean radius) is given by
f ( Hz ) )
12
X 1 foot( ( mile )) = ----------------------------------------------------100,0 0,20223653 log GMR
( inches ) )
X 1 foot( ( mile )) = ( 2 1,609344 10
=2
12
) log -------------------------------------GMR ( inches ) )
GMR (geometric mean radius) of the conductor, in units of:

millimetres
if UNITS = "METRIC" and INOPT = 1
centimetres if UNITS = "METRIC" and INOPT = 2
inches
if UNITS = "ENGLISH"
9 - 17
Line Constants
9.4.6
=3
Dimensionless ratio GMR/r, where "r" is the conductor outer

radius. For solid conductors, this ratio is equal to 0.7788.
=4
This is the standard case described before. The internal

reactance is calculated from the geometry of the tubular
conductor as defined by SKIN, field 'REACT' is used to enter the
relative permeability r of the conductor. Default value is of r
is 1.0.
>5
For use with Galloway-Wedepohl's equation (SKIN=1). REACT

is used to indicate the relative permeability r of the outer
strands. Default value of r is 1.0. IXTYPE is equal to the total
number of outer strands in the conductor.
Frequency Cards [5]
There is one frequency card for each frequency at which the line parameters are to be calculated.
The FREQUENCY card contains information on the ground resistivity, electrical frequency,
segmentation of the ground wires, and line length.
There is a special request flag in this card (MUTUAL, column 59) to calculate the coupling between
a transmission line and a communications line.
E8.0
E10.0
6I1
6I1
ALONG
E8.0
4I1 I1 I1
IFILE
IZPRN
MODAL
ICPRN
ISEG
MUTUAL
FMATRIX
IPIPRN
RHO
ICAP
1
2
3
4
5
6
7
8
12345678 9012345678 90123456789 012345 6 789012 3 4 56789012 3 4567 8 9 012345678 90 12 34567890
I2 I2
(1-8)
Resistivity of the ground return path in units of -meter. This applies to

both to "METRIC" and "ENGLISH" options in the UNITS card.
FMATRX
Frequency used in the calculation of the line parameters in Hz.
RHO
(9-18)
9 - 18
Line Constants
ICPRN
(30-35)
Group of 6 flags that control the printout for the capacitance matrices
[C], or the susceptance matrices [C], or the inverse of these. Flag ICAP
in column 44 controls the option of printing either the inverse or normal
matrix. Six possible independent outputs can be requested by entering
"1" punches in the appropriate columns:
"1" in Column No.
Resulting printout
30
Inverse of [C] or [C]
31
Inverse of [CE] or [CE]
32
Inverse of [CS] or [CS]
33
[C] or [C]
34
[CE] or [CE
35
[CS] or [CS]
No subscript indicates the unreduced system. Each physical conductor

has a row and column in the matrix, as does each ground wire.
Subscript "E" stands for "Equivalent phase conductors". (After
elimination of ground wires and bundling of subconductors with the
same phase number.)
Subscript "S" stands for "Symmetrical components" of the equivalent
phase conductors.
Table 9.2: Summary of Output Options for the Capacitance Matrix
IZPRN
(37-42)
Group of 6 flags to control the printout of the series impedance matrices

[Z] = [R] + j[L], or their respective inverse. Six possible independent
printed outputs can be requested by entering "1" in the appropriate
columns.
"1" in Column No.
Resulting printout
37
[Z]
38
[ZE]
39
[ZS]
40
[Z]-1
9 - 19
Line Constants
"1" in Column No.
Resulting printout
41
[ZE]-1
42
[ZS]-1
No subscript indicates the unreduced system. Each physical conductor

has a row and column in the matrix, as does each ground wire.
Subscript "E" stands for "Equivalent phase conductors". (After
elimination of ground wires and bundling of subconductors with the
same phase number.)
Subscript "S" stands for "Symmetrical components" of the equivalent
phase conductors.
Table 9.3: Summary of Output Options for the series impedance
matrix
ICAP
(44)
ALONG
Flag to control over whether it is capacitance [C] or susceptance [C]

that will be outputted by "ICPR" requests in columns 30-35:
=1
capacitance [C] (or its inverse).
=0
susceptance [C] (or its inverse).
Length of the transmission line under consideration, in units of:
(45-52)
kilometres
if UNITS = "METRIC"
miles
if UNITS = "ENGLISH"
This field can usually be left blank in the LINE-PARAMETERS module. It

is only needed if the (approximate) multiphase nominal-pi representation
of the line is requested. The multiphase nominal-pi is requested with the
flag IFILE (columns 71-72).
IPIPRN
(54-57)
Set of 4 flags to request the output of a multiphase nominal-pi circuit

for the indicated line length.
Four possible independent printouts can be requested by entering "1" in
the appropriate columns:
"1" in Column No.
Output
54
[Y]
55
[YS]
9 - 20
Line Constants
"1" in Column No.
Output
56
[Z]
57
[ZS]
Where no subscript indicates the unreduced system. Each physical

conductor has a row and column in the matrix, as does each ground wire.
Table 9.4: Summary of Output Options for the Nominal-pi Model
ISEG
(58)
MUTUAL
(59)
MODAL
Flag indicating whether the specified ground wires are to be modelled as

continuous or segmented.
=0
continuous ground wires (default case).
=1
segmented ground wires (split at the towers).
Flag to request interference calculations with a communication circuit

parallel to the power circuit. (See further discussion in Section 9.5.2.).
=0
no interference calculation (default case).
=1
output of interference to a nearby communication circuit.
Flag to request the output of modal parameters.
(69-70)
=0
or blank
No output of modal parameters.
=1
Exact modal parameters.
Modal parameters are calculated from the exact [Z] and [Y] matrices at
the specified frequency. The printout includes R, X, and C, as well as
the characteristic impedance, wave velocity and attenuation for each
mode. The modal transformation matrix Ti is also listed. In these
calculations, the shunt conductance G is assumed to be zero.
=1
Exact modal parameters with R = 0.

The resistances are set to zero before the modal parameters are
calculated. The modal parameters are then evaluated exactly as in
the case of MODAL = 1. This produces a lossless approximation
that differs from the more conventional one obtained with MODAL
= 2 (see below). One of the differences of using this option
instead of the conventional one below is that the wave velocity of
the zero sequence mode will be less than the wave velocity of the
aerial modes, which is closer to the actual exact case.
9 - 21
Line Constants
=2
Lossless High-Frequency Approximation

Resistances are ignored and modal quantities are calculated from
high-frequency approximations. This approximation is often used
in lightning surge studies. It implies that all modes travel with the
speed of light. The self and mutual surge impedances in phase
quantities become Z ii = 60 log ( 2h i r i ) and
Zik = 60 log ( Dik d ik ) , respectively.
=3
Modal parameters are printed for both cases: MODAL = 1 and

MODAL = 2 above.
=3
Modal parameters are printed for both cases: MODAL = 1 and

MODAL = 2 above.
IFILE
(71-72)
Multiphase nominal-pi model of the line (short-line approximation) is

punched in LU7, in a format that can be used directly by the EMTP.
The nominal-pi model is only a short-line approximation which is not
valid for electrically long lines. A multiphase pi-equivalent model, which
is valid for long lines at one specific frequency, can be requested in the
LINE-MODEL module (see Section 9.6).
=0
or blank
Do not generate the nominal-pi model.
Generate the multiphase nominal-pi model. The model is punched

in LU7. The units for L and C are controlled by the specific value
of IFILE.
IFILE = 1 L in mH, C in F
IFILE = 2 L in mH, C in S
IFILE = 3 X in , C in F
IFILE = 4 X in , C in S
When the nominal-pi model is requested, there is the option of specifying

node names to be punched in the branch card for the model. To specify
the node names, the control card .NODES must immediately follow the
FREQUENCY card where the request is made.
9 - 22
Line Constants
9.4.7
123456
Special Request .NODES Card
1
2
3
4
7890123456789 012345 6789 012345 6789 012345
5
7
8
6789 012345 6789 6012345 6789 012345 67890
.NODES
SEND-1
RECV-1
SEND-2
RECV-2
SEND-3
A6
A6
A6
A6
A6
A6
.NODES
(1-6)
Keyword (left-justified).
SEND-1
Name of sending end of phase number 1.
RECV-3
(20-25)
RECV-1
Name of receiving end of phase number 1.
(30-35)
SEND-2
(40-45)
RECV-2
(50-55)
SEND-3
(60-65)
RECV-3
(70-75)
When there are more than 3 phases, the node names for the next 3 phases are provided on the next
card (with the same format, from column 20 to 75) and so on. The first 6 columns of these
additional cards must be either blank or contain the keyword .NODES.
9.4.8
Example of Data Deck for the Line-Parameters Module
BEGIN NEW DATA CASE

LINE CONSTANTS
c
TEST CASE 2 (LTC2). Line-Parameters calculation.
c .............................................................................
Files
ltc2.out
ltc2.pun
Line-Parameters
c "JOHN DAY-LOWER MONUMENTAL LINE" (222 Km). DATA AS IN BPA'S. Nov. 12, 1985.
c New Conductors Data Format (INOPT=1)
9 - 23
Line Constants
METRIC
1.3636 .03240 4
1.3636 .03240 4
2.3636 .03240 4
2.3636 .03240 4
3.3636 .03240 4
3.3636 .03240 4
0.5000 1.6216 4
0.5000 1.6216 4
BLANK
C Frequency Card
100.
500.
.nodes
k-a
100.
1000.
BLANK
BLANK
BEGIN NEW DATA CASE
BLANK
9.4.9
(A)
1
40.6908-6.3246
40.6908-5.8674
40.6908-0.2286
40.6908 0.2286
40.6908 5.8674
40.6908 6.3246
9.8044-3.9319
9.8044 3.9319
15.240
15.240
23.622
23.622
15.240
15.240
30.023
30.023
000011 101000 1
m-a
k-b
011000 010000 1
222. 10001
m-b
k-c
222. 00101
1 1
m-c
1 0
Description of Output from the Line-Parameters Module

Listing of Conductor Characteristics:
The information contained on the conductor cards of the input data deck is
printed for the record more or less in its original form, with the following
exceptions
(B)
(1)
In place of height at tower and midspan, the average height is listed

as y-coordinate.
(2)
The order of the conductor cards in the input data deck is arbitrary,
while the order in the listing will always be as follows: conductors
first encountered with phase numbers 1,2,3, ... , followed by
conductors with already-existing phase numbers (= 2nd, 3rd, 4th, ...
conductors in bundles or parallel circuits), followed by ground wires
(phase number = 0).
(3)
While a single conductor card may specify M conductors with the

BUNDLE data option, all M conductors will be listed separately in the
output.
Line Parameters:
Since all matrices are symmetric, only values in and below the diagonal are
printed, as indicated below.
9 - 24
Line Constants
Only lower triangular matrix is printed
All matrices are complex, except the susceptance (or capacitance) matrices for
the system of physical conductors and for the system of equivalent phase
conductors. Real and imaginary parts are printed above each other, as
indicated below.
real part
imaginary part
Impedance Matrices: The matrix elements of the impedance matrices per

kilometre or per mile are defined as follows:
Zi,k =
mutual impedance between i and k,
Zi,i =
self impedance of i, with current returning through ground (and

through ground wires if there are any and if they have been
eliminated).
Capacitance Matrices: The matrix elements of the susceptance (or

capacitance) matrices per kilometre or mile are defined as follows:
Ci,k =
negative value of susceptance between i and k,
Ci,i =
sum of all susceptances from i to all other conductors and to ground.
Symmetrical Components Matrices: Note that the matrices for symmetrical

components have their rows ordered in the sequence "zero (0), positive (1),
negative (2) of first three-phase circuit, (0), (1), (2) of second three-phase
circuit, etc.", whereas the columns have (1) and (2) exchanged and are thus
ordered "(0), (2), (1) of first circuit, (0), (2), (1) of second circuit, etc.". This
trick makes these matrices symmetrical again, as indicated below.
9 - 25
Line Constants
(0),(2),(1)...(0),(2),(1)
(0)
(1)
(2)
...
(0)
(1)
(2)
From this modified row and column numbering, it follows that

Z1,1 = Z2,2 within any three-phase circuit.
Z1,0 = Z0,2 within any three-phase circuit etc., but
Z1,0 = Z0,1, etc.
If there are only two equivalent phase conductors, a two-pole DC line is
assumed. In this case, zero sequence refers to the operation where equal
currents go into both poles and return through ground (and through ground
wires if they exist and were eliminated), and positive sequence refers to the
operation where the current goes into one pole and returns through the other.
For three or more equivalent phase conductors, only three-phase circuits are
assumed, with numbers 1,2,3 forming the first circuit, numbers 4,5,6 forming
the next circuit, etc. If the number of phases were 7 or 8, the last one or two
phases would simply be ignored. If the number were 9, then three three-phase
circuits would be assumed.
9.5 Special Calculation Options

9.5.1
Calculation of Electric Field Strength at Ground Level
The Line Parameters Module has an option to calculate electric field strength at ground level, and
its usage is described in Section 9.4.4.
The electric field strength at ground level is difficult to evaluate if the terrain is irregular or if
objects such as vehicles or buildings are close to the line. In the following, it is assumed that the
terrain is perfectly flat, that the conductors are perfectly horizontal, and that there are no nearby
objects.
The charges on the conductors are given by:
9 - 26
Line Constants
(9.1)
[ Q ] = [ Q ] 1 [ V ] = [ C ] [ V ] kC/km
or
n
Cj =
Cjk Vk
(9.2)
kC/km)
k=1
where n is the number of conductors, Cjk are the elements of the capacitance matrix in F/km, and
Vk is the root-mean-square phasor value of the line-to-ground voltage of conductor k in kV.
Qj
Ej
Hj
Lj
-Qj
Figure 9.4:
Contribution of Conductor j to Field Strength in P
The contribution from charge Qj on conductor j and from charge -Qj on the image of this
conductor to the field strength at point P (Figure 9.4 above) is
Q j Hj
E i = -----------------------------------2
2
o ( H + L )
kV/m
(9.3)
if 0 = 10-6/(36) in F/km, and Hj and Lj in m.
9 - 27
Line Constants
The magnitude of the total electric field strength at point P on the ground is
n
E total =
Ej
(9.4)
kV/m
j=1
which is the value printed by the program. Note that Etotal is a root-mean-square value since the
voltages were given as root-mean-square values. The instantaneous value of the field strength
would therefore be
E total = 2 E total cos ( t + )
(9.5)
with being the angle of the phasor value

n
Ej
(9.6)
j=1
9.5.2
Mutual Impedance with Communications Lines
When flag MUTUAL in the FREQUENCY CARD (Col. 59) is set to 1, the mutual impedances from the
equivalent phase conductors 1,...N-1 to the N-th last equivalent phase conductor will be printed, as
well as the impedance of the system of equivalent conductors [ZE]. This is useful to study
interference with communication lines, where the N-th equivalent phase conductor must represent
the communication line (any type of conductor can be used for it, because the conductor type has
no influence on mutual impedances). The longitudinally induced voltage in the N-th equivalent
phase conductor is then,
3 ...
11
22
33
power line
-VN
N
communication line
9 - 28
Line Constants
V N = Z N1 I1 + ZN2 I2 +... + ZN,N-1 IN-1
(9.7)
In addition, it is assumed that equivalent phase conductors 1,2,3 belong to three-phase circuit I;
4,5,6 to three-phase circuit II, etc. The mutual impedances are then also given for currents
expressed in symmetrical components, or
V N =
ZzeroI IzeroI + ZposI IposI + ZnegI InegI

+ ZzeroII IzeroII + ZposII IposII
+ ZnegII InegII + ...
with IzeroI, IposI, InegI being the zero, positive, negative sequence currents of circuit I, etc. The
symmetrical components are unnormalized,
I zero
1 1 1
I1
I pos = 1 3 1 a a 2 I 2 with a = ej120

I neg
1 a
I3
(for normalized symmetrical components, the factor in the above equation would be 1/ 3 instead
of 1/3).
9.6 Line-Model Module

The Line-Model module can produce transmission line models for steady state and for transients
studies. The data deck for this module follows the general structure shown in Table 9.1 (Section
9.3). The required data cards for this module are described next.
9.6.1
Line-Model Control Card
The Line-Model Control Card determines which model will be generated (e.g. frequency
dependent, constant parameter, etc.), and which input option will be used.
9 - 29
Line Constants
The normal required information to characterize the transmission line is the geometric location of
the conductors and their electrical characteristics. This information is specified in the
CONDUCTOR data cards. However, when the line geometry is not known, it is still possible to
generate reasonably accurate line models from the 60-Hz (or any other specific frequency) positive
and zero sequence impedances. This input option, sometimes referred to as the "poor mans
frequency dependent model" can be accessed by using the keyword "LBUILD" in the 'matrix' field
(columns 30-39) of the LINE-MODEL control card:
The description of the fields and the specific meanings for each line model are given below as
follows:
Card Format
FD-LINE
CP-LINE
PI-EXACT
SCAN
Line-Rebuild Option MATRIX = LBUILD

Fixed-Parameters Option MATRIX=CONSTANT
Card Format
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
LINE-MODEL
Model
Matrix
Scale
FMIN
NPDEC/
DELF
NDEC/
FMAX
A10
A10
A10
A10
E10.0
I10/E10.0
I10/E10.0
Set the keyword "LINE-MODEL" in columns 1-10.

The LINE-MODEL card specifies the characteristics of the line model to be generated by the
program. Output information on the processing of the model is listed in the output file. The model
itself stored in the punch file in a format that can be read directly by the EMTP.
The following line models are available:
FD-LINE
Frequency dependent line model for transients simulations (also

known as JMARTI line model).
CP-LINE
Constant-parameter line model for transients simulations (also

known as Dommel line model).
9 - 30
Line Constants
PI-EXACT
Exact frequency domain representation of a line at a given

frequency, for steady-state and frequency scan simulations.
SCAN
Generate line parameters and other information. No model is

generated.
FD-LINE Model
The frequency dependent line model is used to represent the line in transients simulations with the
EMTP. This model represents the true nature of a transmission line by modelling the line
parameters as distributed and frequency dependent. The line resistance and inductance are
evaluated as functions of frequency, as determined by skin effect and ground return conditions.
The capacitance is assumed constant. A non-zero constant shunt conductance G (default value =
0.2x10-9 S/km) is included in the model.
The structure for the "fd-line" model request in the LINE-MODEL card is as follows:
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
LINE-MODEL
Model
Matrix
Scale
FMIN
NPDEC
NDEC
A10
A10
A10
A10
E10.0
I10/E10.0
I10/E10.0
Keyword
(1-10)
Set to "LINE-MODEL".
MODEL
Set MODEL = FD-LINE.
(20-29)
MATRIX
See Section 9.6.8 for additional comments.
(30-39)
= QREAL
In the general case of an untransposed line, the elements

of the transformation matrix relating phase and modal
quantities are complex numbers. Also, the matrix has
different values at different frequencies. The fd-line
model, however, makes the approximation of using a
constant transformation matrix of real numbers for all
frequencies in the modelling interval (default).
9 - 31
Line Constants
The optimum frequency at which to evaluate this real
constant transformation matrix is determined
automatically by the program (default option), or it can
be specified externally by the user in the FREQUENCY
control card.
After calling the eigenanalysis routines to evaluate the
exact (complex) transformation matrix of the line at one
frequency, the matrix is rotated and normalized. The
imaginary part is then discarded and the remaining real
part is taken as the "correct" transformation matrix to
evaluate the line parameters and propagation functions at
all frequencies. The errors due to this approximation are
estimated in the Q-Error Table listed in the output file
= BALANCED
The line is modelled as perfectly transposed. The

diagonal and the off-diagonal elements in the reduced
Zphase and Yphase matrices are averaged out. The
balanced-line transformation matrix used is the
generalized Clarke transformation (,,) for an mphase line.
Unless the line is actually transposed, the results using
this option are usually poorer than with the default
"qreal" option. Check the Q-Error table in the output
listing for error indicators.
= ZDOUBLE
This option applies only to double-circuit lines. The line

is modelled as consisting of two separate circuits, each
circuit perfectly balanced with respect to itself and to the
other circuit.
Under these conditions, the only coupling between the
two circuits is zero sequence coupling. A special
transformation matrix corresponding to this condition is
used.
Unless the transposition scheme of the line approaches
the ideal zero coupling condition on which this option is
based, better results are usually obtained with the default
"qreal" option. Check the Q-Error Table in the output
listing for an indication of the errors
SCALE
(40-49)
The only available option is "LOG" (which is the default option). A log(f)
scale is used to generate the frequency dependent functions in the model.
9 - 32
Line Constants
FMIN
(50-59)
NPDEC
(60-69)
NDEC
(70-79)
Lower limit of the frequency interval in which the line propagation

function e-l and the line characteristic impedance Zc are synthesized
(fitting interval). Default value is FMIN=0.1Hz.
Number of equally spaced points (on a log scale) in each decade of the
fitting interval. NPDEC can be 10, 20, ..., up to 90 points per decade. Inbetween values (e.g., 15) are not permitted. Default value NPDE=10.
Number of decades (Default = 8). Defines the length of the fitting interval
FMAX = FMIN x 10NDEC.
CP-LINE Model
This option produces a constant parameters line model for transients simulations. The parameters
are evaluated at the frequency specified in the FREQUENCY control card (default = 60 Hz). In this
model the line is represented as a lossless, distributed LC line, with the total series resistance
lumped as R/2 in the middle of the line and R/4 at each end. The shunt conductance G is taken as
zero.
The data for the "cp-line" model in the LINE-MODEL card are as follows:
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 01234567890123456789012345678901234567890
LINE-MODEL
Model
Matrix
A10
A10
A10
Keyword
(1-10)
MODEL
Set MODEL = CP-LINE.
(20-29)
MATRIX
(30-39)
= QREAL
A constant real transformation matrix, determined

automatically by the program (default) or specified in the
FREQUENCY card, is used for the model.
9 - 33
Line Constants
= BALANCED
The line is modelled as balanced (perfectly transposed).

The generalized Clarke (,,) transformation is used for
the model.
= ZDOUBLE
Applies only to double-circuit lines. It is assumed that

there is only zero-sequence coupling between the two
circuits.
PI-EXACT Line Model

The "pi-exact" line model is used for STEADY-STATE or FREQUENCY SCAN solutions; it is not
valid for transients simulations. This model is an exact lumped-impedance multiphase
representation of the line as seen from its end points. This model is not adequate for transients
simulations because, even assuming constant line parameters, the parameters of the model are
different for different frequencies.
If a lumped-parameter model is desired for transients studies, it is better to use cascaded short-line
sections of nominal pi-circuits. Nominal pi-circuits can be obtained as an option of the LINEPARAMETERS module (see Section 9.4). In general, however, if a simple line model is desired for
transients analysis, it is better to use the "CP-LINE" model (see above). This model gives much
faster and much more accurate results than cascaded nominal pi-circuits.
The pi-exact equivalent is the line model that should be used for steady-state solutions and for
frequency scans. The punched output for this model is given in terms of a Y-matrix representation
that includes the series and shunt branches of the multiphase pi model.
The pi-exact model is produced for the frequency range specified in the LINE-MODEL card.
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
LINE-MODEL
Model
Matrix
Scale
FMIN
NPDEC/
DELF
NDEC/
FMAX
A10
A10
A10
A10
E10.0
I10/E10.0
I10/E10.0
Keyword
(1-10)
Set to LINE-MODEL.
MODEL
Set MODEL = PI-EXACT.
(20-29)
MATRIX
(30-39)
9 - 34
Line Constants
= QCOMPLEX Since the pi-exact model is produced in phase quantities,
it is not subject to the transformation matrix modelling
constraints of transients line models. The exact complex
transformation matrix can then be correctly used at each
frequency at which the pi model is requested (default).
= QREAL
In the general case of an untransposed line, the elements

of the transformation matrix relating phase and modal
quantities are complex numbers. Also, the matrix has
different values at different frequencies. The fd-line
model, however, makes the approximation of using a
constant transformation matrix of real numbers for all
frequencies in the modelling interval.
The optimum frequency at which to evaluate this real
constant transformation matrix is determined
automatically by the program (default option), or it can
be specified externally by the user in the FREQUENCY
control card.
After calling the eigenanalysis routines to evaluate the
exact (complex) transformation matrix of the line at one
frequency, the matrix is rotated and normalized. The
imaginary part is then discarded and the remaining real
part is taken as the "correct" transformation matrix to
evaluate the line parameters and propagation functions at
all frequencies. The errors due to this approximation are
estimated in the Q-Error Table listed in the output file
SCALE
= BALANCED
The line is assumed perfectly transposed. The line series

impedances and shunt admittances, as determined from
the line geometry, are averaged out.
= ZDOUBLE
For double circuit lines. The line is assumed zerosequence coupled only.
= LIN
The pi-exact matrices are generated over a linear

frequency range defined by the interval [FMIN, FMAX] at
discrete increments DELF.
= LOG
The pi-exact matrices are produced over a logarithmic

frequency range in the interval [FMIN, FMAX], where
FMAX = FMIN x 10NDEC, and NPDEC is the number of
points per decade. (See below for the specification of
FMIN, NPDEC, and NDEC.)
(40-49)
9 - 35
Line Constants
SCALE = LIN
FMIN
(50-59)
Initial frequency of frequency range in Hz. The built-in

default value is 0.0 Hz.
If FMIN < 0 then only one pi-exact equivalent, at
frequency |FMIN|, is generated. This option is useful for
steady-state solutions at one frequency (when this option
is requested, the fields 'scale', NPDEC/DELF, and NDEC/
FMAX are ignored).
SCALE = LOG
Initial frequency of the frequency range in Hz. The

built-in default value is 0.1 Hz.
If FMIN < 0 then only one pi-exact equivalent, at
frequency |FMIN|, is generated. This option is useful for
steady-state solutions at one frequency (when this option
is requested, the fields 'scale', NPDEC/DELF, and NDEC/
FMAX are ignored).
NPDEC/
DELF
SCALE = LIN
DELF: Frequency increment in Hz. The built-in default

value is 100.0 Hz.
SCALE = LOG
NPDEC: Number of equally spaced points in each decade
(60-69)
of the logarithmic frequency range. Use multiples of 10.
The built-in default value is 10, maximum value is 90.
NDEC/FMAX
SCALE = LIN
(70-79)
FMAX: Last frequency in the frequency range in Hz.
The built-in default value is 5000.0 Hz.

SCALE = LOG
Number of decades (Default = 6). Defines the length of

the frequency range as: FMAX = FMIN x 10NDEC.
Parameter SCAN Option

The parameters scan option is requested with the keyword "scan" in the 'MODEL' field of the LINEMODEL card. This option generates an output listing showing the transformation matrices, line
parameters, and wave functions over the requested frequency range. No line model is generated.
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 01234567890123456789012345678901234567890
LINE-MODEL
Model
Matrix
A10
A10
A10
9 - 36
Line Constants
Keyword
(1-10)
Set to "LINE-MODEL"
MODEL
MODEL = SCAN
(20-29)
MATRIX
= QREAL
Real transformation matrix from Q exact at one

frequency.
= QCOMPLEX
Complex transformation matrix, evaluated exactly at

each frequency.
= BALANCED
Clarke transformation for balanced lines.
= BALPAR
Same as "balanced" but more compact output listing.
= ZDOUBLE
Transformation matrix for double circuit lines with only

zero-sequence coupling
(30-39)
SCALE
(40-49)
FMIN
(50-59)
NPDEC
(60-69)
NDEC
(70-79)
The only available option is "LOG" which is the default option. The line
parameters and wave functions are generated over a logarithmic
frequency range.
Lower limit of the frequency interval in which the line propagation
function e-l and the line characteristic impedance Zc are synthesized
(fitting interval). Default value is FMIN=0.1Hz.
Number of equally spaced points (on a log scale) in each decade of the
fitting interval. NPDEC can be 10, 20, ..., up to 90 points per decade. Inbetween values (e.g., 15) are not permitted. Default value NPDE=10.
Number of decades (Default = 8). Defines the length of the fitting
interval FMAX = FMIN x 10NDEC.
Line-Rebuild Option MATRIX = LBUILD

The line-rebuild option is requested with the keyword "LBUILD" in the 'matrix' field of the LINEMODEL card. This is a data input option that can be used with any of the available line models.
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 01234567890123456789012345678901234567890
LINE-MODEL
Model
Matrix
A10
A10
A10
9 - 37
Line Constants
Keyword
(1-10)
Set to "LINE-MODEL"
MODEL
Set MODEL to any of the available line models described above.
(20-29)
MATRIX
= LBUILD
(30-39)
The CONDUCTOR data cards for this option do not contain

the line geometry as in the standard case, but the 60 Hz (or
any other specific frequency) values of the zero and
positive sequence impedances: R0, L0, G0, C0, and R1, L1,
G1, C1. Optionally, for extra accuracy, the conductor's DC
resistance can be supplied. The format of the CONDUCTOR
cards is described in Section 9.6.4.
From the 60 Hz parameter information, the program builds
an equivalent balanced arrangement of phase conductors
that matches the specified 60 Hz sequence parameters. It
also estimates the skin effect characteristics of the
conductors. (this estimate is better if the conductor's DC
resistance is specified.)
After rebuilding the line geometry and conductor
characteristics, the requested line model is processed as in
the ordinary case. Since the reconstructed line is assumed
to be balanced, the transformation matrix is the same one
used for balanced lines (generalized Clarke).
Fixed-Parameters Option MATRIX = CONSTANT

The fixed-parameters option is requested with the keyword "CONSTANT" in the 'matrix' field of the
LINE-MODEL card. As in the "LBUILD" case, this option can be used with any of the available line
models.
1
1234567890
2
3
4
5
6
7
8
123456789 0123456789 0123456789 01234567890123456789012345678901234567890
LINE-MODEL
Model
Matrix
A10
A10
A10
Keyword
(1-10)
9 - 38
Line Constants
MODEL
Set MODEL to any of the available line models described above.
(20-29)
MATRIX
(30-39)
= CONSTANT
As in the case of the line-rebuild option (subsection

8.8.1.5) the line data is specified in terms of the 60 Hz
(or any other frequency) zero and positive sequence
parameters, except that no DC resistance is used in this
case. The CONDUCTOR cards for this option are
described in Section 9.6.5.
As opposed to the line-rebuild option that rebuilds the
line geometry in order to be able to consider the
frequency dependence of the line parameters, the fixedparameters option assumes that the line parameters R, L,
G, and C remain constant as the frequency changes.
This option can be used in combination with the fd-line
model to create what is mathematically equivalent to a
BALANCED cp-line model where R is truly distributed
rather than lumped in three places.
In general, the line-rebuild option gives much better
results than the fixed-parameters option, and is the
recommended model when only one-frequency data
(e.g., 60 Hz) is available.
9.6.2
Units Control Card [3]
Specifies whether the S.I. metric or the English (British) system of units is used for the conductors
and line data.
UNITS
INOPT
1
2
3
4
5
6
7
8
12345678 9012345678901234567890123456789 0 1234567890123456789012345678901234567890
A8
I1
UNITS
Flag identifying units used for conductor and line data.
(1-8)
9 - 39
Line Constants
= METRIC
The S.I. system of units is used for conductor and line

data (mm or cm are used for conductor diameter and
bundle spacing depending on the specified data input
option INOPT (described below)).
Units in conductor cards
Resistance
/km
Diameter
mm if INOPT = 1
cm if INOPT = 2
Height & spacing
Bundle spacing
mm if INOPT = 1
cm if INOPT = 2
Units in the frequency card

Line length
= ENGLISH
or BRITISH
km
The English system of units is used for conductor and

line data.
Units in conductor cards
Resistance
/mile
Diameter
inches
Height & spacing
feet
Bundle spacing
feet
Units in the frequency card

Line length
INOPT
miles
Default indicating option is used in the input format.
(40)
= 2 (default)
This option corresponds to the traditional EMTP format

for line constants.
=1
This new format allows for electric field strength

calculations in the LINE-PARAMETERS module.
9 - 40
Line Constants
9.6.3
Conductor Data Cards [4]
The conductor cards contain the geometrical data of the line and the characteristics of the
individual conductors.
The meaning of the different fields in the CONDUCTOR cards is basically the same as explained in
Section 9.4 for the LINE-PARAMETERS module. A summary of the required information and the
differences in the present module are presented next.
As indicated in the UNITS card, two input formats are supported for the conductors data. The
option INOPT = 2 (default) corresponds to the traditional EMTP format. The new format INOPT = 1
allows for two additional fields (columns 73-80) which are used for electric field calculations in
the LINE-PARAMETERS module.
In the LINE-MODEL module, frequency dependence due to skin effect is always calculated. The
approximation for tubular conductors is used for these calculations. Flag IXTYPE (columns 17-18)
is assumed to be 4 by default and variable SKIN (columns 4-8) must be positive. The voltage
specification fields for electric field calculations, columns 73-80 for INOPT = 1, are not used in this
module.
Format for INOPT = 2 (Default):
E8.0
IXTYPE
REACT
I2 E8.0
DIAM
E8.0
HORIZ
E8.0
VTOWER
E8.0
VMID
E8.0
SEPAR
E8.0
ALPHA
NBUND
E5.0
RESIS
E6.0
I2
Format for INOPT = 1 (New Option)
SKIN
RESIS
I3
E5.0
E8.0
REACT
DIAM
I2
E8.0
E8.0
HORIZ VTOWER
E7.0
78.0
5
9012345
7
8
678 789012 34567890
VMID
NBUND
1
2
3
4
123 45678 90123456 78 90123456 78901234 5678901 2345678
IXTYPE
I3
SKIN
IPHASE
IPHASE
1
2
3
4
5
6
7
8
123 45678 90123456 78 90123456 78901234 56789012 34567890 12345678 90123456 789012 345678 90
ALPHA
E7.0
I3
E6.0
There must be one conductor card for each physical conductor in the line. This includes each
subconductor in a bundle and also the ground wires. There is no conductor card associated with
9 - 41
Line Constants
the earth return path. Bundled conductors specified using the bundle input option SEPAR (59-66)
and ALPHA (67-72) use a single conductor card.
To facilitate data entry of identical conductors, if any or all of the first six fields (IPHASE, SKIN,
RESIST, IXTYPE, and REACT) is left blank, the value of the corresponding variable will be assumed
to be the same as in the previous card.
In the following description, N conductors are assumed (column numbers refer to INOPT=2 above):
IPHASE
(1-3)
The phase number to which the conductor belongs. If more than one
conductor is given the same phase number, this means that the
conductors are electrically connected (connected in parallel). This is the
case, for instance, of individually specified conductors in a bundle
(fields SEPAR and ALPHA left blank). It could also be used, for instance,
to internally combine two parallel lines when it is not desired to
preserve their individual identity. Phase numbers for conductors must
follow the sequence 1, 2, 3, ..., N with no missing phases.
Set IPHASE = 0 for a ground wire (ground is phase number zero, by
definition).
SKIN
(4-8)
RESIS
(9-16)
Flag for skin effect correction defined as Thickness/Diameter (T/D) of

an equivalent tubular conductor (see Figure 9.5 below)
>0
Assume a tubular conductor with T/D = SKIN.
= 0.0 or
BLANK
Neglect the skin effect correction (see also the

description of RESIS below).
= 0.5
Assume a solid conductor.
DC resistance of the conductor in units of /km or /mile, according

to UNITS.
(17-18)
KTYPE = 4 or blank. Internal reactance is corrected for skin effect from

the geometry of the tubular conductor defined by SKIN.
REACT
REACT is the relative permeability and it is automatically defaulted to
KTYPE
(19-26)
r = 1.0 if left blank.
DIAM
Outside diameter of the conductor in units of:
(27-34)
HORIZ
(35-42)
centimetres
millimetres
inches
if UNITS = BRITISH
Horizontal distance of the conductor from the reference point x = 0 (see

diagram in Figure 9.5 below)), in units of:
9 - 42
Line Constants
metres
if UNITS = METRIC
feet
if UNITS = BRITISH
The reference point x = 0 can be located anywhere since distances are

relative. Distances to the right of x = 0 are positive and distances to the
left are negative.
VTOWER
(43-50)
Vertical height of the conductor above the ground, at the tower in units
of:
metres
if UNITS = METRIC
feet
if UNITS = BRITISH
When both VTOWER and VMID are specified, an average height is

calculated by the program using the formula:
VTOWER VMID
2 VMID + VTOWER
VMID + -------------------------------------------------- = ---------------------------------------------------------3
3
If only VTOWER is specified, then this height will be assumed uniform

along the line.
VMID
(51-58)
Midspan height of the conductor above the ground, at the tower in units
of:
metres
if UNITS = METRIC
feet
if UNITS = BRITISH
Example of a conductor card:

C
1
2
3
4
5
6
7
8
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
1.3871 .03480 2
.7092
1.802
-.75
69.0
39.0
1.3871 .03480 2
.7092
1.802
.75
69.0
39.0
2.3871 .03480 2
.7092
1.802
40.0
69.0
39.0
18.0
0.0
2
0 0.5
3.10 1
.484
.495
20.0 133.0
9 - 43
Line Constants
Conductors
a, b, c, d:
earth wire
GMR = 0.7092"
DIAMETER = 1.802"
phase 2
c d
phase 1
a b
R = 0.0398 /mile
T/D = 0.3871
103 at midspan
133 at tower
.75
69 at tower
-.75
39 at midspan
Ground wire R = 3.1 /mile
reference
Figure 9.5:
Reactance for 1
spacing at 60 Hz =
0.484
DIAMETER = 0.495"
T/D = 0.5 (solid)
Example of Conductor Card (INOPT = 2)
Bundle Input Option:

Specified using parameters NBUND, SEPAR, ALPHA
Leave NBUND blank for single conductors. For a bundle with K conductors (see Figure 9.6 below)
there are three options available to enter the bundle data:
(A)
Enter a normal conductor card for each of the K conductors in the bundle, and
leave NBUND blank. This may be too time-consuming for a regular
symmetrical bundle, but it is the only option available for asymmetrical
bundles.
(B)
Convert the bundle into a single equivalent conductor using the distance
averaging formulas available for this purpose and enter the equivalent
conductor in a normal conductor card (leave NBUND blank). This alternative is
less accurate than options (A) and (C).
(C)
Specify a symmetrical bundle using the fields NBUND, SEPAR, and ALPHA.
There should be only one CONDUCTOR card per bundle.
The bundle fields for option (C) are specified as follows:
9 - 44
Line Constants
NBUND
Number of conductors in the bundle.
(79-80)
Spacing between adjacent conductors in the bundle, in units of:
SEPAR
(59-66)
ALPHA
(67-72)
centimetres
millimetres
inches
UNITS = "ENGLISH"
Angular position of the first conductor (or any conductor) of the bundle,
in units of degrees. Positive angles are measured counter-clockwise.
SEPAR
1
ALPHA
2
4
3
Figure 9.6:
9.6.4
Sample sketch of a bundle with NBUND = 6 conductors, with angle

ALPHA = 30o
Conductor Cards for the Line-Rebuild Option
The Line-Rebuild option is described in Section 9.6.1. The format for the conductor cards is
indicated next.
9 - 45
Line Constants
KCIRCT
1
2
3
4
5
6
7
8
12 3456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 01234567890
FPARAM
RZERO
LZERO
GZERO
CZERO
RDC
I2
E10.0
E10.0
E10.0
E10.0
10.0
E10.0
KCIRCT
Number of phases.
(1-2)
FPARAM
(10-19)
RZERO
Frequency in Hz at which the sequence parameters are given. The builtin default value is 60 Hz.
Zero sequence resistance in /km, for UNITS = METRIC, or /mile for
(20-29)
UNITS = ENGLISH.
LZERO
Zero sequence inductance in mH/km, for UNITS = METRIC, or mH/mile

for UNITS = ENGLISH.
(30-39)
GZERO
Zero sequence conductance in S/km, for UNITS = METRIC, or S/mile for
(40-49)
UNITS = ENGLISH.
CZERO
(50-59)
Zero sequence capacitance in F/km, for UNITS = METRIC, or F/mile

RDC
=0
DC resistance of the conductors in /km or /mile (according

to UNITS) will be estimated by the program.
>0
DC resistance of the conductors in /km or /mile (according

to UNITS). This parameter is optional for greater accuracy. If
(60-69)
not given, it will be estimated by the program.
2
3
4
5
6
7
8
1234567890123456789 0123456789 0123456789 0123456789 0123456789 012345678901234567890
RPOS
(20-29)
RPOS
LPOS
GPOS
CPOS
E10.0
E10.0
E10.0
10.0
Positive sequence resistance in /km, for UNITS = METRIC, or /mile

9 - 46
Line Constants
Positive sequence inductance in mH/km, for UNITS = METRIC, or mH/
mile for UNITS = ENGLISH.
LPOS
(30-39)
Positive sequence conductance in S/km, for UNITS = METRIC, or S/mile

GPOS
(40-49)
Positive sequence capacitance in F/km, for UNITS = METRIC, or F/

CPOS
(50-59)
9.6.5
Conductor Cards for the Fixed-Parameters Option
The Fixed-Parameters option is described in Section 9.6.1. The format for the conductor cards is
indicated next.
KCIRCT
1
2
3
4
5
6
7
8
12 3456789 0123456789 0123456789 0123456789 0123456789 0123456789 012345678901234567890
FPARAM
RZERO
LZERO
GZERO
CZERO
I2
E10.0
E10.0
E10.0
E10.0
10.0
KCIRCT
Number of phases.
(1-2)
FPARAM
(10-19)
RZERO
(20-29)
LZERO
(30-39)
GZERO
(40-49)
CZERO
(50-59)
Frequency in Hz at which the sequence parameters. The built-in default

value is 60 Hz.
Zero sequence resistance in /km, for UNITS = METRIC, or /mile for
UNITS = ENGLISH.
Zero sequence inductance in mH/km, for UNITS = METRIC, or mH/mile
Zero sequence conductance in S/km, for UNITS = METRIC, or S/mile for
UNITS = ENGLISH.
Zero sequence capacitance in F/km, for UNITS = METRIC, or F/mile
9 - 47
Line Constants
2
3
4
5
6
7
8
1234567890123456789 0123456789 0123456789 0123456789 0123456789 012345678901234567890
RPOS
(20-29)
LPOS
(30-39)
GPOS
(40-49)
CPOS
(50-59)
9.6.6
RPOS
LPOS
GPOS
CPOS
E10.0
E10.0
E10.0
10.0
Positive sequence resistance in /km, for UNITS = METRIC, or /mile

Positive sequence inductance in mH/km, for UNITS = METRIC, or mH/
Positive sequence conductance in S/km, for UNITS = METRIC, or S/mile
Positive sequence capacitance in F/km, for UNITS = METRIC, or F/
Frequency Card [5]
6
7
8
34567 8 9012345678901234567890
RHO
FMATRX
ALONG
ISEG
1
2
3
4
5
12345678 9012345678 90123456789012345678901234 56789012
E8.0
E10.0
E8.0
I1
RHO
Ground return resistivity in ohm-m.
(1-8)
FMATRX
Frequency, in Hz, at which to evaluate the transformation matrix.
(9-18)
Default = Internally determined by the program for the fd-line model.
60 Hz for the cp-line model.
Frequency at which to evaluate the transformation matrix Ti in line
models with 'matrix' = "qreal" option.
9 - 48
Line Constants
For the frequency dependence line model (fd-line), the program will
automatically select an optimum value of FMATRX for the range of
switching transients. This value is based on asymptotic conditions for
the particular line under consideration. Typical values are in the range
from 500 Hz to 5 kHz with a mean around 1000 Hz. The selection of an
optimum value is based on constancy of Ti within the typical frequency
range for switching transients. For studies involving other frequency
ranges (e.g., lightning studies), FMATRX should be supplied externally.
For the constant parameters line model (cp-line), the default value for
FMATRX is 60 Hz. If FMATRX < 0 is specified for this model, the
program will determine FMATRX internally, using the same procedure as
for the fd-line model.
ALONG
Line length in km, for UNITS = METRIC, or miles for UNITS = ENGLISH.
(45-52)
ISEG
=0
Ground wires are not segmented (default).
=1
Ground wires are segmented.
(58)
9.6.7
Optional Control Cards [6]
Optional control cards allow the user to specify additional information on the transmission system
(e.g. node names and transposition scheme) and to have a greater degree of control over internal
processes (e.g. rational functions fitting). They can also be used to request additional output and
debugging information.
These cards are optional and can appear in any order after the FREQUENCY card. Their presence is
signalled by a dot '.' in column one joined to a key word.
The following optional control cards are available:
(A) Associated with the processing of the line parameters and functions:
.dbgline
.imbal
.outline
.phase
.nodes
.transp
(B) Associated with the rational functions fitting for frequency dependence line models:
.ctlfit
.dbgfit
.outfit
Unless otherwise indicated, control flags are either on or off, where 1 = on and 1 = off.
9 - 49
Line Constants
(A) Line Processing Control Cards

.NODES (Node Names)
Allows input of node names at the sending and receiving ends of each phase. The node names will
be printed in the line model branches of the punch file.
1
2
3
4
5
6
7
8
123456 7890123456789 012345 6789 012345 6789 012345 6789 012345 6789 012345 6789 012345 67890
.nodes
k-a
m-a
k-b
m-b
k-c
m-c
A6
A6
A6
A6
A6
A6
A6
1
2
3
4
5
6
7
8
1234567890123456789 012345 6789 012345 6789 012345 6789 012345 6789 012345 6789 012345 67890
k-d
m-d
k-e
m-e
k-f
m-f
A6
A6
A6
A6
A6
A6
Node names are assigned in sending-receiving end pairs. Additional cards can be used for node
specification of more than three phases.
.TRANSP (Transposition Option)

This option produces a line model for transposed lines based on the averaging of the series
impedance and shunt admittance matrices. This procedure is only an approximation to the correct
solution of modelling each transposition segment explicitly and then specifying the appropriate
node connections in the EMTP. Averaging is valid only when the length of the transposition
sections is several times smaller than the wavelength of the propagating signals.
Span Length Cards
.TRANSP
A9
NSPAN
1
2
3
4
8
123456789 01 23456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
spanl1
spanl2
spanl3
spanl4
spanl5
spanl6
I2
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
9 - 50
Line Constants
1
2
3
4
8
1234567890123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
spanl7
spanl8
spanl9
spanl10
spanl11
spanl12
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
.transp
(1-9)
Keyword to flag the transposition option.
nspan
(10-11)
Number of transposition sections.
spanl-j
Span length, sections 1 to 6 for the first card, and 7 to 12 for the second
card. Length in km (METRIC) or miles (ENGLISH).
Additional cards can be added for more that 12 transposition sections.
The sum of the lengths of all the transposition sections must equal the
total length of the line as specified in the FREQUENCY card to an
accuracy of three digits. This provides a check on the correct
specification of the transposition sections. There is no check, however,
on the specification of the phase sequence.
Phase Sequence in Transposition Section:
1
2
1234567890123456789 01 23
45
3
67 89 01
1
n1
1
n2
1
n3
1
n4
1
n5
1
n6
I2
I2
I2
I2
I2
I2
1
2
1234567890123456789 01 23
45
3
67 89 01
2
n1
2
n2
2
n3
2
n4
2
n5
2
n6
I2
I2
I2
I2
I2
I2
4
5
6
7
8
2345678901234567890123456789012345678901234567890
...
...
4
5
6
7
8
2345678901234567890123456789012345678901234567890
...
...
9 - 51
Line Constants
Phase sequence in section 1 is specified in Card 1 (20-79), namely:
1
= phase number as specified in 'iphase' in the CONDUCTOR

cards (see Section 9.6.3)
= phase number as specified in 'iphase' in the CONDUCTOR

cards (see Section 9.6.3)
n 1 , n2 , n 3 , n 4 , n 5 , n 6 , , n n
n 1 , n2 , n 3 , n 4 , n 5 , n 6 , , n n
At least one additional card is needed for each additional transposition section.
.Gphase (Phase Shunt Conductance)

This option allows the specification of values for the line shunt conductance G other than the
internal default value of 0.2x10-9 S/km (for those line models that assume nonzero G). The values
specified in these cards are the diagonal elements of the reduced (not the full conductors matrix)
Gphase matrix. The off-diagonal elements of Gphase are assumed to be zero.
1
2
3
4
8
1234567 890123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
.GPHASE
Gaa
Gbb
Gcc
Gdd
Gee
Gff
A7
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
1
2
3
4
5
6
7
8
1234567890123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
Ggg
Ghh
Gii
Gjj
Gkk
Gll
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
.GPHASE
(1-7)
Keyword to flag the shunt conductance option.
Gxx
(20-79)
Phase conductances (Gaa Gbb ... Gff in the first card and Ggg ...
Gll for the second card) in S/km for UNITS = METRIC or S/mile for
UNITS = BRITISH.
These are the conductances from conductors to ground. Additional
cards can be used as required.
Note:
A value of zero or blank for a given entry will be taken as

Gii = 0. and not as the default value of 0.2x10-9 S/km, which
is assumed when the .gphase option is not specified.
9 - 52
Line Constants
2
3
4
5
6
7
8
01 23 456789012345678901234567890123456789012345678901234567890
.OUTLINE
A8
IMONIT
1
12345678 90123456789
IWAVEF
.OUTLINE (Output from Line Processing)
I2 I2
.OUTLINE
(1-8)
Keyword to flag the output information option.
IWAVEF
Flag controlling the printout status of Z c =
(20-21)
Ap = e
IMONIT
( ZY l )
Z Y and
=1
The wave functions Zc and Ap are printed on the output file for
each line mode over the frequency range specified in the
LINE-MODEL card.
= -1
No output of the line wave functions (default value).
Flag controlling the monitoring of frequency being processed.
(22-23)
=1
The frequency being processed is indicated during the

evaluation of the line wave functions. (This is useful during
slow system processing.)
= 1
No frequency monitoring (default value).
.DBGLINE (Debug Line)
A8
IWD1
IWD
IWQ
IROT
.DBGLINE
IBUG1
1
2
3
4
5
6
7
8
12345678 90123456789 01 23 45 67 89 456789012345678901234567890123456789012345678901234567890
I2 I2 I2 I2 I2
9 - 53
Line Constants
.DBGLINE
(1-8)
Keyword to flag line processing debugging and control.
IBUG1
Flag controlling the amount of printout during line parameter

calculations. Valid range is 0,1,2, or 3 (default = 0).
(20-21)
IROT
(22-23)
Flag controlling the rotation of the modal transformation matrix Q or TI.

Default value is 1.
=1
Rotate the transformation matrix Ti to satisfy the condition:

modal shunt conductance matrix Gm = 0 when phase shunt
conductance matrix Gph = 0.
=2
Rotate the transformation matrix Ti to minimize the imaginary

part of its elements.
The following flags are valid when MODEL = SCAN and Transformation
Matrix = qcomplex in the LINE-MODEL control card:
IWQ
(24-25)
IWD
(26-27)
IWD1
(28-29)
Flag to control the printing of the modal transformation matrix Q or TI.

=1
Print the transformation matrix Q-complex at each frequency of

the frequency loop.
= 1
Do not print.
Flag controlling the printout of the product [Ymode . Zmode]. Default

value is 1 (no printout).
=1
Print the product [Ymode . Zmode] at each frequency of the

frequency loop. Ideally, for Ti-exact, [Ymode . Zmode] =
diagonal matrix.
= 1
Do not print.
Similar to IWD, above. Default value is 1 (no printout).

=1
Same as above, but with the elements of [Ymode . Zmode]

normalized so that the largest real part of any element equals one.
= 1
Do not normalize.
9 - 54
Line Constants
.IMBAL (Test Sources for Q-error Indicators)

Value of the voltage sources for the open-circuit/short-circuit Q-Error tests. In the open circuit
test, all the phases at the receiving end of the line are open. In the short circuit test, all the phases
at the receiving end of the line are short circuited.
1
2
3
4
5
6
7
8
12345678 90123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0123456789 0
.IMBAL
VOLT-A
PHASE-A
VOLT-B
PHASE-B
VOLT-C
PHASE-C
A8
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
.IMBAL
(1-8)
Flag to request the override of the test sources for the Q-error indicators.
VOLT-J
(20-29)
...
Voltages at the sending end. Magnitude of the voltage source VJ

connected from phase J to ground at the sending end of the line. (A short
circuit at the sending end can be simulated by making VOLT-J = 0.)
PHASE-J
Phase angle (in degrees) of voltage source VJ.
(30-39)
...
Additional cards can be added as required for additional phases. The internal default values (if the
.IMBAL option is not used) are as follows:
Va = 1 /0o
Vb = 1 /0o
Vc = 1 /120o
Vd = 1 /0o
Ve = 1 /0o
Vf = 1 /120o
(B) Fitting Control Cards

.CTLFIT (FIT-S Control)
Allows for additional user control over the fitting of the line wave functions Zc and Ap = e-l.
Additional output information can also be requested.
9 - 55
Line Constants
A7
IFITAL
IFITZC
IXDYN
.CTLFIT
IQUICK
NORMAX
1
2
3
4
5
6
7
8
1234567 890123456789 01 23 45 67 89 456789012345678901234567890123456789012345678901234567890
I2 I2 i2 i2 i2
.CTLFIT
(1-7)
Keyword to control the fitting process of fd-line models.
NORMAX
Maximum number of poles in the synthesis of the line wave functions

Zc and Ap. Default value is 25.
(20-21)
IQUICK
(22-23)
IXDYN
(24-25)
IFITZC
(26-27)
IFITA1
(28-29)
Flag controlling the accuracy of the fit of Zc and Ap.

Default value is 1.
=1
Much faster (fewer iterations) approximation of Zc and Ap is

produced. Some accuracy is sacrificed with respect to the
normal procedure.
= 1
Normal iteration procedure is used.
Flag controlling the low frequency approximation of Ap.

Default value is 1.
=1
Extra dynamics (extra poles and zeroes) are added to the

approximation of the low frequency region of the propagation
function Ap. This allows a more accurate simulation of very
short line sections (e.g., for breaker re-ignition studies) and of
very low frequencies (e.g., for trapped charge conditions).
= 1
No extra dynamics are added to the approximation of Ap. This

results in a lower order approximation but less reliable for short
line sections or very low frequencies.
Flag controlling the fitting of Zc in different modes.

=N
Only one mode "N" of the Zc function is fitted. Mostly for

testing purposes.
=0
Fit all modes (Default).
Flag controlling the fitting of Ap in different modes.
9 - 56
Line Constants
=N
Only one mode "N" of the Ap function is fitted. Mostly for

testing purposes.
=0
Fit all modes (Default).
To produce the fd-line model, synthesis of all modes is required. This is the normal case with the
default flags IFITZC = 0 or blank and ifital = 0 or blank.
.OUTFIT (FIT-S Output)

Controls the amount of output information on the fitting of the line functions.
A7
IPRAT
.OUTFIT
IPLOTF
ICOMPF
1
2
3
4
5
6
7
8
1234567 890123456789 01 23 45 6789012345678901234567890123456789012345678901234567890
I2 I2 I2
.OUTFIT
(1-7)
Keyword to control the amount of output generated during the fitting

process of fd-line models.
ICOMPF
Flag controlling the printout of a comparison table. Default value is -1

(no printout).
(20-21)
IPLOTF
(24-25)
IPRAT
(26-27)
=1
An output table is produced comparing the data functions Zc

and Ap as produced by the line constants routines and the
approximating rational functions produced by FIT-S.
= 1
No comparison table is produced.
Flag controlling the printout of a printer plot.

Default value is 1 (print).
=1
A printer plot is produced comparing the data functions and the

approximations.
= 1
No printer plot is produced.
Flag controlling the printout of poles and zeros tables. Default value is 1
(print).
9 - 57
Line Constants
=1
Tables are produced showing the location of the poles and

zeros of the rational function approximations of Zc and Ap.
Also shown are the RC equivalent network for Zc and the time
domain exponential representation of the approximating
functions.
= 1
No tables are produced.
.DBGFIT (FIT-S Debug)

Controls the amount of internal processing output from FIT-S.
.DBGFIT
IBUGF
1
2
3
4
5
6
7
8
1234567 890123456789 01 23456789012345678901234567890123456789012345678901234567890
A7
I2
.DBGFIT
(1-7)
Keyword to request debug information during the fitting process of fdline models.
IBUGF
Flag controlling the level of diagnostic printout. Valid range is 1-3

(default = 0). The higher the number, the larger the amount of
debugging output from FIT-S.
(20-21)
9.6.8
Transformation Matrices and Q-Error Indicators
For a perfectly balanced line, the modal transformation matrices to relate modal and phase
quantities do not change with frequency (constant transformation matrices) and can be chosen to
be real (e.g. generalized Clarke, as used by the program).
In the general case of the untransposed line, however, the transformation matrices change with
frequency. The line currents transformation matrix Ti is the matrix that diagonalizes the product
Yphase.Zphase where Yphase is the shunt admittance matrix in phase quantities and Zphase is the
series impedance matrix in phase quantities. The resulting Q or Ti matrix, determined by the
eigenanalysis routines, is complex. To standardize the results, Ti is normalized, using the
Euclidean Norm (whereby each column j is divided by k j =
Q2ij .
The voltages
transformation matrix Tv (which diagonalizes the reverse product Zphase.Yphase) is not determined
9 - 58
Line Constants
by the eigenanalysis routines but calculated directly from the relationship Tv = Ti-t (where the
superscript means inverse transposed).
Processing of the transients line models in the EMTP requires real transformation matrices Ti and
Tv. To obtain approximate Ti and Tv matrices, the columns of Ti-complex can be rotated to make
the imaginary parts of its elements small and then retain only the real parts.
In the case of the pi-exact model, the final form of the model is expressed in terms of self and
mutual phase quantities, and there is no impediment in using exact complex transformation
matrices at each frequency at which the model is produced. This model, however, is a onefrequency model, valid for steady-state solutions but not for transients simulations.
The cp-line model does not take into account the frequency dependence of the line parameters.
The model is formulated in terms of modal quantities, with the modal parameters R, L, and C
calculated exactly at only one frequency using the exact complex transformation matrix at that
frequency. Since the model assumes zero modal conductances (Gm = 0), the columns of the
transformation matrix Ti are rotated to satisfy this condition. As a result of this rotation, the
imaginary parts of the elements of Ti usually become very small. Since the EMTP requires Ti to be
purely real, only the real part of Ti (after the indicated rotation) is retained in the model (the punch
files has Ti = real).
The fd-line model takes into account the frequency dependence of the line parameters and the
distributed nature of the losses (including a finite inductance G). As in the case of the cp-line
model, however, the fd-line model is formulated in terms of modal quantities, and also has the
constraint of requiring a real constant transformation matrix Ti. Even though the fd-line model
does not assume zero modal conductances, the recommended criterion to rotate Ti is the same as
for the cp-line model, that is, Ti is rotated to satisfy the condition Gmode = 0. for Gphase = 0. This
default rotation can be overridden with the optional control card .DBGLINE (field 'irot'). Since G is
normally very small, the results obtained with both rotation criteria are very similar. It is
nonetheless believed that the default rotation gives more physically consistent results.
Q-error Indicators
A Q-Error table is printed out by the Line-Model module. This table gives an indication of the
possible errors when using a constant real transformation matrix Q or Ti instead of the exact
complex one at each frequency. A constant real Ti is used in the fd-line and in the cp-line models.
(An exact complex Ti at each frequency is used in the pi-exact model.)
The errors shown in the Q-Error table correspond to single-frequency steady-state comparisons for
unbalanced combinations of open and short circuit conditions. In these tests, all phases at the
receiving end of the line are open or all phases are shorted. Unbalanced sources are connected at
the sending end of the line (see Section 9.6.7 on the .IMBAL optional control card for the values of
these sources).
9 - 59
Line Constants
The percent errors shown in the Q-Error table for a given frequency correspond to the phase
voltage or current that has the largest error.
The Q-Error table is a qualitative guide and does not include all possible factors. As the frequency
goes higher than about 1000 Hz, the resonant peaks in the open and short circuit response curves
are relatively sharp and small phase errors can result in relatively larger magnitude differences.
Another factor that must be considered in these evaluations is that small open circuit currents can
be in relatively large error under unbalanced conditions. To give qualitatively meaningful results,
the error comparisons in the Q-Error table do not include currents or voltages smaller than 5% of
the largest values.
9.6.9
Examples of Data Decks for the Line-Model Module
The following benchmark example generates a frequency dependent line model with default
options.
BEGIN NEW DATA CASE
LINE CONSTANTS
c
TEST CASE 1A (LTC1A). Frequency Dependence Line-Model
c
NORMAL DATA DECK WITH ALL DEFAULT OPTIONS
c .............................................................................
FILES
ltc1a.out
ltc1a.pun
Line-Model
FD-LINE
c .............................................................................
C BPA'S 50O KV, 174-MILE, COULEE-RAVER DOUBLE CIRCUIT LINE
C (Original Deck of 12/10/73)
c .............................................................................
BRITISH
1.3636 .05215 4
1.602 -17.1875 49.06
49.06
1.3636 .05215 4
1.602 -18.25
48.0
48.0
1.3636 .05215 4
1.602 -19.3125 49.06
49.06
2.3636 .05215 4
1.602 -27.1875 85.06
85.06
2.3636 .05215 4
1.602 -28.25
84.0
84.0
2.3636 .05215 4
1.602 -29.3125 85.06
85.06
3.3636 .05215 4
1.602 -17.1875 121.06 121.06
3.3636 .05215 4
1.602 -18.25
120.0
120.0
3.3636 .05215 4
1.602 -19.3125 121.06 121.06
4.3636 .05215 4
1.602
17.1875 121.06 121.06
4.3636 .05215 4
1.602
18.25
120.0
120.0
4.3636 .05215 4
1.602
19.3125 121.06 121.06
5.3636 .05215 4
1.602
27.1875 85.06
85.06
5.3636 .05215 4
1.602
28.25
84.0
84.0
5.3636 .05215 4
1.602
29.3125 85.06
85.06
6.3636 .05215 4
1.602
17.1875 49.06
49.06
6.3636 .05215 4
1.602
18.25
48.0
48.0
6.3636 .05215 4
1.602
19.3125 49.06
49.06
0.5
2.61
4
.386 -9.0
163.96 163.96
0.5
2.61
4
.386
9.0
163.96 163.96
BLANK
C FREQUENCY CARD
C
rho
length
iseg
100.
174.
1
.nodes
k-a
m-a
k-b
m-b
k-c
m-c
9 - 60
Line Constants
BLANK
BLANK
BLANK
BEGIN NEW DATA CASE
BLANK
The following example is for a pi-exact model for steady-state frequency scans
BEGIN NEW DATA CASE
LINE CONSTANTS
c
TEST CASE 3 (LTC3). Pi-Exact steady-state line model.
c .............................................................................
Files
ltc3.out
ltc3.pun
Line-Model
Pi-Exact
log
0.1
1
10
c "JOHN DAY-LOWER MONUMENTAL LINE" (222 Km).
c New Conductors Data Format (INOPT=1)
METRIC
1
1.3636 .03240 4
40.6908-6.3246 15.240
1.3636 .03240 4
40.6908-5.8674 15.240
2.3636 .03240 4
40.6908-0.2286 23.622
2.3636 .03240 4
40.6908 0.2286 23.622
3.3636 .03240 4
40.6908 5.8674 15.240
3.3636 .03240 4
40.6908 6.3246 15.240
0.5000 1.6216 4
9.8044-3.9319 30.023
0.5000 1.6216 4
9.8044 3.9319 30.023
BLANK
C Frequency Card
c rho
length
iseg
100.
222.
1
BLANK
BLANK
BEGIN NEW DATA CASE
BLANK
The following example .id for the LBUILD line reconstruction option
BEGIN NEW DATA CASE
LINE CONSTANTS
c
TEST CASE 6 (LTC6). REBUILT LINE GEOMETRY
c
"John Day - Lower Monumental Line"
c .............................................................................
FILES
ltc6.out
ltc6.pun
Line-Model
fd-line
lbuild
c .............................................................................
METRIC
c .......f.........R.........L.........G.........C.........Rdc.................
3
60.
0.18736
3.6012
0.
0.007524 0.0162
0.017413 0.96731
0.
0.012027
BLANK
C Frequency Card (Only rho and length used in this option)
c rho
length
100.
222.
C .outline: iwavef,imonit
C .outfit: icompf,iplotf,iprat
9 - 61
Line Constants
C .ctlfit: normax,iquick,ixdyn,ifitzc,ifita1
.outline
1-1
.outfit
1 1-1
.ctlfit
1
BLANK
BLANK
BEGIN NEW DATA CASE
BLANK
9.7 FIT-S Module

The FIT-S module provides access to the fitting routines FIT-S as a separate program to generate
rational function approximations of user-supplied frequency domain transfer functions.
This option is invoked by the keyword "FIT-S" in the MODULE card of the Line Constants Program
data deck (see Table 9.1 in Section 9.3).
The function to be approximated is read in as a table containing magnitude and phase angle as a
function of frequency on a logarithmic scale (see FIT-S control card below). The rational function
approximation is given in the output file in frequency domain form (poles and zeros) and in time
domain form (k's and poles and time coefficients in a sum of exponential representation). An
equivalent circuit representation can also be obtained, upon request, through the 'ipnet' flag in the
.outfit control card.
9.7.1
FIT-S Control Card
1
2
3
4
5
6
7
8
12345 67890123456789 01234567890123456789 01234567890123456789012345678901234567890
FIT-S
PARFILE
TFORM
A5
A20
A40
The required input data is the magnitude and, optionally, the phase angle of the frequency domain
function to be synthesized. The name of the file containing the data curve is specified "FIT.DAT".
The names for the output and punch files are specified in the FILES card, or assigned externally.
The output and punch files have the same form as the ones for the fd-line model (see Section
9.6.1). The punch file will contain the time domain form of the approximation in terms of a sum of
exponential functions. The output file contains the approximating function in both frequency and
time domain forms, including, for non-delay functions, the value of the elements in an RC synthesis
network.
9 - 62
Line Constants
FIT-S
Keyword to request the external curve fitting option FIT-S.
(1-5)
PARFILE
(20-39)
TFORM
(40-80)
Name of file with the function data (Default - "FIT.DAT"). File

containing the function data, as described in Section 9.7.2.
Fortran READ format statement. This field must contain the Fortran
format specifier (including the external parentheses) for the program to
read the data function from file 'parfile'.
The data function is read point by point as follows:

READ (LU4,TFORM) FREQ, AMAG, APHASE
where:
TFORM
Is the FORMAT specified in columns 40 to 80 of the FIT-S control card.
LU4
The data curve is read from file 'parfile' which is internally connected to
LU4.
FREQ
frequency point in Hz. The data frequency points must be

logarithmically spaced.
AMAG
Magnitude of the data function.
APHASE
Phase angle in degrees (for radians use .ctlfit card).
9.7.2
Data Function File
The data curve file 'parfile' (attached to LU4) has the following structure:
(A)
It can contain any number of comment lines (identified by 'C ' or 'c ' in the first
two columns), as long as these lines are placed before the block of data points.
No comment lines are allowed in between data points.
(B)
Data points. These points will be read according to TFORM.
(C)
"END" (columns 1-3) at the end of the data points.
9 - 63
Line Constants
9.7.3
Data Function Format
The data points must be given on a logarithmic frequency scale as indicated in the following
diagram:
Logarithmically-spaced
frequency interval
fdc
fdc
fmin
fmax
finf
DC frequency point in Hz (e.g., 10-15 Hz).
This frequency must be less than or equal to fmin. The data at this point
is used by the program to match the asymptotic behaviour when f=0.
finf
Infinite frequency in Hz (e.g., 108 Hz).

This frequency must be greater than or equal to fmax. The data at this
point is only used for comparison (output) purposes (it is not used in the
processing of the approximation).
The data points between fmin and fmax must be logarithmically spaced and there must be 10 or a
multiple of 10 (up to 90) points per decade.
9.7.4
Type Of Fit
Unless otherwise specified in the .CTLFIT card (Section 9.7.5 below), the rational function
synthesis assumes that the data function is minimum phase shift with no time delay. Both
magnitude and phase angle are used in the approximation process.
9.7.5
Optional Control Cards
As in the case of the fd-line model in the LINE-MODEL module, optional control cards are allowed
in order to change the internally set default conditions. These cards are:
.ctlfit
.dbgfit
.outfit
9 - 64
Line Constants
.CTLFIT (FIT-S Control)

Allows for additional control over the fitting of the data function. Fields NORMAX, IQUICK, and
IXDYN in this card have the same meaning as for the fd-line model (Section 9.6.1). Fields IDELAY
and IPHASE are new in this module.
A7
IPHASE
IDELAY
IXDYN
.CTLFIT
IQUICK
NORMAX
1
2
3
4
5
6
7
8
1234567 890123456789 01 23 45 67 89 456789012345678901234567890123456789012345678901234567890
I2 I2 I2 I2 I2
.CTLFIT
(1-7)
Keyword to control the fitting process of fd-line models.
NORMAX
Maximum number of poles in the synthesis of the line wave functions Zc

and Ap. Default value is 25.
(20-21)
IQUICK
(22-23)
IXDYN
(24-25)
IDELAY
(26-27)
Flag controlling the accuracy of the fit of Zc and Ap.

Default value is 1.
=1
Much faster (fewer iterations) approximation of Zc and Ap is

produced. Some accuracy is sacrificed with respect to the
normal procedure.
= 1
Normal iteration procedure is used.
Flag controlling the low frequency approximation of Ap. Default value

is 1.
=1

approximation of the low frequency region of the IDELAY =
1 functions (see below). Extra dynamics (extra poles and
zeros) are added to the approximation of the low frequency
region.
= 1
No extra dynamics are added.
Flag indicating the type of the data function.

Default value is 0.
=0
Corresponds to a data function with no time delay (e.g., the

Zc function in the fd-line model).
9 - 65
Line Constants
=1
IPHASE
Corresponds to a data function with a pure time delay t (e.g.,

the propagation function Ap = e-l in the fd-line model).
When this option is specified, flag IPHASE (below) is
ignored.
(only for 'idelay'=0)
(28-29)
>0
(Default) Both magnitude and phase data are used in the

fitting procedure.
<0
The phase angle of the data function is ignored. A

minimum-phase shift approximation based only on the
magnitude of the data function is performed.
= 2 or 2
Phase angle of the data function is in radians (otherwise it is

assumed to be in degrees).
.DBGFIT (FIT-S Debug)

Controls the amount of internal processing output from FIT-S.
.DBGFIT
IBUGF
1
2
3
4
5
6
7
8
1234567 890123456789 01 23456789012345678901234567890123456789012345678901234567890
A7
I2
.DBGFIT
(1-7)
Keyword to request debug information during the fitting process of fdline models.
IBUGF
Flag controlling the level of diagnostic printout. Valid range is 1-3

(default = 0). The higher the number, the larger the amount of
debugging output from FIT-S.
(20-21)
.OUTFIT (FIT-S Output)

Controls the amount of output information on the fitting process. Fields ICOMPF, and IPLOTF in
this card have the same meaning as for the fd-line model (Section 9.6.1).
9 - 66
Line Constants
A7
IPNET
.OUTFIT
IPLOTF
ICOMPF
1
2
3
4
5
6
7
8
1234567 890123456789 01 23 45 6789012345678901234567890123456789012345678901234567890
I2 I2 I2
.OUTFIT
(1-7)
Keyword to control the amount of output generated during the fitting

process of fd-line models.
ICOMPF
Flag controlling the printout of a comparison table. Default value is 1

(no printout).
(20-21)
IPLOTF
(22-23)
IPNET
(24-25)
9.7.6
=1
An output table is produced comparing the data functions Zc

and Ap as produced by the line constants routines and the
approximating rational functions produced by FIT-S.
= 1
Flag controlling the printout of a printer plot.

Default value is 1 (print).
=1
A printer plot is produced comparing the data functions and

the approximations.
= 1
Flag indicating the type of equivalent network representing the

approximating function. This applies only to 'idelay'=0 functions.
=0
(Default) No equivalent network is produced.
=1
Series R//C blocks equivalent.
=2
Parallel R+C branches equivalent.
=3
Series R//L blocks equivalent.
Examples of Data Deck for the FIT-S Module
Sample data deck for FIT-S option follows.

BEGIN NEW DATA CASE
LINE CONSTANTS
9 - 67
Line Constants
c ** New Line Constants Program ** June 1987.
c
TEST CASE 9 (LTC9).
c
EXTERNAL DATA-CURVE OPTION
c ...................................................................
FILES
ltc9.out
ltc9.pun
FIT-S
ltc9.par
(1X,E11.0,44X,2E11.0)
c ...................................................................
C OPTIONAL CONTROL CARDS
C .outfit: icompf,iplotf,iprat,idata
.outfit
0 1 0 1
BLANK
BLANK
BEGIN NEW DATA CASE
BLANK
Sample Data-Curve file follows

c
c
c
c
c
** New Line Constants Program ** June 1987.

TEST CASE 9 (LTC9). EXTERNAL DATA-CURVE OPTION
Data-Curve File. Columns Zc and Phase(dg) will be used.
FREQ
1.0000E-15
1.0000E-01
1.2589E-01
1.5849E-01
1.9953E-01
2.5119E-01
3.1623E-01
R
1.6195E-02
1.6491E-02
1.6568E-02
1.6664E-02
1.6785E-02
1.6937E-02
1.7129E-02
L
1.6993E+01
5.4988E+00
5.4299E+00
5.3609E+00
5.2920E+00
5.2231E+00
5.1542E+00
G
3.0009E-12
2.0006E-10
2.0006E-10
2.0006E-10
2.0006E-10
2.0006E-10
2.0006E-10
C
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
Zc
Phase
7.3463E+04-4.5117E-10
1.8870E+03-3.7872E+01
1.6953E+03-3.6771E+01
1.5279E+03-3.5354E+01
1.3830E+03-3.3609E+01
1.2593E+03-3.1541E+01
1.1553E+03-2.9180E+01
3.9811E+06
5.0119E+06
6.3096E+06
7.9433E+06
1.0000E+07
1.0000E+08
END
9.6203E+02
1.0872E+03
1.2277E+03
1.3854E+03
1.5623E+03
5.0833E+03
1.7091E+00
1.7046E+00
1.7007E+00
1.6971E+00
1.6940E+00
1.6763E+00
2.0006E-10
2.0006E-10
2.0006E-10
2.0006E-10
2.0006E-10
2.0006E-10
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
7.5248E-03
4.7664E+02-6.4456E-01
4.7601E+02-5.8013E-01
4.7544E+02-5.2159E-01
4.7494E+02-4.6851E-01
4.7449E+02-4.2048E-01
4.7199E+02-1.3826E-01
.
.
.
9 - 68
Section 10
Cable Constants
10.1
Introduction to the Cable Constants Routines ..................................... 10-1
10.2
Capabilities of the "Cable-Model" Module ......................................... 10-3
10.3
10.4
10.2.1
Models for Transients Simulations ................................ 10-3
10.2.2
Models for Steady-State Analysis: EXACT-PI ............. 10-5
10.2.3
External Parameters Data ............................................... 10-7
10.2.4
Cross-Bonded Cables ..................................................... 10-7
10.2.5
Input Format Conversion ............................................... 10-8
Data Entry Rules for "Cable-Model" .................................................. 10-9

10.3.1
Cable-Model Control Card ............................................. 10-9
10.3.2
Cable-Type Card ............................................................ 10-11
10.3.3
Conductor Cards for Single-Core (SC) Coaxial Cables . 10-12
10.3.4
Earth/Frequency Card .................................................... 10-22
10.3.5
Optional Control Cards .................................................. 10-26
10.3.6
External Data Cards ....................................................... 10-33
Cable-Model and Cable-Parameters Format Conversion ................... 10-36

10.4.1
Conversion from Cable-Model to Cable-Constants ....... 10-36
10.4.2
Conversion from Cable-Constants to Cable-Model ....... 10-37
10.5
CABLE-MODEL Examples ............................................................... 10-37
10.6
Cababilities of the CABLE-PARAMETERS Module ........................ 10-48

10.6.1
Introduction .................................................................... 10-48
10.6.2
General Structure of "CABLE-PARAMETERS"

Data Cards ..................................................................... 10-48
10.7
Data Entry Rules for CABLE-PARAMETERS .................................. 10-52

10.7.1
Format for "A2", "B2", and "C2" Data .......................... 10-52
10.7.2
Format for "A3", "B3", and "C3" Data .......................... 10-53
10.7.3
Format for "A4" and "B6" Data ..................................... 10-57
10.7.4
Format for "B4" Data ..................................................... 10-57
10.7.5
Format for "C4" Data ..................................................... 10-59
10.7.6
Format for "A5" and "B7" Data ..................................... 10-62
10.7.7
Format for "B5" Data ..................................................... 10-64
Section 10
Cable Constants
10.7.8
Format for "C5" Data ..................................................... 10-65
10.7.9
Format for "A6" Data ..................................................... 10-66
10.7.10
Format for "C6", "A7", and "B9" Data .......................... 10-67
10.7.11
Format for "C7", "A8", and "B10" Data ........................ 10-69
10.7.12
Format for "B8" Data ..................................................... 10-71
10.8
Degenerate Configurations and Special Cases .................................... 10-72
10.9
Approximation Used for the Bundling of Overhead Conductors ....... 10-77
10.10
Crossbonded Cables ............................................................................ 10-78
10.11
10.10.1
Introduction .................................................................... 10-78
10.10.2
Modelling of a Crossbonded Cable ................................ 10-79
10.10.3
Data Format for Using the Crossbonded Cable Model .. 10-80
CABLE-PARAMETERS Example ..................................................... 10-88
Section 10
Cable Constants
10.1
Introduction to the Cable Constants Routines
By means of the special request word "CABLE CONSTANTS", an AUX user gains access to the Cable
Constants set of routines. The Cable Constants routines calculate the resistance, inductance,
conductance, and capacitance matrices for underground cables (e.g., Single-Core (SC) or PipeType (PT)). These routines can also be used to generate EMTP models for these cables, both for
transient and frequency scan simulations.
As in the case of the overhead line calculation module of AUX, there are two independent modules
which can be accessed via keywords:
(A)
Cable Model module. Keyword: "CABLE-MODEL"
(B)
Cable Parameters module: Keyword: "CABLE-PARAMETERS"
The "Cable Model" module is used primarily for the generation of underground cable models
(e.g., constant parameters, frequency dependent, nominal and exact pi, etc.). This module can
also be used to calculate cable parameters at any given frequency, and in a variety of forms (e.g.,
phase, modal, or sequence parameters). This module was introduced in version 3.0, and it is not a
re-written version of the CABLE CONSTANTS module available prior to version 3.0, but rather, a
new program with new models, capabilities and more robust numerical algorithms. Its
predecessor, is still available in the "Cable Parameters" module. The input data format has been
enhanced to reflect the new modelling capabilities and it is not directly compatible with the format
used prior to version 3.0. A keyword-directed conversion routine is provided.
The "Cable Parameters" module is the "old" CABLE CONSTANTS support routine (prior to version
3.0). It has some functionality and modelling capabilities which have not been added to the
"Cable Model" module, namely, stratified earth modelling and overhead line modelling.
Otherwise, the "Cable Parameters" is a subset of the newer "Cable Model" module.
The general structure of the input data file for running the cable constants program in shown in
Table 10.1 below.
Comment lines are ignored by the program. They can be inserted anywhere in the data deck.
These lines are identified by "C " or "c " (letter C or c in the first column, followed by a space in
the second column). They can contain any alphanumeric characters in the rest of the line. As
shown in Table 10.1, a Cable Constants data case may contain more than one group of CableParameters and/or Cable-Model data cards (in any order). Each such group is a separate,
independent case within the Cable Constants routine. The blank card at the end of Cable10 - 1
Cable Constants
Parameters data is the same card that marks the end of frequency cards, as described in Section
10.2. The blank card at the end of Cable-Model data is the same card that marks the end of
optional control cards, as described in Section 10.2.
[1]
"BEGIN NEW DATA CASE"

keyword
[2]
"CABLE CONSTANTS" request card
[3]
"CABLE-PARAMETERS"
[3.1]
Cable-Parameters control card
[3.2]
Cable data cards
[3.3]
Frequency data cards
[3.4]
BLANK
[4]
card to indicate the end of "cable

parameters" data
"CABLE-MODEL"
[4.1]
Cable-Model control card
[4.2]
Cable data cards
[4.3]
Frequency data cards
[4.4]
Optional Control Data Cards
[4.5]
BLANK
card to indicate the end of "cable

model" data
Cards for Another Cable-Parameters or Cable-Model Data Case
[5]
BLANK
card to end "Cable

Constants" requests
[6]
BLANK
[7]
"BEGIN NEW DATA CASE"

keyword
[8]
BLANK
card to end AUX requests
card to indicate end-of-
run
Table 1: General Structure of the Input Data File
10 - 2
Cable Constants
10.2
Capabilities of the "Cable-Model" Module
This module reads physical layout and characteristics of a cable and produces cable models that
will be used by the EMTP for transient, and frequency scan simulations. Single-Core as well as
Pipe-Type cables are supported.
10.2.1
Models for Transients Simulations
FD-Model Class
The Frequency-Dependent or FD-Model class provides an accurate representation of the
distributed nature of all the cable parameters: R, L, G, and C, as well as their frequency
dependence in modal quantities. In this model class it is assumed that the characteristic
admittance and propagation function matrices [Yc,phase] and [Aphase] can be diagonalized (by a
modal transformation matrix Q).
[ Y c, phase ] = [ Y phase ] [ Z phase ]
= [ Gphase + jC phase ] [ R phase + jL phase ]
(7.1)
(7.2)
[ A phase ] = exp ( [ phase ] l )
[ phase ] = [ Y phase ] [ Z phase ] = [ G phase + jC phase ] [ R phase + jL phase ]
(7.3)
The modal transformation matrix [Q] is the eigenvector matrix that diagonalizes the product
[Yphase][Zphase]; that is,
[ Q ] [ Y phase ] [ Zphase ] [ Q ] = [ Y mode ] [ Z mode ]
[ Q ] [ Y c, phase ] [ Q ]
= [ Y c, mode ]
[ Q ] [ Z phase ] [ Q ] = [ Z mode ]
and
(7.4)
( diagonal )
(7.5)
( diagonal )
[Q]
[ Yphase ] [ Q ]
10 - 3
= [ Y mode ]
( diagonal )
(7.6)
Cable Constants
[ Q ] [ A phase ] [ Q ] = [ Amode ]
(7.7)
( diagonal )
The elements of [Yc,mode] and [Amode] are scalar functions of frequency and are approximated in
the frequency domain with rational functions. In the time-step loop of the EMTP these rational
functions become sums of exponential functions.
The modal transformation matrix Q is, in general, frequency dependent, and its elements can be
approximated with rational functions (FDQ option). There are instances, however, when it is
desirable to assume that the modal transformation matrix is constant and real (QREAL option). In
this case, Q is calculated at a given frequency, its columns are rotated to minimize the imaginary
parts of their elements, and the resulting imaginary terms are discarded. The resulting matrix Qreal
is the used to calculate the modal parameters as shown in equations (10.4) to (10.7).
The FDQ option should be used when the highest accuracy is desired (see Reference 3). There are
instances, however, when some of the elements of Q are ill-conditioned in some frequency ranges.
In these cases it is preferable to use the QREAL option. There are a number of known situations
where the FDQ option (as presently coded) will occasionally result in ill-conditioned Q functions.
These situations are flagged by the program and an FDQ model is not produced (this automatic
detection can be overridden by the user by entering "-1" in columns 22-23 of the ".dbgfit" control
card). In such cases it is preferable to use the QREAL option. The situations where this illconditioning takes place are under investigation, and it is expected that future releases of the
program will not have this limitation.
CP-Model
The CP-Model (constant-parameter model) assumes that the cable parameters R, L, and C are
constant, and they are calculated at a user-supplied frequency. This model considers L and C to be
distributed ("ideal cable") and R to be lumped at three places (cable ends and cable middle). The
shunt conductance G is assumed to be zero.
Taking into account the frequency dependence of the cable parameters (as modelled by the FDModel class of cable models) is an important factor for the accurate simulation of transients in the
EMTP. However, the CP-Model is computationally fast and it is generally used as an alternative to
model secondary lines or cables.
Note:
In this implementation of AUX, the "Cable-Model" module does not generate nominalpi circuit models. To generate nominal pi models for transient simulations, the "CableParameters" module must be used.
10 - 4
Cable Constants
10.2.2
Models for Steady-State Analysis: EXACT-PI
This model provides an exact single-frequency representation of the cable in terms of a multiphase
pi-equivalent. The model can be read directly by the EMTP when frequency scans or singlefrequency steady-state solutions are requested.
The model is punched out as a Y-matrix that includes the series and shunt branches of the
multiphase exact-pi circuit. This Y matrix is produced for each frequency point in a logarithmic
or linear frequency interval.
It is important to realize that the exact-pi model is only a frequency domain representation of a
cable at a given frequency. It is not an EMTP model in the same sense the as FDQ model. The
exact-pi model can only be used in EMTP frequency scans.
To run a frequency scan in the EMTP, the special request card FREQUENCY SCAN must be included
in the EMTP data deck. The format and usage of this card is described in Section 4.1.1 RuleBook 1,
and summarized below.
EMTP Miscellaneous Data Card
1
2
3
4
5
6
7
8
123456789012345678901234 56789012 34567890 12345678 90123456 7890123456789012 34567890
FREQUENCY SCAN
A24
fmin
fmax
NPDEC
MODSYM
E8.0
E8.0
E8.0
I8
A16
fmin
(2532)
Minimum scan frequency in Hz.
f
(3340)
Frequency increment for linear frequency increments. Leave blank for

logarithmic increments.
fmax
(41-48)
Maximum frequency in Hz.
NPD
Number of points per decade for logarithmic frequency increments.

Leave blank for linear increments.
(4956)
MODSYM
(57-62)
Flag indicating the type of line/cable model to be used in the frequency

scan. Must be set to FREQUENCY-DOMAIN to use the EXACT-PI model.
It is important to match the frequency point generated in AUX (see Section 10.3.4 ) with the
frequency points expected in the EMTP (i.e., fmin, fmax, f and NPDEC in columns 25 to 56 above).
Otherwise, an error condition will occur.
10 - 5
Cable Constants
Additionally, the appropriate cards to describe the cable itself in the EMTP must be included. For
the EXACT-PI model the FDQ or LMARTI cable model cards should be used, as explained in Section
7.5 of RuleBook 1. A brief description is also included below:
I2
A6
A6
A6
A6
ITYPE
(1-2)
BUS1
Filename
A28
I2 I2
I2 I2
IOUT
BUS4
IDEBUG
BUS3
ICHOP
BUS2
NPHS
BUS1
IMODEL
ITYPE
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 78901234567890123456789012 34 56 78 90 1234567890123456789 0
A13
I1
Set ITYPE to 1, 2, 3, , N for conductors 1, 2, 3, , N. If

10 N 35 use A, B, C, , Z. If 36 N 61 use a, b, c, ,
z. Note that the ordering of the cable conductors in the Cable-Model
module is determined by parameter KPH in the Conductor/Insulator
cards (see Section 10.3.3). On the other hand, output from CableParameters module is fixed and ordered core-1, core-2, core-3, sheath1, sheath-2, sheath-3.
(3-8)
BUS2
(9-14)
BUS3
(3-8)
BUS4
(9-14)
FILENAME
(25-52,
61-74)
The names of the nodes of the reference branch (the node names of the
first branch of a previously defined LMARTI cable model.
Name of the file which contains the cable data in rational-functions

form. This file is obtained directly from the punched output of the
EMTP support program AUX.
If the field within columns 27 to 52 is not sufficient for the name of the
file, then columns 61 to 64 can be used as a continuation field (blank
characters within the file name are not allowed). If FILENAME is left
blank, then the fitted-functions data will be read immediately after the
last branch card of the current line/cable unit.
IMODEL
(53-54)
NPHS
(55-56)
Set to -4 for EXACT-PI or for FDQ model options. Set to 3 if Q is

constant.
Number of phases of the current line/cable unit. For example, if a
3-phase cable has two concentric conductor per phase (e.g., core and
sheath), then set NPHS = 6.
10 - 6
Cable Constants
Does not apply to the EXACT-PI model
ICHOP
(57-58)
(59-60)
This flag controls the amount of printed output pertaining to the

processing of LMARTI lines/cables at input time. Normally left blank.
IOUT
Standard EMTP branch output request flag.
IDEBUG
(80)
IOUT
=1
Branch currents will be included on output.
IOUT
=2
Branch voltage vk(t)-vm(t) will be included on output.
IOUT
=3
Branch currents and branch voltages will be included on

output.
Note:
The EXACT-PI model cannot be used in transient or single-frequency steady-state

simulations. It can only be used in frequency scans. To use the EXACT-PI option in a
frequency scan, the FREQUENCY-SCAN miscellaneous data card must be present, and
MODSYM must be set to FREQUENCY-DOMAIN
10.2.3
External Parameters Data
The impedance and admittance matrices of a cable are normally computed from the physical data
specified in the Cable-Model data cards. Additionally, the Cable-Model module can read these
matrices directly from a separate external data file and use them to produce different cable models
without the need for the physical data. The user may obtain impedance and admittance matrices in
a range of frequencies from any external source (such as a program based on finite element
method) and provide them to Cable-Model routine through the external data file. The format of
this file is described in Section 10.3.6.
10.2.4
Cross-Bonded Cables
In order to model a cross-bonded cable accurately, each major section must be modelled in detail.
This means that each minor section of the cable must be modelled (preferably with an FDQ model),
and the sheath bonding and sheath grounding connections must be made explicitly using the EMTP
node names.
Such a detailed representation can be computationally intensive because modelling short cable
segments of the order of 400 meters or so, requires a very small time step (a fraction of the travel
time of the fastest propagation mode). Furthermore, a number of these major sections must be
connected to represent the entire cable. For example, a 12 km cable with 400 m minor sections,
would require a total of 30 6-phase FDQ cable models. Nevertheless, this type of detailed
10 - 7
Cable Constants
representation is necessary when sheath currents and voltages have to be assessed (see
Reference [4] and reference [5])
The detailed representation of each minor section of a cross-bonded cable is in some ways
analogous to modelling a transposed overhead transmission lines by representing each
transposition section explicitly, and connecting the sending and receiving node names accordingly
with EMTP node names. In the case of transmission lines this situation can be approximated by
assuming that the line is balanced, and using a single line where the elements of impedance and
admittance matrices have been averaged to account for the effect of transposition.
A cross-bonding option is available in the Cable-Model module to provide this type of
approximation. If parameter IXBD in the "Cable-type" card is set to 1, then the elements of the
impedance and admittance matrices of the cable are averaged to reflect the effect of cross-bonding.
The grounding of the sheaths is then controlled using the KPH parameter in the "Conductor/
Insulator" cards. Setting KPH = 0 for the sheaths, is equivalent to assuming that the sheaths are
continuously grounded (at zero potential throughout the entire cable length). In this case, the
sheaths can be eliminated and a three-conductor approximation of a cross-bonded cable is
obtained. This three-conductor approximation compares quite favourably with the detailed
modelling of each minor section of a cross-bonded cable, and it is ideally suited for switching
transient studies of cross-bonded cables, because of its computational speed and accuracy.
10.2.5
Input Format Conversion
The data entry rules and format for "Cable-Model" and "Cable-Parameters" modules are different,
mostly because of the added functionality of the "Cable-Model" module. To facilitate crossvalidation of both modules, and migration from old to new formats, an automatic input data
conversion option has been provided.
To enable data conversion, set either IPCH or IRUN to "1" in the appropriate fields of the "CableModel Control" Card (see Section 10.3.1). If IPCH = 1, the input data file will be converted and
then stored into the standard punch file (i.e., logical unit 7). Additionally, if IRUN is set to 1, AUX
will execute immediately after data conversion. Any combination of the two flags is acceptable.
The direction of the format conversion is detected automatically (i.e., from "Cable-Model" to
"Cable-Parameters", and vice versa).
Depending on the cable module keyword used, the program will do one of the following:
1.
Keyword = "CABLE-MODEL", IPCH = IRUN = 0. Normal Cable-Model run with

Cable-Model input format.
2.
Keyword = "CABLE-MODEL", IPCH = 1 or IRUN = 1. AUX will try to run/convert a

Cable-Model case with Cable-Parameters input format.
3.
Keyword = "CABLE-PARAMETERS", IPCH = IRUN = 0. Normal Cable-Parameters

run with Cable-Parameters input format.
10 - 8
Cable Constants
4.
Keyword = "CABLE-PARAMETERS", IPCH = 1 or IRUN = 1. AUX will try to run/

convert a Cable-Parameters case with Cable-Model input format.
When using the input format conversion option, a certain amount of caution is required because
the two modules do not have the same functionality. In those cases where a one-to-one
relationship between data and/or modelling requests is not possible, some assumptions and
defaults have been made. These assumptions are summarized in Section 10.4.
10.3
Data Entry Rules for "Cable-Model"
The following section describes the format of a Cable-Model data case.
10.3.1
Cable-Model Control Card
1
2
3
12345678901 234567890 1234567890
4
5
6
7
1234567890 1234567890 1234567890 123 4567890 12345 12345
CABLE-MODEL
Model
Q-Optn
FREQ-Q
LENGTH
Ext
IPCH
IRUN
A11
A10
E10.0
E10.0
E10.0
A7
I5
I5
CABLE-MODEL
Keyword for "Cable-Model" module
(1-11)
Model
(21-30)
Model keyword. It can be one of the following

"FD-MODEL"
R, and L are assumed to be frequency-dependent.

Modal transformation matrix may or may not be
constant, depending on Q-Optn below (see
Section 10.2.1). Produces punch file for the LMARTI or
FDQ cable model.
"CP-MODEL"
R, L, and the modal transformation matrix Q are

assumed to be constant (see Section 10.2.1). Produces
output for the Dommel or constant-parameter line
model.
10 - 9
Cable Constants
Q-Optn
(31-40)
"EXACT-PI"
Pi-circuits are calculated at a given number of

frequencies. If Q-Optn is set to FDQ (default), the
EXACT-PI becomes a correct representation of the cable
at a given frequency (see Section 10.2.2). Produces
punched for frequency scan simulations only.
"SCAN"
Cable parameters in either phase, modal or sequence

quantities computed at specified frequencies. No
model is generated.
Type of modal transformation matrix Q. This keyword can be one of

the following
"FDQ"
Q is assumed to be frequency-dependent (see

Section 10.2.1). In the EMTP data cards for the LMARTI/
FDQ model, parameter "imodel" must be set to 4.
"QREAL"
Q is evaluated at a frequency FREQ-Q, its columns are

rotated (to minimize imaginary parts), and it is also
stripped of its imaginary part to make Q real and
constant. In the EMTP data cards for the LMARTI/FDQ
model, parameter "imodel" must be set to 3.
"QCMPLX"
Q is evaluated at a frequency FREQ-Q, its columns are

rotated (to minimize imaginary parts), and it its
imaginary part is retained to make Q complex and
constant. EMTP models cannot use a complex, constant
modal transformation matrix. This option is meant,
mostly, for research purposes.
The following table shows valid combinations of Model and Q-Optn keywords:
"FD-MODEL"
"FDQ" (default), "QREAL", "QCMPLX"
"CP-MODEL"
"QREAL" (default), "QCMPLX
"EXACT-PI"
"FDQ (default), "QREAL", "QCMPLX"
"SCAN"
"FDQ" (default), "QREAL", "QCMPLX"
FREQ-Q
(41-50)
Frequency in Hz at which the constant modal transformation matrix Q is

computed (default is 1000 Hz). In the case of the CP-MODEL, this is also
the frequency at which R, L, and C are evaluated. This field is ignored
with the "FDQ" option
.LENGTH
(51-60)
Cable length in km (default is 1.0 km)
Ext
(64-70)
Keyword for read Z and Y matrices from an external file. Normally left
blank
10 - 10
Cable Constants
[blank]
No external data. Impedance and admittance matrices are

calculated by the program from physical data.
"EXTDAT"
external data. Impedance and admittance matrices (for a

range of frequencies) are provided in a separate data file
(see Section 10.3.5 for a detailed description of this option)
Flag to control the storage of input data files, after input format
conversion has been completed. Converted data files are stored into the
standard "punch" file (i.e., logical unit 7).
IPCH
(71-75)
=0
Do not store converted input data file (only if IRUN = 1)
=1
Store converted input data file into punch file.
Flag to control program execution after input format conversion
IRUN
(76-80)
Note:
=0
Do not run AUX after input data file has been converted and
stored into the punch file
=1
Run AUX after data conversion is completed, whether or not

storage of the converted input data file is requested. Note
that if both IRUN = 1 and IPCH = 1, then any punched output
which results from an AUX run (e.g., punched file for an
FDQ model) will be appended to the record of the converted
input data file
Input data conversion between Cable-Model and Cable-Parameters formats will only
take place if either IPCH or IRUN are equal to 1.
If IPCH and IRUN are zero the program will not attempt to provide any form of input
format conversion, and using the wrong format rules will result in an invalid run.
10.3.2
Cable-Type Card
TYPE
1
2
3
4
5
6
7
8
123 45 67890 12345 67890123456789012345678901234567890123456789012345678901234567890
A2
NCBL IXBD
I5
I5
10 - 11
Cable Constants
Type
(4-5)
Keyword describing cable type. Valid keywords are:

"SC"
Single-Core coaxial cables.
"PT"
Pipe-type cable.
(6-10)
Number of component coaxial cables (or single-phase units) which make

up the SC or PT cable. For example, NCBL=3 for a three-phase SC cable.
IXBD
Cross-bonding flag.
NCBL
(11-15)
=0
Not cross-bonded (default).
=1
Sheaths (second conductor) are cross-bonded.
If IXBD = 1, parameter LENGTH corresponds to the length in km of a

major section. The connection of the cross-bonded sheaths (i.e., kept
separate, joined together or grounded) is determined by their Phase
Numbers which are specified in the conductor data cards described in
Section 10.3.3 for SC cables and for PT cables.
10.3.3
Conductor Cards for Single-Core (SC) Coaxial Cables
This subsection describes the following cards:

SC
Cable: Individual Cable Card
SC
Cable: Conductor/Insulator Cards
SC
Cable: Phase Numbers Card
Pipe-Type Cable: Individual Cable Card

Pipe-Type Cable: Conductor/Insulator Card
Pipe-Type Cable: Phase Numbers Card
Pipe-Type Cable: Pipe-Data Cards
The conductor data cards required to describe a pipe-type cable are somewhat different than the
conductor data cards required for an SC coaxial cable. Although many data fields are similar their
description will be presented separately to facilitate readability.
10 - 12
Cable Constants
SC Cable: Individual Cable Card:

This card and the following cards describe each coaxial cable. A total of NCBL Individual Cable
Cards are required. No special ordering sequence is required.
1
2
3
4
5
6
7
8
12345 67890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
NCN
VRT
HRZ
ROUT
I5
E10.0
E10.0
E10.0
NCN
(1-5)
Number of concentric tubular conductors in this cable (or -K as

described below). For example, set NCN=3 for a cable with core, sheath
and armour.
If NCN is positive, the Individual Cable Card must be followed by NCN
Conductor/Insulator cards which describe the concentric conductors and
their insulation.
If NCN is negative, then NCN=K, where K is the Kth conductor entered.
This option is used to copy conductor data in the case of identical cables.
VRT
(11-20)
Vertical distance (depth) measured from the of the centre of this cable to
the earth's surface. This is a positive number. Units = meters
VRT-3
VRT-2
VRT-1
reference
HRZ-2
HRZ-3
HRZ
(21-30)
HRZ-1
Horizontal distance measured from the centre of this cable to an

arbitrary point of reference. Units = meters.
10 - 13
Cable Constants
Outside radius of the insulation layer surrounding the cable. Leave
blank if there is no surrounding insulation. Units = meters.
ROUT
(31-40)
Note that if NCN is negative (data is being copied from an earlier cable) then ROUT is ignored.
SC Cable: Conductor/Insulator Cards:

When NCN in the previous card is positive, the next NCN cards describe the tubular conductors and
their surrounding insulation. They must be ordered from inside out (the core conductor comes
first, followed by sheath, etc.).
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345 67890
RIN
ROUT
RHO
MUE
MUE-I
EPS-I
LFCT
KPH
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
I5
RIN
Inside radius of the conductor. Units = meters.
(1-10)
ROUT
Outside radius of the conductor. Units = meters
(11-20)
ROUT
ROUT-3
RIN-3
ROUT-2
RIN-2
ROUT-1
RIN-1
reference
reference
10 - 14
Cable Constants
RHO
(21-30)
MUE
Resistivity of the conductor. Units = m.

Relative permeability of the conductor.
(31-40)
MUE-I
Relative permeability of the surrounding insulation.
(41-50)
EPS-I
Relative permittivity of the surrounding insulation.
(51-60)
LFCT
Loss-factor of the surrounding insulation.
(61-70)
KPH
(71-75)
Phase-number of the conductor. Conductors of all cables must be given

phase numbers starting from 1, with no gaps in phase numbering. For
example, for a three-conductor cable KPH = 1, 2, 3 is a legitimate
numbering arrangement, while KPH = 1, 3, 4 is not. Conductors with
KPH = 0 will be grounded and all conductors with identical phase
number will be bundled into a single equivalent conductor
Note: Use KPH to ground conductors which are not needed explicitly in a transient simulation.
For example, if the sheaths of a submarine cable are in contact with water, they become
effectively grounded. In this case set KPH = 0 for all sheaths to obtain a simpler 3-conductor
model.
Note that the phase numbering sequence must start at zero and must have no gaps, but the order in
which KPH appears is arbitrary. In other words, Conductor/Insulator cards do not have to be
ordered according to KPH. By the same token, the order of the conductors in printed or punched
output will be made according to the sequence defined by KPH.
SC Cable: Phase Numbers Card:

When NCN in the Individual Cable Card is negative, the data duplication function is enabled.
Setting NCN to a negative number is equivalent to saying "make the conductor data for this cable
identical to the data of cable number NCN". For example, if all three (single-phase) cables in a
three-phase cable system are identical, it is sufficient to enter the Conductor/Insulator cards for the
first cable, and duplicate the rest. In this case NCN = 1 for cables 2 and 3.
Since phase number assignment is independent of the physical characteristics of the cable, it must
be specified using the Phase Numbers Card.
10 - 15
Cable Constants
1
12345 67890 12345
2
3
4
5
6
7
8
67890 12345 6789012345678901234567890123456789012345678901234567890
KPH1 KPH2 KPH3 KPH4 ...

I5
KPH1
I5
I5
I5
...
Phase number of the first conductor (core).
(6-10)
KPH2
Phase number of the second conductor (sheath).
(11-15)
KPH3
Phase number of the third conductor (armour).
(16-20)
...
...
If the cable has more than 15 conductors, the phase numbers of the next 15 conductors are read
from a second card and so on.
Note: If NCN in the Individual Cable Card is positive, the it must be followed by NCN Conductor/
Insulator Cards and no Phase Numbers Card.
If NCN is negative, the Conductor/Insulator Card will be followed by a single Phase Numbers Card
(unless there are more than 15 concentric conductors
Pipe-Type Cable: Individual Cable Card:

This card and the following cards describe each coaxial cable within the pipe. A total of NCBL
Individual Cable Cards are required. No special ordering sequence is required.
1
2
3
4
5
6
7
8
12345 67890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
NCN
DST
ANG
ROUT
I5
E10.0
E10.0
E10.0
NCN
(1-5)
Number of concentric tubular conductors in this cable (or K as

described below). For example, set NCN = 3 for a cable with core, sheath
and armour.
10 - 16
Cable Constants
NCN
(1-5)
If NCN is positive, the Individual Cable Card must be followed by NCN

Conductor/Insulator cards which describe the concentric conductors and
their insulation.
If NCN is negative, then NCN = K, where K is the Kth conductor entered.
This option is used to copy conductor data in the case of identical cables.
DST
(11-20)
Distance measured from the of the centre of this cable to the centre of the
pipe. Units = meters.
VDPTH
REXT
ROUT
RIN
reference
DIST
ANG
ANG
(21-30)
ROUT
(31-40)
Angle measured from the line joining the centre of this cable and the
centre of the pipe, and an arbitrary reference axis. Units = degrees.
Outside radius of the insulation layer surrounding the pipe. Leave blank
if there is no surrounding insulation. Units = meters.
Note that if NCN is negative (data is being copied from an earlier cable)
ROUT is ignored.
10 - 17
Cable Constants
Pipe-Type Cable: Conductor/Insulator Cards:

When NCN in the previous card is positive, the next NCN cards describe the tubular conductors and
their surrounding insulation. They must be ordered from inside out (the core conductor comes
first, followed by sheath, etc.).
1
2
3
4
5
6
6
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345 67890
RIN
ROUT
RHO
MUE
MUE-I
EPS-I
LFCT
KPH
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
I5
RIN
(1-10)
ROUT
Outside radius of the conductor. Units = meters
(11-20)
ROUT
ROUT-3
RIN-3
ROUT-2
RIN-2
ROUT-1
RIN-1
reference
RHO
(21-30)
MUE
reference
Resistivity of the conductor. Units = m.

Relative permeability of the conductor.
(31-40)
10 - 18
Cable Constants
MUE-I
Relative permeability of the surrounding insulation.
(41-50)
EPS-I
Relative permittivity of the surrounding insulation.
(51-60)
LFCT
Loss-factor of the surrounding insulation.
(61-70)
KPH
(71-75)
Conductors of all cables must be given phase numbers starting from 1,

with no gaps in phase numbering. For example, for a three-conductor
cable KPH = 1, 2, 3 is a legitimate numbering arrangement, while KPH =
1, 3, 4 is not. Conductors with KPH = 0 will be grounded and all
conductors with identical phase number will be bundled into a single
equivalent conductor
Note: .Use KPH to ground conductors which are not needed explicitly in a transient simulation.
For example, if the sheaths of a pipe-type cable are in contact with the pipe, then an approximate
4-conductor system can be obtained by bundling the sheaths and the pipe together (e.g., setting
KPH = 4 on sheath and pipe cards. If the pipe can also be assumed to be continuously grounded,
setting KPH = 0 on sheath and pipe cards result in a 3-conductor system. For a description of the
approximations involved in bundling and conductor elimination, please refer to the EMTP Theory
Book.
Note that even though the phase numbering sequence must start at zero and must have no gaps, the
order in which KPH appears is arbitrary. In other words, Conductor/Insulator cards do not have to
be ordered according to KPH. By the same token, the order of the conductors in printed or punched
output will be made according to the sequence defined by KPH.
Pipe-Type Cable: Phase Numbers Card:

When NCN in the Individual Cable Card is negative, the data duplication function is enabled.
Setting NCN to a negative number is equivalent to saying "make the conductor data for this cable
identical to the data of cable number NCN". For example, if all three (single-phase) cables in a
three-phase cable system are identical, it is sufficient to enter the Conductor/Insulator cards for the
first cable, and duplicate the rest. In this case NCN = 1 for cables 2 and 3.
Since phase number assignment is independent of the physical characteristics of the cable, it must
be specified using the Phase Numbers Card.
10 - 19
Cable Constants
1
2
3
4
5
6
7
8
12345 67890 12345 67890 12345 6789012345678901234567890123456789012345678901234567890
KPH1 KPH2 KPH3 KPH4 ...

I5
KPH1
I5
I5
I5
...
Phase number of the first conductor (core).
(6-10)
KPH2
Phase number of the second conductor (sheath).
(11-15)
KPH3
Phase number of the third conductor (armour).
(16-20)
...
...
If the cable has more than 15 conductors, the phase numbers of the next 15 conductors are read
from a second card and so on.
Note: If NCN in the Individual Cable Card is positive, the it must be followed by NCN Conductor/
Insulator Cards and no Phase Numbers Card.
If NCN is negative, the Conductor/Insulator Card will be followed by a single Phase Numbers Card
(unless there are more than 15 concentric conductors in a given cable).
Pipe-Type Cable: Pipe-Data Cards:

For Pipe-type cables, the next two cards describe the tubular pipe and its inside and outside
insulation.
First card:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 12345 67890123456789012345678901234567890
RIN
ROUT
REXT
VDPTH
KPH
E10.0
E10.0
E10.0
E10.0
I5
10 - 20
Cable Constants
RIN
Inside radius of the PIPE. Units = meters.
(1-10)
ROUT
Outside radius of the pipe. Units = meters.
(11-20)
VDPTH
REXT
ROUT
RIN
reference
DIST
ANG
REXT
(21-30)
Outside radius of the tubular insulator surrounding the pipe.

Units = meters.
(31-40)
Vertical distance (depth) of the pipe's centre from the surface of the
earth. Units = meters.
KPH
Phase-number of the pipe (zero if it is grounded.
VDPTH
(71-75)
10 - 21
Cable Constants
Second card:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
RHO
MUE
MUE-IN
LFTC-IN-IN
MUE-I
MUE-OUT
EPS-OUT
LFTC-OUT
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
E10.0
RIN
(1-10)
Resistivity of the pipe. Units = m.
RHO
(1-10)
MUE-IN
Relative permeability of the insulation inside the pipe.
(21-30)
EPS-IN
Relative permittivity of the insulation inside the pipe.
(31-40)
LFCT-IN
Loss-factor of the insulation inside the pipe.
(41-50)
MUE-OUT
Relative permeability of the insulation surrounding the pipe.
(51-60)
EPS-OUT
Relative permittivity of the insulation surrounding the pipe.
(61-70)
LFCT-OUT
Loss-factor of the insulation surrounding the pipe.
(71-80)
Note: When REXT on the previous card is left blank (or zero), a very thin insulating layer is
assumed around the pipe with MUE-OUT=1.0,
10.3.4
EPS-OUT=1.0
Earth/Frequency Card

Earth/Frequency Card: Logarithmic Scale
Earth/Frequency Card: Linear Scale
Earth/Frequency Card: User-Supplied Frequencies
Discrete-Frequency Cards
10 - 22
and LFCT-OUT=0.0.
Cable Constants
The Earth/Frequency card specifies the values for the resistivity and relative permeability of the
earth. It also specifies whether the cable parameters will be calculated over a linear, logarithmic or
user-defined frequency ranges. The frequency range is controlled with the keyword "F-Scl" in
columns 35-40 of this card.
Earth/Frequency Card: Logarithmic Scale:

This form of frequency scaling is required for the FD-Model class of models, and it can also be
used for EXACT-PI and SCAN calculations (see Section 10.3.1).
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 12345 67890 1234567890 12345 67890 1234567890 1234567890
RHO-E
MUE-E
FG0
F-Scl
FMIN
NPD
NDC
FDC
E10.0
E10.0
E10.0
A5
E10.0
I5
I5
E10.0
RHO-E
(1-10)
MUE-E
Earth resistivity. Units = m.

Relative permeability of the earth (default is 1.0)
(11-20)
FG0
(21-30)
Breakpoint frequency (in Hz) of the shunt conductances for all insulating
layers. For all insulators, the shunt conductance G is described with the
following function of
frequencyG = 2 (FG0 + Frequency) (Loss Factor) (Capacitance).
When FG0 is left blank (not zero), a default value of 100.Hz is assumed.
F-Scl
(36-40)
FMIN
(41-50)
NPD
(51-55)
Set F-Scl = "LOG" for logarithmically-spaced frequencies (default). See

"Earth/Frequency Card: Linear Scale", which follows, for other valid
keywords
Minimum or starting frequency in Hz. The default value of FMIN is
0.01 Hz for logarithmic spacing, 0.001 Hz for linear spacing, and
1000 Hz for user-defined discrete frequencies.
Number of frequency points per decade. Default is 10 for the FD-Model
class, and 1 otherwise. Logarithmic scale only.
10 - 23
Cable Constants
Number of decades. Default is 8 for the FD-Model class, and 6
otherwise. Logarithmic scale only. The program automatically selects
NPDNDC+1 logarithmically spaced frequencies in the range of FMIN to
FMIN10-NDC. For the FD-Model class, one point at FDC is also computed
for use by the rational function fitting routine.
NDC
(56-60)
Near-DC frequency in Hz. For "FD-Model" and "SCAN" options, default

value of FDC is the lesser of 10-5 Hz and FMIN/10. Otherwise FDC is
ignored. Logarithmic scale only.
FDC
(61-70)
Note: .For the FD-Model class, only LOG scale is allowed. For the CP-Model option all frequency
specifications made in this card are ignored.
Earth/Frequency Card: Linear Scale:

This form of frequency scaling can be used for EXACT-PI and SCAN calculations (see Section 9.2.1).
It cannot be used for the FD-Model class of models.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 12345 67890 1234567890 1234567890 1234567890 1234567890
RHO-E
MUE-E
FG0
F-Scl
FMIN
FDLT
FMAX
E10.0
E10.0
E10.0
A5
E10.0
E10.0
E10.0
RHO-E
(1-10)
MUE-E

Relative permeability of the earth (default is 1.0).
(11-20)
FG0
(21-30)
Breakpoint frequency (in Hz) of the shunt conductances for all

insulating layers. For all insulators, the shunt conductance G is
described with the following function of
F-Scl
Set F-Scl = "LIN" for linearly-spaced frequencies (default).
(36-40)
10 - 24
Cable Constants
(41-50)
Minimum or starting frequency in Hz. The default value of FMIN is

0.01 Hz for logarithmic spacing, 0.001 Hz for linear spacing, and
1000 Hz for user-defined discrete frequencies.
FDLT
Frequency-increment in Hz (default is 100.0 Hz). Linear scale only.
FMIN
(51-60)
FMAX
(61-70)
Maximum frequency in Hz (default is 5000.0 Hz). Linear scale only.

The program automatically selects FMIN, FMIN+FDLT, FMIN+2.FDLT, ...
FMAX as the frequency points.
Note: For the CP-Model option all frequency specifications made in this card are ignored.
Earth/Frequency Card: User-Supplied Frequencies:
This form of frequency scaling can only be used SCAN calculations (see Section 10.3.1). This card
must be followed by one or more Discrete Frequency Cards.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 12345 67890 1234567890123456789012345678901234567890
RHO-E
MUE-E
FG0
F-Scl
E10.0
E10.0
E10.0
A5
RHO-E
(1-10)
MUE-E

Relative permeability of the earth (default is 1.0).
(11-20)
FG0
(21-30)
Breakpoint frequency (in Hz) of the shunt conductances for all

insulating layers. For all insulators, the shunt conductance G is
described with the following function of
F-Scl
(36-40)
Set F-Scl = "DSC" for user-supplied frequency specification. Note that

RHO-E and MUE-E are only necessary if a CP-Model is requested, or if
RHOi and MUEi are left blank in the Discrete-Frequency cards below.
10 - 25
Cable Constants
Discrete-Frequency Cards
When F-Scl in the previous card is set to "DSC", the user must provide one data card for each
frequency, followed by a blank card to indicate the end of the Discrete Frequency Cards.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 12345678901234567890123456789012345678901234567890
FRQi
RHOi
MUEi
E10.0
E10.0
E10.0
FRQi
Frequency in Hz.
(1-10)
RHOi
(11-20)
MUEi
(21-30)
Earth resistivity. Units = m. Default is RHO-E of the Earth/Frequency

card above.
Earth relative permeability at FRQi.Default is MUE-E of the Earth/
Frequency card above.
Note: The last discrete-frequency card is followed by a blank card to indicate the end of the
discrete-frequency card set. Information provided by the Discrete-Frequency cards are ignored in
the case of CP-Models.
10.3.5
Optional Control Cards

Debug Card
Printout Control Card
Node Names Card
Fitting Control Card
Fitting Printout Card
Fitting Debug Card
End Card
Optional control cards allow the user to specify additional information (e.g., node names for
punched model output), and to have a greater degree of control over internal processes (e.g.,
rational functions fitting). They can also be used to request additional output and debugging
information on the internal solution process.
10 - 26
Cable Constants
These cards are optional and can appear in any sequence after the earth/frequency card (or discrete
frequency cards or external data cards, as the case may be). Their presence is flagged by a dot '.' in
column one joined to a keyword.
Debug Card:
2
3
4
5
6
7
8
01 23 456789012345678901234567890123456789012345678901234567890
.DBGCBL
A19
IROTQ
1
1234567890123456789
IDBG
This card controls debugging printouts and rotation of the Transformation matrix Q.
I2 I2
.DBGCBL
(1-19)
IDBG
(20-21)
Parameter controlling the amount of printout during calculation. Higher

values generate more printout (default is 0).
IROTQ
Parameter controlling the rotation of Q (does not apply to FDQ).
(22-23)
=1
Rotate Q so that modal capacitances Cm = diag (Q-1C Q-t) remain

real (default).
=2
Rotate Q to minimize the imaginary part of its element.
Printout Control Card:

This card controls the printout of modal quantities, as well as impedance and admittance matrices
at every frequency.
.
A19
I2 I2
I WM
.OUTCBL
I WY
I WZ
1
2
3
4
5
6
7
8
1234567890123456789 01 23 45 6789012345678901234567890123456789012345678901234567890
I2
10 - 27
Cable Constants
.OUTCBL
(1-19)
IWZ
Parameter controlling printout of impedance matrices.
(20-21)
IWY
=0
no printout (default).
=1
printout of impedance matrix Z = R + jL.
=2
printout of resistance and inductance matrices R and L.
Parameter controlling printout of admittance matrices.
(22-23)
IWM
=0
=1
printout of admittance matrix Y = G + jC.
=2
printout of conductance and capacitance matrices G and C.
Parameter controlling the printout of modal quantities.
(24-25)
=0
=1
printout of modal impedances, admittances, characteristic

admittances and propagation functions.
=2
same as IWM = 1, plus the transformation matrix Q.
=3
same as IWM = 2, with the addition of:

[Q-1 YZ Q - diag (YmodeZmode)] = E
(to verify the validity of Q). Ideally, E should be the zero
matrix.
10 - 28
Cable Constants
Node Names Card

This card provides node names for the sending and receiving ends of the cable. These names
appear in the punched output for the models of the cable, and correspond to phase numbers 1 to N
specified in KPH in the conductor/insulator cards (see Sections 9.2.3.2 and 9.2.4.2).
1
2
3
4
5
6
7
8
1234567890123456789 012345 6789 012345 6789 012345 6789 012345 6789 012345 6789 012345 67890
.NODES
A19
SEND-1
RECV-1
SEND-2
RECV-2
SEND-3
RECV-3
A6
A6
A6
A6
A6
A6
.NODES
(1-19)
SEND-1
(20-25)
RECV-1
(30-35)
SEND-2
(40-45)
RECV-2
(50-55)
SEND-3
(60-65)
RECV-3
(70-75)
When there are more than 3 phases, the node names for the next 3 phases are provided on the next
card (with the same format, from column 20 to 75) and so on. The first 6 columns of these
additional cards must be either blank or .NODES.
Fitting Control Card:

This card changes the value of parameters which control the fitting of modal characteristic
admittances Yc,mode, propagation functions Amode, and elements of transformation matrix Q.
10 - 29
Cable Constants
.
A19
IFITQ
IFITA
IFITYC
IXDYN
.CTLFIT
IQUICK
NORMAX
1
2
3
4
5
6
7
8
1234567890123456789 01 23 45 67 89 01 2345678901234567890123456789012345678901234567890
I2 I2 I2 I2 I2 I2
.CTLFIT
(1-19)
Keyword (left-justified)
NORMAX
Maximum number of poles in the synthesis of Yc,mode, Amode and Q

(default is 25).
(20-21)
IQUICK
(22-23)
IXDYN
(24-25)
IFITYC
(26-27)
IFITA
(28-29)
Flag controlling the accuracy of the fit of Yc,mode, Amode, and Q (default
is 1).
=1
Much faster, but less accurate, approximation of Yc,mode, Amode

and Q is produced.
= 1
Accurate fitting procedure is used.
Flag controlling the approximation of Amode in the low frequency range

(default is 1).
=1

approximation of the low frequency region of the propagation
function Amode. This allows a more accurate simulation of very
short cable sections and of very low frequencies.
= 1
No extra dynamics are added to the approximation of Amode.

This results in a lower order approximation but less accurate fit
for short cable sections or very low frequencies. In some cases,
IXDYN = 1 can produce low frequency oscillations (e.g., 1 Hz or
so) in the simulation of trapped charge.
Flag controlling which modes of Yc,mode will be fitted.

=0
fit all modes (default).
=N
fit only mode "N" of Yc,mode .
= 1
do not fit any mode of Yc,mode .
Flag controlling which modes of Amode shall be fitted.

=0

10 - 30
Cable Constants
=N
fit only mode "N" of Amode .
= 1
do not fit any mode of Amode .
Flag controlling which modes of Q will be fitted.
IFITQ
(30-31)
=0
=K
fit only element "K" of Q. Elements of Q are numbered columnwise, so for an NxN matrix, Q (i,j) is number i+(j-1)N.
= 1
do not fit any element of Q.
Note: To produce FD-Models, the synthesis or fitting of all modes is required. This is the
normal case where IFITYC, IFITA and IFITQ are 0 or blank.
Fitting Printout Card:

This card controls the amount of intermediate output information on the fitting on Yc,mode, Amode
and Q.
.
A19
IPRAT
.OUTFIT
IPLOTF
ICOMPF
1
2
3
4
5
6
7
8
1234567890123456789 01 23 45 6789012345678901234567890123456789012345678901234567890
I2 I2 I2
.OUTFIT
(1-19)
ICOMPF
Flag controlling the printout of a comparison table between calculated

and fitted functions (default is 1).
(20-21)
=1
An output table is produced comparing the data functions

Yc,mode, Amode and Q as produced by the Cable-Model routine
and the approximating rational functions produced by the fitting
routine.
= 1
10 - 31
Cable Constants
IPLOTF
(22-23)
IPRAT
(24-25)
Flag controlling the printout of a printer plot out the output file (default
is 1).
=1
A printer plot is produced comparing the calculated cable

functions (Yc,mode, Amode, and Q) with their corresponding
rational function approximations.
= 1
Flag controlling the printout of poles and zeros from the rational
function approximations (default is 1).
=1
Tables are produced showing the location of the poles and zeros
of the rational function approximations of Yc,mode, Amode and Q.
Also shown are the RC equivalent network for Yc,mode, and the
time domain exponential representation of the approximating
functions.
= 1
No tables are produced.
Fitting Debug Card:

This card controls the amount of printout of internal processing information in the fitting routine.
IBUGF
1
2
3
4
5
6
7
8
1234567890123456789 01 23456789012345678901234567890123456789012345678901234567890
.OUTFIT
A19
I2
.OUTFIT
(1-19)
IBUGF
Flag controlling the level of diagnostic printout. Allowed values are 0,

1, 2, or 3 (default = 0). The higher the value, the larger the amount of
debugging output.
(20-21)
End Card:
The end-of-data card of a Cable-Model case is indicated by a blank card (or the keyword
"BLANK"). Note that the end of discrete frequency cards was also indicated by a blank card. This
10 - 32
Cable Constants
means that in the case of discrete frequencies, and in the absence of optional control cards, there
will be 2 blank cards at the end of the data case.
10.3.6
External Data Cards
It is possible to link the Cable Model routines to an external cable constants program as long as
data are transmitted via an ASCII file in the format described below. This option would permit the
creation of frequency dependent cable models using parameter calculation techniques different
from those used in this module (e.g., finite elements).
External Data File Control Cards:

To use the external data option, all the cards following the Cable-Model control card are replaced
by two cards. One card specifies the name of the external file:
1
2
3
4
5
6
7
8
1234567890123456789023456789012345678901234567890123456789012345678901234567890
EXTERNAL FILENAME
A80
External Filename
(1-80)
Name of the file containing the impedance and admittance

matrices
Since AUX converts all characters to uppercase, the name of this
file must be in uppercase. This is important with operating
systems with case-sensitive syntax
The next card provides the following information:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 12345 67890 1234567890123456789012345678901234567890
NPHS
FDC
FMIN
NPD
NDC
I10
E10.0
E10.0
I5
I5
10 - 33
Cable Constants
Number of phases (size of Zphase and Yphase matrices).
NPHS
(1-10)
Near-DC frequency in Hz.
FDC
(11-20)
Starting frequency in Hz.
FMIN
(21-30)
Number of points per decade for frequency scale.
NPD
(31-34)
Number of decades for frequency scale.
NDC
(35-40)
Note : FDC, FMIN, NPD and NDC need to be specified only for the FD-Model class. Otherwise,
they are ignored.
Also, the number of points and frequency scale (spacing) of the externally-supplied data must
match FDC, FMIN, NPD and NDC exactly. The Cable-Model routine does not provide a sanity check
to this effect.
External Data File Structure:

The structure of the external data must adhere to the following rules:
1.
Any number of comment lines at the beginning of the file. Each line must start
with "C " or "c " in columns 1 and 2. Comment lines among the data lines are not
allowed.
2.
First data line containing the number of phases (integer). This number must
match NPHS on the external data card of Cable-Model data case.
3.
One record containing the frequency F in Hz, followed by data records

containing the row-wise lower triangle matrices of Zphase in /km and Yphase in
S/km at evaluated at F (real and imaginary parts). For example, if NPHS = 3, then
F
Zr(1,1)
Zi(1,1)
Zr(2,1)
Zi(2,1)
Zr(2,2)
Zi(2,2)
Zr(3,1)
Zi(3,1)
Zr(3,2)
Zi(3,2)
Zr(3,3)
Zi(3,3)
10 - 34
Cable Constants
Yr(1,1)
Yi(1,1)
Yr(2,1)
Yi(2,1)
Yr(2,2)
Yi(2,2)
Yr(3,1)
Yi(3,1)
Yr(3,2)
Yi(3,2)
Yr(3,3)
Yi(3,3)
Zphase and Yphase could be in units of /m and S/m, in which case the cable length
must be specified in units of meters on the Cable-Model control card.
Since data are read in free format, it does not matter how many data lines are used to
produce Zphase and Yphase, as long as the correct number of terms is used.
4.
Repeat c) as many times as there are frequency points defined by FDC, FMIN,
NDC, and NDC. There should be no blank lines or comment lines between valid
data.
The total number of frequency points depends on the type of model to be generated. If NPHS = 1,
there should be exactly NDC NPD + 2 points, organized as follows:
FDC
F(1) = FMIN
F(2) = F(1) DELTAF
F(3) = F(2) DELTAF
...
where DELTAF = 10-NPD.

If NPHS > 1, there should be exactly NDC NPD + 3 points, organized as follows:
FSEED
FDC
F(1) = FMIN
F(2) = F(1) DELTAF
F(3) = F(2) DELTAF
...
where DELTAF = 10-NPD.

In the case of FD models with constant modal transformation matrix Q, FSEED is the frequency at
which Q constant and real will be evaluated (from Z and Y at F = FSEED). In the case if an FDQ
model, FSEED is the frequency at which the seeding Q is calculated. FSEED = FMIN usually gives
reasonable results.
Note: Following the second External Data File Control Card, any number of special control (i.e.,
dot cards) can be specified, followed by the appropriate number of End Cards (i.e., blank cards).
10 - 35
Cable Constants
10.4
Cable-Model and Cable-Parameters Format Conversion
The format conversion option between Cable-Model and Cable-Parameters modules provided in
this version of AUX, does not always produce equivalent data cases because the capabilities of both
modules are different.
The recommended approach to data conversion is to set IPCH = 1, and IRUN = 0 in the Cable-Model
Control Card (see Section 10.3.1). This has the effect of producing a punch file which contains the
converted input data file without actually running the converted data case.
Nevertheless, efforts have been made to produce meaningful translations by providing defaults and
built-in assumptions when necessary.
10.4.1
Conversion from Cable-Model to Cable-Constants
This conversion takes place when either IRUN=1 (columns 71-75) or IPCH=1 (columns 67-80) of
the Cable-Model Control Card (see Section 10.3.1) and the "CABLE-PARAMETERS" keyword is used
to request a Cable-Parameters simulation.
MODEL
Q-OPTN
F-Scl
FDQ
LOG
Approximate conversion to Frequency-Dependent

model with constant Q (JMARTI line model). IRUN set
internally to zero.
QREAL
LOG
Equivalent conversion to Frequency-Dependent model

with constant Q (JMARTI line model). Q calculated at
FREQ-Q.
QREAL
N/ A
Equivalent conversion to constant-parameter model

calculated at f=FREQ-Q.
QCMPLX
N /A
Approximate conversion to constant-parameter model

calculated at f=FREQ-Q. Modal transformation matrix
is assumed to be real. IRUN set to zero.
SCAN
FDQ
LOG
LIN
DSC
Equivalent conversion. Parameters calculated at each

frequency within the range specified.
EXACT-PI
FDQ
LOG
LIN
DSC
Approximate conversion. Cross-bonding option is

used for non-cross-bonded cable, and homogeneous
(Discrete) pi, depending on conductor and grounding
arrangement. Discrete pi is used only if all the sheaths
are connected together but not grounded. In this case,
the discrete pi is used with a 1000 grounding
resistance. IRUN is set internally to 1.
FD-MODEL
CP-MODEL
COMMENTS
10 - 36
Cable Constants
In most cases, if a conversion option is not available, an appropriate error message is issued. For
example, SCAN and EXACT-PI options with QREAL or QCMPLX are not converted.
10.4.2
Conversion from Cable-Constants to Cable-Model
This conversion takes place when either IRUN=1 (columns 71-75) or IPCH=1 (columns 67-80) of
the Cable-Model Control Card (see Section 10.3.1) and the "CABLE-MODEL" keyword is used to
request a Cable-Model simulation.
Option
Cross-bonding
Comments
Only those cases for which NCROS is not zero are converted (i.e.,
sheaths are not cross-bonded). In such cases, all grounding codes are
ignored, and all sheaths (armours and pipe, if any) are grounded.
Also, all sheaths are assumed to be connected together (ISEP ignored)
and grounding resistances (ISG) are assumed to be zero.
FrequencyDependence
FD-Model
with QREAL is used. The modal transformation matrix is

calculated at the frequency specified in the first frequency card.
Otherwise, 1 kHz is assumed.
ConstantParameters
SCAN option with FDQ is used. Discrete frequency scale (DSC) is used
to produce output for each frequency card provided.
In most cases, if a conversion option is not available, an appropriate error message is issued, for
example, cables in air, stratified earth, and overhead lines are not converted.
10.5
CABLE-MODEL Examples
The following examples shop a few data cases to illustrate some of the most commonly-used
features of the Cable-Model module. Please note that text in italics (such as this) is used as an
annotation and it is not part of the data file.
Case 1: Setting up a pipe-type cable case

BEGIN NEW DATA CASE
C
C ---------------------------------------------------------------------C
C
Pipe Type cable with 3 identical, 2-conductor cables
C
Frequency dependent model.
10 - 37
Cable Constants
C
Modal transformation matrix Q is frequency dependent.
C
Sheaths grounded
C
C ---------------------------------------------------------------------CABLE CONSTANTS
C Cable-model Control Card
C --key----|
|<--model--<--Q-optn-<-freq-q--<--length-| |<--ext-<ipch<irun
cable-model
fd-model fdq
1000.
1.
0
0
C
C Cable-type card
C TY<NCBL<IXBD
each cable has a core and a sheath, no cross-bonding
PT
3
0
C -----------------------------------------------------------------------------C Data for SC cable # 1 follows
C
C PT Cable: Individual Cable Card
C NCN| |<----DST---<----ANG--<--ROUT--2
0.02442
0.0
.03809
C
C PT Cable: Conductor/Insulator Card
C --RIN---<--ROUT---<---RHO---<---MUE---<--MUE-I--<--EPS-I--<--LFCT---<-KPH
.00001
.02197 2.945E-8
1.
1.
3.5
.005
1
.033274
.03744 1323.E-8
1.
1.
2.0
.001
0
C -----------------------------------------------------------------------------C Data for PT cable # 2 follows (copy data from PT cable # 1)
C NCN| |<----DST---<----ANG--<--ROUT--NCN=-1 to indicate copy data from cable #1
-1
0.061042
137.7
C
C PT Cable: Phase Number Card
C
<KPH1<KPH2<KPH3
KPH2=0 to indicate that the sheath is grounded
2
0
C -----------------------------------------------------------------------------C Data for PT cable # 3 follows (copy data from PT cable # 1)
-1
0.061042
-137.7
C
C PT Cable: Phase Number Card
C
<KPH1<KPH2<KPH3
3
0
C -----------------------------------------------------------------------------C
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
C Pipe-Data Cards
C --RIN---<--ROUT---<---REXT--<--VDPTH--<-KPH
KPH=0 to ground the pipe
.1024
.1094
.10954
1.0
0
C --RHO---<---MUE---<-MUE-IN--<-EPS-IN--<-LFCT-IN-<-MUE-OUT-<-EPS-OUT-<-FCT-OUT3281.
9.
1.
3.0
.002
1.
2.0
.002
C -----------------------------------------------------------------------------C
C Earth/Frequency Card (log scale)
C -RHO-E--<--MUE-E--<---FG0---|
|<FScl<---FMIN--<-NPD<-NDC<---FDC--50.
1.
0.
LOG
.1
10
8
.1
C -----------------------------------------------------------------------------C
C Optional cards
C ---- nodes ------<-snd1| |<-rcv1| |<-snd1| |<-rcv1| |<-snd1| |<-rcv1|
.nodes
snd1
rcv1
snd2
rcv2
snd3
rcv3
C
1
2
3
C 3456789 123456789 123456789 1
.dbgcbl
0
.outcbl
0 0 0
10 - 38
Cable Constants
.ctlfit
25-1 1 0 0 0
no printer plot or equivalent network requested
.outfit
1-1-1
.dbgfit
0
C
C Summary of control flags:
C Keyword
COL
FLAG
C .dbgcbl
20-21
IDBG
C
22-23
IROTQ
C .outcbl
20-21
IWZ
C
22-23
IWY
C
24-25
IWM
C .nodes
20-25
send-1
C
30-35
recv-1
C
40-45
send-2
C
50-55
recv-2
C .ctlfit
20-21
normax
C
22-23
iquick
C
24-25
ixdyn
C
26-27
ifityc
C
28-29
ifita
C
30-31
ifitq
C .outfit
20-21
icompf
C
22-23
iplotf
C
24-25
iprat
C .dbgfit
20-21
ibugf
C
BLANK CARD ENDING CABLE-MODEL DATA
BLANK CARD ENDING CABLE CONSTANT DATA CASE
BEGIN NEW DATA CASE
BLANK
Case 2: Setting up a Single-Core Cable

Sheaths are not grounded so we will obtain a six-conductor model. This could be used, for
example, as an explicit model of a minor section of a cross-bonded cable. An FDQ model is
requested.
BEGIN NEW DATA CASE
C -----------------------------------------------------------------------------C
3 SC identical cables, Frequency dependent model. Modal transformation
C
matrix Q is frequency dependent. Sheath ungrounded
C -----------------------------------------------------------------------------CABLE CONSTANTS
C --key----|
cable-model
fd-model fdq
1.
0
0
C
C Cable-type card
C TY<NCBL<IXBD
SC
3
0
C
C SC Cable: Individual Cable Card
10 - 39
Cable Constants
C NCN| |<----VRT---<----HRZ--<--ROUT--2
1.1
0.0
.029335
C
KPHS in this card defines EMTP phases
C SC Cable: Conductor/Insulator Card
.003175
.01254
.17D-7
1.
1.
3.5
.001
1
.022735
.026225
.21D-6
1.
1.
2.0
.001
2
C -----------------------------------------------------------------------------C Data for SC cable # 2 follows (copy data from SC cable # 1)
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.25
C
C SC Cable: Phase Number Card
C
<KPH1<KPH2<KPH3
KPH will go from 1 to 6 to retain all conductors in the final model
3
4
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.50
C
C
<KPH1<KPH2<KPH3
5
6
C -----------------------------------------------------------------------------C
C -RHO-E--<--MUE-E--<---FG0---|
1.
0.
log
.1
10
8
.1
C -----------------------------------------------------------------------------C
C Optional cards
C
.dbgcbl
.outcbl
0 0 0
c ---- nodes ------<-snd1| |<-rcv1| |<-snd1| |<-rcv1| |<-snd1| |<-rcv1|
.nodes
snd1
rcv1
snd2
rcv2
snd3
rcv3
snd4
rcv4
snd5
rcv5
snd6
rcv6
C
1
2
3
C 3456789 123456789 123456789 1
.ctlfit
35-1 1 0 0 0 Extra poles are requested. default is 25.
Full output request
.outfit
1 1 1
.dbgfit
0
C
C Keyword
COL
FLAG
C .dbgcbl
20-21
idbg
C
22-23
irotq
C .outcbl
20-21
iwz
C
22-23
iwy
C
24-25
iwm
C .nodes
20-25
send-1
C
30-35
recv-1
C
40-45
send-2
C
50-55
recv-2
C .ctlfit
20-21
normax
C
22-23
iquick
10 - 40
Cable Constants
C
24-25
ixdyn
C
26-27
ifityc
C
28-29
ifita
C
30-31
ifitq
C .outfit
20-21
icompf
C
22-23
iplotf
C
24-25
iprat
C .dbgfit
20-21
ibugf
C
BLANK CARD ENDING CABLE CONSTANTS DATA CASE
C BLANK
C
C
C
Case 3: Setting up a Single-core cable with Q constant.

Also, in this case, the sheaths will be grounded to produce a 3-conductor model.
BEGIN NEW DATA CASE
C -----------------------------------------------------------------------------C
3 SC identical cables, Frequency dependent model.
C
Modal transformation matrix Q is real. it is computed at 1 KHZ.
C
Sheath grounded
C --key----|
cable-model
fd-model QREAL
1000
1.
0
0
C
C Cable-type card
C TY<NCBL<IXBD
SC
3
0
C
C NCN| |<----VRT---<----HRZ--<--ROUT--2
1.1
0.0
.029335
C
C SC Cable: Conductor/Insulator Card KPH=0 on sheath specification to ground it
.003175
.01254
.17D-7
1.
1.
3.5
.001
1
.022735
.026225
.21D-6
1.
1.
2.0
.001
0
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.25
C
C
<KPH1<KPH2<KPH3
2
0
10 - 41
Cable Constants
C
C
C
C
C
-----------------------------------------------------------------------------Data for SC cable # 3 follows (copy data from SC cable # 1)

SC Cable: Individual Cable Card
NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.50
C
C
<KPH1<KPH2<KPH3
3
0
C -----------------------------------------------------------------------------C
C -RHO-E--<--MUE-E--<---FG0---|
1.
0.
log
.1
10
8
.1
C -----------------------------------------------------------------------------C
C Optional cards
C
.nodes
snd1
rcv1
snd2
rcv2
snd3
rcv3
C
1
2
3
C 3456789 123456789 123456789 1
Most control cards are absent: default values are assumed
.outfit
1 1-1
C Keyword
COL
FLAG
C .outfit
20-21
icompf
C
22-23
iplotf
C
24-25
iprat
BEGIN NEW DATA CASE
BLANK
C
C
Case 4: Setting up a Single-core cable with a CP model

This case also illustrates the specification of non-identical conductors. Note that in a constantparameters model, R and L are calculated at the same frequency as the modal transformation
matrix Q.
BEGIN NEW DATA CASE
C -----------------------------------------------------------------------------C
3 SC non-identical cables, Constant parameter Model computed at 1 KHZ
C
Sheath ungrounded
C --key----|
cable-model
cp-model QREAL
1000
1.
0
0
C
C Cable-type card
C TY<NCBL<IXBD
SC
3
0
C
10 - 42
Cable Constants
C NCN| |<----VRT---<----HRZ--<--ROUT--2
1.1
0.0
.029335
C
.003175
.01254
.17D-7
1.
1.
3.5
.001
1
.022735
.026225
.21D-6
1.
1.
2.0
.001
2
C
C NCN| |<----VRT---<----HRZ--<--ROUT--NCN is positive. Data is not copied.
2
1.1
0.25
.028335
C
.004175
.01154
.17D-7
1.
1.
3.7
.001
3
.020735
.024225
.21D-6
1.
1.
2.5
.001
4
C
C NCN| |<----VRT---<----HRZ--<--ROUT--Data is copied from cable # 1 only
-1
1.1
0.50
C
C
<KPH1<KPH2<KPH3
KPH from 1 to 6 to indicate all condutors are retained
5
6
C -----------------------------------------------------------------------------C
C -RHO-E--<--MUE-E--<---FG0---|
|<FScl<---FMIN--<-NPD<-NDC<---FDC--Anything beyond MUE-E is meaningless for a CP model
250.
1.
C -----------------------------------------------------------------------------C
C Optional cards
C
.nodes
snd1
rcv1
snd2
rcv2
snd3
rcv3
snd4
rcv4
snd5
rcv5
snd6
rcv6
Print Y, Z in phase quantities (IWM=0) at f = FREQ-Q.
.outcbl
1 1 0
.ctlfit
35-1 1 0 0 0
.outfit
1 1 0
C
C Keyword
COL
FLAG
C .outcbl
20-21
IWZ
C
22-23
IWY
C
24-25
IWM
C .ctlfit
20-21
normax
C
22-23
iquick
C
24-25
ixdyn
C
26-27
ifityc
C
28-29
ifita
C
30-31
ifitq
C .outfit
20-21
icompf
C
22-23
iplotf
C
24-25
iprat
10 - 43
Cable Constants
C
BEGIN NEW DATA CASE
BLANK
C
C
Case 5: Setting up a Single-core cable. No model generated.

This example illustrates the calculation of cable parameters without generating a model.
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
BEGIN NEW DATA CASE
C -----------------------------------------------------------------------------C
3 SC identical cables, to compute cable parameters (not model)
C
at specified frequencies.
C --key----|
cable-model
scan
fdq
1000
1.
0
0
C
C Cable-type card
C TY<NCBL<IXBD
SC
3
0
C
C NCN| |<----VRT---<----HRZ--<--ROUT--2
1.1
0.0
.029335
C
.003175
.01254
.17D-7
1.
1.
3.5
.001
1
.022735
.026225
.21D-6
1.
1.
2.0
.001
2
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.25
C
C
<KPH1<KPH2<KPH3
3
4
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.50
C
C
<KPH1<KPH2<KPH3
10 - 44
Cable Constants
5
6
C -----------------------------------------------------------------------------C
C Earth/Frequency Card
C -RHO-E--<--MUE-E--<---FG0---|
|<FScl
250.
1.
0.
dsc
C Discrete-Frequency cards
C --FRQ--<---RHO---<---MUE--Parameters will be printed at 10 Hz and 1 kHz
60.
1000.
BLANK CARD TO TERMINATE DISCRETE FREQUENCY CARDS
C -----------------------------------------------------------------------------C
C Optional cards
C
1
2
3
C 3456789 123456789 123456789 1
IWM = 2 to request modal output
.outcbl
1 1 2
C
C Keyword
COL
FLAG
C .outcbl
20-21
IWZ
C
22-23
IWY
C
24-25
IWM
BEGIN NEW DATA CASE
BLANK
C
C
Case 6: Data conversion request from Cable-Mode to CableParameters

Conversion request is indicated by IPCH=1 in the Cable-Model Control Card. The direction of the
conversion is determined by the fact that this is a Cable-Model data case, and it can only be
converted to a Cable-Parameters data case.
C
1
2
3
4
5
6
7
8
C 3456789 123456789 123456789 123456789 123456789 123456789 123456789 123456789
BEGIN NEW DATA CASE
C -----------------------------------------------------------------------------C
3 SC identical cables, Frequency dependent model. Modal transformation
C
matrix Q is constant
C
Data conversion is requested
C
Sheath ungrounded
C --key----|
cable-model
fd-model qreal
1.
1
0
C
C Cable-type card
C TY<NCBL<IXBD
SC
3
0
C ------------------------------------------------------------------------------
10 - 45
Cable Constants
C Data for SC cable # 1 follows
C
C NCN| |<----VRT---<----HRZ--<--ROUT--2
1.1
0.0
.029335
C
.003175
.01254
.17D-7
1.
1.
3.5
.001
1
.022735
.026225
.21D-6
1.
1.
2.0
.001
2
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.50
C
C
<KPH1<KPH2<KPH3
5
6
C
C NCN| |<----VRT---<----HRZ--<--ROUT---1
1.1
0.25
C
C
<KPH1<KPH2<KPH3
3
4
C -----------------------------------------------------------------------------C
C -RHO-E--<--MUE-E--<---FG0---|
1.
0.
log
.1
10
8
.1
C -----------------------------------------------------------------------------C
C Optional cards
C
.nodes
snd1
rcv1
snd2
rcv2
snd3
rcv3
snd4
rcv4
snd5
rcv5
snd6
rcv6
C BLANK
C
C
C
C
Case 7: Data conversion from Cable-Parameters to Cable-Model

Conversion request is indicated by IPCH=1 in the Cable-Model Control Card. The direction of the
conversion is determined by the fact that this is a Cable-Model data case, and it can only be
converted to a Cable-Parameters data case.
10 - 46
Cable Constants
C
C -----------------------------------------------------------------------------C
Cable-Parameters data case (old cable constants data format)
C
Conversion requested
C
3 SC identical cables
C
Sheath ungrounded
C
CABLE CONSTANTS
C
C
This data file is in "old" CABLE CONSTANTS format
C --key--------|
|<ipch<irun
CABLE-PARAMETERS
1
C
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
C ITY ISYS NPC IEAR KMOD IZFL IYFL NPP NGRN
2
-1
3
0
1
-1
-1
0
C
C NPC NPCC NPCC NPCC NPCC NPCC ...
2
2
2
C
C CONDUCTOR
C
R1
R2
R3
R4
R5
R6
R7
C
RC
UC
UI1
EI1
RS
US
UI2
EI2
0.003175 0.012540 0.022735 0.026225 0.029335
.17000E-07
1.000
1.000
3.500.21000E-06
1.000
1.000
2.000
C
C CONDUCTOR
C
R1
R2
R3
R4
R5
R6
R7
C
RC
UC
UI1
EI1
RS
US
UI2
EI2
0.003175 0.012540 0.022735 0.026225 0.029335
.17000E-07
1.000
1.000
3.500.21000E-06
1.000
1.000
2.000
C
C CONDUCTOR
C
R1
R2
R3
R4
R5
R6
R7
C
RC
UC
UI1
EI1
RS
US
UI2
EI2
0.003175 0.012540 0.022735 0.026225 0.029335
.17000E-07
1.000
1.000
3.500.21000E-06
1.000
1.000
2.000
C
C
C
VERT1
HORIZ1
VERT2
HORIZ2
VERT3
HORIZ3
. . .
1.100
0.000
1.100
0.250
1.100
0.500
C
C
REARTH
FREQ
C
REARTH
FREQ IDEC IPNT
DIST2
0.250000E+03
0.100000E+04
.FIT-S
0.250000E+03
0.100000E+00
8
10 1000.0
BEGIN NEW DATA CASE
BLANK
10 - 47
Cable Constants
10.6
Cababilities of the CABLE-PARAMETERS Module
10.6.1
Introduction
The cable parameters module calculates the resistance, inductance, and capacitance matrices for an
arbitrary configuration of single-core (SC) coaxial cables. A pipe-type configuration, where the
aforementioned SC coaxial cables are all enclosed in a conducting pipe, is also allowed. Line
constants for conventional overhead transmission lines can also be calculated using "CABLE
PARAMETERS": this calculation is completely independent of the "LINE CONSTANTS" routine of
Section 9.
Pipe
external
insulator
armour
sheath
sheath
insulator
core
main
insulator
Figure 10.1:
Cross-Section of Most General SC

Coaxial Cable Geometry
Figure 10.2:
Illustrative Cross-Section
of 0verhead Pipe-Type
Cable System which
contains three SC coaxial
cables
The "CABLE PARAMETERS" (formerly CABLE CONSTANTS) module of the EMTP was superseded
with the "CABLE-MODEL" module (version 3.0 and higher). However, since "CABLE-PARAMETERS"
still has some little-used capabilities such as stratified earth modelling, it has not been removed
from AUX at this point in time.
10.6.2
General Structure of "CABLE-PARAMETERS" Data Cards
There are classes of cables considered in the "CABLE PARAMETERS" module:
Class A
A system of single-core coaxial cables without any enclosing pipe.

10 - 48
Cable Constants
Class B
A system of single-core coaxial cables within an enclosing pipe.
Class C
A system of conventional overhead conductors.
Class A Data Structure:

This is the case of a system of single-core coaxial cables that have no conducting-pipe enclosure.
The input data structure for this class is as follows:
A1
First comes a "BEGIN NEW DATA CASE" card, followed by a "CABLE

CONSTANTS" request card.
A2
Next comes a "CABLE PARAMETERS" request card.
A3
Next comes miscellaneous data card.
A4
Next comes one (or possibly more) cards upon which is punched the
number of conductors which make up each SC coaxial cable of the
system. One card will suffice for a system of up to sixteen cables; two
cards are required for 1732 cables, etc.
A5
Next come two (or possibly three) cards of geometrical and physical data
for each SC coaxial cable in the system, e.g,. for three SC coaxial cables, a
maximum of nine cards are required.
A6
Next comes one (or possibly more) card(s) which gives the horizontal
and vertical location of the centres of all SC coaxial cables in the system.
A single card can handle up to four SC coaxial cables; two cards are
required for 58, etc.
A7
Next comes a frequency card, which specifies a new earth resistivity and
frequency (or range of frequencies) for which cable constants are to be
calculated.
A8
Only if the stratified earth modelling is being used for an overhead cable
system, two additional cards are required, to give all parameters of this
more complex earth model. If the earth is instead modelled as
homogenous, no such cards exist.
A9
Repeat the data of Point A7 and Point A8 for as many different discrete
frequencies or frequency ranges as desired. Signal the end of such data
by means of a blank card.
A10
Repeat the data of Points A2-A9 as often as may be desired. Each of

such grouping is a separate, distinct, independent data case within the
"CABLE PARAMETERS" routine. Signal the end of such data by means of
a blank card, which transfers control back to the main AUX module,
ready to read in a new AUX data case.
10 - 49
Cable Constants
Class B Data Structure

This is the case of a system of SC coaxial cables which are all enclosed by a conducting pipe. The
input data structure for this class is as follows:
B1

CONSTANTS" request card.
B2
B3
Next comes a miscellaneous data card.
B4
Next will come one card which gives parameters of the pipe.
B5
Next will come one (or possibly more) card(s) which specifies the
location of each SC coaxial cable within the pipe. One card will suffice
for up to four SC coaxial cables, two will be required for 58 SC coaxial
cables, etc.
B6
Next comes one (or possibly more) card(s) upon which is punched the
number of conductors which make up each SC coaxial cable of the
system. One card will suffice for a system of up to sixteen cables; two
cards are required for 1732 cables, etc.
B7
Next come two (or possibly more) card(s) of geometrical and physical
data for each SC coaxial cable in the system, e.g., for three SC coaxial
cables, a maximum of nine cards are required.
B8
Next comes one card which gives the horizontal and vertical location of
the centre of the pipe.
B9
frequency (or range of frequencies) for which cable constants are to be
calculated.
B10
Only if the stratified earth modelling is being used for an overhead cable
system, two additional cards are required, to give all parameters of this
more complex earth model. If the earth is instead modelled as
homogeneous, no such cards exist.
B11
Repeat the data of Point B9 and Point B10, for as many different discrete
frequencies or frequency ranges as desired. Signal the end of such data
by means of a blank card.
B12
Repeat the data of Points B2-B9 as often as may be desired. Each of

a blank card, which transfers control back to the main AUX module,
ready to read a new AUX data case.
10 - 50
Cable Constants
Class C Data Structure

This is the case of a system of conventional overhead conductors, as is usually solved using the
"LINE CONSTANTS" routine of Section 9. The input data structure for this class is as follows:
C1

CONSTANTS" request card
C2
C3
Next comes a miscellaneous data card.
C4
Next come three cards for each circuit which belongs to the overhead
conductor system. Parameters specified include the number of phases,
the number of ground wires, the number of conductors in a bundle,
geometrical data, conductor resistivity, etc., e.g., considering a system
which consists of a single-circuit 500-kV transmission line and a doublecircuit 230-kV transmission line all on the same right-of-way, nine data
cards would be involved.
C5
Next comes one (or possibly more) data card(s) which gives the height,
sag, and horizontal location for the centre of each bundle of each circuit
of the system. One card will suffice for 1 or 2 bundles, two cards are
required for 3 or 4 bundles, etc., e.g., two coupled single circuits, each of
which is supported by its own towers and has a single ground wire,
would require four cards (because there are eight bundles total -- four for
each circuit).
C6
frequency (or range of frequencies) for which line constants are to be
calculated.
C7
Only if the 3layer Nakagawa stratified earth model is being used, two
additional cards are required, to complete the parameters of this more
complex model of the earth. If the earth should instead be modelled as
homogeneous, no such cards exist.
C8
Repeat the data of Point C6 and C7 for as many different discrete

frequencies or frequency ranges as may be desired. Signal the end of
such data by means of a blank card.
C9
Repeat the data of Points C2-C9 as often as may be desired. Each of

a blank card, which serves to transfer control back to the main AUX
module, ready to ready a new AUX data case.
10 - 51
Cable Constants
10.7
Data Entry Rules for CABLE-PARAMETERS
The preceding section outlined the structure of a "CABLE PARAMETERS" data case in general terms.
The format and meaning of the associated data cards are now described in detail. Unscaled MKS
units are consistently used throughout, it may be noted (e.g., distance in meters, voltage in volts,
capacitance in farads, etc.).
10.7.1
Format for "A2", "B2", and "C2" Data
The data cards of Points A1-A2, B1-B2 and C1-C2 have the following format:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
BEGIN NEW DATA CASE
Followed by
1
2
3
4
5
6
7
8
123456789012345 67890123456789012345678901234567890123456789012 34 5678901234567890
CABLE CONSTANTS
A15
I2
N
(63-64)
= 1
Transmission line (C2 data).
=0
Excluding pipe, the majority of cables have two or less

conductors (B2 or A2 data).
= +1
Excluding pipe, the majority of cables have more than 2

conductors (B2 or A2 data).
1
2
3
4
5
6
7
8
12345678901234567 89012345678901234567890123456789012345678901234567890 12345 67890
IPCH IRUN
CABLE-PARAMETERS
A17
I5
10 - 52
I5
Cable Constants
CABLE-PARAMETERS
Keyword.
(1-17)
Flag to control the storage of input data files, after input format
conversion has been completed. Converted data files are stored
into the standard "punch" file.
IPCH
(71-75)
=0
Do not store converted input data file (only if IRUN = 1).
=1
Store converted input data file into punch file IRUN.
Flag to control program execution after input format conversion.
IRUN
(76-80)
10.7.2
=0
Do not run AUX after input data file has been converted
and stored into the punch file.
=1
Run AUX after data conversion is completed, whether or

not storage of the converted input data file is requested.
Note that if both IRUN = 1 and IPCH = 1, then any punched
output which results from an AUX run (e.g., punched file
for an FDQ model) will be appended to the record of the
converted input data file.
Format for "A3", "B3", and "C3" Data
The miscellaneous data card of Points A3, B3 and C3 has the following format:
12345
1
2
67890 12345 67890 12345
ITYPEC ISYST
I5
NPC
I5
I5
IEARH
I5
3
67890
KMODE IZFLAG
I5
I5
12345
4
67890 12345
IYFLAF
NPP
I5
I5
5
6
7
8
67890123456789012345678901234567890
NGRND
I5
(1-5)
Flag which indicates the class of data case which is presently being
created.
IPCH
CLASS A:
ITYPE
CLASS B:
ITYPE
ITYPEC
(71-75)
= 2, which implies a system of SC coaxial cables

without any surrounding pipe.
= 3, which implies a system of SC coaxial cables
which are enclosed within a conducting pipe.
10 - 53
Cable Constants
CLASS C:
ITYPE = 1, which implies a system of conventional overhead
conductors.
ISYST
(6-10)
For CLASS A and CLASS B cases "ISYST" indicates whether the cable
system is underground or in the air.
= 1
Underground (buried) cable system.
= 0
Cable system is in the air, but is touching the earth's surface.

See further comment elsewhere.
= +1
Cable system is in the air, above the surface of the earth.
For CLASS C cases, "ISYST" indicates possible transposition:
NPC
(11-15)
IEARTH
=0
Untransposed overhead line.
=2
Continuously-transposed overhead line.
CLASSES
A and B
Number of SC coaxial cables which make up the system of

interest.
CLASS C
The number of transmission circuits which make up the

overhead system of interest, e.g., the most common case
will have just NPC=1 circuit, which consists of three phaseconductor bundles (for a 3-phase line) and possibly one or
more ground-wire bundles
Flag indicating which model for the earth is to be used:
(16-20)
KMODE
(21-25)
IZFLAG
(26-30)
=0
for homogeneous (Carson) earth model.
= 99
for 3-layer stratified (Nakagawa) earth model. This is

allowed only for overhead systems (miscellaneous data
parameter "ISYST" equal to 0 or 1).
Flag used to request the calculation and output of various modal

quantities of interest, as described later.
=0
No modal calculation or output.
=1
Modal quantities will be calculated and printed during the

calculation at individual frequencies. If the automatic
looping is requested (see Section 10.7.10). These quantities
are calculated but not printed.
=2
Print modal quantities during automatic frequency looping.
Flag which indicates the format of series-impedance output (inductance

or reactance or both) in the phase domain.
= 1
Don't print anything.
= 0
Print matrices [R] and [L].

10 - 54
Cable Constants
= 1
Print matrices [R] and [L].
= 2
Print both of the above.
The diagonal elements are self-impedances of the conductors and the

off-diagonal elements are the mutual impedances. For overhead lines,
the order of the printout is the same as that of the phase conductors
entered. For SC coaxial cables, the printout starts from the inner-most
conductor of each cable in the order of input, then continues to the next
outer layer of the conductor of each cable, etc. For a pipe-type cable
system, the last series-impedance in the printout is for the pipe.
IYFLAG
(31-35)
Flag which indicates the format of shunt-admittance output (capacitance,

or susceptance, or both) in the phase domain.
= 1
Don't print anything.
= 0
Print matrices [G] and [C].
= 1
Print matrices [G] and [C].
= 2
Print both of the above.
The order of this printout is same as that for the series-impedance.

NPP
(36-40)
CLASSES
A&C
Unused (leave blank).
CLASS B
NGRND
(41-45)
=1
finite pipe thickness.
=2
pipe of infinite thickness. Miscellaneous data parameter

"ISYST" must also be zero, in this case.
This parameter describes the grounding conditions of the cable system,

i.e., for data in Class A and Class B.
CLASS A
= 0 or 1
None of the conductors is grounding.
=2
All armours, if any, are grounded.
=3
sheaths and armours, if any are grounded.
=4
See note below.
CLASS B
=0
None of the conductors is grounded.
=1
The pipe is grounded.
=2
All armours and pipe are grounded.

10 - 55
Cable Constants
=3
All armours and all sheaths, if any, and pipe are grounded.
=4
See note below.
CLASS C
Leave blank.
Note: If the grounding conditions are different for different cables in the system, or not all the
outer conductors of the cables are grounded, set NGRND = 4 and add one extra data card with the
format: 2X,78I1 right before the frequency card(s). (See Section 10.7.10.)
Input an integer code number (I1), based on the following rules, for each of the cables according to
their input ordering; and enter the pipe, if any, in the very last entry.
For all cables excluding pipe:

ngrnd
=0
None of the conductors of the cable is grounded.
=1
The core of cable is grounded.
=2
The sheath of the cable is grounded
=3
The armour of the cable is grounded.
=4
The sheath and the armour of the cable are grounded.
=5
The core and the sheath of the cable are grounded.
=6
The core and the armour of the cable are grounded.
=7
The core, the sheath, and the armour of the cable are grounded.
For pipes only:

NGRND
=0
Not grounded.
=1
Grounded.
10 - 56
Cable Constants
10.7.3
Format for "A4" and "B6" Data
For cable systems (either Class A or Class B), the number of conductors which make up each SC
coaxial cable of the system must be indicated; the following format is used:
1
2
3
4
5
6
7
8
12345 67890 12345 67890 12345 67890 12345 67890 1234567890123456789012345678901234567890
NCPP1 NCPP2 NCPP3 NCPP4 NCPP5 NCPP6 NCPP7 NCPP8
I5
I5
NCPPK
I5
I5
I5
I5
I5
I5
The k-th SC coaxial cable of the system has this many conductors in it:
=3
For the SC coaxial cable which has all three conductors: core,
sheath and armour.
=2
for the SC coaxial cable which has only two conductors: core and
sheath.
=1
for the situation of a core only.
For purposes of this input, it may be noted that the SC coaxial cables have been numbered between
"1" and "NPC". Such numbering is arbitrary, except that the string of NCPPk must be nonincreasing (that is, all 3-conductor SC coaxial cables must precede any 20-conductor cables and all
2conductor cables must precede any core-only cables). This ordering, once established, is
applied to the rest of the data case.
10.7.4
Format for "B4" Data
Point B4 data consists of a single card, upon which the user is to punch various parameters of the
pipe. The following format applies:
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
RP1
RP2
RP3
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
10 - 57
Cable Constants
RP1
Inner radius of the pipe, in units of meters.
(1-10)
RP2
Outer radius of the pipe, in units of meters.
(11-20)
RP3
(21-30)
Outer radius of the tubular insulator which surrounds the pipe, in units
of meters.
(31-40)
Resistivity of the pipe, in units of ohm-meters.
r
(41-50)
Relative permeability of the pipe. This is a dimensionless number: the

ratio of the permeability of the pipe to the permeability of free space.
Pipe insulation 2
inside
insulation 1
RP3
RP1
RP2
air or earth
surrounding
tubular insulator or pipe
Figure 10.3:
Cross-Section of a Pipe-Type Cable
1
(51-60)
Relative permittivity of the insulating medium which is inside the pipe

(between the pipe and SC coaxial cables which are contained). This is a
dimensionless quantity: the ratio of the permittivity of the inner
insulating medium to the permittivity of free space.
2
(61-70)
Relative permittivity of the insulating tube which surrounds the pipe.

This is a dimensionless quantity.
10 - 58
Cable Constants
10.7.5
Format for "C4" Data
For each circuit of the system of conventional overhead conductors which is being studied, three
data cards of the following format are to be punched.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
NP
NG
KBP
KBG
I5
I5
I5
I5
NP
(1-5)
NG
(6-10)
KBP
(11-15)
KBG
(16-20)
The number of phase-wire bundles which belong to the circuit in

question, e.g., for a 3phase circuit, "NP" will equal three.
The number of ground-wire bundles which belong to the circuit in
question.
The number of individual physical conductors which compose each
phase-wire bundle of the circuit in question. If there is no bundling of
phase wires, "KBP" will equal unity.
The number of individual physical conductors which compose each
ground-wire bundle of the circuit in question. If there is no bundling of
ground wires, "KBG" will equal unity.
g1
g2
NG = 2
SEPG
KBG = 3
NP = 3
SEPP
Figure 10.4:
KBP = 4
Illustrative single-circuit 3-phase overhead transmission

line (as seen in cross-section). The three phase-wire
bundles are of four conductors each; there are two
ground-wire bundles, of three conductors each.
10 - 59
Cable Constants
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
ROUTP
RINP
ROUTG
RING
SEPP
SEPG
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
ROUTp
(1-10)
Outer radius of the tubular conductors which are used for each phasewire bundle of the circuit in question. Units are meters.
RINP
RINp
(11-20)
ROUTG
(21-30)
Inner radius of the tubular conductors which are used for each phasewire bundle of the circuit in question. Units are meters.
Outer radius of the tubular conductors which are used for each groundwire bundle of the circuit in question. Units are meters.
RING
RING
(31-40)
SEPP
(41-50)
ROUTP
ROUTG
Inner radius of the tubular conductors which are used for each groundwire bundle of the circuit in question. Units are meters.
Separation between centres of two adjacent conductors of any one of the
phase-wire bundles. Units are meters. The "KBP" conductors of the
bundle are assumed to be uniformly spaced around the circumference of
a circle.
10 - 60
Cable Constants
SEPG
Separation between centres of two adjacent conductors of any one of the

ground-wire bundles. Units are meters. The "KBG" conductors of the
bundle are assumed to be uniformly spaced around the circumference of
a circle.
SE
P
(51-60)
SEP
Figure 10.5:
p
(1-10)
Illustration of two different bundles; that on the left has

four conductors, while the bundle on the right has only
two. Note the uniform spacing.
Resistivity of the material used in each tubular conductor of each phasewire bundle of the circuit under consideration. Units are ohm-meters.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
E10.1
E10.1
E10.1
E10.1
p
(1-10)
Resistivity of the material used in each tubular conductor of each phasewire bundle of the circuit under consideration. Units are ohm-meters.
p
(11-20)
Relative permeability of the material used in each tubular conductor of

each phase-wire bundle of the circuit under consideration. This is a
dimensionless quantity.
G
(21-30)
Resistivity of the material which is used in each tubular conductor of

each ground-wire bundle of the circuit under consideration. Units are
ohm-meters.
10 - 61
Cable Constants
G
(31-40)
10.7.6
Relative permeability of the material which is used in each tabular

conductor of each ground-wire bundle of the circuit under consideration.
This is a dimensionless quantity.
Format for "A5" and "B7" Data
For each SC coaxial cable of the system, a maximum of three data cards are to be punched
according to the following format. Such cards are to be stacked in the circuit order which was
defined for Point "A4" data.
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
r1
r2
r3
r4
r5
r6
r7
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
r1
(1-10)
Inner radius of the tubular core, for the SC coaxial cable under
consideration. Units are meters.
r2
(11-20)
Outer radius of the tubular core (equal to the inner radius of the first
insulating layer), for the SC coaxial cable under consideration. Units are
meters.
r3
(21-30)
Inner radius of the sheath (equal to the outer radius of the first insulating
layer), for the SC coaxial cable under consideration. Units are meters.
r4
(31-40)
Outer radius of the sheath (equal to the inner radius of the second
meters.
r5
(41-50)
Inner radius of the armour (equal to the outer radius of the second
meters.
r6
(51-60)
Outer radius of the armour (equal to the inner radius of the third
meters.
r7
(61-70)
Outer radius of the third (outer-most) layer of insulation, for the SC

coaxial cable under consideration. Units are meters.
10 - 62
Cable Constants
r7
r6
r5
r4
r3
r2
r1
reference
a
reference
s
c c
Figure 10.6:
Cross-Section of an SC Coaxial Cable
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
i1
i1
i2
i2
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.0
c
(1-10)
Resistivity of the tubular core, for the SC coaxial cable under

consideration. Units are ohm-meters.
c
(11-20)
Relative permeability of the tubular core, for the SC coaxial cable under
consideration. This is a dimensionless quantity.
I1
(21-30)
Relative permeability of the first insulating layer. This is a

I1
(31-40)
Relative permittivity of the first insulating layer. This is a

s
(41-50)
Resistivity of the tubular sheath, for the SC coaxial cable under

10 - 63
Cable Constants
S
(51-60)
Relative permeability of the tubular sheath. This is a dimensionless

quantity.
I2
(61-70)
Relative permeability of the second insulating layer. This is a

I2
(71-80)
Relative permittivity of the second insulating layer. This is a

1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
i3
i3
E10.1
E10.1
E10.1
E10.1
a
(1-10)
Resistivity of the tubular armour, for the SC coaxial cable under

a
(11-20)
Relatively permeability of the tubular armour. This is a dimensionless

quantity.
I3
(21-30)
Relative permeability of the third insulating layer. This is a dimensionless

quantity.
I3
(31-40)
Relative permittivity of the third insulating layer. This is a dimensionless

quantity.
10.7.7
Format for "B5" Data
The location of each of the "NPC" SC coaxial cables within the surrounding conducting pipe is
specified by one (or possibly more) card of the following format:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
DIST1
THETA1
DIST2
THETA2
DIST3
THETA3
DIST4
THETA4
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
10 - 64
Cable Constants
DISTk
Distance between the centre of the pipe and the centre of the kth SC
coaxial cable, in units of meters.
THETAk
Angular position of the k-th SC coaxial cable, in units of degrees.
DIST
10.7.8
THETA
Format for "C5" Data
For each bundle of the overhead conductor system, a triplet of numbers giving the horizontal and
vertical location is to be supplied, according to the following format:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
VTOWER1
VMID1
HORIZ1
VTOWER2
VMID2
HORIZ2
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
VTOWERk
Height above the earth's surface of the centre of the k-th bundle; this is
height in meters, at the tower (the maximum height).
VMIDk
Height above the earth's surface of the centre of the k-th bundle; this is
height in meters, at mid-span (the minimum height).
HORIZk
The centre of the k-th bundle is this far to the right of an arbitrarily
chosen reference line.
10 - 65
Cable Constants
With regard to the ordering of the bundles which belong to the system under study (i.e., index "k"),
two rules must be observed:
Rule 1:
First come all phase-wire bundles of the system, in order of the circuit number to
which they belong (as established by Point C4 data), i.e., start with all phase-wire bundles of circuit number two, etc.
Rule 2:
Then come all (if there are any) ground-wire bundles of the system, in order of the circuit number to which they belong (as established by Point C4 data), i.e., start with all
ground-wire bundles of circuit number one, if any; then consider all of the ground-wire
bundles of circuit number two, etc.
Within any one circuit, ordering of the phase-wire bundles and the ground-wire bundles is
arbitrary. Rows of the resulting line constants matrices [R], [L] and [C] will be based on this
ordering, however, it might be noted.
10.7.9
Format for "A6" Data
For each of the "NPC" SC coaxial cables of the system, horizontal and vertical coordinates which
locate the centre must be specified, as follows:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 1234567890
VERT1
HORIZ1
VERT2
HORIZ2
VERT3
HORIZ3
VERT4
HORIZ4
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
VERTk
Vertical separation between the centre of the k-th SC coaxial cable and
the surface of the earth, in units of meters. This is always a positive
number, whether the system of SC coaxial cables is below the ground or
not.
HORIZk
The centre of the k-th SC coaxial cable is this far (in meters) to the right
of an arbitrary reference line.
10 - 66
Cable Constants
VERT3
VERT2
VERT1
reference
HORIZ3
HORIZ2
HORIZ1
Here the ordering of the SC cables (i.e., index "k") is as established in Section 10.7.3.
10.7.10 Format for "C6", "A7", and "B9" Data

Parameters can be calculated at discrete frequencies or over automatic loops over a certain
frequency range.
Discrete frequencies:
The "frequency card" of all three classes of data has the same format:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345678901234567890123456789012345678901234567890
RHO
FREQ
E10.1
E10.1
RHO
(1-15)
Resistivity of the top (i.e., surface) layer of the earth, in units of ohmmeters. If the earth is assumed to be homogeneous (parameter "IEARTH"
equal to zero; see Section 10.7.2), "RHO" is the resistivity of the entire
uniform earth.
10 - 67
Cable Constants
FREQ
(16-30)
Frequency in Hertz at which cable constants (for "A7" or "B9" data) or

line constants (for "C6' data) are to be calculated, should only one
frequency be desired.
A blank or zero field will be defaulted to the synchronous power system
frequency (generally 50 Hz or 60 Hz).
Frequency Looping:
For the automatic looping over logarithmically-spaced frequencies three cards are needed.
Card1:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345678901234567890123456789012345678901234567890
RHO
FREQ
E15.6
E15.6
RHO
(1-15)
uniform earth.
Frequency in Hz at which the transformation matrix is to be calculated.
FREQ
(16-30)
Card 2:
Enter the keyword ".FIT-S" in columns 1-6. This will cause the transfer of control to the fitting
routines of LINE CONSTANTS (see Section 9). Optionally, the user can also add the .NODES cards of
LINE CONSTANTS (see Section 9).
2
3
4
5
6
7
8
123456 67890123456789012345678901234567890123456789012345678901234567890
.FITS
A6
10 - 68
Cable Constants
Card 3:
1
2
3
4
5
6
7
8
123456789012345 678901234567890 12345 67890 12345678 90123456789012345678901234567890
RHO
FREQ
E15.6
E15.6
RHO
(1-15)
FREQ
IDEC IPNT
DIST
uniform earth.
Beginning frequency of the scan (frequency looping) Hz.
(16-30)
IDEC
(31-35)
Number of decades of frequency space which are to be spanned during

the automatic frequency looping.
(36-40)
Number of points per decade of frequency space at which [R], [L], and
[C] are to be calculated. There must be 10 or a multiple of 10 (up to 90)
points per decade.
DIST
Length of transmission circuit under consideration in units of meters.
IPNT
(41-48)
Note: The fitting of the frequency-dependent characteristic is presently available only for the
untransposed circuit configuration.
10.7.11 Format for "C7", "A8", and "B10" Data

If the 3-layer stratified (Nakagawa) earth modelling is being used, then the just-described
"frequency card" is to be immediately followed by two cards of the following format:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890123456789012345678901234567890
DEP12
DEP23
RHO2
RHO3
E10.1
E10.1
E10.1
E10.1
10 - 69
Cable Constants
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
E10.1
E10.1
E10.1
E10.1
E10.1
E10.1
DEP12
(1-10)
DEP23
(11-20)
RHO2
(21-30)
Distance below the surface of the earth at which layer number 1 ends
and layer number 2 begins. Units are in meters. See sketch below.
Distance below the surface of the earth at which layer number 2 ends
and layer number 3 begins. Units are meters. Recall the layer number 3
is infinitely deep.
Resistivities of layer number 2of the earth,. Units are ohm-meters.
Recall that "RHO" of the frequency card is used to specify the resistivity
of layer number 1.
(31-40)
Resistivities of layer number 3 of the earth,. Units are ohm-meters.

Recall that "RHO" of the frequency card is used to specify the resistivity
of layer number 1.
1
(1-10)
Relative permeability of layer number 1 of the earth. This is a

2
(11-20)
Relative permeability of layer number 2 of the earth. This is a

3
(21-30)
Relative permeability of layer number 3 of the earth,. This is a

1
(31-40)
Relative permittivity of layer number 1 of the earth. This is a

2
(41-50)
Relative permittivity of layer number 2 of the earth. This is a

3
(51-60)
Relative permittivity of layer number 3 of the earthy. This is a

RHO3
10 - 70
Cable Constants
surface
Layer 1
DEP12
DEP23
1, 1, RHO1
Layer 2
2, 2, RHO2
Layer 3
3, 3, RHO3
Figure 10.7:
Stratofoed Earth Model
10.7.12 Format for "B8" Data

The vertical distance between the centre of the pipe (which enclosed the SC coaxial cables) is to be
punched according to the following format:
1
2
3
4
5
6
7
8
1234567890 1234567890123456789012345678901234567890123456789012345678901234567890
CENTER
E10.1
CENTRE
is always positive, whether the pipe is below the ground or not. Units are meters.
center
10 - 71
Cable Constants
10.8
Degenerate Configurations and Special Cases
It is the purpose of the present section to describe how the user of "CABLE-PARAMETERS" can
handle special configurations which are less general than those described in Section 10.7. The
topics covered are as follows:
Pipe Without Tubular Insulator Around It
Infinitely-Thick Pipe
Solid Core for SC Coaxial Cable
Solid Overhead-Line Conductors
No Bundling of Conductors
SC Coaxial Cable Without Outer Insulator
SC Coaxial Cable with No Armour and Outer Insulator
SC Coaxial Cable With No Armour and No Outer Insulators
SC Coaxial Cable with No Sheath and No Outer Insulators
SC Coaxial Cable Having Core 0nly (all three insulators, armour and sheath missing)
Pipe Without Tubular Insulator Around It

Suppose that a pipe-type (Class B) configuration is involved, but without any insulating tube
around the outside of the conducting pipe. Then Point B4 data is to be treated as follows:
1.
Leave RP3 (columns 21-30) blank.
2.
Leave the data field for E2 (columns 61-70) blank
air or earth
Infinitely-Thick Pipe
While physically unrealizable, a pipe of infinite thickness is nonetheless useful in certain situations
as a modelling approximation. This is a special case of the Class B situation. By definition, there
is no earth for this case, and all zero-sequence current of the enclosed SC coaxial cables must return
through the pipe. Data requirements for this special case are as follows:
10 - 72
Cable Constants
1.
On the Point B3 miscellaneous data card, two parameters are to be punched unusually:
"ISYST" of columns 610 is to be punched zero;
"NPP" of columns 46-40 is to be punched zero (rather than the usual value of unity).
2.
On the Point B4 data card, three data fields can be left blank:
"RP2" of columns 11-20 and "RP3" of columns 21-30 --- since the outer radii of the pipe
and the surrounding insulator are not even finite.
"2" of columns 61-70 --- since the surrounding insulator does not even exist (or if it
does, it is infinitely far away!).
On the "frequency card" of Point B9 data, field "RHO" of columns 1630 can be left
blank, since the earth no longer exists.
pipe extends
to infinity
Solid Core for SC Coaxial Cable

For either Class A or Class B cable systems, the core of any SC coaxial cable can be made solid
rather than tubular, if so desired. The first of two Point A5 or Point B7 data cards has columns 110 used for punching the inner radius r1 of the tabular core. Simply set this parameter to zero, to
produce a solid core.
Solid Overhead-Line Conductors

For a conventional overhead transmission line of Class C, the conductors of either the phase-wire
bundles or the ground-wire bundles can be made solid, rather than tubular, if so desired. Recall
10 - 73
Cable Constants
that the second of the Point C4 data cards is punched with an inner radius RINp for phase-wire
conductors, and RING for ground-wire conductors.
1.
Set RINp of columns 11-20 equal to zero, in order to obtain solid conductors for the phase
wires.
2.
Set RING of columns 31-40 equal to zero, in order to obtain solid conductors for the ground
wires.
No Bundling of Conductors
For a conventional overhead transmission line of Class C, it is Point C4 data which specifies
whether phase-wire or ground-wire conductors are to be bundled.
1.
If a phase-wire bundle only consists of a single tubular conductor (i.e., if there is no

bundling of phase conductors), the "KBP" of columns 11-15 of the first Point C4 data card
will be punched equal to unity, by definition. Field SEPp of columns 41-50 of the second
Point C4 data card can then be left blank -- since interconductor separation within a phasewire bundle does not exist.
2.
If a ground-wire bundle only consists of a single tubular conductor (i.e., if there is no

bundling of ground conductors), then "KBG" of columns 16-20 of the first Point C4 data
card will be punched equal to unity, be definition. Field SEPG of columns 51-60 of the
second Point C4 data card can then be left blank -- since interconductor separation within a
ground-wire bundle does not exist.
SC Coaxial Cable Without Outer Insulator

For a Class A or Class B data case, the outer (or third) layer of insulation of any SC coaxial cable
can be omitted. In this case, the three Point A5 or Point B7 data cards are handled as follows:
1.
Leave r7 (columns 61-70 of the first data card) blank.
2.
Leave data fields mI3 and eI3 (columns 31-40 of the third card) blank -- since such
parameters do not exist.
10 - 74
Cable Constants
r7
r6
r5
r4
r3
r2
r1
SC Coaxial Cable with No Armour and Outer Insulator

For a Class A or Class B data case, both the armour and the outer (or third) layer of insulation of
any SC coaxial cable can be omitted. In this case, the three Point A5 or Point B7 data cards are
handled as follows:
1.
Leave r7 and r6 (columns 51-70 of the first card) blank.
2.
Omit the third data card because of the nonexistence of ra, mma, mI3, and eI3.
r7
r6
r5
r4
r3
r2
r1
SC Coaxial Cable With No Armour and No Outer Insulators

For a Class A or Class B data case, the outer (second and third) layers of insulation and armour of
any SC coaxial cable can be omitted. In this case, the three Point A5 or Point B7 data cards are
handled as follows:
10 - 75
Cable Constants
1.
Leave r7, r6 and r5 (columns 41-70 of the first card) blank.
2.
Leave data fields mI2 and eI2 (columns 61-80 of the second card) blank and omit the third
card -- since such parameters do not exist.
r7
r6
r5
r4
r3
r2
r1
SC Coaxial Cable with No Sheath and No Outer Insulators

For a Class A or Class B data case, both the sheath, armour and the outer (second and third) layers
of insulation of any SC coaxial cable can be omitted. In this case, the three Point A5 or Point B7
data cards are handled as follows:
1.
Leave r7, r6, r5 and r4 (columns 31-70 of the first card) blank.
2.
Leave blank all other data fields which are used for parameters of the nonexistent sheath
and outer insulator. In particular, there are four: rs, ms, mI2, and eI2 (columns 41-80 of the
second card).
3.
Omit the third card -- since parameters for armour and its outer insulator do not exist.
r7
r6
r5
r4
r3
r2
r1
10 - 76
Cable Constants
SC Coaxial Cable Having Core 0nly (all three insulators, armour and
sheath missing)
For a Class A or Class B data case, the armour, sheath and all three layers of insulation of any SC
coaxial cable can be omitted, leaving just the tubular conductor core. In this case, the three
Point A5 or Point B7 data cards are handled as follows:
1.
Leave r7, r6, r5, r4 and r3 (columns 21-70 of the first card) blank.
2.
Leave blank all other data fields which are used for parameters of the nonexistent sheath
and layers of insulation. In particular, there are six: mI1, eI1, rs, ms, mI2, and eI2 (columns
21-80 of the second card).
3.
Omit the third card -- since the parameters for armour and its outer insulator do not exist.
r7
r6
r5
r4
r3
r2
10.9
r1
Approximation Used for the Bundling of Overhead

Conductors
It is important for the user to be aware that "CABLE PARAMETERS" treats bundled conductors of
conventional overhead transmission lines quite differently than does "LINE CONSTANTS" of
Section 9. The reader may already have realized this, since the Point C4 data is not sufficient to
uniquely specify the geometry of a bundle. No angular position of any one conductor of the
bundle is specified, it will be noted.
The "LINE CONSTANTS" code of Section 9 calculates line constants for the system of physical
conductors first. This is then reduced, as conductors are paralleled (the bundling operation). On
the other hand, "CABLE PARAMETERS" does the bundling at data-input time. The geometric mean
radius of the bundle is immediately calculated, and then a single approximately equivalent
10 - 77
Cable Constants
conductor is used to represent each bundle for the calculation of line constants. There never is any
set of line constants for the system of physical conductors, then, when using "CABLE
PARAMETERS".
Needless to say, this bundling of conductors at data-input time simplifies the calculation
considerably, and speeds it up. But an approximation is involved.
10.10 Crossbonded Cables

10.10.1 Introduction
It is common practice to crossbond a three-phase PV or CV cable, i.e., to wrap around single-core
coaxial cable with a polyethelene or oil-immersed paper insulation, so that the circulating current
within the sheaths is reduced. A schematic diagram of a crossbonded cables is given in Figure
10.8, below. This cable consists of 3 or more cascaded major sections. One major section consists
of three minor sections as illustrated in Figure 10.8.
The sheaths are crossbonded at the ends of the first and second minor sections. As a common
practice, the length of one minor section is between 300m and 500m, thus the length of one major
section is about 1 km to 1.5 km. The sheaths of the three phases are short-circuited and grounded
at the junction of each major section. Because of the existence of a resistance at the grounding
point due to a poor conductivity of soil, it appears as if the sheaths are grounded through the
resistance of Rs as shown in Figure 10.8.
10 - 78
Cable Constants
sending end
receiving end
a
1st major
section
2nd major
section
Rs
node j
Rs
nth major
section
Rs
Rs
node k
core
sheath
core
sheath
core
sheath
minor section
major section
Figure 10.8:
A Schematic Diagram of a Crossbonded Cable
At the sending and receiving ends of a cable, the resistances are lower than those at each major
section, because of higher degree of grounding. (The details of grounding should be found in the
standard of electrical apparatus or machinery, possibly in the ANSI.) Thus, the resistances at the
sending and receiving ends are given by R's, which is different from Rs. In practice, Rs is 1 ohm
to some tens ohm, and R's 0.1 ohm to 10 ohm depending on the method of grounding.
10.10.2 Modelling of a Crossbonded Cable

A crossbonded cable can be modelled as a uniform distributed parameter line. An equivalent
circuit for one major section of the crossbonded cable is shown in Figure 10.9, below. Z" and Y"
are the equivalent series impedance and shunt admittance respectively of one major section. Note
that this equivalent circuit is a four-conductors system but not six-conductors system. This is due
to the fact that the effect of the short-circuit of the three-phase sheath has been taken into account
in this equivalent circuit, and thus three sheaths are reduced to one sheath. The sheath voltage in
the equivalent circuit of Figure 10.9 is the same as the voltages of three sheaths in the original
10 - 79
Cable Constants
circuit of Figure 10.9, and the sheath current in the equivalent circuits is the sum of the three-phase
sheath currents in the original circuit.
core
1
2
3
[Z"], [Y"]
sheath
G
Figure 10.9:
An Equivalent Uniform Distributed-Parameter Line for One Major

Section
Also, it should be noted that the grounding conductance G is not taken into account in this
equivalent circuit. Thus, the conductance G should be included as a boundary condition between
two major sections.
The above equivalent circuit can be adopted to the EMTP, because it is a uniform distributedparameter line.
A pi-circuit equivalent is quite often used to represent a distributed-parameter line, although it
cannot take into account the frequency-dependent effect of the distributed-parameter line.
1.
Accurate Modelling: Each minor section is represented by a six-phase pi-circuit, and is

connected as shown in Figure 9.8. Then, each major section is connected as shown in
Figure 9.8 including a grounding resistance.
2.
Uniform Line Modelling: Each major section is represented by a four-phase pi-circuit of

which R, L and C are given by Z" and Y" of Figure 9.9, i.e., Z" = R+ jwL and y"=jwC.
Then, each major section is connected as shown in Figure 9.8 including a grounding
resistance.
10.10.3 Data Format for Using the Crossbonded Cable Model

The data structure for the crossbonded cable model is the same as that for the usual (not
crossbonded) cable (see Section 10.6.2), except the following two additional data cards are needed:
1.
A card with "PUNCH" inputted in columns 1-5 should be put right after the "CABLE
PARAMETERS" card (see "A2" and "B2" of Section 10.7.1).
10 - 80
Cable Constants
2.
Then, the following card should be inputted after the "miscellaneous data card" which was
described as "A3" and "B3" data is Section 10.7.2.
I5
IRSEP
XMAJOR
RSG
CNAME
NPAIS
NCROS
1
2
3
12345 67890 12345 6789012345 6789012345 6
I5
E10.1
E10.1
A1
I5
4
5
6
7
8
78901234567890123456789012345678901234567890
(1-5)
Number of pi-sections a user should define except "NPAIS=0". The

detailed explanation for 'NPAIS' will be given later.
NCROS
=0
the usual (not crossbonded) cable.
crossbonded cable.
NPAIS
(6-10)
For the overhead like case, leave BLANK.

IRSEP
=0
(or blank) all the cable sheaths are short-circuited and grounded
through resistance RSG between pi-sections.
each sheath is grounded separately from each other through

resistance RSG between pi-sections.
(11-15)
does not apply to the overhead line case and to the cable case
when "NPAIS > .0". In these cases, leave IRSEP blank.
IRSEP
XMAJOR
(16-25)
Length of one major section for "NPAIS.LT.0" in the cable case. It is the
length of one pi-section for "NPAIS.GT.0" in the cable case, and in the
overhead line case.
XMAJOR should equal to the total length of
NPAIS, or total length = XMAJOR * NPAIS.
RSG
(26-35)
CNAME
(36)
a cable or line divided by
Sheath grounding resistance at the end of a major section for a cable.

RSG does not apply to the overhead line case, thus leave blank.
Node name of pi-circuits modelling a user should define in the case of
NPAIS.NE.0. For NPAIS = 0 leave blank
More detailed explanation for NPAIS and the related variables is given here:
10 - 81
Cable Constants
Case A: NPAIS 0: Data cards of a line or cable for pi-circuit modelling will be punched
out.
(A-1)
For the cable case (ITYPE 1):

(1-1)
NPAIS.GT.0
and NCROS = 0 or BLANK: Uniform pi-circuit modelling of

usual non-crossbonded cable.
cores
3
R
1st pi
2nd pi
nth pi
sheaths
Figure 10.10: Uniform Pi-Circuit Modelling for a Non-Crossbonded Cable
a)
A user will get a cascade pi-circuit modelling shown in Figure 9.10

for a given distributed-parameter line. There is no grounding
resistance and no connection other than the cascade connection
between two pisections. Thus, IRSEP and RSG should be BLANK.
b)
The number of pi-sections is given by NPAIS (n = NPAIS in Figure

10.10, above). NPAIS is arbitrary and thus a user should define it. But
the following relation should always be satisfied:
total length of a cable = NPAIS * XMAJOR
c)
The user should define the node name of the pi-circuit modelling by
CNAME. The data for CNAME is one alphabetic letter and is read by
format "A1". Then, all the node names in this pi-circuit modelling
are internally determined in the following form.
At the sending-end:
column
1 2 3 4 5 6
I N
phase number
CNAME
At an intermediate node:
10 - 82
Cable Constants
column
1 2 3 4 5 6
I N
CNAME
phase number
major or pi
section number
(minor section number only in
the crossbonded cable case,
otherwise "1")
At the receiving end:

column
1 2 3 4 5 6
O U T
phase number
CNAME
(1-2)
NPAIS.GT.0 and NCROS = 0:
Uniform pi-circuit modelling of a crossbonded
cable.
cores
3
R
1st pi
2nd pi
nth pi
sheaths
RS
RE
RSG
RSG
RSG
RSG
Figure 10.11: Uniform Pi-Circuit Modelling for a Crossbonded Cable
a)
A user will get a pi-circuit modelling shown in Figure 9.11 for a

given crossbonded cable. Each pisection corresponds one major
section of the crossbonded cable, and R, L and C are calculated in the
method explained in Section 10.7. There is a grounding resistance
RSG, and thus the user should define the value of RSG in his second
Miscellaneous Data Card. But no need to define IRSEP because there
exists only one sheath as explained in Section 10.7.
10 - 83
Cable Constants
b)
For each pi-section corresponds to one major section, the number of

pi-sections not arbitrary, and should be identical to the number of the
major sections in the given crossbonded cable, i.e.,
NPAIS
c)
= total length of the cable / XMAJOR
The data for CNAME is one alphabetic letter and is read by
CNAME.
d)
A user should add the parallel resistances RS and RE to the RSG at the
sending- and receiving-ends as shown in Figure 10.11, above after he
gets the punched out data cards for his pi-circuit modelling, because
the grounding resistances at both ends are, in general, different from
RSG. In other words, the user can get the correct grounding
resistances by adding RS and RE, i.e.,
correct resistance at the sending-end = (1/RSG + 1/RS)-1
correct resistance at the receiving-end = (1/RSG + 1/RE)-1
(2-1)
NPAIS.LT.0
and NCROS = 0 or BLANK: Discrete pi-circuit modelling of a
usual cable
a)
A user will get a pi-circuit modelling shown in Figure 10.12 below

(which is the same as Figure 10.11 in fact) for a given cable. Each picircuit corresponds to one major section of the cable. There is a
grounding resistance RSG, and thus, the user should define the
value of RSG. Also, IRSEP should be defined, although, in most
practical cases, the sheaths are short-circuited and grounded, i.e.,
"IRSEP = 0".
b)
For each pi-section corresponds to one major section, the number of

pi-sections not arbitrary, and should be identical to the number of the
major sections in the given crossbonded cable, i.e.,
NPAIS
c)
= total length of the cable / XMAJOR
10 - 84
Cable Constants
d)
sending- and receiving-ends as shown in Figure 10.11 after he gets
the punched out data cards for his pi-circuit modelling, because the
grounding resistances at both ends are, in general, different from RSG.
In other words, the user can get the correct grounding resistances by
adding RS and RE, i.e.,
cores
S
3
R
1st pi
2nd pi
nth pi
sheaths
Figure 10.12: Discrete Pi-Circuit Modelling for a Usual Cable
(2-2)
NPAIS.LT.0 and NCROS 0:
Discrete pi-circuit modelling of a crossbonded
cable
a)
In this case, a user will get a pi-circuit modelling shown in

Figure 10.13, below. One major section consists of three pi-circuits,
of which each pi-circuit corresponds to one minor section. Within
one major section, crossbonding of three-phase sheaths are carried
out. Since there is a grounding resistance, the user should define its
value. Also, IRSEP should be defined.
10 - 85
Cable Constants
Overall Circuit Diagram

cores
S
sheaths
1st major
2nd major
n-th major
section
section
section
core
sheath
1st pi
2nd pi
3rd pi
minor section
major section
Figure 10.13: Discrete Pi-Circuit Modelling for a Crossbonded Cable
b)
The number of major sections is given by NPAIS and is not arbitrary.

The user should give the actual number of the major sections of his
crossbonded cable as NPAIS in his data card, but the following
condition should be kept.
total length of the cable = NPAIS * XMAJOR
c)
10 - 86
Cable Constants
d)
sending- and receiving-ends as shown in Figure 10.11 after he gets
the punched out data cards for his pi-circuit modelling, because the
grounding resistances at both ends are, in general, different from RSG.
In other words, the user can get the correct grounding resistances by
adding RS and RE, i.e.,
(A-2)
For the overhead line case (ITYPEC = 1):

In this case, NPAIS should be greater than '0' ("NPAIS.GT.0."). The model circuit
configuration and the data input are the same as those explained in the case of (1-1):
NCROS = 0 of (A-1) for the cable case.
Case B: NPAIS = 0 or Blank:

No data card will be punched out.
(B-1)
NCROS
= 0 or BLANK
In this case, a user will get exactly the same version of the CABLE
PARAMETERS as that in the 1980 version. Thus, leave BLANK all the
data in the second Miscellaneous Data Card.
(B-2)
NCROS
0: Crossbonded Cable
This is only for the cable case (ITYPEC = 1). In this case, XMAJOR
should be defined. A user will get printouts of various cable
parameters for his crossbonded cable the same as those in the case of
NCROS = 0.
Summarizing all the above explanation for the second Miscellaneous Data Card, the following
table is obtained
NPAIS
<0
need all the data
NPAIS
> 0 & NCROS 0
need all the data
NPAIS
> 0 & NCROS = 0
need XMAJOR and CNAME
NPAIS
= 0 & NCROS = 0
need only XMAJOR
NPAIS
= 0 & NCROS 0
no data, just one BLANK card
ITYPEC
= 1: NPAIS < 0 cannot be used.

10 - 87
Cable Constants
10.11 CABLE-PARAMETERS Example

Following is an example of pi-circuit modelling of a crossbonded cable with one major section.
The cable configuration is illustrated in Figure 9.14. The cable consists of core, sheath and
armour. The armours are solidly grounded and the sheaths are crossbonded. To eliminate the
grounded armours, NGRND (No. of solidly grounded conductors) is taken as 2. Since the cable is
crossbonded, NCROS = 1 (or not equal to zero). The sheaths being short-circuited and grounded to
both ends of the major section, IRSEP = 0 and RSG = 0.1 ohm.
In the same manner, one can handle cable of which the both armours (or pipe) and sheaths are
grounded, using the Discrete Pi-Circuit Modelling, i.e., use NGRND for grounding the armours or
pipe and ground the sheaths by RSG and IRSEP.
armour
sheath
core
Figure 10.14:
A Three-Phase Crossbonded Cable
First comes the listing of the data case:

C tcase9.dat
C
C -----------------------------------------------------------------------------C
Cable-Parameters data case
C
TEPCO OF 275 KV CABLE ****
C
CABLE CONSTANTS
C
C
C --key--------|
|<ipch<irun
CABLE-PARAMETERS
PUNCH
10 - 88
Cable Constants
C
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
C ITY ISYS NPC IEAR KMOD IZFL IYFL NPP NGRN
2
-1
3
0
1
1
1
0
2
C
C Data for cross-bonded model
C
N
N
I
X
RC
C
P
C
R
M
SN
C
A
R
S
A
GA
C
I
O
E
J
M
C
S
S
P
R
E
-1
1
0
1.E3
1.E-1A
C
C NPC NPCC NPCC NPCC NPCC NPCC ...
3
3
3
C
C CONDUCTOR
C
R1
R2
R3
R4
R5
R6
R7
C
RC
UC
UI1
EI1
RS
US
UI2
EI2
0.013200 0.024900 0.054200 0.057000 0.063000
.0660
0.720
.18900E-07
1.000
1.000
2.300
.3E-7
1.000
1.000
3.500
.3E-7
1.000
1.000
3.500
C
C CONDUCTOR
C
R1
R2
R3
R4
R5
R6
R7
C
RC
UC
UI1
EI1
RS
US
UI2
EI2
0.013200 0.024900 0.054200 0.057000 0.063000
.0660
0.720
.18900E-07
1.000
1.000
2.300
.3E-7
1.000
1.000
3.500
.3E-7
1.000
1.000
3.500
C
C CONDUCTOR
C
R1
R2
R3
R4
R5
R6
R7
C
RC
UC
UI1
EI1
RS
US
UI2
EI2
0.013200 0.024900 0.054200 0.057000 0.063000
.0660
0.720
.18900E-07
1.000
1.000
2.300
.3E-7
1.000
1.000
3.500
.3E-7
1.000
1.000
3.500
C
C
C
VERT1
HORIZ1
VERT2
HORIZ2
VERT3
HORIZ3
. . .
2.000
0.000
1.8095
0.110
2.000
0.220
C
C
REARTH
FREQ
C
REARTH
FREQ IDEC IPNT
DIST2
100.
1000.
BEGIN NEW DATA CASE
BLANK
The pi-circuit branch cards resulted on the punched file LUNIT are:
$VINTAGE, 1
AIN 4
1AIN 1A 11 1
2AIN 2A 11 2
3AIN
3A 11 3
0.10000E+00
0.31876E-01
0.32590E-04
0.31876E-01
0.32595E-04
10 - 89
0.65708E-01
0.51255E-05
0.65708E-01
0.51274E-05
0.54835E-01
0.00000E+00
0.54835E-01
0.00000E+00
Cable Constants
4AIN
4A 11 4
5AIN
4A 11 5
6AIN
4A 11 6
1A 11 1A 12
2A 11 2A 12
3A 11 3A 12
4A 11 6A 12
5A 11 4A 12
6A 11 5A 12
1A 12 1AOUT
2A 12 2AOUT
3A 12 3AOUT
4A 12 6AOUT
5A 12 4AOUT
6A 12 5AOUT
AOUT 4
$VINTAGE, 0
1AIN
2
3
4
5
6
1AIN
2
3
4
4
4
AIN
0.32590E-04
0.31876E-01
0.10876E-01
0.32590E-04
0.32595E-04
0.20291E-01
0.32590E-04
0.10876E-01
0.32590E-04
0.32590E-04
0.20291E-01
0.32595E-04
0.32590E-04
0.10876E-01
0.32595E-04
0.32590E-04
0.20291E-01
1A 11 1
1A 11 1
10 - 90
0.51255E-05
0.65708E-01
0.92798E-02
0.51255E-05
0.51274E-05
0.87483E-02
0.51255E-05
0.92798E-02
0.51255E-05
0.51255E-05
0.87483E-02
0.51274E-05
0.51255E-05
0.92798E-02
0.51274E-05
0.51255E-05
0.87483E-02
0.00000E+00
0.54835E-01
-0.54835E-01
0.00000E+00
0.00000E+00
0.70334E+00
0.00000E+00
-0.54835E-01
0.00000E+00
0.00000E+00
0.70334E+00
0.00000E+00
0.00000E+00
-0.54835E-01
0.00000E+00
0.00000E+00
0.70334E+00
Section 11
FDNE
11.1
Introduction ......................................................................................... 11-1
11.2
Dimensioning Limits ........................................................................... 11-2
11.3
Structure of an FDNE Data Case ........................................................ 11-3
11.4
Example ............................................................................................... 11-15
11.5
Error Messages .................................................................................... 11-23
11.6
Admittance Curve Option ................................................................... 11-25
11.7
Program Description ........................................................................... 11-27
Section 11
FDNE
The Frequency Dependent Network Equivalent program (FDNE) is a stand-alone program that
reduces a network of arbitrary size (within dimensioning limits) to a compact equivalent network.
The response of the reduced network (i.e., poles and zeros) are matched or fitted to the response of
the original network.
11.1
Introduction
The accurate calculation of electromagnetic transients may require a detailed representation of

transmission lines several buses away from the point where the system disturbance takes place. In
many situations the number of transmission lines that should be represented is prohibitively large
and a suitable reduction must be made.
The FDNE program was developed to model an extensive network of transmission lines, linear
branches, and shunts by a limited number of linear branches with lumped resistances, inductances
and capacitances. The admittance matrix of the reduced linear model approximates the
admittance matrix of the original network over a range of frequencies.
The FDNE is a stand-alone pre-processing program, and it generates the reduced equivalent
network as a punch file for inclusion into EMTP data cases (see Section 5.5, RuleBook 1).
Note that FDNE is different from the network reduction option described in Section 22,
RuleBook 1:
1.
FDNE
is a fiited model, and the resulting equivalent network is much smaller

than the original one. In the network reduction option, no fitting is involved.
However, the "reduction" is only from the point of view of output management.
Inside the EMTP, the entire network is retained and solved using Kron reduction
techniques.
2.
The network reduction option is restricted to coupled RLC branches. A network

to be approximated with the FDNE, can have lumped linear elements as well as
overhead transmission lines. Transmission lines are located in right-of-ways
which may carry one or several three-phase circuits. Normally right-of-ways
are divided into longitudinal sections; each longitudinal section has its own
unique geometrical arrangement of circuit conductors, groundwires, etc. This
program treats each longitudinal section of a right-of-way as a separate entity.
11 - 1
FDNE
11.2
Dimensioning Limits
The present dimensioning limits of the program are shown below:
Variable
11.2.1
Maximum number
Reference ports
Right-of-Ways
60
Circuits per R/W
20
Conductors per R/W
100
Circuits, total
250
Linear Branches
100
Shunt Branches
50
Network Buses
250
Frequency points
1000
Summary of Files Created by the Program
FILEIDENT.LPS
Transmission Line Coefficients (binary). Used during case re-runs.
FILEIDENT.LZR
Transmission Line Coefficients (binary). Used during case re-runs.
FILEIDENT.ADZ
Zero sequence Admittance Values (binary). Used during case reruns.
FILEIDENT.ADP
Positive sequence Admittance Values (binary). Used during case

re-runs.
FILEIDENT.PLZ
Zero sequence plot file, same format as the EMTP pl4 plot file
(binary).
FILEIDENT.PLP
Positive sequence plot file, same format as the EMTP pl4 plot file
(binary).
FILEIDENT.PCH
Punch file for inclusion into EMTP (ASCII).
FILEIDENT.OUT
Output listing (ASCII).
11 - 2
FDNE
11.3
Structure of an FDNE Data Case
The input data cards for the FDNE program are to be structured as follows:
1.
First come three cards with case description (study identification) and
specification of output files. Section 7.3.1
2.
Next comes an optional card defining the steady-state frequency.

Section 7.3.2: Cards with Study Parameters
11.3.1
3.
Next comes a single card specifying parameters for the maxima and minima
selection.
4.
Next comes a single Program Options Card.
5.
Next comes a single card specifying the reference buses to be used in the study.
6.
Next comes a single card specifying the frequency range over which the
equivalent is to approximate the network admittance.
7.
Next come the cards to specify transmission line geometry. This grouping of
data cards is to be terminated by a blank card.
8.
Next come the cards to specify linear series branches. This grouping of data
cards is to be terminated by a blank card.
9.
Next come the cards to specify linear shunt branches. This grouping of data
cards is to be terminated by a blank card.
Case Description Cards
There are three cards in this grouping. The first two cards are used to provide study identification.
The text on these cards will appear as a header in the output file as well as in the plots.
Cards 1 and 2:
1
2
3
4
5
6
7
8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
Label
11 - 3
FDNE
Label
(1-80)
Study identification.
Card 3:
123456789
1
2
3
4
5
6
7
8
01234567890123456789012345678901234567890123456789012345678901234567890
FILEIDENT
A9
FILEIDENT
(1-9)
Name of the study files produced during the computations; up to 9

characters.
Optional card: (card 3a)
1
2
3
4
5
6
7
8
123456789012345 67890123456789012 34567890 1234567890123456789012345678901234567890
POWER FREQUENCY
STATFR
A15
F8.0
Enter the keyword "POWER FREQUENCY" in columns 1-15
STATFR
Steady-state frequency in Hz. The built-in default is 60Hz.
(33-40)
11 - 4
FDNE
11.3.2
Cards With Study Parameters
Card 4:
1
2
3
4
5
6
7
8
1234567890 1234567890 1234567890 1234567890 1234567890 1234567890 12345678901234567890
XOPT
COPT
REJECTZ
REJECTP
XOPTO
COPTO
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
XOPT
(1-10)
COPT
(11-20)
REJECTZ
(21-30)
REJECTP
(31-40)
XOPTO
(41-50)
COPTO
(51-60)
Parameter indicating whether inductance or inductive reactance is to be

entered.
XOPT
=0
All inductance values are interpreted as L in mH.
XOPT
=f
All inductance values are interpreted as L=2fL in ohms

at frequency f.
Parameter indicating whether capacitance or capacitive susceptance is to

be entered.
COPT
=0
All capacitance values are interpreted as C in F.
COPT
=f
All capacitance values are interpreted as C=2fC in S

at frequency f.
Level of rejection of maxima of the zero sequence admittance. Default

value is 0.1.
Level of rejection of maxima of the positive sequence admittance.
Default value is 0.2.
Parameter indicating whether inductance or inductive reactance is to be
produced on output.
XOPTO
=0
All inductance values are interpreted as L in mH.
XOPTO
=f
All inductance values are interpreted as L=2fL in ohms

at frequency f.
Parameter indicating whether capacitance or capacitive susceptance is to

produced on output.
COPTO
=0
All capacitance values are interpreted as C in F.
COPTO
=f
All capacitance values are interpreted as C=2fC in S at

frequency f.
11 - 5
FDNE
11.3.3
Program Options Card
Card 5:
1
12345 67890 12345
2
3
4
67890 12345 67890 12345 67890
5
6
7
8
12345 67890 123456789012345678901234567890
PLOT SCALE RERUN REFIT PRNT1 PRNT2 PRNT3 CHECK

I5
I5
PLOT
I5
I5
I5
I5
I5
I5
KILL
PL4
I5
I5
Plot request card
(1-5)
PLOT
=0
No plot
PLOT
=1
Plot as a function of frequency.

a) calculated and optimized model conductance C.
b) calculated and optimized model susceptance B.
PLOT
SCALE
(6-10)
RERUN
=2
Also plot the nonoptimized G and B curves.
Flag indicating the type of spacing of the data points of the admittance
table.
SCALE
=0
logarithmic division.
SCALE
=1
linear division.
Flag indicating the status of transmission line data.
(11-15)
REFIT
RERUN
=0
New run.
RERUN
=1
Use previously-calculated transmission line data.
Flag indicating the status of network admittance data.
(16-20)
PRNT1
REFIT
=0
New run.
REFIT
=1
Refit the existing network admittance data.
Output print control for the main program.
(21-25)
PRNT1
=0
PRNT1 =
minimum printout, normal option.
N additional information, marked "level N" will be printed out

(where N=1,2,...,5).
11 - 6
FDNE
Note: The additional information is primarily of interest during the

initial stage of a study. The use of print control larger than 1 is not
recommended.
PRNT2
Same as PRNT1 but for subroutine NETWRK (see Section 11.7)
(26-30)
PRNT3
(31-35)
Same as PRNT1 but for subroutine MINIMNS, FITTER and OCTCHK (see
Section 11.7)
CHECK
Flag controlling the calculation of the open-circuit admittance.
(36-40)
KILL
CHECK
=0
No open-circuit admittance calculations.
CHECK
=1
Calculate and plot the open-circuit admittance, and write to

plot files.
CHECK
=2
Calculate and plot the open-circuit admittance, and write to

plot files, and produce Postscript plot.
Short run flag.
(41-45)
PL4
(46-50)
KILL
=0
Normal run including the model derivation.
KILL
=1
Program execution is stopped after the calculation of the

network admittances.
PL4
file request flag. These pl4 files are listed in Section 11.2.1 and are
compatible with any EMTP output processing program
PL4
= 1
PL$4
=1
No pl4 files are requested

Generate pl4 files.
Note: When REFIT option is active (REFIT=1 in columns 16-20), no more data cards are required.
11 - 7
FDNE
11.3.4
Card With Frequency Range Parameters
Card 6
1
1234567890 12345
TOPFR
2
3
4
5
6
7
8
67890 12345 6789012345678901234567890123456789012345678901234567890
NNPNT JQFR JHIFR
F10.0
I5
I5
I5
Highest frequency of the frequency range in Hz.
TOPFR
(1-10)
NNPNT
(11-15)
JQFR
(16-20)
JHIFR
(21-25)
Number of frequency points to be calculated. Maximum allowed is

1000. If set to zero or left blank, only the network admittance at the
reference frequency will be calculated.
Frequency in kHz at which the transformation matrix for right-ofway is calculated in kHz (entered as an integer). The built-in default
value the frequency point closest to 10 kHz. The maximum allowed
value is 100 kHz.
Highest frequency for admittance calculation in kHz (entered as an
integer). The built-in default value is 500 kHz. The maximum
allowed value is 501 kHz.
Note:
In general, JQFR and JHIFR should not be specified by the user.
The frequency range begins at the steady-state frequency and ends at TOPFR
11 - 8
FDNE
11.3.5
Card With Reference Bus Data
Card 7:
NR
(1-2)
REFNAM
(3-8)
(15-20)
(27-32)
(39-44)
(51-56)
REFKV
(9-14)
A6
F6.0
A6
F6.0
REFKV
REFNAM
F6.0
REFKV
REFNAM
A6
REFKV
REFNAM
I2
REFKV
REFNAM
NR
REFNAM
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 789012 345678 901234 567890 123456 789012 345678901234567890
REFKV
A6
F6.0
A6
F6.0
Number of reference ports. The built-in default value is 1. The

maximum allowed number is 5.
Name of a Reference Port. The Reference Port must be one of the
system buses.
Note: Bus names are up to six characters. The program superimposes

characters A, B or C on the sixth character of the reference port name
when it prepares the output punch file. Reference port names therefore
should be unique in the first five characters. Better still would it be if the
reference port names are unique in the first four characters, as writing to
the PL4 files superimposes identifying characters on the fifth and the
sixth character of the bus name.
Reference Port base voltage in kV. The built-in default value is 220 kV.
All reference ports must belong to the same voltage level.
(21-26)
(33-38)
(45-50)
(57-62)
11 - 9
FDNE
11.3.6
Right-of-Way Cards
The right-of-way (R/W) cards describe the transmission lines. Each R/W starts with a header card.
R/W Header card:
1
2
3
4
5
6
7
8
12 345678901234 567890 123456 789012 345678 901234567890123456789012345678901234567890
R/W NAME
LENGTH
A12
F6.0
R/W NAME
R/W
NCCT DATA LPRNT

I6
I6
I6
identification name.
(3-14)
(15-20)
R/W
Length, km or miles, depending on parameter DATA

(columns 27-32).
NCCT
Number of circuits. Maximum permitted is 20.
LENGTH
(21-26)
Flag to distinguish between metric and British units.
DATA
(27-32)
DATA
= 0 Data in metric units.
DATA
= 1 Data in British units.
Local print control; print control in effect during the processing of the R/
W data. LPRINT = 0, 1, 2, etc. (see Section 11.3.3, Program Options
Card).
LPRNT
(33-38)
The R/W header card must be followed by one card for each circuit.
Circuit card:
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456789012345678901234567890123456789012345678901234567890
FROM
TO
BASEKV
A6
A6
F6.0
11 - 10
FDNE
"From" bus name.
FROM
(3-8)
"To" bus name.
TO
(9-14)
BASEKV
Base voltage in kV.
(15-20)
Line input data in the AUX "Line Constants" format (see Section 9.4.4) is to follow the circuit
cards, including a blank terminator card to terminate conductor data and a ground resistivity card.
The total number of conductors on a right-of-way is limited to 100.
Bus names will be used to identify network transmission lines. Transmission lines terminated at
both ends by network buses are called internal lines. Transmission lines connected at only one
end to network buses are called external lines. They will be represented in the network admittance
matrix by their surge impedances. A transmission line will be considered an external line when
the To-bus name is left blank. Mixing of internal and external circuits on one right-of-way is not
permitted.
The right-of-way data is to be terminated by a blank card.
11.3.7
Cards for Linear Series Branches
A linear branch consists of a series of connected resistance, reactance, and capacitance. At least
one of the R, L, C values of the branch must be nonzero. Two cards per linear branch are needed.
First card:
1
2
3
4
5
6
7
8
12 345678 901234 567890 123456 7890123456789012 3456789012345678 9012345678901234 567890
IT FROM
A6
IT
TO
RFROM
RTO
R0
X0
B0
BASEKV
A6
A6
A6
F16.0
F16.0
F16.0
F6.0
Sequence flag. Set IT = 51 for first card of zero sequence data.
(1-2)
11 - 11
FDNE
FROM
"From" bus name.
(3-8)
TO
"To" bus name.
(9-14)
REFOM
"From" bus name of reference branch.
(15-20)
RTO
"To" bus name of reference branch.
(21-26)
If a branch has identical data (both zero and positive sequence) as a previously specified branch,
specify the RFROM and RTO bus names and leave the remaining fields in columns 27-68 blank of
the first card blank. Set IT = 52 (in columns 1-2) of the second card and leave the remaining fields
blank. Otherwise, leave the fields RFROM and RTO blank and specify the remaining data on cards
1-2.
R0
Zero sequence resistance, in units depending on BASEKV.
(27-42)
X0
Zero sequence reactance, in units depending on BASEKV.
(43-58)
B0
Zero sequence susceptance, in units depending on BASEKV.
(59-74)
BASEKV
(75-80)
Base Voltage in kV.

= 0 All values have to be specified in p.u. based on 100 MVA,
calculated at the steady-state frequency.
BASEKV
> 0 All values have to be in ohms with inductances and

capacitances in units according to XOPT and COPT.
BASEKV
Second card:
1
2
3
4
5
6
7
8
12 345678901234567890123456 7890123456789012 3456789012345678 9012345678901234 567890
IT
R1
X1
B1
I2
F16.0
F16.0
F16.0
11 - 12
FDNE
Set IT = 52 for positive sequence data.
IT
(1-2)
R1
(27-42)
Positive sequence resistance in the same units as R0 (see basekv on

the first card above).
X1
(43-58)
Positive sequence reactance in the same units as R0 (see basekv on

the first card above).
B1
(59-74)
Positive sequence susceptance in the same units as R0 (see basekv

on the first card above).
Note:
A maximum of 100 linear branches is allowed. Terminate the linear branch cards
with a blank card.
11.3.8
Cards For Shunt Branches
A shunt branch consists of a series connected resistance, inductance and capacitance. At least one
of the R, L, C values of the branch must be non-zero. Two cards per shunt branch are required.
First card:
1
2
3
4
5
6
7
8
12 345678 901234567890123456 7890123456789012 3456789012345678 9012345678901234 567890
IT FROM
I2
A6
IT
R0
X0
B0
BASEKV
F16.0
F16.0
F16.0
F6.0
Sequence flag. Set IT = 51 for first card of zero sequence data.
(1-2)
FROM
"From" bus name.
(3-8)
TO
"To" bus name.
(9-14)
R0
(27-42)
Zero sequence resistance, in units depending on BASEKV.
11 - 13
FDNE
X0
(43-58)
Zero sequence reactance, in units depending on BASEKV.
B0
Zero sequence susceptance, in units depending on BASEKV.
(59-74)
BASEKV
(75-80)
Base Voltage in kV.

= 0 All values have to be specified in p.u. based on 100 MVA,
calculated at the steady-state frequency.
BASEKV
BASEKV > 0 All values have to be in ohms with inductances and

capacitances in units according to XOPT and COPT.
Second Card
1
2
3
4
5
6
7
8
12 345678901234567890123456 7890123456789012 3456789012345678 9012345678901234 567890
IT
R1
X1
B1
I2
F16.0
F16.0
F16.0
IT
Set IT = 52 for positive sequence data.
(1-2)
R1
(27-42)
X1
(43-58)
B1
(59-74)
Positive sequence resistance in the same units as R0 (see basekv on the

first card above).
Positive sequence reactance in the same units as R0 (see basekv on the
first card above).
Positive sequence susceptance in the same units as R0 (see basekv on the
first card above).
Shunt-branch from-buses must be network buses, i.e. previously provided either as a circuit frombus or to-bus, or as a linear branch from-bus or to-bus.
More than one shunt can be connected to a bus. A maximum of 50 shunts is allowed. The shunt
branch cards are to be terminated by a blank terminator card.
11 - 14
FDNE
11.4
Example
See Figure 10.1 for a one-line diagram of the sample case. All options on the option card are zero,
including the print suppression options. The printer output is therefore minimal. The frequency
range for which the network admittance is calculated ranges from 60 Hz to 100 Hz and will have
500 points in a geometric progression.
SS-CLARXBRAM
R21TC
BRAMH2
PLEAS1
CLAIRA
Linear branch
R22TC
BRAMH1
PLEAS9
CHRWDB
S2-JLEAXCHER
BRAM76
ST-HANLXPLEA
CLAIRB
S2-HANLXPLEA
Figure 11.1: Test network with 5 right-of-ways, one linear branch, and one shunt
The listing of the input data file is shown below:

C LABEL CARDS:
TESTING OF MULTIPORT FDNE PROGRAM WITH SMALL NETWORK
TEST RUN FOR TWO PORTS, 60 - 11000 HZ, 500 POINTS
C ------>--- CARD 3, FILES IDENTIFIER:-----------------------------------------TESTRUN
C -----------------------------------------------------------------------------C
1
2
3
4
5
6
7
8
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
C -----------CARD 4, PARAMETERS:-----------------------------------------------C -XOPT-->---COPT-->--REJCTZ->--REJCTP->
60.0
60.0
C -----------CARD 5, OPTIONS:--------------------------------------------------C PL>-SC->-RRN>-RFT>-P1->-P2->-P3->-CHK>-KL->----------------------------------0
0
0
0
0
0
0
0
0
C -----------CARD 6, FREQUENCY RANGE:------------------------------------------C -TOPFR->-PNT>-QFR>-HFR>-------------------------------------------------------
11 - 15
FDNE
11000.00 500
C ---- REFERENCE PORTS --------------------------------------------------------C <-NM-><-KV-><-NM-><-KV-><-NM-><-KV-><-NM-><-KV-><-NM-><-KV-><-NM-><-KV->-----2R21TCA 220.0R22TCA 220.0
C ---RIGHT-OF-WAYS-------------------------------------------------------------C ---R/W HEADER CARD-----------------------------------------------------------C <---NAME---><-L--><-CCT><-DAT-<-PRT-<----------------------------------------S1-HANLXPLEA 30.0
2
0
0
C ---CIRCUITS------------------------------------------------------------------C FROM->TO--->BASEKV<----------------------------------------------------------R21TCABRAMH2 220.0
R22TCABRAMH1 220.0
1.3682 .07170 4
2.81430 -5.182 30.632 22.860 .00000
.00*D5W** 0
2.3682 .07170 4
2.81430 -8.230 24.536 16.764 .00000
.00*D5W** 0
3.3682 .07170 4
2.81430 -5.182 18.440 10.668 .00000
.00*D5W** 0
4.3682 .07170 4
2.81430
5.182 30.632 22.860 .00000
.00*D4W** 0
5.3682 .07170 4
2.81430
8.230 24.536 16.764 .00000
.00*D4W** 0
6.3682 .07170 4
2.81430
5.182 18.440 10.668 .00000
.00*D4W** 0
0.5000 .74280 4
1.38680 -4.267 38.405 32.156 .00000
.00*D5W** 0
0.5000 .74280 4
1.38680
4.267 38.405 32.156 .00000
.00*D4W** 0
32.00
S2-HANLXPLEA 18.04
2
0
0
BRAMH2PLEAS1 220.0
BRAMH1PLEAS9 220.0
1.3682 .07170 4
2.81430 -5.182 30.632 22.860 .00000
.00*D5W** 0
2.3682 .07170 4
2.81430 -8.230 24.536 16.764 .00000
.00*D5W** 0
3.3682 .07170 4
2.81430 -5.182 18.440 10.668 .00000
.00*D5W** 0
4.3682 .07170 4
2.81430
5.182 30.632 22.860 .00000
.00*D4W** 0
5.3682 .07170 4
2.81430
8.230 24.536 16.764 .00000
.00*D4W** 0
6.3682 .07170 4
2.81430
5.182 18.440 10.668 .00000
.00*D4W** 0
0.5000 4.10100 4
.95250 -4.267 38.405 32.156 .00000
.00*D5W** 0
0.5000 4.10100 4
.95250
4.267 38.405 32.156 .00000
.00*D4W** 0
32.00
S3-CLARXBEAV 10.71
2
0
0
CLAIRBBRAM76 220.0
CLAIRAPLEAS1 220.0
1.3682 .07170 4
2.81430 -5.182 30.632 22.860 .00000
.00*D5W** 0
2.3682 .07170 4
2.81430 -8.230 24.536 16.764 .00000
.00*D5W** 0
3.3682 .07170 4
2.81430 -5.182 18.440 10.668 .00000
.00*D5W** 0
4.3682 .07170 4
2.81430
5.182 30.632 22.860 .00000
.00*D4W** 0
5.3682 .07170 4
2.81430
8.230 24.536 16.764 .00000
.00*D4W** 0
6.3682 .07170 4
2.81430
5.182 18.440 10.668 .00000
.00*D4W** 0
0.5000 4.10100 4
.95250 -4.267 38.405 32.156 .00000
.00*D5W** 0
0.5000 4.10100 4
.95250
4.267 38.405 32.156 .00000
.00*D4W** 0
32.00
S4-CLARXBRAM 9.99
2
0
0
BRAMH2PLEAS9 220.0
BRAMH1BRAM76 220.0
1.3682 .07170 4
2.81430 -5.182 30.632 22.860 .00000
.00*D5W** 0
2.3682 .07170 4
2.81430 -8.230 24.536 16.764 .00000
.00*D5W** 0
3.3682 .07170 4
2.81430 -5.182 18.440 10.668 .00000
.00*D5W** 0
4.3682 .07170 4
2.81430
5.182 30.632 22.860 .00000
.00*D4W** 0
5.3682 .07170 4
2.81430
8.230 24.536 16.764 .00000
.00*D4W** 0
6.3682 .07170 4
2.81430
5.182 18.440 10.668 .00000
.00*D4W** 0
0.5000 4.10100 4
.95250 -4.267 38.405 32.156 .00000
.00*D5W** 0
0.5000 4.10100 4
.95250
4.267 38.405 32.156 .00000
.00*D4W** 0
32.00
S5-JLEAXCHER 11.39
1
0
0
PLEAS9CHRWDB 220.0
1.2174 .04560 4
3.39600 -4.267 35.204 21.641 .00000
.00*Q30M* 0
11 - 16
FDNE
2.2174 .04560 4
3.39600 -7.315 29.718 16.154 .00000
.00*Q30M* 0
3.2174 .04560 4
3.39600 -4.267 24.232 10.668 .00000
.00*Q30M* 0
0.5000 5.89100 4
.79380 -2.286 42.672 30.632 .00000
.00*Q30M* 0
32.00
BLANK TERMINATOR CARD OF THE RIGHT-OF-WAYS
C ---LINEAR BRANCHES-----------------------------------------------------------C FROM->TO--->FROM->TO---><------R0------><------X0------><------B0------>KVBASE
51PLEAS1CHRWDB
9.74
31.98
220.00
52PLEAS1CHRWDB
.56
7.01
220.00
BLANK TERMINATOR OF THE LINEAR BRANCHES
C ---SHUNT BRANCHES------------------------------------------------------------C FROM->TO--->FROM->TO---><------R0------><------X0------><------B0------>KVBASE
51CLAIRA
.81
3.34
220.00
52CLAIRA
.12
2.24
220.00
BLANK TERMINATOR OF THE SHUNT BRANCHES
C ---END OF THE CASE------------------------------------------------------------
(i)
The program output starts from the study description (cards 1 and 2)
which is printed at the head of the first page.
(ii)
The rejection and frequency range parameters are listed, and also the
frequency for which the right-of-way zero sequence transformation
matrices are computed.
(iii)
The network input is listed in tabular form. All elements are numbered
for identification in error messages. The Bus Table uses a warning flag
to alert the user when buses have only one element attached.
Misspelling of a bus name will trigger printing of this flag.
(iv)
The element tables are followed by a list which reflects the network
connection and the order in which the elements are used in forming the
admittance matrix. Additional details will be printed when the print
option PRINT2 is set.
(v)
When print option PRINT3 is set to 1, all admittance curves will be

given in tabular form. All data points are numbered for further
reference. A summary of the maxima, minima and zero crossings will
always be listed, showing point number and frequency. FITTER selects
model branch values and lists these values. Some information on the
optimization process and the final value of the model's branch
parameters is also listed.
(vi)
The main output of the program is a punch file containing the model
branch parameters in EMTP format.
(vii)
If the open circuit admittance calculation is requested the printed

output will show a message, followed by plot information on the o.c.
admittance curves.
11 - 17
FDNE
(viii) A "STUDY FINISHED" message concludes the printout.
For brevity, only a few selected parts of the program output are shown below.
First is the beginning of the case through the first R/W circuit cards.
ONTARIO HYDRO FREQUENCY DEPENDENT NETWORK EQUIVALENT PROGRAM

============================================================
STUDY IDENTIFICATION:
TESTING OF MULTIPORT FDNE PROGRAM WITH SMALL NETWORK
TEST RUN FOR TWO PORTS, 60 - 11000 HZ, 500 POINTS
STUDY - DATE:
8.18.98 10. 6
CURVE CONTROL PARAMETERS:
REJCTZ= 0.100, REJCTP= 0.200
OUTPUT: INDUCTANCES WILL BE OUTPUTTED IN MILLI-HENRIES.
OUTPUT: CAPACITANCES WILL BE OUTPUTTED IN MICRO-FARADS.
TABLE OF ADMITTANCE POINTS RANGES FROM
60.0 HZ TO
11000.0 HZ
TABLE HAS
500 POINTS; LOGARITHMIC DIVISION
POWER FREQUENCY =
60.0 HZ
EIGENVECTORS COMPUTED FOR
11000.0 HZ
RECORD OF SORTED INPUT DATA
RIGHT-OF-WAY
1 S1-HANLXPLEA
PHASE NUMBER R-TYPE RESISTANCE X-TYPE X OR GMR DIAMETER X-COORD. Y-COORD.
1
1
0.3682 0.11539
4
0.00000 1.10797 -17.001
83.498
2
2
0.3682 0.11539
4
0.00000 1.10797 -27.001
63.498
3
3
0.3682 0.11539
4
0.00000 1.10797 -17.001
43.499
4
4
0.3682 0.11539
4
0.00000 1.10797
17.001
83.498
5
5
0.3682 0.11539
4
0.00000 1.10797
27.001
63.498
6
6
0.3682 0.11539
4
0.00000 1.10797
17.001
43.499
7
0
0.5000 1.19544
4
0.00000 0.54597 -13.999 112.330
8
0
0.5000 1.19544
4
0.00000 0.54597
13.999 112.330
IMPEDANCES ARE BASED ON EARTH RESISTIVITY OF
32.00 OHM-M AND CORRECTION FACTOR=
0.000001
Next comes bus table and the listing of the remaining network elements.
1 BUS TABLE
=========
* = PORT, * * * = SINGLE ELEMENT WARNING
BUS
CONNECTED TO:
IT. NAME
BASE KV
CCTS LBCS SHNT
--- ------ ---------- ---- ---1 R21TCA 220.000 *
1
0
0 * * *
2 BRAMH2 220.000
3
0
0
3 R22TCA 220.000 *
1
0
0 * * *
4 BRAMH1 220.000
3
0
0
5 PLEAS1 220.000
2
1
0
6 PLEAS9 220.000
3
0
0
7 CLAIRB 220.000
1
0
0 * * *
8 BRAM76 220.000
2
0
0
9 CLAIRA 220.000
1
0
1
10 CHRWDB 220.000
1
1
0
CIRCUIT TABLE
=============
TYPE: 1-INTERNAL CCT, 2-EXTERNAL CCT, 3-LINEAR BRANCH
CODE: RRCC RR=R/W NUMBER, CC=R/W SEQUENTIAL CCT NUMBER
CCT TYPE CODE ---FROM-------TO---------------------------------------------
11 - 18
FDNE
1
1 101
1
2
1 102
3
3
1 201
2
4
1 202
4
5
1 301
7
6
1 302
9
7
1 401
2
8
1 402
4
9
1 501
6
10
3
5
LINEAR BRANCH TABLE
===================
R21TCA
R22TCA
BRAMH2
BRAMH1
CLAIRB
CLAIRA
BRAMH2
BRAMH1
PLEAS9
PLEAS1
2
4
5
6
8
5
6
8
10
10
BRAMH2
BRAMH1
PLEAS1
PLEAS9
BRAM76
PLEAS1
PLEAS9
BRAM76
CHRWDB
CHRWDB
REACTANCE VALUES AT POWER FREQUENCY IN P/U ON 100 MVA BASE

IT.
CIRCUIT
R0
XL0
XC0
R1
XL1
XC1
----------------------------------------------------------------------------------------------------------1
10
0.20124E-01
0.66074E-01
0.00000E+00
0.11570E-02
0.14483E-01
0.00000E+00
SHUNT TABLE
===========
REACTANCE VALUES AT POWER FREQUENCY IN P/U ON 100 MVA BASE
IT. BUS# NAME
R0
XL0
XC0
R1
XL1
XC1
----------------------------------------------------------------------------------------------------------1
9 CLAIRA
0.16736E-02
0.69008E-02
0.00000E+00
0.24793E-03
0.46281E-02
0.00000E+00
RIGHT-OF-WAY TABLE
==================
Next comes the listing of selected minima and maxima of the zero-sequence network admittance:
ZERO SEQUENCE CURVE(S)

MAXIMUM OF BIG-Y CURVE: S.C. ADMITTANCE FROM R21TCA TO GROUND
ZERO SEQ
10.708
MAXIMUM OF BIG-Y CURVE:
ADMITTANCE FROM R22TCA TO R21TCA
ZERO SEQ
-7.425
MAXIMUM OF BIG-Y CURVE: S.C. ADMITTANCE FROM R22TCA TO GROUND
ZERO SEQ
7.458
MAXIMUM OF ALL BIG-Y G VALUES = 10.70761, GRJCT =
1.07076
ANALYSIS OF S.C. ADMITTANCE FROM R21TCA TO GROUND
ZERO SEQUENCE
-------------------------------------------------------------------18 MINMAX; (-1) = G-MIN, ( 1) = G-MAX, ( 2) = ZERO CROSSING
1
60.0 ( 1)
274
980.0 (-1)
302
1298.5 ( 1)
325
1650.3 (-1)
356
387
3116.6 (-1)
419
4305.3 ( 1)
427
4679.6 (-1)
430
4831.0 ( 1)
447
464
6810.3 ( 1)
475
7637.6 (-1)
481
8130.4 ( 1)
486
8476.6 (-1)
493
497
9506.3 (-1)
503
10119.8 ( 1)
511
11000.0 (-1)
NUMBER OF POINTS = 511 ; MAXIMUM G =
10.708 AT POINT
419
14 SELECTED MINMAX;
(-1) = G-MIN, ( 1) = G-MAX, ( 2) = ZERO CROSSING, (4) = MIN, ORIGINALLY MAX
1
60.0 ( 1)
274
980.0 (-1)
302
1298.5 ( 1)
325
1650.3 (-1)
356
387
3116.6 (-1)
419
4305.3 ( 1)
427
4679.6 (-1)
430
4831.0 ( 1)
447
464
6810.3 ( 1)
475
7637.6 (-1)
503
10119.8 ( 1)
511
11000.0 (-1)
2279.7 ( 1)
5764.2 (-1)
9118.1 ( 1)
2279.7 ( 1)
5764.2 (-1)
Next comes the listing of the branches of the zero sequence equivalent network:
INITIAL ERROR
0.40709957
0.03801966 (P/U OF GMAX)
FINAL ERROR=
0.13850176; SKIP ITERATION 5
OPTIMIZATION ENDED WITH 4 ITERATIONS
FINAL ERROR AFTER OPTIMIZATION
0.13839708
0.01292512 (P/U)
BRANCHES AFTER ITERATION:
R(P/U)
L(P/U)
C(P/U)
OMG RES
1
0.135665E-01
0.222373E-03
0.000000E+00
R2= 0.119549E+02
2
0.212699E+01
0.338625E-02
0.440773E-05
0.818528E+04
3
0.283037E+00
0.308997E-03
0.159106E-04
0.142620E+05
4
0.952200E-01
0.149120E-03
0.915798E-05
0.270603E+05
5
0.382182E+00
0.607061E-03
0.177323E-05
0.304790E+05
6
0.188006E+00
0.176003E-03
0.309122E-05
0.428722E+05
7
0.236783E+00
0.130619E-03
0.190430E-05
0.634059E+05
8
0.162600E+01
0.000000E+00
0.769131E-05
OPTIM G CURVE ERROR RANGE
0.5797
-0.2176, B CURVE ERROR RANGE
0.6045
-0.3126
11 - 19
FDNE
INITIAL AND FINAL ERROR IN P/U OF GMAX:
INITIAL/FINAL ERROR:
2.9
0.038020
0.012925
The data of the last two points is then repeated for the positive sequence network. A listing of the
resulting punch file follows:
C TESTING OF MULTIPORT FDNE PROGRAM WITH SMALL NETWORK

C TEST RUN FOR TWO PORTS, 60 - 11000 HZ, 500 POINTS
C TIME OF TEST: 8.19.98 13.41
C FOR CONVERSION XOPTO =
0.00 & COPTO =
0.00 ARE USED.
C -------------------------------------------------------C S.C. ADMITTANCE FROM R21TCA TO GROUND
ZERO SEQUENCE
51SRC01AR21TCA
-6666. -6666.
52SRC01BR21TCB
-6666. -6666.
53SRC01CR21TCC
-6666. -6666.
BRANCHES
C
16 BRANCHES
C <-----R (OHMS)<-------[L] (MH)<[C](MICROFARAD)<------R2 (OHMS)
0.65661850E+01 0.10762853E+03 0.00000000E+00 0.57861782E+04
0.10294651E+04 0.16389435E+04 0.91068745E-02
0.13698972E+03 0.14955466E+03 0.32873195E-01
0.46086498E+02 0.72174019E+02 0.18921452E-01
0.18497601E+03 0.29381772E+03 0.36637033E-02
0.90994827E+02 0.85185257E+02 0.63868132E-02
0.11460278E+03 0.63219486E+02 0.39345040E-02
0.78698462E+03 0.00000000E+00 0.15891137E-01
-0.93405542E+01 -0.20807150E+03 0.00000000E+00 -0.36753752E+08
0.32164941E+03 0.39448004E+03 0.38128678E-01
-0.25540991E+03 -0.26476379E+03 -0.18584330E-01
-0.64707260E+02 -0.11742371E+03 -0.11668603E-01
-0.43332877E+03 -0.92045947E+03 -0.11706959E-02
-0.15537872E+03 -0.12967139E+03 -0.42135348E-02
-0.11611839E+03 -0.70869835E+02 -0.35385762E-02
-0.74228574E+04 0.00000000E+00 -0.88860345E-03
9999
POS SEQUENCE
C
23 BRANCHES
0.34826255E+01 0.65733902E+02 0.00000000E+00 0.11784418E+05
0.22934036E+02 0.15422610E+03 0.74987195E-01
0.22601698E+02 0.11138622E+03 0.35324182E-01
0.51113674E+02 0.16376405E+03 0.83937859E-02
0.11009645E+02 0.38709305E+02 0.30230630E-01
0.99473717E+02 0.45744946E+03 0.19688669E-02
0.18457144E+02 0.51157150E+02 0.98968400E-02
0.51601929E+03 0.18523064E+03 0.17148334E-02
0.18488791E+02 0.52920502E+02 0.52788407E-02
0.52490533E+03 0.31770370E+03 0.76247304E-03
0.38278494E+03 0.00000000E+00 0.11064400E-02
-0.12854956E+02 -0.21060136E+03 0.00000000E+00 -0.30250000E+06
0.15154126E+02 0.10243711E+03 0.11289921E+00
-0.36805851E+02 -0.18018417E+03 -0.21835849E-01
-0.28007389E+02 -0.97751701E+02 -0.14061074E-01
0.18123829E+02 0.62687958E+02 0.18665565E-01
-0.55135365E+02 -0.22195740E+03 -0.40595978E-02
-0.14053844E+04 -0.66231982E+03 -0.82378369E-03
11 - 20
FDNE
-0.21485539E+03 -0.59772791E+03 -0.84693625E-03
0.86239418E+02 0.24488747E+03 0.13055651E-02
0.19157334E+04 0.99346914E+04 0.27920960E-04
-0.75481819E+02 -0.15808243E+03 -0.15233965E-02
0.92453943E+03 0.00000000E+00 0.36802122E-02
9999
ZERO SEQUENCE
51SRC02AR22TCA
-6666. -6666.
52SRC02BR22TCB
-6666. -6666.
53SRC02CR22TCC
-6666. -6666.
BRANCHES
C
17 BRANCHES
-0.93405542E+01 -0.20807150E+03 0.00000000E+00 -0.36753752E+08
0.32164941E+03 0.39448004E+03 0.38128678E-01
-0.25540991E+03 -0.26476379E+03 -0.18584330E-01
-0.64707260E+02 -0.11742371E+03 -0.11668603E-01
-0.43332877E+03 -0.92045947E+03 -0.11706959E-02
-0.15537872E+03 -0.12967139E+03 -0.42135348E-02
-0.11611839E+03 -0.70869835E+02 -0.35385762E-02
-0.74228574E+04 0.00000000E+00 -0.88860345E-03
0.11207032E+02 0.16687210E+03 0.00000000E+00 0.10245266E+06
0.10122321E+03 0.12434045E+03 0.12252698E+00
0.48340045E+03 0.51780927E+03 0.95048556E-02
0.66339348E+02 0.60796040E+02 0.22478348E-01
0.10746614E+04 0.23983931E+04 0.44978532E-03
0.34956335E+03 0.11419439E+03 0.58910856E-02
0.26480106E+03 0.24007510E+03 0.22830456E-02
0.95315659E+02 0.51331348E+02 0.49365461E-02
0.69759222E+03 0.00000000E+00 0.16692964E-01
9999
POS SEQUENCE
C
22 BRANCHES
-0.12854956E+02 -0.21060136E+03 0.00000000E+00 -0.30250000E+06
0.15154126E+02 0.10243711E+03 0.11289921E+00
-0.36805851E+02 -0.18018417E+03 -0.21835849E-01
-0.28007389E+02 -0.97751701E+02 -0.14061074E-01
0.18123829E+02 0.62687958E+02 0.18665565E-01
-0.55135365E+02 -0.22195740E+03 -0.40595978E-02
-0.14053844E+04 -0.66231982E+03 -0.82378369E-03
-0.21485539E+03 -0.59772791E+03 -0.84693625E-03
0.86239418E+02 0.24488747E+03 0.13055651E-02
0.19157334E+04 0.99346914E+04 0.27920960E-04
-0.75481819E+02 -0.15808243E+03 -0.15233965E-02
0.92453943E+03 0.00000000E+00 0.36802122E-02
0.46760225E+01 0.88093803E+02 0.00000000E+00 0.35057844E+05
0.10133679E+02 0.68436211E+02 0.16899464E+00
0.60221584E+02 0.29693033E+03 0.13250149E-01
0.16007254E+02 0.56394165E+02 0.24372561E-01
0.29362171E+02 0.10362984E+03 0.11291543E-01
0.31067492E+02 0.12671633E+03 0.71108276E-02
0.20120203E+02 0.61587608E+02 0.89218942E-02
0.29247164E+02 0.74179932E+02 0.43130005E-02
0.15736946E+02 0.35418812E+02 0.68007251E-02
0.39151276E+03 0.00000000E+00 0.22514369E-01
11 - 21
FDNE
9999
C -------------------------------------------------------C
ZERO SEQUENCE
51R22TCAR21TCA
-6666. -6666.
52R22TCBR21TCB
-6666. -6666.
53R22TCCR21TCC
-6666. -6666.
BRANCHES
C
8 BRANCHES
0.93405542E+01 0.20807150E+03 0.00000000E+00 0.36753752E+08
-0.32164941E+03 -0.39448004E+03 -0.38128678E-01
0.25540991E+03 0.26476379E+03 0.18584330E-01
0.64707260E+02 0.11742371E+03 0.11668603E-01
0.43332877E+03 0.92045947E+03 0.11706959E-02
0.15537872E+03 0.12967139E+03 0.42135348E-02
0.11611839E+03 0.70869835E+02 0.35385762E-02
0.74228574E+04 0.00000000E+00 0.88860345E-03
9999
C -------------------------------------------------------C
POS SEQUENCE
C
12 BRANCHES
0.12854956E+02 0.21060136E+03 0.00000000E+00 0.30250000E+06
-0.15154126E+02 -0.10243711E+03 -0.11289921E+00
0.36805851E+02 0.18018417E+03 0.21835849E-01
0.28007389E+02 0.97751701E+02 0.14061074E-01
-0.18123829E+02 -0.62687958E+02 -0.18665565E-01
0.55135365E+02 0.22195740E+03 0.40595978E-02
0.14053844E+04 0.66231982E+03 0.82378369E-03
0.21485539E+03 0.59772791E+03 0.84693625E-03
-0.86239418E+02 -0.24488747E+03 -0.13055651E-02
-0.19157334E+04 -0.99346914E+04 -0.27920960E-04
0.75481819E+02 0.15808243E+03 0.15233965E-02
-0.92453943E+03 0.00000000E+00 -0.36802122E-02
9999
C --------------------------------------------------------
11 - 22
FDNE
11.5
Error Messages
There are two types of error messages:

(i)
internal dimensions are exceeded;
(ii)
input information is inconsistent.
Program generated error messages are helpful in the first stages of a study. Messages associated
with errors of the first type are mostly self explanatory; other messages may need some
explanation.
In the following section the text of the messages will be printed in capitals, followed by the
explanation. Messages are separated by a dashed line.
R/W________is external; circuit ____ to-bus must be left blank.
Internal and external circuits are mixed on a right-of-way (see Section 11.3.6 for an explanation of
internal lines).
-------------------------------------------------------------------------------NUMBER OF Q-ARRAY ENTRIES ___ EXCEEDS THE PROGRAM LIMITS OF ____
There are too many right-of-ways with too many circuits.

-------------------------------------------------------------------------------BRANCH ___ FROM BUS ___, TO BUS ___ REFERRED TO BUT CANNOT BE LOCATED.
Error in a Linear Branch specification.

-------------------------------------------------------------------------------ERROR ENCOUNTERED IN BUILDING THE NETWORK.
ALREADY ENGAGED.
REVERSAL OF CIRCUIT ___ REQUESTED; R/W ___
Inconsistency in the labelling of the 'FROM' and 'TO' busses of a right-of-way's circuits.
-------------------------------------------------------------------------------LUA WRITTEN FOR ___ CIRCUITS, REFERENCE FREQUENCY OF ___ HZ, TOP FRQUENCY OF ___ HZ AND
HIGH POINT FREQUENCY OF ___ HZ.
11 - 23
FDNE
LUB WRITTEN FOR ___ CIRCUITS, REFERENCE FREQUENCY OF ___ HZ, TOP FREQUENCY OF ___ HZ AND
HIGH POINT FREQUENCY OF ___ HZ.
SYSTEM HAS ___ CIRCUITS, REFERENCE FREQUENCY OF ___ HZ, TOP FREQUENCY OF ___ HZ AND HIGH
POINT FREQUENCY OF ___ HZ.
These messages will appear when the rerun flag is set (see Section 11.3.3, Program Options Card),
and the previously calculated line parameters are not corresponding to the rerun case
specifications.
* * * * * * WARNING * * * * * *
ERRORS IN THE SUBROUTINE MINIMS, RESULTS INCOMPLETE
The fitting of one or more of the curves was skipped due to errors. The .RLC file will be
incomplete.
-------------------------------------------------------------------------------ERROR IN MINIMS ROUTINE.
NUMBER OF TABLE ENTRIES EXCEEDS THE PROGRAM LIMIT
The number of frequency points, including added zero-crossing points, exceeds the built-in
program limit. Reduce the number of points.
-------------------------------------------------------------------------------ERROR IN MINIMS ROUTINE.
NUMBER OF MIN-MAX EXCEEDS THE PROGRAM LIMIT OF ___
ERROR IN MINIMS ROUTINE.
TOO MANY BRANCHES REQUIRED; STUDY TERMINATED
Reduce the frequency range and/or increase the rejection level.

-------------------------------------------------------------------------------ERROR IN MINIMNS ROUTINE
ADMITTANCE FILE ___ FOR ___; THIS STUDY FOR ___
ERROR IN MINIMS ROUTINE
ADMITTANCE FILE ___ FOR ___ WITH ___ PORTS AND HIGHPOINT FREQUENCY OF ___ HZ;
THIS STUDY FOR ___ WITH ___ PORTS AND HIGHPOINT FREQUENCY OF ___ HZ.
ERROR IN MINIMS ROUTINE
THE STEADY-STATE FREQUENCIES IN .ADMP AND .ADMZ ARE NOT THE SAME.
WARNING.
POWER FREQUENCY IS ___
These messages appear when the Refit flag is set (see Section 11.3.3, Program Options Card) and
the admittance file does not agree with either the read-in case or each other.
11 - 24
FDNE
11.6
Admittance Curve Option
This option provides for using the results of field tests (measurements) and bypassing the built-in
calculations of network admittance.
11.6.1
Format and Rules
Two data cards are to be inserted before the Study Parameters Card (see Section 11.3.2).
Card 1:
Enter the keyword ADMITTANCE CURVE in columns 1 through 16. This will serve as a flag for use
of this special option.
1
2
3
4
5
6
7
8
1234567890123456 7890123456789012345678901234567890123456789012345678901234567890
ADMITTANCE CURVE
A16
Card 2:
Define the names of two files in columns 1-40 and 41-80, respectively.
1
2
3
4
5
6
7
8
1234567890123456789012345678901234567890 1234567890123456789012345678901234567890
FILENAME
(41-80)
FILENAME 2
A40
A40
The first file is to contain the zero sequence admittance data. The file
description is given below.
The second file is to contain the positive sequence admittance data. The
file description is given below.
(1-40)
FILENAME
FILENAME 1
11 - 25
FDNE
File Format
The file is a plain ASCII file. Data is read with a "free format" read statement. The file can have a
maximum 1000 records or frequency points. The number of records in the file is given by the
parameter NNPNT (see Card 6 of Section 11.3.4). The last data record in the file must contain the
admittance data at the extreme frequency (JHIFR, in Card 6). The structure of each record is as
follows:
Record Structure: Freq, (G(i), B(i), i = 1, k)
where
k
(NR*(NR + 1))/2
NR
(Number of reference ports (see Card 7 of Section 11.3.5)
Freq
Frequency (Hz)
Conductance (p.u.)
Susceptance (p.u.)
Notes:
11.6.2
1.
In this mode, the parameters (i) SCALE, (ii) RERUN, (iii) REFIT, and (iv) PRINT2 in
Card 5 are ignored.
2.
Minimum frequency is given through the optional card "POWER FREQUENCY".

However, the program assumes that the first point has to be fitted.
3.
Network element cards are not needed.
Example
The following input data file uses the admittance curve option
C
TESTING ADMITTACE CURVE OPTION OF FDNA PROGRAM AND XOPTO/COPTO
SMALL NETWORK, RUN THREE DECADES
FDNATC
C ************* CURVE ADMITTANCE OPTION ***********************
ADMITTANCE CURVE
GBCURVE.ZRO
GBCURVE.POS
C *************
C
1
2
3
4
5
6
7
C 345678901234567890123456789012345678901234567890123456789012345678901234567890
C
XOPT
COPT
REJCTZ
REJCTP
XOPTO
COPTO
60.0
60.0
0.080
0.060
0.0
0.0
11 - 26
FDNE
0
0
0
10000.00 792
1R21TC* 220.
The first few lines of file GBCURVE.ZRO follow

60.00000
60.39500
60.79200
61.19200
61.59500
62.00000
3.687398
3.675772
3.664084
3.652328
3.640525
3.628664
-2.018326
-2.023720
-2.029080
-2.034386
-2.039628
-2.044830
The first few lines of file GBCURVE.ZRO follow

60.00000
60.39500
60.79200
61.19200
61.59500
62.00000
62.40800
11.7
8.551723
8.524687
8.497450
8.469993
8.442402
8.414577
8.386629
-4.136479
-4.156137
-4.175629
-4.195065
-4.214297
-4.233411
-4.252367
Program Description
This section describes how the internal components of the FDNE work. This could be useful to
help interpret error messages, and other output messages.
The main program FDNEMAIN reads the input data, prepares transmission coefficient tables for
later use and controls the computation. It calls subroutines NETWRK, MINIMS, FITTER, OPTIM and
OCTCHK during the execution of the program. The user controls the printout produced by the
program. Separate print control options are available for the main program and the different
subroutines.
11.7.1
Main Program
The main program reads input data consisting of:
case description,
study parameters,
program options,
11 - 27
FDNE
frequency range parameters,
reference ports, and
description of network elements.
Input data are organized and printed out in form of tables of system buses, right-of-ways, circuits,
linear branches and shunts.
Transmission lines are located on right-of-ways which may carry one or several three-phase
circuits. Normally right-of-ways are divided into longitudinal sections; each longitudinal section
has its own unique geometrical arrangement of circuit conductors, groundwires, etc. This program
treats each longitudinal section of a right-of-way as a separate entity.
When right-of-way information is read in, the main program prepares a list of transmission
coefficients associated with that right-of-way. After information is read in and processed,
transmission coefficients are written to a file and control is passed to subroutine NETWRK. Storing
transmission coefficients on file allows the user to rerun the program using the pre-calculated
values. Reruns are limited to the same frequencies as the original case.
Subroutine NETWRK
This subroutine organizes the network elements by an internal numbering scheme and calculates
the admittances over a range of frequencies. The frequency range begins with the steady-state
frequency (reference frequency), and ends with the user specified high frequency.
A list of frequencies is generated, and for each frequency the admittance of individual network
elements are computed and the overall network admittance matrix is formed. Matrix reduction
techniques are used to calculate the equivalent admittances of the network from reference ports to
ground and between reference ports. The admittance values are written to the admittance files.
The procedure is carried out both for the zero and positive sequence networks. Control is then
passed on to subroutine MINIMS via the main program. By having admittance files available, the
user can rerun the cure fitting part of the program using different fitting criteria.
Subroutine MINIMS
Subroutine MINIMS processes the sequence admittances per group. MINIMS determines the
maximum of all zero sequence admittances and then identifies acceptable maxima on the
individual curves of the zero-sequence group. It then sets up a loop to read in and process the
individual admittance-versus-frequency curves, one at a time. Maxima, minima and zero crossings
of the admittance curves are determined as each curve is read in. Interpolation is used to determine
the zero crossing frequencies.
Significant maxima and minima then are selected, a table of maxima and minima is established,
and control is passed on to subroutine FITTER.
11 - 28
FDNE
When control is returned, MINIMS writes the model branch parameters in EMTP format to the result
file. The entire process is then repeated for the positive sequence admittance, after which control
is passed on to the main program.
Subroutines FITTER, OPTIM, and OCTCHK

Subroutine FITTER selects the linear branch parameters whose frequency response will
approximate the calculated network frequency response. The model matches the network
response at peak admittance values. Control is then passed on to subroutine OPTIM.
Subroutine OPTIM calculates the error in matching the frequency response of the model and the
given admittance curve. OPTIM then adjusts the branch parameters to reduce the error, using a
least squares approach. Finally control is returned to MINIMS.
This subroutine calculates the equivalent open-circuit nodal admittance-versus-frequency curve
for each reference port. The plotted curves provide a visual check on the stability of the model.
Use of this subroutine is optional.
11 - 29
Section 12
References
[1]
R.C. Degeneff, A Method for Constructing Terminal Models for Single-Phase NWinding Transformers", Paper A-78539-9; presented at IEEE PES Summer Meeting, Los
Angeles, CA, July 1978.
[2]
H.W. Dommel, Transformer Models in the Simulation of Electromagnetic Transients,

Fifth Power Systems Computation Conference held in Cambridge, England, September 15, 1975.
[3]
L. Mart, "Simulation of Transients in Underground Cables with Frequency-Dependent

Modal Transformation Matrices". IEEE Transactions on Power Delivery, Vol. 3, No. 3,
pp.1099-1110, July 1988.
[4]
L. Mart, R.H. Brierley, T.E. Grainger "Analysis of Electromagnetic Transients in CrossBonded Cable Systems Using Frequency Dependent Cable Models", Proceedings of the
International Conference of Power Systems Transients IPST'95, Lisbon, Portugal
September 3-7 1995. )
[5]
L. Mart, T.E. Grainger, A.S. Morched, "Sheath Overvoltages in HV Underground

Cables", CIGRE, Paris, 1996.
[6]
K. Schlosser, "An Equivalent Circuit for N-Winding Transformers Derived From a

Physical Basis", Brown Boveri-Nachrichten, vol. 45, pp. 107-132, March 1963, and
"Application of the equivalent circuit of an N-winding transformer", Brown BoveriNachrichten, vol. 45, pp. 318-333, June 1963 (in German)
[7]
Tarasiewicz, Morched, Narang, Dick, "Frequency Dependent Eddy Current Models for
Nonlinear Cores", PWRS, Vol. 8, No. 2, May 1993, pp.588-597
[8]
M. Kh. Zikherman, "Magnetizing Characteristics of Large Power Transformers",

Elektrichestvo No. 3, pp. 79-82, 1972 (in Russian)
12 - 1

RuleBook Full PDF

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

RuleBook Full PDF

Transféré par

Droits d'auteur :

Formats disponibles

Rule Book Vol 1

Rule Book Vol 2

EMTP Rule Book Vol. 1

Introduction to the EMTP

Computer Aspects of the EMTP

Structure of EMTP Input Data

Cards to Begin A New Data Case

Lumped Linear Impedance Models

Frequency-Dependent Transformer Model

Distributed Parameter Transmission Line Models

User-Supplied Initial Conditions

Output-Variable Specification Cards for the Time-Step Loop

User-Supplied Source Functions (Type 1-10 Sources)

"TACS"Transient Analysis of Control Systems

Description of Program Output

Informative and Error Messages

EMTP Data Modules

Static Var Compensator

LDFL: Three-Phase Load Flow

Air Gap Arc Model

References and Bibliography

Program Development - Historical Notes ............................................... 1-1

Applications Of The EMTP ...................................................................... 1-2

Program Capabilities - Overview ........................................................... 1-3

Basic Elements Represented In The Program ......................................... 1-4

Program Input And Output ..................................................................... 1-8

Support Programs ................................................................................... 1-9

Supported Hardware Platforms ............................................................... 1-10

User Support And Maintenance .............................................................. 1-10

EMTP Documentation .............................................................................. 1-11

1.2 Applications of The EMTP

Shaft Torsional Stress

Static VAR Compensation

Carrier Frequency Propagation

1.3 Program Capabilities - Overview

1.4 Basic Elements Represented In The Program

The specific EMTP component used to represent single-phase lumped inductances is

Multiphase pi-equivalents of transmission lines, where the preceding scalar R, L,

Frequency-dependent network equivalents for approximate modelling of large

Multiphase distributed-parameter transmission lines, wherein propagation time

Single-circuit transposed line model of Section 7.3.1.

Double-circuit transposed line model with zero-sequence coupling of

Untransposed line model of Section 7.3.1.

J. Marti line model of Section 7.4.

L. Marti cable model of Section 7.5.

Corona model of Section 7.6.

Nonlinear resistors, the v-i characteristics of which must be single-valued.

Type-92 true nonlinear element of Section 8.8.

Type-99 pseudo-nonlinear element of Section 8.3.

Nonlinear inductors, either with the conventional single-valued characteristics

Type-93 true nonlinear inductance of Section 8.10.

Type-98 pseudo-nonlinear inductance of Section 8.4.

Type-91 true nonlinear resistance of Section 8.7.

Type-97 staircase time-varying resistance of Section 8.2.

Switches, used to simulate circuit breakers, lightning-arrester flashover, or any

Ordinary switches of Section 9.2.1 and Section 9.2.2.

Switches with randomly or systematically varied closing/opening times

Permanently closed switches of Section 9.2.5.

TACS controlled switches (valves, diodes, etc.) of Section 9.5.

Sophisticated circuit breaker model of Section 8.10.

Voltage or current sources. In addition to standard mathematical functions

Type-11 step function of Section 10.3.