2021 IEEE PES/IAS PowerAfrica

Impact of designed line parameters on Distance

protection - Investigation on a 110 kV Double
Circuit Transmission Line Fault
Anas Abdulkhader,
Regional Application Specialist
OMICRON electronics middle East
Manama, Bahrain
anas.abdulkhader@omicronenergy.com

Abstract - Measuring the impedances in overhead lines simplification steps are taken which do not apply in the
and cables is absolutely essential for ensuring selectivity real world. Therefore, the results are often not precise
and accurate fault location by means of distance relays. enough.
This paper investigates a double-circuit line fault with non- In particular, line-to-ground impedances, and therefore
selective tripping caused by the parameterization of the
relay based on inaccurate/lack of knowledge of the line
Z0, are influenced by the physical characteristics that are
impedances. Analysis of the fault is based on the measured present, such as metal pipes or neighbouring cables in
line impedances and the simulation/system-based testing the ground. This is why they are not sufficiently accurate
using RelaySimTest. An appropriate procedure was when calculated. Furthermore, it is assumed that the soil
adopted and documented to measure the line impedances of has uniform resistivity which is usually not the case due
double circuit lines in a minimally invasive manner. to different soil properties in different layers of soil.
However, a measurement reflects all factors and obtains
Keywords - Double circuit lines, mutual coupling, true values.
distance protection, line impedance measurement,
Therefore, engineers started to measure line impedances
selectivity, Relay SimTest, system-based testing
to verify and optimize distance relay parameterization
I. INTRODUCTION
Distance protection and impedance-based fault location III. FAULT DESCRIPTION
on overhead lines and power cables is being used at
utilities all over the world. The Z1 positive-sequence
impedance and Z0 zero-sequence impedance are
significant characteristics of the protected line which
must be understood precisely in order to make the relay
decide whether to trip or not to trip with reliability.
Sending out the line crew requires accurate knowledge of
the fault location in order to keep the outage short.
Therefore, fault locators that also depend on accurate line
parameters are being used. Fig 1: Topology of the double circuit line
The distance protection relay, discussed in this paper,
tripped by a ground fault non-selectively, which was The double-circuit line discussed in this paper consists of
parameterized solely based on estimated Z1 and Z0 two identical electric circuits “Line 1” and “Line 2”
values. However, as the mutual coupling impedance Z0M (Figure 1) and connects the two busbars A and B. It is part
of this double circuit line is significant, it must also be of a solidly grounded urban distribution network with a
considered. nominal voltage of 110 kV.
Here we describes the details of the fault, how to measure Busbar A is a gas-insulated switchgear with a cable run
the line impedances Z1, Z0, and Z0M using the of 160 m to the overhead line gantry. Busbar B opposite
conventional method and compares with the simulation is a cable section 1.1 km in length. The over-head lines
carried out in RelaySimTest with the fault recording. are located on the same poles, which explains why a
Finally describes the minimally invasive measurement of significant zero-sequence coupling impedance Z0M
Z1, Z0, and Z0M as an alternative to the conventional exists.
measurement-method.

II. MEASUREMENTS OR CALCULATIONS?

Line parameters can either be measured or calculated,
based on the geometry of the conductors and soil
properties. When calculating line impedances, multiple Fig 2: Switchgear State 1: Δt1 = 53ms

978-1-6654-0311-5/21/$31.00 ©2021 IEEE

 The fault shown in Figure 2 occurred on the cable of the through the incorrect setting of the impedance-related
overhead line gantry on phase L1 of line 2. It was caused parameters in zone 1 and tripped instantaneously. At that
by sawing of the cable following unauthorized access to moment three lines were therefore disconnected instead
the overhead line gantry. The sole infeed of the fault was of one.
busbar A via three 220kV/110kV transformers. One of
these three transformers was destroyed by the fault, since
it was not designed to withstand the fault current. IV. MULTI-FUNCTIONAL TEST SYSTEM
Most of the fault current initially flowed directly via the For the measurement, CPC 100 as a multifunctional main
feeder of line 2. Only a small portion flowed via line 1 unit which generates frequency variable test signals and
and busbar B. The distance and differential protection of measures current and voltage by applying digital filtering
CB1 and CB2 tripped correctly. CB1 was the first to open, for effective noise suppression is used. This main unit was
53ms after fault inception. combined with the CP CU1 accessory, a coupling unit
which ensures galvanic isolation between the line under
test and the main unit. Additionally, the CP GB1
Grounding Box was used which comes with surge
arrestors capable of diverting high fault currents in case
of unexpected over-voltages from the line side. CPC 100
+ CP CU1 test system can also use to verify the
effectiveness of its ground grids by means of measuring
the Ground Potential Rise, Step & Touch Potentials and
Fig 3: Switching state 2: Δt2=20ms Point-to-Point resistances

The fault is now fed from line 1 and busbar B. CB2 has
not yet opened, as it’s trip time is little longer than that of V. PRECISE MEASUREMENTS
CB1. This switching state lasted 20ms, or one cycle at 50 When measuring Z1, Z0, and Z0M, both circuits of the
Hz. double-circuit line was de-energized at the same time so
Switching state 2 resulted in the distance relay of CB3, a that the official test template for the measuring device CP
Siemens 7SA513, detecting the fault in zone 1 and tripped CU1 could be used. In order to determine the Z1 positive-
immediately. sequence impedance, the Z0 zero-sequence impedance
and Z0M mutual coupling impedance, eight loop
impedances are measured. This includes all three phase-
to-phase loops, all three phase-to-ground loops and one
loop where all three phases are shortened when the
system II is grounded and one loop where all three phases
are shortened when the system II is open. The test took
less than one hour. Figure 6 is the overall setup.
Fig:4 Switching state 3:Δt3 = 50ms

Fig:5 Switching state 4

Once CB2 is open, as shown in Figure 4, CB3 opens

shortly afterwards (Figure 5) as a reaction to the incorrect
trip command in switching state 2. As a result, the load at
busbar B was not supplied anymore.
Although outside of the scope of this paper, it is worth
mentioning that when attempting to reconnect the load on L2
busbar B using one of the two lines, the faulty line 2 was L1
L3
connected. Before this connection was made, the load
flow was optimized and adapted to the new grid E
conditions. As a consequence of the new in-feed
configuration, an additional outgoing line from busbar A
(to another busbar C) was disconnected, as the distance Fig:6 Set up for the line impedance measurement
protection on busbar C of this line had detected the fault
 geometry of the conductor arrangement. However, quite
L1-L2 high deviations has been observed for the zero-sequence
impedance. The X value of Z0, which is the more crucial
component aside from the R value, was almost 90%
higher than the estimated design values. The difference
can be due to the inaccurate knowledge on soil
properties. During the calculation of Z0, uniform soil is
Fig 7: Phase to phase and Phase to ground measurement considered, and a typical value of soil resistivity is used.
But soil resistivity can be different over the run of the
line, can be different in different depths. Other buried
conductors such as pipes or shields can have a huge
impact on the current return path. A proper measurement
provides an accurate impedance value, because the
current distribution during testing will be same as in the
event of fault, with a lower amplitude.
Here, the fault is positive, which tends to result in
overreaching protection. There was no estimate of the
Fig 8: Three phase to ground measurement coupling impedance to compare with the measured
value.
During the measurements, earth switch of the near end is
opened. While changing the connection on GB1, for VI. SIMULATIONS IN RELAYSIM TEST
every measurement, earth switch must be closed for
safety reason. Figure 7 shows the six measurement, 3 line The simulation of the voltages, currents, and impedances
to line and 3 line to ground with system II grounded. that occurred during the fault, and that are required for
Figure 8 shows the connection for mutual coupling analysis purposes, was carried out using the Relay-
impedance with system II grounded and system II open. SimTest software. First, the double-circuit line with
single sided infeed was entered in the software, see
The measurement took place at the overhead line gantry Figure 9.
of busbar A – the line on busbar B was grounded. The
overhead line and the cable on busbar B were thus
considered for the measurement. The short cable section
from the overhead line gantry to the switchgear of busbar
A was ignored. Table 1 compares the measured values
for Z1, Z0 and Z0M which were derived from the
measurement of the eight loop impedances with the
design values based on the conductor geometry and soil
resistivity. The design values were available with the Fig.9: Line design in RelaySim Test
sub-station protection team.
The double-circuit line depicted in Figure 9 contains the
Table 1: Comparison of measured and design values 3 sections of each of the two circuits, which were
parameterized as follows:
Z1 Z0 Z0M
• Busbar A cable
Measured R (Ω) 0.849 2.131 1.144
o Z1 and Z0 are identical to the values of the
value
busbar B cable
X (Ω) 2.776 9.132 5.779
• Overhead line
Design value R (Ω) 0.940 3.07 NA o Z1 and Z0 represent 96% of the measured
from values
substation X (Ω) 2.780 17.2 NA o Z0M corresponds to the measured value

Error (ref. to R (Ω) 10.85% 44.71% NA • Busbar B cable

measured o Z1 and Z0 represent 4% of the measured values
value ) X (Ω) 0.13% 88.29% NA
The 96%:4% split of the measured impedances assumes
that Z1 and Z0 of an overhead line are 4 times greater-
than the impedances of a cable. The fact that cable
As expected, the deviation of the positive-sequence impedances have a smaller angle and it is ignored in this
impedance is negligible as the positive-sequence instance. A further constraint is that the fault is fed
impedance can be accurately calculated from the exclusively from busbar A.
Figure 9 shows the simulation of the fault (L1-G) at the
actual fault location (overhead line gantry = 200%). State
2 (from 53ms to 73ms) is studied in more detail as below,
since the relay is mis-operated as a consequence of this
state. As this state only lasts 20ms, the time do-main
depiction for voltage and current was used for comparing
the simulation and fault recording, as a steady-state
impedance does not occur owing to the short duration of
state 2.

Fig 10: Fault inception angle

The inception angle of the fault has a major impact on Fig 13: Simulation of the fault (adapted to fault
the transient response of the fault current. It must be read recording) voltage vs time
from the fault recording as accurately as possible. Figure
10 shows the input of the fault inception angle in
RelaySimTest; in this case the angle is 204°. VII. MINIMALLY INVASIVE MEASUREMENT OF THE
LINE IMPEDANCE
A procedure proposed at the OMICRON user meeting,
2017 in Friedrichshafen allows the zero-sequence
impedance Z0 and the coupling impedance Z0M of a
double-circuit line with only one line taken out of service
to be determined by measurement. [2] The simultaneous
disconnection of two coupled electric circuits is difficult
to arrange once the line has been commissioned.
However, the conventional method of measuring a
Fig 11: Source impedances double-circuit line requires simultaneous de-energization,
which is why the alternative, minimally invasive
The internal impedances Z1s and Z0s of the source procedure for the retrospective measurement of double-
determine the amplitude of the voltage and current. To circuit lines is of such interest.
plot the simulated current as accurately as possible Table 2 presents the results of both measurements. As
against the actual fault current, the parameters shown in expected, the deviation in respect of Z1 is negligible. The
Figure 11 were determined through trial and error. As deviation of less than 5% in the case of Z0 is still within
can be seen from Figure 12 and 13, the actual fault acceptable limits, whereas the deviation of more than
current can be simulated very precisely. Similarly, the 13% for Z0M requires further analysis.
simulated voltage closely matches the voltage from the
fault recording. The close match between the simulated Table 2: results of the minimally invasive line
values and those from the fault recording indicates that impedance measurement
the measured line impedances (see Table 1) are
Z1 Z0 Z0M
extremely accurate.
Measured R (Ω) 0.849 2.131 1.144
value
X (Ω) 2.776 9.132 5.779

Minimally R (Ω) 0.863 2.200 1.25

invasive
X (Ω) 2.776 8.690 5.01

Error in % R (Ω) 1.65% 3.24% 9.27%

X (Ω) 0% -4.84% -13.3%

The accuracy of the procedure depends on two variables:

Fig 12: Simulation of the fault (adapted to fault • Current Ip in the in-service circuit and the
recording) current vs time derived current factor fsp.
 [2] Pikisch. M, “The significance of mutual coupling in the line
model”, OMICRON User Meeting 2017; Friedrichshafen.,
• Auxiliary impedance Publication 2017
[3] U. Klapper, "Reliability of Transmission by Means of Line
Impedance and k-Factor Measurement", 18th International
Conference and Exhibition on Electricity Distribution, 6-9 June
2005, Turin - Italy
[4] Konzelmann, S.; Pikisch, M.: Measuring the impedance on
double-circuit lines with the parallel system in operation.
OMICRON User Meeting; Berlin, Publication 2018
[5] S. Kaiser, "Different Representation of the Earth Impedance
Matching in Distance Protection Relays", Proceedings
OMICRON User Conference in Germany, Publication 2014.
[6] Paulino M.E.C; Gutierrez,M.; Carvalheira, E. "Measurement of
Transmission Line Impedance, Ground Factor (K) Determination
and Improved Reliability of Distance Protection”, CIGRE study
committee B5 August 2013.
Fig 14: measurement of Ip with 4 Rogowski coils

The measurement of Ip was carried out on the cable of

the overhead line gantry, see Figure 14. Ip is
0.5818<7.28. The errors in Table 2 are all to do with the
inaccuracy of the auxiliary impedance. When
determining the auxiliary impedance, the geometry of the
six conductors of the two circuits was available at that
instance. No further examination into the accuracy of this
data was carried out.
In the case of the fault under discussion here, the coupling
has the maximum possible effect, as the coupling
impedance has an impact along the entire length of the
line.
It can also be seen that the accuracy of Z0 is less
dependent on the auxiliary impedance than Z0M. Despite
all the above, an attempt should be made to estimate the
auxiliary impedance as accurately as possible.

VIII. CONCLUSION
The measurement demonstrates the lack of reliability of
distance protection relays if inaccurate line impedances
have been used to parameterize the relays. Inaccuracies in
the design value of Z0 is due to unknown properties of
soil. A proper measurement provides an accurate
impedance value, since the current distribution during
testing will be same as in the event of fault, with a lower
amplitude. Accurate line impedances will increase the
precision of calculating fault distances from stations in all
types of scenarios. This paper demonstrates that by
measuring Z1, Z0, and Z0M and carrying out a simulation
in RelaySimTest, the currents and voltages associated
with a fault can be simulated accurately. OMICRON also
suggests a minimally invasive measurement of Z1, Z0,
and Z0M with CP CU1. Minimally invasive means that
only one circuit must be de-energized. With this method
also, we can produce good results.
36

REFERENCES

[1] Luxenburger, R.; de Villiers, W “ Calculation and verification of

distance protection settings based on line impedance
measurements” OMICRON User Meeting; Germany
Publication 2015