2012 IEEE International Power Engineering and Optimization Conference (PEOCO2012), Melaka, Malaysia: 6-7 June 2012

Effective Two-Terminal Single Line to Ground

Fault Location Algorithm
M. H. Idris, Member, IEEE, M. W. Mustafa, Member, IEEE and Y. Yatim

Abstract-This paper presents an effective algorithm to locate Travelling wave propagation was first adopted by
Single Line to Ground (SLG) fault at a transmission line. Post Dommel and Michels (1978) to detect faults in transmission
fault voltages and currents from both substation terminals were line [3]. Then, other papers were published to employ this
used as the input parameters to the algorithm. Discrete Fourier technique to locate fault [4]-[8]. In travelling wave method, an
Transform (DFT) was used to extract the magnitudes and phase electrical pulse is sent along the transmission line. The time of
angles of three phase voltages and currents. The modeling of the
transmission line along with the algorithm was performed using
the pulse return back indicates the distance to fault point. The
Matlab/Simulink package. The results of fault location for SLG disadvantage of travelling wave method is the propagation can
faults along the transmission line demonstrated the validity of the be significantly affected by system parameters and network
algorithm used even for high resistance earth fault. configuration. Besides that, it also difficult to locate fault that
is near to the bus or fault that occurs near zero voltage
inception angle [9].
Index Terms-Discrete Fourier Transform, Fault Location,
Matlab/Simulink, Two-terminal, Single Line to Ground Impedance based technique uses the fundamental
frequency of voltage and current phasors from installed
transducers such as numerical relays and fault/disturbance
I. INTRODUCTION recorders [2]. This method is widely used because of its
simplicity and low cost to be adapted to electronic devices in

T RANSMISSION line is one of the most important the substations. This method can be classified into two
component in a power system. Fault has higher possibility categories depending on the measurements they used. The
to happen at transmission line rather than other power system categories are one terminal data algorithms [10]-[16] and two
components because it is exposed to the environment. Such terminal data algorithms [17]-[25].
conditions which can trigger faults at the transmission line are
Commonly, one terminal data algorithms are the most
danger tree, lightning strike, cross arm fall, crane
widely used to estimate fault location because of limited
encroachment, current transformer explosion, animals and access of data between two substation terminals. This
insulator pollution. All this faults require fast restoration of the algorithm requires many mathematical assumptions to
transmission line. The accurate fault location is very important eliminate the effects of fault resistance on the accuracy of fault
because a small error in fault location estimation may similar location estimation. There are many algorithms to estimate
to several kilometers and this would delay the time for the fault location have been developed using one terminal data
maintenance peoples to locate the actual fault location and only. Q. Zhang et al. (1999) designed an algorithm to estimate
make appropriate repair. fault location of a SLG fault on a two paralleled transmission
line [13]. The algorithm uses the faulted phase circuit and zero
Most of the faults are unbalanced faults. From Tenaga sequence circuit for fault location estimation and doesn’t
Nasional Berhad’s (TNB) five year tripping statistics (2001 to require the source impedance data from remote end substation.
February 14, 2006), more than 90% of the tripping was single The technique assumed that, in general, the load is not
line to ground (SLG) fault [1]. For SLG fault, the fault contained in the zero sequence circuit of the faulted phase.
resistance is the combination of arc resistance, tower Thus, the effect of load disturbance on the accuracy of fault
resistance and tower footing resistance (TFR). location can be eliminated. The technique shows accurate fault
location estimation with less than 1 % error when the
Many methods have been used by other researches to parameters such as source impedance of substations, load
estimate fault location at the transmission line. Fault locating capacity and fault resistance values is changing. However, the
techniques are classified into two categories depending on technique assumed accurate line parameters and sampling
their basic essence [2]. The categories are travelling wave data. Inaccurate line parameters and sampling data produced
method and impedance based method. Besides that, fault about 10 % error of fault location estimation.
locating technique also can be classified into one terminal or
two terminal data algorithms. T. Adu (2001) developed a technique which is based on
reactance method to estimate fault location using one terminal
data [14]. The technique compensates the effects of load flow

978-1-4673-0662-1/12/$31.00 ©2012 IEEE

Authorized licensed use limited to: ExxonMobil. Downloaded on 246
April 05,2022 at 03:56:57 UTC from IEEE Xplore. Restrictions apply.
 2012 IEEE International Power Engineering and Optimization Conference (PEOCO2012), Melaka, Malaysia: 6-7 June 2012

and fault resistance on the accuracy of fault location In this paper, the author presents a new two-terminal
estimation. Fault location is estimated by rearranging the fault algorithm to locate the SLG fault at a transmission line. The
distance from the equation of bus voltage for the faulted phase algorithm does not require the knowledge of source
and only the imaginary part of the equation is required. The impedances and the estimated fault location is not influenced
technique has been tested with actual fault data that have been by fault resistance value. By knowing the fault location, the
reported and the results prove the accuracy of the technique. fault resistance can be calculated directly which aid in
determining the nature of the fault occurrence.
Another one terminal data algorithm is by Carlos Eduardo
de Morais Pereira and Luiz Cera Zanetta (2004). The authors
used fault location calculation based on steady state measured
II. THEORY OF TWO-TERMINAL ALGORITHM
phasors from one substation terminal [15]. The data required
for the algorithm are line length, line series impedance matrix,
line shunt admittance matrix, local and remote equivalent
admittance matrix, equivalent admittance matrix, local
terminal post fault and pre fault three phase voltages and local
terminal pre fault three phase currents. The most significant
advantage of this technique is it doesn’t require post fault
current values which might be inaccurate due to current
transformer saturation.
Fig. 1. Ground Fault Condition
Two terminal data algorithm was proposed to improve the
impedance based method that use one terminal data. This Fig. 1 shows the SLG fault condition at a phase line. The
algorithm uses voltage and current phasors from local and fault location is at m distance from substation A and (1-m)
remote end terminal either the data are synchronized [17]-[19] distance from substation B where m is in per unit (p.u.). Fault
or unsynchronized [20]-[25]. In term of accuracy, two resistance, RF, fault voltage, VF and m distance are unknown
terminal data algorithm are more accurate from one terminal parameters. When a fault occurs at any point at the
data algorithm because the algorithm is not affected by the transmission line, currents from both substations will flow
fault resistance value and reactance effect. This algorithm toward the fault point and return back to the substations
require communication channel to transfer the data between through the ground path. So, fault current, IF is the
both ends or to a main substation. E. Schweitzer (1982) combination of current from both substations.
utilized the pre fault phasors where the pre fault phasors are
compared to determine the synchronization error between the The representation of SLG fault at a phase line can be
two ends [20]. Using the synchronization error, the fault extended to its’ sequence components as seen from both
location is calculated in direct manner. However, this substations. Fig. 2(a) and Fig. 2(b) show the equivalent
technique requires the user to manually specify the pre fault sequence network connection seen from substation A and
data, compute the synchronization error and finally calculate substation B respectively.
the fault location. Besides that, this technique is not
appropriate for fault without pre fault load. M. Sachdev and R.
Agarwal (1988) used the post fault phasors in their algorithm
[21]. Cartesian coordinates are used to solve the voltage drop
equations which yields a quadratic equation for the unknown
synchronization error. However, the algorithm requires fault
type identification to compute the fault location.

An example of algorithm that uses synchronized phasors

was suggested by D. J. Lawrence et al. (1992). ABCD
parameters are used to define the transmission line and derive
the equation for fault location [17]. Different equations are Fig. 2(a). Sequence network connection seen from substation A
derived based on fault types and the distance to fault location
is derived from Newton’s technique for non-linear equations.

Two-terminal fault location algorithms only require low

speed communication to transmit the data between both
substations or between the substations and utility’s office. The
fault estimation can be done online or offline. For offline
method, the waveforms of post fault voltages and currents can
be extracted from fault recorders or numerical relays. Two-
terminal fault location algorithms are expected to expand in
the future because they are more accurate than one-terminal Fig. 2(b). Sequence network connection seen from substation B
algorithms.

Authorized licensed use limited to: ExxonMobil. Downloaded on 247

April 05,2022 at 03:56:57 UTC from IEEE Xplore. Restrictions apply.
 2012 IEEE International Power Engineering and Optimization Conference (PEOCO2012), Melaka, Malaysia: 6-7 June 2012

Where; III. MODELING TRANSMISSION LINE USING MATLAB/SIMULINK

m = 1, 2, .., line length (in per unit) Matlab is powerful analysis software which has the
Z1 = Positive sequence impedance capability of simulating power system flow using Simulink
Z2 = Negative sequence impedance package. In the package, it has a toolbox called
Z0 = Zero sequence impedance SimPowerSystems block set which consist of many power
VA = Phase to ground voltage seen from substation A system components that can be used by just click and drag into
VB = Phase to ground voltage seen from substation B the main file.
VA1 = Positive sequence component of fault voltage
VA2 = Negative sequence component of fault voltage The transmission line can be modeled using available
VA0 = Zero sequence component of fault voltage Distributed Parameter Line or Three Phase Pi Section Line
VF = Phase to ground fault voltage blocks. By using any of these blocks, the user should only
IA0 = Zero sequence component of phase current from enter the values of line parameter. Fig. 3 shows the
substation A transmission line which has been modeled using
IB0 = Zero sequence component of phase current from Matlab/Simulink. All the line parameters used are listed in
substation B Table I. It should be noted that the line modeled is short line,
ZF = Fault impedance therefore the effect of charging current was neglected.
However, for medium and long transmission line, the charging
Phase to ground fault voltage, VF seen from both substations current must be considered in fault location calculation.
can be written as;
TABLE I
VF = VA-m(Z1+Z2+Z0) (1) TRANSMISSION LINE PARAMETERS
VF = VB-(1-m)(Z1+Z2+Z0) (2) Line Parameters Value Unit
Length 47 km
Zero sequence components of phase currents from both Nominal Frequency 50 Hz
substations are equal to; Phase to phase Voltage 132,000 Volt
Positive Sequence Resistance 0.045531917 Ω/km
IA0 = IA/3 (3) Zero Sequence Resistance 0.151489359 Ω/km
IB0 = IB/3 (4) Positive Sequence Inductance 0.000617657 Henry/km
Zero Sequence Inductance 0.001533983 Henry/km
Equalizing (1) and (2);
During fault, there are harmonic components which are
VA-m(Z1+Z2+Z0) = VB-(1-m)(Z1+Z2+Z0) (5) added into the fundamental components of voltage and current
waveforms. These harmonic components must be filtered out
Arranging for m; from the waveforms by using simple low pass filter. Inside
I
VA -VB + B 3 (Z1 +Z2 +Z0 ) Simulink, the filtration is done by using the available Analog
m= (IA +IB ) (6) Filter Design block. The passband edge frequency selected is
3
(Z1 +Z2 +Z0 )
628.3185 rad/s.
Equation (6) is in the complex form. To get the magnitude
of m, the values of magnitude and phase angle of VA, VB, IA, Fig. 4 shows the completed model combining the
IB, Z1, Z2 and Z0 must be available first. Magnitudes and transmission line, filtration, Fast Fourier Transformation
(FFT) and fault location calculation blocks. The FFT block
phase angles of Z1, Z2 and Z0 can be taken directly from the
functioned to convert the sampled voltage and current
settings available in the protection relay while the magnitudes
and phase angles of VA, VB, IA and IB which represent the post waveforms into magnitude and phase angle at every samples.
fault values can be extracted from the fault recorders or
numerical protection relays.

Fig. 3. Transmission line modeled using Matlab/Simulink

Authorized licensed use limited to: ExxonMobil. Downloaded on 248

April 05,2022 at 03:56:57 UTC from IEEE Xplore. Restrictions apply.
 2012 IEEE International Power Engineering and Optimization Conference (PEOCO2012), Melaka, Malaysia: 6-7 June 2012

Fig. 44. Transmission line model with fault location calculation

IV. SIMULATION RESULT TABLE III

FAULT LOCATION ERRORS FOR RF = 50 Ω
This section presents the analysis of fault location Actual Fault Estimated Fault
F
Error (% )
estimations by the developed algorithm. S SLG faults were location Location
n
initiated along the transmission line. The errrors between the 5 5.882 1.877
estimated fault locations and actual faultt locations were 10 10.63 1.340
calculated using (7). 15 15.4 0.851
20 20.16 0.340
25 24.93 -0.149
Estimated location − Actual loccation 30 29.69 -0.660
Error (%) = x100
Line Length 35 34.47 -1.128
(7) 40 39.26 -1.574
45 44.03 -2.064
Table II and Table III show the results of fault location
estimation for faults along the transmissionn line with fault
resistances, RF equal to 5 Ω and 50 Ω respectively. By
comparing both tables, it proves that thhe fault location
estimation by using the algorithm is not suubjected by fault
resistance value. Fig. 5 shows the plotted errrors for both fault
resistance values.
TABLE II
FAULT LOCATION ERRORS FOR RF = 5 Ω
Actual Fault Estimated Fault
E
Error (% )
location Location
5 5.756 1.609
10 10.55 1.170
15 15.34 0.723
20 20.14 0.298
25 24.94 -0.128
30 29.74 -0.553
35 34.53 -1.000
40 39.3 -1.489 Fig. 5. Fault location errors for different fault resistances and locatio
45 44.15 -1.809

Authorized licensed use limited to: ExxonMobil. Downloaded on 249

April 05,2022 at 03:56:57 UTC from IEEE Xplore. Restrictions apply.
 2012 IEEE International Power Engineering and Optimization Conference (PEOCO2012), Melaka, Malaysia: 6-7 June 2012

[14] T. Adu. “A New Transmission Line Fault Locating System”, IEEE

Transactions on Power Delivery, Vol. 16, No. 4, October 2001.
[15] C. E. M. Pereira, L. C. Zanetta and Jr. “Fault Location in Transmission
V. CONCLUSION Line Using One-Terminal Postfault Voltage Data”, IEEE Transactions
on Power Delivery, Vol. 19, No. 2, April 2004.
[16] S. Guobing, S. Jiale and G. Yaozhong. “An Accurate Fault Location
The algorithm proposed to estimate fault location for SLG Algorithm for Parallel Transmission Line Using One-Terminal Data”,
fault in this paper presents quite accurate result. The algorithm Electrical Power and Energy System, 31.124-129; 2009.
is not subject to fault resistance value. Fault location [17] D. J. Lawrence, L. Cabeza and L. Hochberg. “Development of an
Advanced Transmission Line Fault Location System Part I1 - Algorithm
estimation can be carried out offline at the utility’s office or
Development and Simulation”, IEEE Transaction on Power Delivery,
main substations where there are facilities to extract fault Vol. 7, No. 4, October 1992, pp. 1972-1983.
waveforms from fault/disturbance recorders or numerical [18] A. T. Johns and S. Jamali. “Accurate fault location technique for power
protection relays using available communication channels. transmission lines”, IEE Proceedings, vol. 137, no. 6, pp. 395-402, Nov.
1990.
Some advantages of the algorithm are listed below: [19] Joe-Air Jiang, Jun-Zhe Yang, Ying-Hong Lin, Chih-Wen Liu and Jih-
i. Fault resistance and in feed from both substations do Chen Ma. “An adaptive PMU based fault detection/location technique
not affect fault location estimation. for transmission lines part I: theory and algorithms”, IEEE Transactions
on Power Delivery, vol. 15, no. 2, pp. 486-493, Apr. 2000.
ii. The different source impedance values will not [20] E. Schweilzer. “Evaluation and Development of Transmission Line
influence the fault location estimation and the data is Fault-Locating Techniques Which Use Sinusoidal Steady-State
Information”, Ninth Annual Western Protective Relay Conference,
not required. Spokane, Washington, October 1982.
iii. By knowing the value of m, fault resistance can be [21] M. Sachdev and R. Agarwal. “A Technique for Estimating Transmission
Line Fault Locations From Digital Impedance Relay Measurements”,
calculated directly by using (1) or (2). IEEE Transaction on Bower Delivery, Vol. 3, No. 1, January 1988,
pp.121-129.
VI. REFERENCES [22] A. A. Girgis, D. G. Hart and W. L. Peterson. “A new fault location
technique for two- and three-terminal lines”, IEEE Transactions on
Power Delivery, vol. 7, no. 7, pp. 98-107, Jan. 1992.
[1] A. M. Zin, S. P. A. Karim, “The Application of Fault Signature Analysis [23] D. Novosel, D. G. Hart, E. Udren and J. Garitty. “Unsynchronized Two-
in Tenaga Nasional Berhad Malaysia”, IEEE Transactions on Power Terminal Fault Location Estimation”, IEEE Transaction on Power
Delivery, Vol. 22, No. 4, October 2007. Delivery, Vol. 11, No. 1, January 1996.
[2] Y. Cheng, J. Suonan, G. Song and X. Kang. “One- Terminal [24] I. Zamora, J. F. Miambres, A. J. Mazn, R. Alvarez-Isasi and J. Lazaro.
Impedance Fault Location Algorithm for Single Phase to Earth Fault of “Fault location on two-terminal transmission lines based on voltages”,
Transmission Line”, IEEE 2010. IEE Proceedings-Generation, Transmission and Distribution, vol. 143,
[3] H. W. Dommel and J. M. Michels. “High Speed Relaying Using no. 1, pp. 1-6, Jan. 1996.
Travelling Wave Transient Analysis”, IEEE PES Winter Power [25] P. Balcerek and J. Izykowski. “Improved unsynchronized two-end
Meeting, New York, Jan. 29-Feb. 3, 1978, pp.1-7. algorithm for locating faults in power transmission lines”, Proc. IEEE
[4] M. Aurangzeb, P. A. Crossley and P. Gale. “Fault location on a Bologna PowerTech Conference, Bologna, June 2003.
transmission line using high frequency travelling waves measured at a
single line end”, PES Winter Power Meeting, 2000, IEEE, Vol. 4, 2000,
pp.2437-2442.
[5] P. F. Gale, J. Stokoe and P. A. Crossley. “Practical experience with VII. BIOGRAPHIES
travelling system”, Development in power system protection. 25-27th
March 1997, Conference Pub. No.434, pp.192-196.
[6] P. F. Gale, P. V. Taylor, P. Naidoo, C. Hitchin and D. Clowes.
“Travelling wave fault locator experience on Eskom’s transmission Muhd Hafizi Idris was born in Alor Star,
network”, Developments in Power System Protection, Seventh Kedah, Malaysia on November 4, 1983. He
International Conference on (IEE), Published 2001, pp.327-330. received the B.Eng degree in Electrical and
[7] Z. Q. Bo, A. T. Johns and R. K. Aggarwal. “A novel fault locator based Electronic Engineering from Universiti
on the detection of fault generated high frequency transients”, Kebangsaan Malaysia (UKM) in 2006 and
Developments in Power System Protection, Sixth International M.Eng Electrical (Power) from Universiti
Conference on (Conf. Publ. No. 434), 1997, pp.197-200. Teknologi Malaysia (UTM) in 2011. His fields
[8] M. A. Street, I. P. Thurein and K. E. Martin. “Positioning System of interest including power system protection,
applications at the Bonneville Power Administration”, Northcon 95 fault location, stability and reliability. He joined
IEEE Technical Applications Conference and Workshops Northcon 95, IEEE in June 2011. Currently, he is a lecturer at
1995, pp.244-251. School of Electrical System Engineering,
[9] T. Kawady and J. Stenzel. “Investigation of practical problems for Universiti Malaysia Perlis (UniMap).
digital fault location algorithms based on EMTP simulation”, Asia-
Pacific Transmission and Distribution Conference and Exhibition, 2002,
Vol. 1, pp.118-123. Mohd Wazir Mustafa received his B.Eng
[10] L. Eriksson, M. M. Saha and G. D. Rockefeller. “An accurate fault degree (1988), M.Sc (1993) and PhD (1997)
locator with compensation for apparent reactance in the fault resistance from University of Strathclyde. His research
resulting from remote-end infeed”, IEEE Trans. Power App. Syst., interest includes power system stability,
vol.PAS-104, no. 2, pp. 424-436, Feb. 1985. FACTS, wireless power transmission and
[11] Ge Yaozhong. “New types of protective relaying and fault location: their power system distribution automation. He is
theory and techniques”, Xi’an: Xi’an Jiaotong University Press, 2007. currently an Associate Professor at Faculty of
[12] K. Takagi, Y. Yomakoshi, M. Yamaura, R. Kondow and T. Matsushima. Electrical Engineering, University Teknologi
“Development of a new type fault locator using the one-terminal voltage Malaysia. Dr. Mustafa is also a member of
and current data”, IEEE Trans. Power App. Syst., vol. PAS-101, no. 8, Institution of Engineers, Malaysia (IEM) and a
pp. 2892-2898, Aug. 1982. member of IEEE.
[13] Q. Zhang, Y. Zhang, W. Song and Y. Yu. “Transmission Line Fault
Location for Phase-to-Earth Using One Terminal Data”, IEE Proc.-
Gener. Transm. Distrib., Vol. 146, No. 2. March 1999.

Authorized licensed use limited to: ExxonMobil. Downloaded on 250

April 05,2022 at 03:56:57 UTC from IEEE Xplore. Restrictions apply.
2012 IEEE International Power Engineering and Optimization Conference (PEOCO2012), Melaka, Malaysia: 6-7 June 2012

Yazhar Yatim was born in Dungun,

Terengganu, Malaysia on August 29, 1982. He
received his B.Eng inn Electrical System
Engineering from Univeersiti Malaysia Perlis
(UniMAP) in 2007 and M Msc. Power Enginering
from Brandenburg Univeersity of Technology,
Cottbus, Germany in 20100. His research interest
in renewable energy, powwer system, protection
system and energy managgement. Currently, he
is a lecturer at School oof Electrical System
Engineering, Universitii Malaysia Perlis
(UniMAP).

Authorized licensed use limited to: ExxonMobil. Downloaded on 251

April 05,2022 at 03:56:57 UTC from IEEE Xplore. Restrictions apply.