Académique Documents
Professionnel Documents
Culture Documents
scheme
Mehmet Tümay
Electrical and Electronics Engineering Department, University of Gaziantep, Turkey
E-mail: tumay@gantep.edu.tr
Abstract This paper presents the modelling and analysis of unit protection of feeders. The protection system module
is linked to the primary power system module in such a way that it acts as a feedback loop. The primary power
system module provides the three-phase voltages and/or currents at the transducer locations. The protection system
module then processes this data, and outputs a circuit breaker status, which can then be feedback to the primary
power system module. The paper also describes the operating principles of the protection unit. Finally, practical
applications of the model under fault conditions are examined.
Transducer modelling
Dynamic modelling of the protective current transformer has been developed
for the transient analysis of the equivalent circuit of the current transformer
using the core model previously described.8 The developed routine is fully
stable and used to simulate the transient response of the current transformer
to asymmetrical fault currents. The core has been represented including
magnetic non-linearity and saturation effects. The line currents are fed to the
CT subroutines from the primary power system module. Secondary currents
are calculated and stored for the power system protection signalling. Figure 4
shows current transformer connections.
Relay modelling
Relay structure
The relay model is divided into following four sub-modules;
1 analogue signal conditioner,
2 analogue to digital converter,
3 relay algorithm including signal conversion,
4 relay status processor.
The analogue signal conditioner models the relay’s pre-filter, filter and inte-
grator. These models have digital transfer functions which represent the ana-
logue transfer functions of the physical components. The analogue to digital
converter conditions the filtered signals so that they can be applied to the
protection algorithm. This will require reducing the sample rate to the relay
sample rate, converting floating point samples to integers, and clipping the
signal if necessary. The inputs to the relay algorithm in general will be d.c.
(Plant Status) inputs as well as the conditioned current and/or voltage samples.
The d.c. (Plant Status) inputs will be stored in the protection unit record. The
algorithm will output a trip or no trip operation signal. The relay status
processor controls the implementation of the microprocessor time delay.
In order to control microprocessor and circuit breaker delays both the relays
and the circuit breaker are given an operating status, and delays are controlled
by the combination of these states. At relay model sample No. 1 the algorithm
outputs a no operation signal thus the relay model status stays in ‘no operation’
and the circuit breaker model remains ‘closed’. Since, in real time, the micropro-
cessor would still be processing data.
Only after a specified duration will the trip decision be taken. After this
delay (three power system model samples) the relay model status becomes ‘o/p
trip’ and the circuit breaker model status becomes ‘operating’. The circuit
breaker model remains in operating mode until the circuit breaker algorithm
determines that it should close. This method of controlling these delays allows
the inclusion of a delayed auto reclose scheme, since the method works equally
well in reverse.
Filter evaluation is required to check that the digital filter is as accurate a
model of the physical filter as possible. A program has been developed to
determine the frequency response, of recursive and non-recursive digital filters.
This program inputs the transfer function of the digital filter and outputs the
gain and phase response over a predetermined range of frequencies. However
this program only tests the steady state response of the digital filter, a step
response test will be performed to determine its transient response.7
Relay operation
On a healthy feeder the vector sum of the terminal currents flowing into the
protected zone should equate to zero in accordance with Kirchhoff ’s first law,
any difference current representing either capacitive or fault current. The a.c.
current input module contains an analogue to digital converter. Each line
currents is sampled at the certain rate. Positive (IPPS) and negative phase
sequence (INPS) component values are derived from these and added together
to form a single representative signal, the local end relaying current I (I =
M M
5I −I ), which is then frequency modulated onto a carrier in the VF
NPS PPS
modem module and transmitted via the communication link to the remote
end. A similar signal is received from the remote end and demodulated.
Differential comparison is carried out on the local and remote I signals.
M
The differential comparison process works on a vector summation principle
which takes into account both the waveform amplitude and phase angle. The
local and remote I current waveforms are cross-correlated with reference sine
M
and cosine waveforms to calculate the polar and rectangular coordinates of
their phasors. From this the differential and restraining quantities are calculated.
The differential comparator operates when the following conditions are met:
|I1MR+I2MR|>differential setting
and
|I1MR|+|I2MR|
|I1MR+I2MR|> >bias slope
2
Where
I1MR is the protection unit 1 differential relaying current
I2MR is the protection unit 2 differential relaying current
The relaying currents are amplitude limited to 5xdifferential setting which
results in a true linear comparison at low levels of current to phase comparison
at high levels of current. Calculations are performed on a complete set of
periodic samples which are updated every sample period. If the differential
comparison indicated a trip condition a trip counter is incremented towards a
user-set-target level. For a stable condition or a communication error detector
output the trip counter is reset.
Simulation examples
The two parallel feeders are connected to the busbar for the simulation example.
Figures 5 and 6 show the schematic diagram of the system studied. The current
transformers are connected to the each protection unit of the parallel feeder.
Different fault conditions can be applied to the each of the line and busbars.
The generator, current transformer and the relay settings are given in
Appendix A.
Each graphical overview shows I1 output waveforms together. These are
defined as follows:
I1A, I1B, I1C the line currents of the protection unit one (PU1).
I1PPS, I1NPS, I1ZPS the sequence currents.
I1M the relaying current of the PU1.
I1MD the delayed relaying current of the PU1.
I2M the relaying current of PU2.
IDIFF the final differential current obtained from PU1
and PU2.
IRES restraint current.
COUNT trip count of the relay.
Case A
Various types of faults are applied to the busbar 2. The current passing from
PU1 and PU2 will be the same for that case. The maximum peak to peak
current of each case is shown in Table 1.
Case B
The faults are applied between protection unit 1 and 2. In this case, the current
passing from PU1 and PU2 will be different. These graphs are shown in Figs
7–12 for three different fault conditions (single-phase, phase-to-phase, three-
phase faults).
Conclusions
In this paper, the modelling and analysis of a unit protection of the parallel
feeders were explained. The modelling strategy of the system was also described.
The modelling strategy of the primary power system and protection module
was described. Four protection system units were used for the protection of
the parallel feeders. Different fault conditions were applied to the different
sections of the system. From the simulation results the maximum fault current
Fig. 9 PU1 waveforms (line-to-line fault is applied between PU1 and PU2).
occurred when the line-to-line fault was applied between the two protection
units.
This type of protection system can be applied to a complicated protection
system. The advantage of the method is that it effectively protects the units
without any interaction with the other systems.
Fig. 10 PU2 waveforms (line-to-line fault is applied between PU1 and PU2).
Fig. 11 PU1 waveforms (three-phase fault is applied between PU1 and PU2).
Acknowledgements
Special thanks are due to Professor J. R. Smith and Professor J. R. McDonald
for their help, encouragement and guidance. Thanks are due to the staff in the
group, Dr. R. R. Simpson, Dr F. P. Flynn and Mr D. M. Grant for their
invaluable discussion and technical assistance.
Fig. 12 PU2 waveforms (three-phase fault is applied between PU1 and PU2).
References
1 Protective Relays Application Guide (GEC Alsthom Measurements Ltd., Stafford, 1990).
2 Applied Protective Relaying (Westinghouse Electric Corporation Relay and Telecommunication
Division, Coral Springs, FL, 1990).
3 IEEE Committee Report, ‘Excitation System Models for Power System Stability Studies’,
IEEE T rans. Power Apparatus and Systems, PAS-100 (1981), 494–509.
4 Meng-Jen, ‘Phase coordinate representation of interconnected power system’, Ph.D. Thesis,
University of Strathclyde, 1992.
5 K. L. Lo, M. S. Buamud, M. Tümay and H. El-Khatroushi, ‘Recovering the secondary voltage
from the single and three-phase capacitive transformer’s primary waveform’, ACEMP’95 Int.
Aegean Conf. Electrical Machines and Power Electronics 1995, Kuadas, Turkey, pp. 204–209.
6 K. L. Lo, M. Tümay, H. El-Khatroushi and M. S. Buamud, ‘Digital computation of the
transient behaviour of current transformer under system fault conditions’, ACEMP’95 Int.
Aegean Conf. Electrical Machines and Power Electronics 1995, Kuadas, Turkey, pp. 198–203.
7 Mehmet Tümay, ‘The performance of power system protection under transient operating
conditions’, Ph.D. Thesis 1995, Strathclyde University, pp. 5–58.
8 Mehmet Tümay, R. R. S. Simpson and Hassan El-Khatroushi, ‘Dynamic model of a current
Transformer’, Int. J. Elec. Engng. Educ., 37(3) (2000), 268.