Journal of Electrical Engineering

www.jee.ro

DISTANCE RELAY PERFORMANCE EVALUATION ON SERIES

COMPENSATED TRANSMISSION LINE UNDER FAULTED CONDITIONS
saptarshi ROY1 Dr. p suresh BABU2
1
Research Scholar, Dept., of Electrical Engineering, NIT Warangal, INDIA, email id: [email protected]
2
Assistant Professor, Dept., of Electrical Engineering, NIT Warangal, INDIA, email id: [email protected]

Abstract- A compensated line imposes problems to compensation implies a modification in the electrical
directional relaying schemes due to Voltage and characteristic of the transmission line with the
current inversion situations and operation of metal objective of increase power transfer capability [4]. In
oxide varistor (MOV) protecting series capacitor, the case of series compensation, the objective is to
reactance modulation issues .In this paper the cancel part of the reactance of the line by means of
behavior of series compensated EHV transmission
lines during faults is simulated. The use of series series capacitors [9]. This result in an enhanced
capacitors for compensating part of the inductive system stability, which is evidenced with an increased
reactance of long transmission lines increases the power transfer capability of the line, a reduction in
power transmission capacity. Emphasis is given on the transmission angle at a given level of power
the impact of modern capacitor protection techniques transfer and an increased virtual natural load [10-11-
(MOV protection).A novel methodology is proposed 12].
to identify faulty phases based on correlation factor
computation. Under various fault conditions the Fast and accurate determination of a fault in electrical
proposed method is tested for its validation. The power system is a vital part in power restoration [13].
proposed method is tested on series capacitor In Power system majority of the faults are happened
compensated transmission lines (SCCTLs) with their to be single line to ground fault [14]. Other important
different configurations and contingency types of faults are LLG, LLL, LL, LLLG faults. The
combinations and performance is observed with presence of the capacitor in the circuit immediately
transmission line both end voltage profiles. Distance after a fault is very important, because it helps in
characteristics are also drawn for various zones of improvement the transient stability of the system.
protection. Simulation results show that proposed Also in case of unbalanced faults, only the protection
method has identified correct fault location. devices of the faulted phases operate leaving the
capacitor of the other phases on line. It is
Index Terms-Distance Protection, Distance relay
indispensable to be able to model such devices in a
Performance, Variation of R&X, Transmission Line,
fault analysis program and predict the level of short
Series Compensation, Series Capacitor Protection
circuit currents as well as the energy absorbed by the
Unit, correlation, faulty phase identification.
conducting MOV.
I.INTRODUCTION Some of the advantages of series compensation of
transmission line are listed below:
Use of series capacitors for compensating inductive 1) Reduces line voltage drop.
reactance of long transmission lines increases the 2) Limits load-dependant voltage drops.
power transmission capacity [3-8]. It also increases 3) Influences load flow in parallel transmission
transient stability margins, optimizes load-sharing lines.
between parallel transmission lines and reduces 4) Increases power transfer capability
system losses [5-6-7]. Transmission line

1
 Journal of Electrical Engineering
www.jee.ro

5) Highly effective in maintaining the desired for positive linear correlations and negative linear
voltage profile along the transmission line correlations, respectively.
interconnecting two busses of the ac system
and providing support to the end voltage of Positive Correlation: If A and B have a strong
radial lines in the face of increasing power positive linear correlation, r is close to +1.
demand [21-25] . An r value of exactly +1 indicates a perfect positive
fit. Positive values indicate a relationship
This paper proposes a novel methodology to between A and B variables such that as values for A
identify faulty phases of a transmission line increases, values for B also increase.
which is tested on a series compensated
transmission line through PSCAD/EMTDC Negative Correlation: If A and B have a strong
simulation . This faulty phase identification negative linear correlation, r is close to -1.
algorithm gives better result compare to An r value of exactly -1 indicates a perfect negative
detection of faulty phase by imposing the fit. Negative values indicate a relationship
tolerance limit method and the polar plot between A and B such that as values for A increase,
analysis gives more insight about the zone of values for B decrease.
the fault and chance of mal-operation.
No-Correlation: If there is no linear correlation or a
II. SERIES CAPACITOR EFFECT ON weak linear correlation, r is close to 0. A value near
DISTANCE MEASUREMENT: zero means that there is a random, nonlinear
relationship between the two variables. r is a
Distance relays are designed to perform correctly on a dimensionless quantity; It does not depend on the
resistive/inductive system. When SCs are introduced, units employed.
the normal voltage/current relationships are affected, A Perfect Correlation of ± 1 occurs only when the
especially when the fault levels are not sufficient to data points all lie exactly on a straight line. If r = +1,
flash-over the gaps or to produce significant the slope of this line is positive. If r = -1, the slope of
conduction in the MOV's [21] . this line is negative.

III.SINGLE AND MULTI-PHASE FAULT A correlation greater than 0.8 is generally described
DETECTION : as strong, whereas a correlation less than 0.5 is
generally described as weak. These values can vary
Linear correlation coefficient r, measures the based upon the type of data being examined. A
strength and the direction of a linear relationship study utilizing scientific data may require a
between two variables. It is a measure of how similar stronger correlation than a study using social science
the two signals or variables are. The data. These statistical concepts will be used for
mathematical formula for computing r is: detection of the phases involves fault.

𝑛 𝐴𝐵−( 𝐴)( 𝐵)
𝑟= (1)
𝑛( 𝐴2 −( 𝐴2 ) 𝑛 𝐵 2 −( 𝐵)2 )
Algorithm for Proposed method:

Step1: Start
Where n is the number of pairs of data. The value of
r is such that -1 < r < +1. The + and – signs are used
Step2: Sample the faulted voltage or current
waveform during fault duration.

2
 Journal of Electrical Engineering
www.jee.ro

Step3: From the samples find out greatest change of Type of rab rbc rca Faulty
current or least change in voltage containing sample. fault phases
created identified
The phase which containing greatest change of LLL 0.14002289 -0.84169 -0.6524 All faulty
AB-G -0.51618 -0.49752 -0.4861 A,B
current or least change in voltage must contain fault.
faulty
The change is considered to eliminate ZCD (Zero LLLG -0.77963 -0.99386 -0.9938 A,B,C
Crossing Detector) Problem. faulty
Ph-CA 0.4555593 -0.455791 -0.9999 A,C
Now check for whether two or more phases contain faulty
fault or not. This paper is concentrated particularly on
the aspect when two or more phases contain fault or
not. The computational steps as continues:
TABLE II
Step4 :Take Ia ,Ib,Ic or Va,Vb,Vc sample data during
fault time and store it in three different variables Analysis with Voltage variable: Fault Created at 140 KM
A,B,C. distance. Faulted time (0.34-0.38sec).Series compensation
20% employed.
Step5: Now compute correlation between
A&B,B&C,C&A. Type of rab rbc rca Faulty
fault phases
created identified
Step6: Stop
LLLG -0.77959 0.723784 -0.99386 A,B,C
faulty
If at least two values comes between -0.5 to -1 then Ph-CA -0.46896 0.790685 -0.92395 C,A
all the phases involves fault(LLL or LLLG fault), or faulty
if any one value comes between -0.5 to -1 then that LLL -0.54237 0.492593 -0.93517 A,B,C
two phases involve with fault . faulty

Test results obtained from the algorithm are enlisted

in TABLE I and TABLE II . IV.IMPLEMENTATION

Fault location is computed using faulted voltage and

current samples that described as below:
Fault impedance Zf= (Vf/If) Ohm (2)
Fault Location= (Zf/ Z 1) (3)
TABLE I
Emperical formula of fault current
TEST SYSTEM-Series Capacitor Compensated IFi = (Isi + IRi Cos(h𝛾il) - VRi/Zci Sin (h𝛾il)) / Cosh(𝛾ild) (4)
Transmission Line with Protection Unit
Fault Voltage
Analysis with current variable: Fault Created at 140 KM VFi = VRi Cos(𝛾il(1-d) -ZciIRi Sin h(𝛾il(1-d))) (5)
distance. Faulted time (0.34-0.38sec).Series compensation
40% employed. Fault impedance
ZFi=(VFi/IFi) (6)

3
 Journal of Electrical Engineering
www.jee.ro

More the length increases the algorithm for fault

location identification gives more accurate value. In
case of fault impedance, during healthy state positive
sequence impedance (Z1 Ohm/Km) present in the
system and the value is given in the paper. Vf and If
are obtained from the sampling of the waveforms
after creation of different faults at different lengths.

Fig2: Performance of the test system during a Ph-B-G fault

at 40% line length

Fig1:series capacitor compensated transmission lines

(SCCTLs) with Protection Unit with capacitor at various
positions

Test system(Fig1)is simulated using PSCAD/EMTDC

software and test different cases with varying fault
location, varying capacitor position, varying
compensation, varying fault resistance, varying fault
duration etc. Then calculated and plotted faulted
Fig3: Performance of the test system during a Ph-AB-G
wave forms and R&X curves(ref. Fig.2 to Fig.12).
fault at 40% line length
Sampling frequency maintained here 4KHZ and
system frequency maintained here 50 HZ.

4
 Journal of Electrical Engineering
www.jee.ro

Fig5: Performance of the test system during a LLLG fault

at 40% line length

Fig4: Performance of the test system during a LLL fault at

40% line length

Fig6: Performance of the test system during a LL fault at

40% line length

5
 Journal of Electrical Engineering
www.jee.ro

Fig7: Performance of the test system during a AG fault at Fig10: Variation of R&X in p.u during a LLLG fault
40% line length

Fig11: Variations of R&X in p.u during a LL fault

Fig8: Variations of R &X in p.u during a b-g fault

Fig12: Variations of R&X in p.u during a LLL fault

Fig 9: Variations of R&X in p.u During a bc-g fault
During fault, from the available voltage and current
samples data corresponding resistance and reactance
parameters are estimated to plot the impedance

6
 Journal of Electrical Engineering
www.jee.ro

characteristics and to confirm the zone where the fault

has located. R and X values are shown in TABLE III
,TABLE IV and TABLE V .

TABLE III
Calculation of R&X for a BC-G fault:

Z(units) R(units) X(units)

99.91033 86.52234 49.95516
99.93408 86.54291 49.96704
99.94821 86.55515 49.97411
99.95479 86.56085 49.97739
99.95831 86.5639 49.97916
99.96272 86.56772 49.98136
a)
99.96037 86.56568 49.98018
99.96443 86.5692 49.98221
99.96637 86.57088 49.98319

TABLE IV
Calculation of R&X for a B-G fault

Z(units) R(units) X(units)

99.93755 86.54592 49.96877
99.93737 86.54576 49.96869
99.93712 86.54555 49.96856
99.93663 86.54512 49.96831
99.93924 86.54738 49.96962 b)
99.94398 86.55149 49.97199
99.94644 86.55362 49.97322
99.95059 86.55721 49.9753
99.95314 86.55942 49.97657

TABLE V
Calculation of R&X for a LL fault

Z(units) R(units) X(units)

129.1968 111.8844 64.59838
361.101 312.713 180.55
1015.94 879.807 507.972 c)
2767.65 2396.78 1383.82 Fig13: Fault Impedance locus during a) LG fault
3951.639 3422.119 1975.819 b) LLG fault c) LL fault
3588.58 3107.71 1794.29
894.3192 774.4805 447.1596 The fault impedance characteristics for different types
2604.014 2255.076 1302.007 of faults are plotted on R-X plane(ref. Fig 13). Here it
1063.61 921.087 531.806 is considered up to magnitude 50 unit circle as Zone1,

7
 Journal of Electrical Engineering
www.jee.ro

up to 100 unit circle as zone2 and beyond this 1 and link 3 are active .In that case the data can be
magnitude as zone3.From the above characteristics it extracted in the following ways :
is clear that the fig13(a) LG fault and fig13(b) LLG
fault both located at Zone2.But fig13(c) is located at Let us consider link i voltage is Vi
Zone3. So, in fig13(c) case there may be a possibility Link i current is Ii
for involvement of load encroachment problem and And Resistance( or impedance ) Ri
distance relay may give a trip signal to the circuit
breaker. where i€{ 0,1,2…….n}

i)Implementation of the above method in case of Then in case of communication link 2 failure, there
a communication link failure : will be two parallel paths in the system between 1 to 3
for the above system(same as Fig.15)
Path 1-2-3 and path 1-3
So, R2 & R3 will be in series which will be parallel
with R1
So,
Total impedance offered between 1 & 3 point will be

Z = (R2+R3) || R1 (7 )

= R1(R2+R3) / ( R1+R2+R3) (8)

So, if the net voltage difference between path 1& 3

is V
Fig 14 : Data extraction during a communication link Then net current flowing through the path is
failure of a 3-phase system
I = V/Z (9)

So, according to the current dividers rule,current

flowing in the path 1-2-3 is

I2 =( R1/(R1+R2+R3))* I ( 10 )

= ( R1/(R1+R2+R3)) *( V/Z) (11 )

=( R1/(R1+R2+R3))*(V)*((R1+R2+R3)/ R1(R2+R3))
Fig 15 : Equivalent Circuit for Data Extraction when Link (12 )
2 fails for the above 3-Phase system
= V/(R2+R3) ( 13)
Consider, the case of implementing this method in So, between 1-2 link Sample voltage will be
case of a larger system, where data or samples of
voltage or current wave forms are coming from = V*R2 /( R2+R3) ( 14)
PMUs or Wide Area measurement systems . Suppose
link 2 of the above system(ref. Fig. 14) fails and link Between 2-3 link sample voltage will be

8
 Journal of Electrical Engineering
www.jee.ro

A typical Voltage and current response is shown in

= V*R3 /(R2+R3) ( 15) Fig. 16 (a) & (b).

Now, the application of the above method proposed iii) Stability of the Method :
here is possible .
The method is based on correlation coefficient
ii ) Validity of the above Method During calculation between two signals . It is totally depend
Transient conditions of PMUs : on the sampling process of the signal . It is not
having direct relation with fault levels or phase
The method will work fine during transient conditions angles .Entirely the method depends on the features of
e.g- power swing or electromagnetic transients etc the samples extracted from the voltage or current
.As the method depends on correlation factor signals . So, it is stable and also not going to affect
calculation, which is a measure of how two signals the stability of the system.
are changing together . So, there is no problem with
cycle to cycle or sample to sample comparison which
happens in case of power swing due to frequency V.CONCLUSION
variation of power other than volage or current
signal[22-27] . The main results regarding MOV protected series
capacitor compensated transmission line, obtained by
the fault simulation can be summarized as follows:
The variation of R&X is less during ground faults,
compared to the phase faults. From the figures it is
evident during a bg fault, the maximum value of
R&X is confined in the order of 140 unit(ref.Fig.8),
while in the case of a Phase-B to Phase-C -g fault, the
variation is of the order of 90 units(ref. Fig. 9) and
during a LLLG fault the variation is of maximum 140
(a) Voltage Response ( Voltage vs time ) unit(ref. Fig.10). But in case of phase faults, it is more
than two times of ground faults. In case of LL fault, it
comes beyond 1000 unit(ref. Fig. 11) and LLL fault
it is coming beyond 300 units in negative half (ref.
Fig 12).
From polar plot of the fault impedance, it can be
concluded which zone the locus is confined and
whether distance relay is going to generate any trip
signal to circuit breaker or not and if there is any
chance for evolving load encroachment problem or
not. Faulty phases can be identified when two or more
phases involves fault using correlation statistical
operator.
(b) Current Response ( Current vs time) Several special cases have been discussed in the
context of the Work . e.g- Implementation of the
Fig 16 : Transient Response from PMUs
proposed method in case of a communication link
failure in the system , the transient performance issues

9
 Journal of Electrical Engineering
www.jee.ro

of the PMUs etc . The stability of the proposed Rated current of 2000A, equivalent of (1750MVA), is
method with respect to the system stability is also considered for sizing Series Capacitor Protection
addressed. Unit. MOV rating after considering an overload factor
This work having vivid contributions w.r.t protection of 1.5 is calculated as 273kV (= 1.5 ×2000A×91.1Ω).
of series compensated transmission line .First of all
a novel methodology is proposed for phase detection REFERENCES
of fault based on correlation factor computation is
very simple and effective in correctly predicting the [1] Glover J.D. and Sarma M.S., Power Systems: Analysis
phase of the fault and it is further extended to the and Design. Pacific Grove, CA: Brooks/Cole -
analysis of polar plot which is helpful in the Thomson Learning, 2010.
assessment of system internal states . It can judge the [2] Kundur P., Power System Stability and Control. New
York: Mc-GrawHill, 1994.
state of the fault as well as any chance of mal-
[3] McNabb D., Granger M., Van Q.B, Rousseau M, and
operation of the system .Eventually this methodology Pilot M," Transient Design Studies for the
is helpful not only for predicting of fault but also for Transmantaro Series-Compensated Transmission
the condition monitoring of the system which can System," Proceedings of the International Conference
lead its increase in security and reliability . on Power System Transients (IPST'01), pp. 1-6, June
This faulty phase identification algorithm gives better 2001.
result compare to detection of faulty phase by [4] Rican P., Electric Power Authority Standard of
imposing the tolerance limit method and the polar Construction of Overhead Transmission Lines May
plot analysis gives more insight about the zone of the 1992.
fault and chance of mal-operation. Thus this work is [5] Trad o., Ratta G., and Torres M., "Experiences in
Setting Protection of Series Capacitor Compensated
commendable w.r.t the protection of series
Lines," Proceedings of the International Conference on
compensated transmission line . Power System Transients (IPST'01), pp. 1-6, June 2001.
[6] Bui-Van Q., Portales E., McNabb D., and Gajardo
APPENDIX A V., "Transient Performance of 500-kV Equipment for
the Chilean Series-Compensated Transmission System,"
Series Capacitor Compensated Transmission Line Proceedings of the international Conference on Power
system data System Transients (IPST'03), pp. 1-6, October 2003.
Length=350Km [7] IEEE 14-bus test system data available
Voltage=500KV at:http://www.ee.washington.edu/research/pstca/
Positive sequence impedance=0.0155+j0.3719 Ω /Km [8] Henschel S., Kischner L., and Lima M.C., "Transient
Recovery Voltage at Series Compensated
Zero Sequence Impedance=0.3546+j1.0670 Ω /Km
Transmission Lines in Piauí, Brazil," Proceedings of
Positive-sequence admittance= 0+ j4.4099×10−6 the International Conference on Power System
mho/Km Transients (IPST'05), pp. 1-6, June 2005.
Zero-sequence admittance= (0+ j2.7844 × 10−6) mho [9] Zimmerman R., Gan D., Matpower 3.0, PSERC
per km Cornell University,USA, December 1997. Available
Positive and zero sequence impedances for sending at:http://www.pserc.cornell.edu/matpower/
end source are ZS1 = (1+j15) Ω, ZS0 = (2.4 + j25) Ω [10] ABB Metal Oxide Surge Arresters – EXLIM P, ABB
respectively. Electrical Apparatus, May 2005.
Positive and zero sequence impedances for receiving [11] Miller T.J., Reactive Power Control in Electric
end Source are ZR1 = (1.2 + j18) Ω, ZR0 = (2.6 + j26.5) Systems. New York: Wiley-Interscience, 1982.
[12] Xuan Q.Y., Johns A.T., ”Digital Simulation of Series
Ω, respectively. Load angle is 30 ◦ with receiving end
compensated EHV (Extra High Voltage) Transmission
source voltage lagging. Systems” , IEEE Colloquium on Simulation of Power
System , 1992.

10
 Journal of Electrical Engineering
www.jee.ro

[13 ]Wattanasakpubal C. and Bunyagal T.,”Algorithm for [23] Terzija V , Radojevic Z.M and Preston G ,” Flexible
Detecting, Identifying, Locating and Experience to Synchronized Measurement Technology-Based Fault
Develop the Automate Faults Location in Radial Locator”, IEEE Transactions on Smart Grid,Vol.6 ,Issue
Distribution System”. Journal of Electrical Engineering 2 , pp. 866-873,March 2015 .
andTechnology, Vol.5, No.1, pp.36-44, 2010. [24] Pal T , Bains S and Zadeh M.R.D,” Enhanced Phasor
[14] Elmitwally A.,.Mahmoud S., Abdel Rahman M.H. Estimation Technique for Fault Location in Series-
“Fault Identification of Overhead Transmission Lines Compensated Lines”, IEEE Transactions on Power
Terminated With Underground Cables,” Proceeding of Delivery ,Vol.30,No.4,pp. 2058-2060,August 2015 .
14th International Middle East Power Systems conference [25] Thakre M.P, Kale V.S,”An Adaptive Strategy For
,Dec 19-21,2010,pp.452-458. Three Zone Operation of Numerical Distance Relay with
[15] Brahma S.M., “Distance relay with out-of-step Shunt FACTs, Based on PMU “, Journal of Electrical
blocking function using wavelet transform,” IEEE Engineering ,Vol. 15,edition : 4 ,December 2015 .
Transactions on Power Delivery., vol. 22, no. 3, pp.1360– [26]Kobibi D ,Samir H and Abdeldjalil ,”Effect of the
1366, Jul. 2007. UPFC on a Multi-machine Power System Steady State and
[16] Mohanty S.R., Pradhan A.K., and Routray A, “A Dynamic Performance”, Journal of Electrical Engineering,
cumulative sum-based fault detector for power system Vol. 15, edition : 4 , December 2015 .
relaying application,” IEEE Transactions on Power [27] Boudiaf M , Moudjahed M ,” Improvement of
Delivery., Vol. 23, no. 1, pp. 79–86, Jan. 2008. Transient Stability of Power System by IPFC ,SSSC and
[17] Nekoubin A, ”Simulation of Series Compensated STATCOM “, Journal of Electrical Engineering,Vol.
Transmission Lines Protected with Mov”, World Academy 14,edition : 1, March 2014.
of Science, Engineering and Technology 58,pp.1002-1006 [28]Muneswar S.N, Hasabe R, Shelar D,and Kose P“A
2011. New Adaptive PMU based Protection Scheme for
[18] Nayak P.K., Pradhan A.K., and Bajpai P, “A Fault Interconnected Transmission Network System”, 2014
Detection Technique for the Series-Compensated Line International Conference on Circuit, Power and
During Power Swing”, IEEE Transactions on power Computing Technologies [ICCPCT], 20-21March ,2014;
Delivery, Vol. 28, NO. 2, April 2013,pp.714-722. Nagercoil,India,IEEE,pp.179- 183 .
[19] Nayak P.K., Pradhan A.K., and Bajpai P., [29]Liao Y and Kang N, “Fault-location algorithms
“Detecting fault during power swing for a series without utilizing line parameters based on the distributed
compensated line,” presented at the International parameter line model”, IEEE Transactions on Power
Conference of Energy, Autom., Signal, Bhubaneswar, Delivery , Vol.24,Issue:2,pp579–584 Feb.2009.
Odisha, India, Dec. 28–30,2011. [30] Korkali M, Abur A , “Optimal deployment of wide-
[20] Esmaeilian, A. Ghaderi A., Tasdighi M., and Rouhani area synchronized measurements for fault-location
A., “Evaluation and performance comparison of power observability,” IEEE Transactions on Power Systems , vol.
swing detection algorithms in presence of series 28, no. 1, pp. 482–489, Feb. 2013.
compensation on transmission lines,” in Procceedings of
10th International Conference Environment Elect. Eng.
, May 8–11, 2011, pp. 1842–1848.
[21] Roy S ,Babu P.S,” Power swing protection of series Saptarshi Roy: Received the B.E
compensated transmission line with novel fault detection degree in Electrical Engineering from
technique “ ,IEEE International Conference on Green Jadavpur University, West Bengal,
Computing, Communication and Electrical Engineering India in 2009.Received M.Tech degree
[ICGCCEE]. 2014, 6–8 March, Coimbatore. from NIT WARANGAL in 2014.
[22] Mansour W.M , Abdelhakeem N.SH and ALI M.A,” Currently he is pursuing PhD in
Combinational protection algorithm embedded with smart Electrical Engineering in the
relays”,Journal of Electrical Engineering,Vol.15,edition:2, department of Electrical Engineering, National Institute of
June 2015. Technology, Warangal, India. His areas of interest are
power system protection, Phasor Measurement Unit
applications in power systems, Synchrophasors
applications in power systems.

11
 Journal of Electrical Engineering
www.jee.ro

Dr. Suresh Babu Perli: Currently he is

working as an Assistant Professor in
Department Of Electrical Engineering,
National Institute of Technology,
Warangal. His areas of interest are
Power System Protection with digital
multifunction relays, Development of
Adaptive protection schemes and Digital filtering
algorithms.

12