http://dx.doi.org/10.5755/j01.eee.19.9.5645 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392-1215, VOL. 19, NO.

9, 2013

Transmission Line Protection and Fault

Location Based on Travelling Wave
Measurement
V. Siozinys1, R. Urniezius2
1
Department of Electric Power Systems, Kaunas University of Technology,
Studentu St. 48, LT–51367, Kaunas, Lithuania
2
Department of Control Technologies, Kaunas University of Technology,
Studentu St. 48, LT–51367, Kaunas, Lithuania
[email protected]


Abstract—Last decade power systems have faced blackouts. involved to investigate principally new generation fast fault
Also smart grid concepts are developing to increase overall identification methods. Such models should describe many
system performance. The need and the challenge is high phenomenon which had no necessity to be analysed in
accuracy distance to the fault measurement algorithm in
classical relay protection concept. The electromagnetic
transmission line protection. The article describes differential
travelling wave protection also as distance to the fault transient in network consists of faster or slower current and
measurement principles. The distance to the fault calculation voltage components. Complete and distinct transient process
deviations are investigated, also as classified as the stochastic view could be yielded only with many electrical unit
and systematic. The systematic deviation elimination methods components. Such task could be covered only with fast
are described. electromagnetic transient.
The need of use and complement of existing power
Index Terms—Electromagnetic transient, power system
network special models arise for the investigation of fast
protection.
fault identification models and algorithms.
I. INTRODUCTION
II. OVERVIEW OF THE METHODS OF FAST FAULT
The electrical systems global problem of last ten years is IDENTIFICATION IN TRANSMISSION NETWORK
electric system blackout. Such phenomenon is described as
part or all electric system switches off as the consequences Couple of methods are known [1]–[7] for the fast fault
of electric network equipment failures or operational staff identification in transmission network. However, only few of
errors. them could be considered as practically applicable for the
Smart Grid ideas are involved to electric power grid to protection and fault location.
increase the operational and energy supply reliability, First method applicable for the protection scheme uses an
control flexibility of the grid, also as to increase the electric idea of multiple registration of travelling wave propagated
energy quality and improve relationship between electric from fault [8]. The travelling waves reflect from substation
energy producer and consumer. The concept of Smart Grid busses due to significant difference of wave impedance. The
involves wide variety innovative technical decisions. One of line impedance is much higher than bus bar. Reflected
the Smart Grid concepts task is to increase the electricity travelling wave from bus bar propagate towards fault and
network reliability. Electric network reliability could be reflect as well, due to difference of wave impedance.
described as the probability of failure. The failure power Propagated towards bus bar travelling wave repeatedly are
network part immediate shut down increase the electric recorded. After that the distance from the buses to the fault
network reliability. could be calculated. The distance would be proportional to
New generation technologies instead of existing ones difference between time stamps of nearby registrations. In
should be developed and installed for the faster failures practice the part of travelling wave refract, dissipate due to
identification and clearness to prevent the electrical system line losses. On other hand if the fault is nearby the busses,
from blackouts. The faster disconnection of monitored part the reflected from bus bar wave propagate to fault.
of system also as faster regulating systems start-up could be The part of wave has reflected and the refracted part of
initiated as the element of fault clearness conception in the wave propagates towards other busses. Two travelling waves
transmission network. propagated towards busses are registered. The second wave
Difficult electric power network models should be is refracted at the fault, so if the line length is high
significant error could be encountered.
Registrations of propagated wave are performed in the
both side of line for the second method, as depicted in
Manuscript received February 4, 2013; accepted May 8, 2013. Fig. 1. The time stamps are fixed at the travelling wave

21
 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392-1215, VOL. 19, NO. 9, 2013

   
arrival moments. time difference between registered and exact time moments:

 t1 U trigger  tx   tx  t0 . The deviation t1 U trigger

l
A B c

minimum value at l  xexact  t0  c as depicted in Fig. 2.

CT CT is not constant for all travelling wave rise time and reach's

Travelling waves are measured at the both line ends as it

CC
RPD RPD is depicted in Fig. 1. Then the exact distance to the fault
T could be expressed
VT VT

l  t  c
Fig. 1. The placement of devices in the network. CT – conventional
current transformer; VT – voltage transformer; RPD – relay protection xexact  , (1)
device; CC – communication channel. 2
Then the distance to the fault would be proportional to where t  t0A  t0B  the exact time moments for each
difference of time value. The error of the method depends on
c  295m/s - speed of light; l  exact distance between
travelling wave travelled to substation A and B;
precision of relay internal timer synchronization and known
line length [9]. The fast and separated communication
channel between relay protection devices could be used for substations A and B.
method realization. Exact distance to the fault is calculated using the
Special requirements for the communication channel measured travelling wave time moments tx for each
create inconvenience for the second method implementation. substation A and B

l   txA   t1A  txB   t1B   c

However, GPS synchronization of the timers of relay
xexact  
protection devices could be used. Timer synchronization
error for the case of worst atmospheric condition is about
l  t  c c
2
   t ,
1 μs . So, the error distance to the fault if we assume that
(2)
travelling wave propagation speed v ~ c: 2 2
l  v  t  2.95 108 1106  295 m. Due to travelling
where txA  travelling wave arrival time at the substation A;
txB  travelling wave arrival time at the substation B;
wave losses in the line, the propagation speed of wave is less

 t1A  travelling
than the speed of light. However, the precision of fault
location could not be reached more than 250 m. Besides the wave registration deviation at the
disadvantages the method has an advantage – the general substation A;  t1B  travelling wave registration deviation
purpose communication channel between devices could be
at the substation B.
used to transfer time stamps.
Distance to the fault in (1) is calculated for the normalized
measured values:
III. DISTANCE TO THE FAULT PLACE CALCULATION AND

U
max  U measured 
ERROR EVALUATION
U measured
Network structure, connected to the end of line should be , (3)

 
evaluated for the creation of fast fault identification models.
I
The fault place voltage source magnitude and rise time I measured
. (4)
variation could be neglected using travelling wave max I measured
registration at the both line ends.
Deviation  t   t1A   t1B acquire the highest values
during the fault in the beginning of the line and are

 
negligible for the fault in the middle of the line
0  t  t U trigger .

A. Substation time delay

Transmission line connects substation which configuration
quite rare match each other. The configuration mismatch
leads to different equivalent substation capacitance. The
travelling wave front sharpness is reduced by substation
capacitance. Different capacitance at the both substation A
Fig. 2. The travelling wave registration magnitude influence to the and B leads to different steepness of the travelling wave, as
distance to the fault calculation error.
depicted in Fig. 2.
Travelling wave registration moment depends on trigger Travelling wave induced voltage front time delay at the

 
level and could be described as the function substation could be evaluated using expression

tx  f U trigger . The registration deviation is described as T  Cf  Z B E , (5)

22
 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392-1215, VOL. 19, NO. 9, 2013

where Cf  substation phase capacitance; ZB E  delay changes the accuracy of the algorithm by component

T  TA  TB ,
transmission lines, connected to the substation equivalent
(7)
surge impedance.
where TA  additional travelling wave delay due to
Calculations for the 110 kV transmission lines with tower

substation with capacitance C  1nF and three transmission substation A capacitance;  TB  additional travelling wave
type P110-3 are made (Fig. 3). The network consists of

delay due to substation B capacitance.

Z B  343Ω . The equivalent transmission lines phase-phase
lines with phase-phase channel surge impedance
Travelling waves induce voltages at the substation and
channel surge impedance Z B E  114 Ω . The phase to phase
currents in the lines then reach the substation. The equation
below shows relation between currents, voltages and
voltage additional delay due to substation capacitance is travelling wave as the voltage

which yields T  1nF 114Ω  0,1143μs . U s   0,5  ZB s  I s   0,5  ZB s  ZBE
1
depicted in Fig. 4. The time delay could be found from (5),
s  s 
U , (8)
2,0

where indices (s) denotes parameters as eigenvalues.

2,1 Equation (8) for the travelling wave is valid only for the
4,0

first time moment then there is no reflected waves. Equation

requires information about the lines connected to the
2,1 4,2
substation: number of lines and their surge impedance. So,
25,0

additional information should be supplied to the protection

19,0

device.

deviation component T in case of voltage channels are

Distance to the fault measurement results would comprise

T could be neglected using current channel (6). However,

used for the fault identification. Accordingly, the deviation
Fig. 3. Tower type P110-3, distances are shown in meters.
the highest precision could be reached if the calculations are

travelling wave. In this case T  0 .

not based on currents, or on voltages, but are based on

B. Deviation due to auxiliary transformer

Auxiliary transformer due to non-ideal transient
characteristics changes travelling wave rise time, some
frequency components are amplified, some of them
suppressed. The additional deviation to the calculated
distance to the fault value is added by the auxiliary
measurement transformers which influence could be
evaluated as
 mg   mgA   mgB , (9)

where  mgA  additional travelling wave delay due to

 mgB  additional travelling

measurement transformers in substation A;
Fig. 4. The phase to phase voltage at the substation: 1 – substation
capacitance evaluated and equal to C = 1 nF ; 2 – substation capacitance
wave delay due to
not evaluated. Calculation performed for 110 kV transmission line with measurement transformers in substation B.
tower type P110-3. Line length is 3 km.
C. Deviation due to placement of auxiliary transformer
Travelling wave induced voltage front time delay mainly
Current and voltage auxiliary transformers are placed far
depend on network topology, line surge impedance and
away from registration devices in the high voltage
substation capacitance. The first wave induced voltage and
relay protection device could seek up to 100 m or  0,33μs
substation. The distance from auxiliary transformer to the
current at substation has the relation

I s   Ti1Cf Tu
d t s
d in time domain. So the total method deviation
U , (6)

 0, then  pA   pB,
p  
 p2 , then  pA   pB ,

(10)
where Ti and Tu denotes current and voltage eigenvectors;
Is  and U s  denotes current and voltage eigenvalues
where  pA  additional travelling wave delay in substation
A;  pB  additional travelling wave delay in substation B.
accordingly.
Deviations in (2) could be separated. The substation time

23
 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392-1215, VOL. 19, NO. 9, 2013

Deviation due to different distance from auxiliary Congress Intelligent Control and Automation (WCICA 2010), 2010,
pp. 25–29.
transformers to protection devices could be eliminated [6] C. Yu, L. Dong, X. Bingyin, H. Yuhui, “Wide area travelling wave
during the commissioning of the relay protection devices. fault location in the transmission network”, in Proc. Int. Conf.
Deviation values (7), (9), (10) are fixed and could be Electricity Distribution (CICED 2010), China, 2010, pp. 1–6.
eliminated during the commissioning. The influence mainly [7] V. Siozinys, “The Model of Fast Fault Identification in Transmission
Network”, Ph.D. dissertation, Lithuanian Energy Inst., Kaunas Univ.
depend on distance to the fault: higher distance leads to Technol., 2010.
lower deviation (7), (9), (10) influence to the distance to the [8] V. Siozinys, “Transmission Line Fault Distance Measurement based
on Time Difference between Travelling Wave Reflection and
The deviation  t in (3) and (4) could be written
fault measurement error.
Refraction”, Elektronika ir Elektrotechnika (Electronics and
Electrical Engineering), no. 2, pp. 25–28, 2010.
[9] P. Chen, K. Wang, “Fault Location Technology for High-Voltage

t  T  mg  p .
Overhead Lines Combined with Underground Power Cables Based on
(11) Travelling Wave Principle”, Int. Conf. Advanced Power System
Automation and Protection, 2011, pp. 748–751. [Online]. Available:
http://dx.doi.org/10.1109/APAP.2011.6180498
Deviations (7), (9), (10) could be classified as systemic,
because they could be evaluated during the commissioning
process. However the real line could have additional
parameters deviation, such as: difference in ground
resistivity along line length; the different overhead line wire
height due to relief, different distance between wire, wire
icing, the type of fault. The mentioned deviation could have
stochastic character and the algorithm would have stochastic
errors.
Differential algorithm as the protection function is
absolutely selective and is very fast, because it does not
require any special signal processing and calculation. The
same differential principle is applied for the fault place
calculation, described above.

IV. CONCLUSIONS
Differential travelling wave method is absolutely selective
transmission line protection method. The method itself
eliminate travelling wave rise at the fault place conditions,
i.e. eliminate unknown conditions such as fault impedance
during the fault.
Differential travelling wave distance to the fault
calculation method deviation is influenced by systematic and
stochastic errors. Special filters could be used to eliminate
travelling wave substation delay also as measurement
transformer shape distortion.
Distance to the fault calculation error dependability on
defined trigger level was discussed. The error would be
negligible for the protection function. However, it should be
evaluated for the distance to the fault measurement.

REFERENCES
[1] V. Siozinys, “Investigation of Possibilities of New Relay Protection
Principles Application for Electric Power Transmission Lines”, in
Proc. 18th Int. Conf. Electromagnetic Disturbances (EMD 2008),
vol. 2, 2008, pp. 33–36.
[2] H. Gao, D. Li, G. Zou, Z. Pan, “A Novel Travelling Waves Based
Ultra-High Speed Ratio Directional Protection”, in Proc. IET 9th Int.
Conf. Developments in Power System Protection (DPSP 2008),
2008, pp. 568–572.
[3] G. Zou, H. Gao, “Algorithm for ultra-high speed travelling wave
protection with accurate fault location”, in Proc. Power and Energy
Society General Meeting - Conversion and Delivery of Electrical
Energy in the 21st Century (IEEE 2008), 2008, pp. 1–5.
[4] D. Feng, Z. Xiangjun, C. Nan, L. Zewen, C. Xianghui, “Research on
Optimal Placement of Travelling Wave Fault Locators in Power
Grid”, in Proc. IET 9th Int. Conf. Developments in Power System
Protection, (DPSP 2008), 2008, pp. 573–578.
[5] T. Ji, Y. Wang, Q. Yuan, “Study on travelling waves transform
characteristics of potential transformers”, in Proc. 8th World

24