Vous êtes sur la page 1sur 5

Selection of Pilot buses for VAR Support

Considering N-1 Contingency Criteria


Talpasai Lakkaraju, and Ali Feliachi, Senior Member, IEEE

Abstract This paper proposes a new approach for the


identification of possible bus location, pilot buses, for reactive
power support that will improve voltage stability and security. A
continuation power flow based N-1 contingency analysis is
performed alongside the reactive power management to pinpoint
the locations for the VAR support provided in this study by an
SVC. The approach is through the optimization of a performance
index that account for worst voltage deviations at all load buses
and at the same time maximizing the loading margin. A case
study of the proposed approach is performed on the IEEE 14 bus
system and when compared to existing methods, it shows that the
proposed approach gives better results.
Index Terms voltage stability, loading margin, pilot bus,
VAR support, SVC location

I. INTRODUCTION
In view of the worldwide restructuring of the electricity
industry, it is essential for power systems to operate securely,
under different operating conditions and especially, during
contingencies. Voltage stability plays a major role in keeping
the system operational. Because of its importance an
impressive amount of work has been devoted to it in view of
recent blackouts that have been attributed to it. Voltage
stability is mainly concerned with maintaining acceptable
voltage profile under all operating conditions.
Voltage control requires a source of reactive power and
one of the main challenges is to find the optimal location for
the reactive power support especially with new technology like
FACTS devices in place. Most of the secondary voltage
control schemes currently being implemented in France and
other countries [2], use pilot bus control techniques [4], [5],
wherein the objective is to maintain the voltage of the pilot
bus, which is the critical bus in the system or area and
controlling the pilot bus voltage is a means to have system
wide good voltage levels keeping in mind the voltage
sensitivity of neighboring buses to the pilot bus. But with ever
increasing power demand and necessity of reactive support
through FACTS devices like SVC, the objective of voltage
security and stability can be better achieved by choosing these
pilot buses as sites for reactive power support. Most of the
research performed so far in selection of these pilot buses does

The authors are with the Advanced Power & Electricity Research Center,
West Virginia University, Morgantown, WV 26506-6109, USA,
Fax:(304)293-8602(emails:tlakkara@mix.wvu.edu; alfeliachi@mail.wvu.edu)

142440178X/06/$20.002006IEEE

not consider affect of generation reactive power limits and N-1


criteria which is an important factor to judge the reliability of
the power network under different conditions.
There has also been significant research in the past to
pinpoint locations for placing FACTS devices or reactive
support. In [12] the author presents modal analysis using Eigen
values of reduced Jacobian to locate the weak buses of the
system, which are similar to the singular values and singular
vectors for detecting proximity of voltage collapse but as
shown in [6], they are not good indicators for predicting
voltage collapse bifurcation point because of their high nonlinear behavior near the bifurcation point. In [7], the author
uses a tangent vector index to find the possible locations for
reactive power support, but as shown in [6], this index is not
useful in predicting the critical bus as one needs to be rather
close to the bifurcation point to estimate this bus. In [9], the
authors present an interesting index titled voltage collapse
proximity index for weak bus selection and use simulated
annealing technique for placement of Var sources which is a
very laborious technique and not efficient. Also there is no
discussion on what happens when the generators hit the
reactive limits and how it effects the selection. In [10], a
voltage stability index is defined to find the critical buses of
the system, but the index does not give enough information
regarding the systems margin to bifurcation and does not
consider contingency conditions.
In this work voltage stability problem is considered more
as a reactive power problem than as a simple voltage control
scheme, and an approach is defined to find the optimal
location for the placement of the FACTS device considering
N-1 criteria and loading conditions. The paper is organized as
follows: Section II discusses the background and motivation
for reactive support in power systems. Section III introduces to
the approach of voltage stability and optimal location choice
using continuation power flow and N-1 criterion. A test system
and analysis tools are briefly introduced in Section IV. Section
V presents some results and discussion. Finally, contribution
and conclusions are presented in Section VI.
II. BACKGROUND AND MOTIVATION
A. Voltage Stability and Reactive Power
As defined in [1], Voltage stability is the process by which
the sequence of events accompanying voltage instability leads
to a blackout or abnormally low voltages in a significant part
of the power system. Voltage stability is inherently a dynamic

1513

PSCE2006

problem. But since time domain simulations are time


consuming and also they do not readily provide the sensitivity
information or the degree of stability [1]. For these reasons
generally for bulk system studies the static analysis is preferred
in order to provide more insight into the voltage and reactive
power problem.
In static voltage stability the slow changes in the power
system eventually lead the system into instability situations
with declining voltage and shortage of reactive power. This
can be observed from the P-V curve analysis, wherein it
demonstrates as the power transfer increases the voltage at the
receiving end decreases as show in Fig.1. The nose-point or
bifurcation point pinpoints to the maximum limit beyond
which the system collapses because of lack of enough reactive
support to maintain the voltage profile. One solution to this
problem is to reduce reactive power load or add additional
reactive power prior to the collapse point.
The technique usually used is to place an additional
reactive power support at critical bus or pilot bus of the
system. But the definition of the weakest bus has to be clearly
defined since, to maintain the additional stability and good
voltage profile the reactive support at heavily loaded bus might
not guarantee to be the best solution because the heavily
loaded bus may not be the pilot bus. In this context the pilot
bus is defined as the bus which is strongly tied with other load
buses and the support at this bus ensures good voltage profile
and additional security to the system. As discussed earlier,
there are many techniques in the literature to identify these
locations but most of the indices do not give pin point
accurately under different conditions close to collapse point
and especially during contingency situations. In this work, a
continuation power flow with N-1 contingency analysis is
proposed to perform a thorough analysis considering the
voltage deviations and the loading margin to establish an index
for finding the critical buses for reactive power support.
B. Mathematical Model
The mathematical model for power system is the standard
load flow model, where the variations of active and reactive
powers are assumed to be the main parameter driving the
system to singularity.
The power flow model [3] used here is a typical load-flow
vector nonlinear equation defining active and reactive power
mismatches.

P ( x, )
Q( x, = F ( x, )

(1)

Where x is a state vector of load flow problem and typically


represents V and , i.e., the bus voltages and angles and can
also be used to compute other system variables like for
example generation reactive power injections Q. Also x can be
modified to include more detail models of certain devices like
loads and FACTS. The variable stands for a parameter that
slowly changes in time, so that one equilibrium point to
another until it reaches the collapse point or bifurcation. The

variable typically here represents loading factor used to


simulate the system load changes that drive the system to
collapse in the following way.

v
*

Loading parameter

Fig. 1 Traditional P-V curve

Where

PD = PD 0

(2)

QD = QD 0

(3)

PDO and QD 0 are base case real and reactive loads

respectively.
C. Continuation power flow
Continuation power flow [13], [14] allows the user to
completely trace the complete voltage profile by automatically
changing the value of , the loading parameter. The
continuation power flow uses a predictor corrector scheme to
find the solution for a set of power flow equations
reformulated by adding an additional parameter , this is
necessary so that the singularity of the jacobian can be avoided
and an accurate value of the collapse point can be estimated. In
this work continuation power flow technique is used to
bifurcation point and estimate the load margin.
III. CRITICAL BUS IDENTIFICATION USING VOLTAGE
STABILITY ANALAYSIS AND N-1 CRITERIA
The flowchart for pilot bus analysis is illustrated in
Fig.2. From Fig.2, it can be observed that data from the output
of the Contingency screening is used as the input to the
continuation power flow and N-1 analysis. Selected lines are
used for line outages using contingency screening. This is
necessary because for a realistic size system it is a lengthy
process to consider each and every outage and is not practical.
Since the characteristics of the system change with
different contingency conditions, the screening of the
contingency is done by selecting lines that are heavily loaded
at the collapse point. This information is available from the
continuation power flow.

1514

The main idea is to find the optimal location based on the


voltage constraints of the system and the loading margin. The
N-1 analysis and continuation power flow is run with SVC at
each load bus and the loading margin is evaluated when the
voltage limit violation is recorded. The important point here is
to understand that the ideal location for the SVC, with reactive
limits considered, would be the location which enhances the
voltage at the buses and at the same time ensures a decent
increase in the loading margin compared to the base case, i.e.,
without an SVC.
The objective function for the selection of bus where to
install an SVC is

min J = 1 J1 + 2 J 2

Start

Read Data

Run Continuation Power Flow

Perform Contingency Screening


+
Evaluate Base Max. Loading
Margin
LMB for each contingency

For i =1:PQ.N (load buses)

(4)

i D

For k = 1: Line.n

Subject to

Vmin V Vmax at all buses


QG min QG QG max at VAR generating buses

For j =1:Line.NC

min max
Run N-1 Contingency Analysis

Where D = {load buses}

J1 =

Run Continuation Power Flow (CPF)

1
LM

(5)
Evaluate LM (j) = | max - 0 |

B
Where LM = LM j LM j
jc

(6)
Evaluate J LM = (LM (j) - LMB(j))

C = {Selected Contingencies}

Evaluate JV =
Max (Infinity Norm(Vk))

k = 1:NT, NT -no of buses

J 2 = Max . V
jC

(7)

Compute : J (i) = { 1*Inv(J LM) + 2 JV)}}

. V
Where

= Max Vi

(8)

iall buses

1 , 2

i = i +1

are normalized weights.

Note that in the test case studied,

1 = 3 and 2 = 0.1, so

J* = min. J

that voltage magnitude deviations and loading margins will


have similar effects)

Fig. 2. Flowchart for determining weak buses using CPF and


N-1 Contingency Analysis.

LM j = Loading Margin of jth contingency,

LM Bj - Base case loading margin of jth contingency.

IV. TEST SYSTEM AND ANALYSIS TOOL

Here LM denotes the loading margin of the system. Loading


margin is defined as the distance between the current operating
point and the maximum loading point. It gives the measure of
how much system can be loaded or how far is the system from
instability .The continuation power flow and N-1 contingency
criteria takes into account the reactive power limits of the
generators and the SVC, which here is used as the additional
reactive power support. The weights 1 , 2 are chosen so that

The IEEE 14-bus test system [8] is used throughout the


study. A single line diagram of the IEEE 14 bus test system is
depicted in Fig.3, which consists of five generator buses
located at buses 1, 2, 3, 6 and 8. In the system, there are
twenty branches and fourteen buses with eleven loads totaling
259MW and 81.4Mvar for the base case.

the indices are normalized.

1515

which does not consider these limits is not a realistic one [3].
TABLE I
CRITICAL BUS RANKING FOR THE VAR SUPPORT

Jmin
1.73
1.20
2.28
2.33
2.45
2.50
2.78
3.04

Bus Number
4
5
10
9
11
14
13
12

TABLE II
LOADING MARGIN WITH SVC AT BUS #14 [12], [7] AND WITH SVC AT
BUS#4 (PROPOSED) WITH OUTAGE AND VLIMITS

Outage
Fig. 3. Single line diagram of the IEEE 14 bus system

Line #11
Line #14
Line #12
Line #16
Line #15
Line #17
Line #1

Power System Analysis Toolbox (PSAT) [8] a toolbox


based on MATLAB was used for the analysis.

V. RESULTS AND DISCUSSION

LM
(Base)
0.93
1.15
1.19
1.35
1.32
1.07
1.41

LM (SVC
at bus#4)
1.38
1.88
1.45
1.87
1.80
1.10
1.94

LM (SVC
at bus#14)
1.02
1.44
1.16
1.66
1.67
1.52
1.73

A. Location and Size of FACTS device


The best location of the compensation device of VAR
support for the system would be the one which improves the
loading margin of the system and at the same time maintains a
good voltage profile. Mathematically, it is at the bus that will
minimize the performance index J described by equation (4)
above.
In order to estimate the capacity of the VAR support, [7] a
relationship between the maximum loading factor and the
corresponding capacity of the device was found which would
give the information about how much support the device can
provide against the voltage collapse. Based on this analysis a
value of 3 pu was chosen to be the rating of the device.
Based on the index and approach mentioned in this paper,
the following results were found for the optimal location for
the VAR support. In contrast to the result in [12], which
identifies bus # 14 as the ideal bus, with the proposed
approach, bus number 4 was found to be the ideal bus for the
reactive power support. This is illustrated with Analysis of the
system using PSAT support the claim as shown below.
Buses numbers 3, 4, and 9 are the most loaded buses when
compared to the rest of the buses. So any additional reactive
power support must make sure that these buses operate at good
voltage profile under normal and contingency conditions. In
this study the generator reactive power limits are taken into
account for estimating loading margin since voltage collapse is
very much linked with existing reactive limits and any solution

Table II tabulates the effects of placing an SVC at


bus#14 (as suggested in [12] and bus#4 (as identified by the
proposed algorithm. The loading margin actually improves
much better in the latter case with voltage limits maintained.
As noticed bus#14 might not be the ideal choice for the
location of VAR support considering the importance of
maintaining good voltage profile with generation limits in
place. Also the plots Fig.4 and Fig.5 show us this result,
wherein the loading margin with reactive power limits with
SVC at bus#4 and bus#14 is evaluated. As shown, the result
obtained with SVC at bus#4 gives better performance with
better voltage regulation and increase in loading margin when
compared to SVC at bus# 14.
VI. CONCLUSIONS
The VAR support using FACTS devices is a large
investment problem and has to be carefully analyzed by
considering different operating conditions. The paper
discusses the effect of contingency conditions and affect of
generation reactive power limits and proposes an index to
evaluate the best possible location for the VAR support using
Static Var compensator (SVC) device for maintaining good
voltage profile and loading margin.

1516

[6]

[7]

[8]
[9]

[10]

[11]

[12]
Fig. 4. PV curves at load buses with and without SVC at bus #4
[13]

[14]

C. A. Caizares, A. Z. de Souza, and V. H. Quintana, Comparison of


performance indices for detection of proximity to voltage collapse,"
IEEE Trans. Power Systems, vol. 11, no. 3, August 1996, pp. 14411450.
SodeYome, Arthit and N. Mithulananthan, Comparison of Shunt
Capacitor, SVC and STATCOM in Static Voltage Stability
Improvement, IJEEE, Vol. 41 No. 2, 2004.
F. Milano, (2002) PSAT, Matlab-Based Power System Analysis
Toolbox.[Online] Available: http://thundebox.uwaterloo.ca/~fmilano
Y.L. Chen, Weak Bus-Oriented optimal Multi-objective VAR
Planning, IEEE Transactions on Power Systems, vol.11, no.4,
pp.1885-1890, November 1996.
Thukaram, D. and Abraham Lomi, Selection of Static VAR
Compensator location and size for system voltage stability
improvement, Journal of Electric Power System Research, Vol. 54,
2000, pp 139-150.
Lagonotte, P., Sabonnadiere, J.C., Leost, J.Y., and Paul, J.P.: Structural
analysis of the electrical system: application to the secondary voltage
control in France, IEEE Trans., 1989, PWJXS-l, (2), pp. 479484
G. Morison, B.Gao and P. Kundur, "Voltage Stability Analysis Using
Static and Dynamic Approaches", IEEE Trans. on Power Systems, Vol.
8, No. 3, pp. 1159-1171, Aug. 1993.
C. A. Caizares and F. L. Alvarado, "Point of Collapse and
Continuation Methods for Large AC/DC Systems," IEEE Transactions
on Power Systems, Vol. 8, No. 1, February 1993, pp. 1-8.
V.Ajjarapu. C.Christy, The continuation power flow: A tool for steady
state voltage stability analysis, IEEE Transactions on Power Systems,
Vol.7, No.1, Feb.1992,pp. 416-423.

IX. BIOGRAPHIES
Talpasai Lakkaraju received his Bachelors degree from Jawaharlal Nehru
Technological University, India. He is currently working towards his Masters
degree in the Lane Department of Computer Science and Electrical
Engineering at West Virginia University, Morgantown. His interests are
voltage stability of large scale power systems and power system economics.

Fig. 5. PV curves at load buses with and without SVC at bus #14

VII.

ACKNOWLEDGMENT

This research is sponsored in part by a US DoE EPSCoR WV


State Implementation Award, and in part by grant from the US
DEPSCoR and ONR (DOD/ONR N000 14-031-0660).

Ali Feliachi (SM86) received the Diplme dIngnieur en Electrotechnique


from Ecole Nationale Polytechnique of Algiers, Algiers, Algeria, in 1976, and
the M.S. and Ph.D. degrees in electrical engineering from the Georgia
Institute of Technology, Atlanta, in 1979 and 1983, respectively. Currently,
he is Full Professor and the holder of the endowed Electric Power Systems
Chair position in the Lane Department of Computer Science and Electrical
Engineering at West Virginia University (WVU), Morgantown, where he has
been since 1984. He is also the Director of the Advanced Power and
Electricity Research Center at WVU. He has been working in the field of
large-scale systems and power systems for 25 years, and has appeared in
many publications. Dr. Feliachi is a member of ASEE, Pi Mu Epsilon, Eta
Kappa Nu, and Sigma Xi. He received an ASEE Dow Outstanding Young
Faculty Award in 1987, and the following awards from the College of
Engineering at WVU: Leadership (1989), Research (1991), and Graduate
Teacher (1991), and in 1994, he received the Claude Benedum Distinguished
Scholar Award for the Sciences and Technology from WVU.

VIII. REFERENCES
[1]
[2]

[3]

[4]

[5]

P. Kundur, Power System Stability and Control. New York: McGrawHill, 1994
Lagonotte, P., Sabonnadiere, J.C., Leost, J.Y., and Paul, J.P.: Structural
analysis of the electrical system: application to the secondary voltage
control in France, IEEE Trans., 1989, PWJXS-l, (2), pp. 479-484
C. A. Caizares, Ed., Voltage Stability Assessment: Concepts,
Practices and Tools, IEEE-PES Power System Stability Subcommittee
Special Publication, SP101PSS, Technical Report, 2002
A. Conejo, T. Gomez, and J.I.de la Fuente, "Pilot bus selection for
secondary voltage control", ETEP, vol 3, No.5, pp. 359-366, Sep-Oct.
1993.
Sancha, J.L.; Fernandez, J.L.; Cortes, A.; Abarca, J.T., .Secondary
voltage control: analysis, solutions and simulation results for the
Spanish transmission system. IEEE Trans. Power systems., vol 11,
issue 2, pp 630-638 May 1996

1517

Vous aimerez peut-être aussi