IMPACT OF DOUBLE CIRCUIT LINES OPERATED IN

PARALLEL ON DISTANCE PROTECTION

A. dos Santos

REN -Rede E16ctrica Nacional S.A.

Rua Cidade de Goa n' 4 Sacav~tm., Portugal
andre.santos~oren.pt

Keywords: Power system, protection, parallel lines, double line to connect one of the substations to a third substation as
circuits, distance protection. seen in Figure 2. The period distance to the final topology can
take several years.
Abstract Section A Uprated section B Section C
Single Double circuit line Single
Double circuit lines, operated in parallel, have become circuit line in parallel circuit line
popular in the Portuguese transmission network. When a new
line is constructed between two substations, the towers are
initially built to encompass two circuit lines even if only one
bay is constructed at each substation line ends. The two
circuits are operated in parallel having the advantage of
postpone the cost of constructing new bays in the substations
to the time where the power transfer demand requires.
The present paper describes the impact of such topology on
Figure 1. Initial network configuration, transmission line with one section
the selectivity performance of distance protection used on uprated and converted to double circuit line geometry.
transmission lines. An analytical model is presented to
evaluate the impedance seen by the distance protection for
several types of faults along the line revealing the underreach
effect. A numerical simulation using ATP/EMTP is
conducted to validate the analytical model results.
At the end, some solutions and recommendations are
proposed to mitigate the problems encountered.

I Introduction
Portugal has established the objective of producing 39% of its
electric energy by renewable power by 2010. This value
holds, not only large hydro plants, 5000MW predicted, but
Figure 2. Final network topology. Section B is shared by two transmission
also the growth of wind power generation from 127MW, in
lines.
2001, to 3750MW by 2010.
Most of the new wind power plants will be connected to the The construction of new transmission lines has become a
220kV and 150kV networks located in the interior north and major issue due to costs, environmental constraints and time
centre part of the country, which includes the need of a consumption. The difficulty in stabilising the required number
transmission network expansion to bring this increment of of line corridors has lead the company to construct mainly
electric power from those areas to the consumers along the double circuit lines when traditionally would be constructed
shore. To satisfy' this necessity, two programs are ongoing: single circuit line with a dedicated corridor. Usually the lines
the uprating of transmission lines and the construction of new were constructed in stages along the years determined by the
transmission lines. required power transfer. Today, the new double circuit lines
The uprating of transmission lines is carried out to increase are constructed with both circuits equipped even if there is no
the line power transmission capacity to a maximum allowed need in power transfer requirement. The reason for this is
conductor temperature at steady state of 850 C. Frequently the because the equipage of the second circuit on a double circuit
opportunity of the line outages state when performing the line in operation requires the line to be placed out-of-service
uprate work is used also to change the line geometry. The during a long period, which is not acceptable by the
new geometry can be a double circuit line inserted in the Transmission System Operator.
existing line section as sown in Figure 1. The purpose of this
is to be able to use the line section in the future on a second

104
Since the power transfer demand does not require the circuits and the corresponding apparent impedance seen by
construction of two substation bays in each line e rid, only one the distance protection is evaluated.
bay at each end will be constructed. As a cons equence, the Fault currents and voltages measured by the distance
line circuits are placed in parallel, Figure 3. protection are calculated solving the power system sequence
network equivalent. Finally, an EMTP model and simulations
Substation I Substation 2
are used to validate the results.

2.1 Sequence network equivalent

I In Figures 6 and 7 the positive and zero sequence equivalent
I networks of the power system shown in Figure 5 [1 ][2][3],
Figure 3 -Initial network configuration. Double circuit line operated in are presented. The two network areas (S and Q) are
parallel over two bays.
represented by their positive and zero sequence Thavenin
When the power transfer demand increases, two r tew bays are equivalent ZsthlI, ZsthO, Zqthl and ZqthO. Each line circuit is
constructed and the double circuit line is operat~ edover four represented by a iT circuit with positive and zero sequence
bays. The final network configuration is shown in 4.
Figue impedance Z I and ZO and mutual coupling impedance ZmO.
Figue
4. The line capacitance is neglected. The line parameters were
Substation I Substation 2 taken from an ATP/EMTP Line Parameters Routine for a six
phase circuit nt nominal model [4]. To obtain the equivalent
~j symmetrical component model, it was used the Fortescue
transformation on the 6x6 output matrix.
The negative sequence network equivalent is considered
equal to the positive sequence network with the voltage
n e sources short-circuited.
ni

Figure 4. Final network configuration. Tbe double circuit line is operated

over four substation bays.

Double circuit lines operated in parallel as shown in figures I Vs iVq

and 3 have great influence over the traditional distance
protections. The apparent impedance seen by the distance
protection, when protecting such lines, is different from what
is observed when protecting a single circuit line with the same
equivalent impedance. The study conducted and here
described explains what these differences are and points out
remedial actions to overcome them. Figure 6. Positive sequence network.

2 Methodology
To evaluate the impact of double circuit lines operated in
parallel over distance protection, a steady state analysis of the
power system network, shown in Figure 5, was done. Two
220kV network areas (S and Q) represented by their Th~venin
equivalent are interconnected by a double circuit line of 55km
long.

h.Zo I(1-h).zo

Figure 7. Zero sequence network.

The fault occurs in the line at a per unit distance h. Changing

Figure 5. The network use during the study. the h value from 0 to 1, by a given step, the fault location will
slide along the line.
A distance protection is located at the substation S and To calculate the voltage and current measured by the distance
measures the voltage at the line panel, v, and the current protection, v and i, during a phase-to-ground fault the
flowing through the line, i. Phase-to-phase faults, as well as positive, negative and zero sequence networks are connected
phase-to-ground faults are applied along one of the line in series. During a phase-to-phase fault the positive and

105
negative sequence networks are connected in parallel. For The results taken from the simulations of phase-to-ground
each type of fault, a corresponding equation matrix is used to faults with 002 and 50fl fault resistances with the CBq opened
evaluate all currents present in the circuit, Equation 1. are shown in Figure 8 (squares).
In Figure 9 (squares) the same results for the CBq closed
(1) scenario are shown.

The voltage vector V has all the network voltage sources and 1-
the current vector I are the calculated network sequence 0.8
currents. The admittance matrix Y(h) is dependent on the
fault location h and has 9x9 and 6x6 dimension for phase-to-
ground and phase-to-phase fault, respectively. 0. 0.51- - - -a
The sequence voltages measured by the distance protection Er
are VI, V2 (not shown) and VO from Figures 6 and 7. The
sequence currents measured by the distance protection are 11,
0.2
f
12 (not shown) and 10 from Figures 6 and 7.
0.2 0.5 0.8 1
2.2 EMTP model Fault location Pu

8000 1
- upo -,

0.8 F ;;R6000
.4
0.
CD
S
0)
Sa 0.5 F - -

0 -S ~ *~
0 4000
El 0~
0.2 - - do
*.Jr 1111
I'S, 0 0.2 0.5 0.8 1
0 0.2 0.5 0.8 1 Fault location pu
Fault location pu Figure 9. Measured voltage (p.u.) and current (A) by distance protection for
sliding phase-to-ground fault along the line. Fault resistances Of2 (dashed)
8000[k and 5002 (solid). Same voltages and currents calculated in the ATP/EMTP
simulation (square). Circuit breaker CBq is closed and VqZ-3 00.

6000 Phase-to-phase faults, with Ofl, were simulated and the results
are shown in Figure 10 (squares).
4000 In Figures 8, 9 and 10 the current and voltage obtained by the
a) . .. .. .. . .FR I
analytical model (solid and dashed lines) are also shown to
2000 1 enable a comparison of the two methods. The accuracy of the
analytical model results when compared with the results taken
0 from the EMTfP simulations is acceptable.
0 0.2 0.5 0.8 1
Fault location pu
Figure 8. Measured voltage (p.u.) and current (A) by distance protection for 3 Impact on distance protection
sliding phase-to-ground fault along the line. Fault resistances OQ (dashed)
and 5092 (solid). Same voltages and currents calculated in the EMTP The analytical model developed to evaluate the measured
simulations (square). Circuit breaker CBq is opened. voltage and current by the line distance protection is now
used to evaluate the apparent impendence measured by the
In order to validate the accuracy of the analytical model, an protection for different faults along the line.
ATP/EMTP network model was constructed and simulations
were conducted. The network areas (S and Q) were simulated 3.1 Phase-to-ground faults
by their Th~venin equivalent impedance and the line was
simulated by a distributed frequency constant parameter For phase-to-ground faults simulations, the same two network
model. Seven points were chosen along the line to perform configuration scenarios of the CBq opened and closed were
phase-to-ground and phase-to-phase faults. taken into consideration. On each scenario, a sliding phase-to-
Two network configurations where chosen during the study: ground fault with OQ and 50L2 fault resistances was imposed.
firstly, the circuit breaker (CBq) was considered to be open to Special attention was taken on the zero sequence
evaluate the no load scenario, Figure 12, and secondly, the compensation factor kO used in distance protection phase to
CBq was considered to be closed to evaluate the load scenario ground loops, when calculating the apparent impedance. First,
and fault fed from two sources, Figure 13.

106
it was used the typically kO value corresponding to the one
used in single circuit lines: p1 .5 ... . . . . . . . ... . . .. . . .

i
50)O

k0 =-C~ (2)
CD

Figure 11I presents the reactance location (dashed) seen by the

distance protection phase-to-ground loop using the kO from
CL
Equation 2 for a sliding phase-to-ground fault with Of) and
CU0Oh
50L2 fault resistances. On the x-axis the normalized reactance
of the line and in the y-axis the normalized apparent reactance Z 0
seen by the protection phase-to-ground loop is shown. The 0 0.2 0.5 0.8 1
base reactance used in the normalized reactances is 1/2 XI. A Normn. line reactance p.u.
straight line is drawn with 450 angle and represents the points Figure 11. Normalized apparent reactance measured by the distance
protection for sliding phase-to-ground fault along the line with CBq opened
where the apparent reactance locus is equal to the true
and fault resistances Of) and 500. Impedance calculated using kO from
reactance. Apparent reactances, seen on the top of this line, equation 2 (dashed). Impedance calculated using 1W from equation 7 (solid).
show underreaching effect and, on the bottom, it show
overreaching effect. At the end of the line, the underreached reactance results from
the value of kO derived from Equation (2). The zero sequence
1 mutual coupling is affecting the impedance of the line;
................ therefore, if not present in the kO value, a permanent error
0.8 when evaluating the apparent impedance does exist.
:- For a phase-to-ground fault, at the end of the line (h-i) with
0)
CD 0.51 OCI fault resistance, the sequence voltage at fault location a
S
equals to:

0.2 [ (3)
2
0 0.2 0.'5 0.8 1 v. = 1,IZII (4)
Fault location pu lo' 2
8 0 00 r 1
V2 2 (5)

7000 . .. .. ... . . .. . . . . . The faulty voltage at this location is:

...........
C

CU 6000 . . .. .. .. . .. .. .. .. .. . . . V. = O+V.+V2=0
0
(6)
5000 F .. .. . . ....
. .. .. .. .. ..... .. .
V =V _V(jf+I±(ZO+Zl.O iij0
4000
0 0.2 0.5 0.8 1
Fault location pu V and I are the fault voltage and current measured by the
Figure 10. Measured voltage (p~u.) and current (A) by distance protection for distance protection. 10 is the neutral cur-rent measured by the
sliding phase-to-phase fault along the line. Faulty phase a (solid) and b protection.
(dashed). Same voltages and currents calculated in the EMTP simulation.
Circuit breaker CBq is opened (square). The new zero sequence compensation factor kO is:

In Figure 11, for both fault resistances and for all the line ko (z +Zm0 -1 =I (7)
length, the apparent reactance is underreached. At the end of
the line, the fault reactance is 1. 144 times the true reactance
for M~ and 1.34 timres for 50f). If the distance relay Zone I is Figure I1I presents the reactance location (solid line) seen by
set to 80% of the line reactance, only 41% (M~) to 50% (502)) the distance protection phase-to-ground loop, using the kO
of the line is covered by this zone. from Equation 7. The underreached reactance, at the end of
the line, is no longer present since the apparent reactance
equals the true reactance. For the rest of the line, the
underreaching effect still exists although not so high as the
one observed with kO from Equation 2.
This underreaching effect occurs due to the existing infeed
fault current, Ifcits, that flows through the healthy circuit,
Figure 12. It has its peak value at the centre of line and

107
extinguishes when the fault moves towards the line ends. 3.2 Phase-to-phase faults
Near to the beginning of the line, the infeed from the healthy
circuit, If,, is weaker when compared to the fault current in The analysis has been carried out for phase-to-phase faults
the faulty circuit, Ifcir2, creating a small underreaching effect. with 0OQ fault resistance, Figure 15. The underreach effect due
Near to the line end, both currents Ifi,1. and Ifjfr2, tend to have to fault current distribution between the two line circuits is
the same value and thus reducing the underreaching effect. presented. As for phase to ground faults this effect is created
In case the distance protection Zone I is set to 80% of the line by the existence of both Iforts and Ifcjr2s currents.
reactance and using kO from equation 7, only 67% (Oil) to For Zone I settled to 80% of the line reactance only 56%
51% (5002) of the line is covered by this zone. (MQ) of the line is covered.

Substation S i~feir Is Substation Q

Zsth Zqth
ifcir2sa C,
0.8
C.,

Figure 12. Fault current distribution between the two line circuits. 0.

The same results for the load scenario, CBq closed, are shown Z0

in Figure 14. The voltage phase shift between the two sources
is -30'. In this situation, the underreaching effect is a result of 0 0.2 0.5 0.8 1
the presence of both Kifrts and Ifcirtq currents, Figure 13. When Norm. line reactance p.u.
both sources contribute to the total fault current, in the Figure 15. Normalized apparent reactance measured by the distance
healthy circuit two currents flow in opposite directions, lfeiris protection for sliding phase-to-phase fault along the line with CBq open and
fault resistances 00.
arid Ifcirtq. The current, Ifcirlq is added to the Ifrir2s current in
the faulty circuit and decreases the voltage drop at substation
S bus. This is more relevant the closest the fault is to the 4 Protection schemes
substation Q and, as a consequence, the underreaching
The presence of the underreaching effect in line distance
symmetry around the centre of the line as shown in Figure I11
protection on double circuit lines operated in parallel is of
is no longer present.
great relevance, when designing the line protection system.
Traditional distance protection, used on such line,s is not able
to cover the entire line without the risk of over setting Zone I
and leading to miscoordination problems. In this case, proper
protection schemes may be used.

4.1 Distance protection

Figure 13. Fault current distribution between the two line circuits. Double fed
fault.
When setting distance protection on a double circuit line
operated in parallel, the underreaching effect may be taken
:. 1.2r
into consideration. For this reason, the use of kO factor from
1 Equation 7 will avoid the underreaching observed for phase-
C.,
0a
0.8 to-ground faults at the end of the line.
0
Co At the same time, Zone 1 reach may be extended in such
amount that the true reach is the same as the desired reach on
0.5 F single circuit lines, without overreaching the remote bus. This
0.
a. means that if the desired reach of Zone 1 is 80% of the line,
0 0.2 for phase-to-ground faults, this zone may be set to 90% of the
E line reactance.
0
0 0.2 0.5 0.8 1 For phase-to-phase faults it is impossible to set Zone 1 to
Norm. line reactance p.u. reach faults up to 80% of the line length without loosing
Figure 14. Normalized apparent reactance measured by the distance selectivity with the protections from the remote substation
protection for sliding phase-to-ground fault along the line with CBq closed,
VqZ-30' and fault resistances 0f2. Impedance calculated using kO from
Therefore it may be set to 90% of the line reactance and, in
Equation 7. this case, only 70% of the line is covered.

In case distance protection Zone 1 is set to 80% of the line 4.2 Teleprotection schemes
reactance and using kO from equation 7, only 50% of the line
is covered. If permissive underreaching schemes (PUTT [1]) are used,
the overlapping reach by the carrier sent signal Zone from
opposite protections may not be achived. As result, in case of

108
an internal line fault, it is possible that the teleprotection
scheme may not work and the fault be cleared in Zone 2.
A proper teleprotection scheme can be used to cope with the
underreaching effect.
Schemes based on permissive overreaching criteria, POIT,
are able to correctively trip undelayed both circuit breakers
from the lines end without loosing selectivity. By sending
teleprotection signal when the protections picks-up, in at least
Zone 2, it is guaranteed that for every fault along the line, the
scheme will work.

4.3 Line current differential protection

In order to achieve undelayed fault clearance times, even for
high resistive faults, where not even the distance protection is
able to pickup and send a teleprotection signal, a line current
differential protection may be used. This protection is
unsensitive to the above mention underreaching effect and at
the same time is more sensitive than a distance protection for
low fault currents.
As a drawback it is highly dependent on the
telecommunication system and therefore may not have the
same dependability level as the one achieved with distance
protection with POTT scheme.

5 Conclusions
At REN, the use of double circuit lines operated in parallel is
becoming common when uprating and constructing new lines.
The impact of such lines in the performance of distance
protection is evident due to underreaching effect caused by
the two line circuits. An analytical model was developed to
study the effect and was validated with ATP/EMTP
simulations. The results are able to explain and quantify this
underreaching effect for phase-to-ground and phase-to-phase
faults at every line location.
A modified kO factor is used in distance phase-to-ground
loops to compensate the zero impedance mutual coupling
present in such lines and so eliminating the underreaching
effect for faults at the end of the line.
Finally remedial actions are proposed which include
increasing Zone 1 reactance reach, use of permissive
overreaching schemes and implementing line current
differential protection.

References
[1] J.L. Blackburn, "Applied Protective Relaying",
Westinghouse, Ch. 2 and Ch.6, (1982).
[2] P. M. Andreson, 'Power System Protection", IEEE
Press, pp 547-570, (1999).
[13] S. H. Horowitz, A. G. Phadke, "Power System Relaying
RSP", Press LTD, (2003).
[14] "Alternative Transients Program Rule Book", IV-B, IV-
C IV-D, (June 1992).

109