179

Design of Phase and Amplitude Comparators for Transmission Line Protection

M.S. Sachdev' T.S. Sidhu' I. Uttamchandani' W.O. Kennedy
Consultant
Power Systems Research Group'
University of Saskatchewan
Saskatoon, Saskatchewan
CANADA S7N 5A9
Abstract: This paper describes the design of polyphase
relays by using the amplitude and phase comparison
techniques. Both techniques are used for designing the
phase-to-phase, three-phase and phase-to-ground relays. A
new polyphase amplitude method, which uses inputs from
the phase comparator, is proposed for protecting lines during
phase-to-phase and three-phase faults. A unique amplitude
and phase comparison approach is introduced for use in a
polyphase phase-to-ground relaying unit. The proposed
designs are suitable for implementation on microprocessors.
Performance of the comparators was evaluated by using data
simulated using EMTDC and some results are included i n
the paper.
Keywords: Polyphase relays, distance relay, polyphase
amplitude comparator, polyphase phase comparator, digital
re1ay s.
I. INTRODUCTION
Polyphase measuring units have been used extensively in
electromechanical and electronic realizations of relays for
protecting transmission lines. In 1958, Sonnemann &
Lensner [l] outlined the theory of polyphase distance relays
using the symmetrical components approach. Polyphase
relays use either phase or amplitude comparison in a single
measuring unit [2]. Inputs to these units are signals formed
from voltages and currents of faulted and unfaulted phases.
The characteristics of these relays can be analyzed by either
mathematical analysis or graphical examination. Phasor
diagrams were used to investigate the behavior of KD relays
U].
Although, single phase mho units are capable of protecting
transmission lines, they take considerable space and cost
more than the polyphase units. Also, electric power utilities
have been using electromagnetic and solid-state type
distance relays to protect transmission lines. With the
availability of modern microprocessors and Digital Signal
Processing (DSP) chips, it has become possible to perform
large amounts of calculations and implement logic in
relatively short times. This has led to the availability of new
and complex microprocessor based relays[4,5].
This paper describes the design of polyphase amplitude
and phase comparators for detecting phase-to-phase, three-
phase and single-phase-to-ground faults. Some results
showing their performance, which was checked using data
obtained from simulations performed on the EMTDC, are
included.
II. POLYPHASE COMPAFWTORS
Polyphase distance relays differ from conventional mho
relays which detect one type of fault only. For example, three
mho elements are needed for phase to ground faults and
another three are needed to detect phase-to-phase faults. A
complete single zone protection system, therefore, requires
six mho elements. Compared to this, only one polyphase
element is needed to detect three types of phase-to-phase
faults and another to detect the three single phase to ground
faults. This section describes the design of phase and
amplitude comparators for detecting phase-to-phase, three-
phase and single-phase-to-ground faults.
A. Phase-to-Phase Design
Polyphase relays which detect phase-to-phase faults, such
as A-B, B-C and C-A faults, use the phase comparison
technique which is most commonly used by industry. For
example the phase-to-phase fault element of KD4, marketed
by the Westinghouse Electric Corporation (now ABB), is
designed to perform phase comparison. The inputs to this
comparator, Su b and s b c [3], are:
In these equations, V,, v b , V , lu, Ib and IC are the voltages
and currents and Z, is the replica impedance used in the
relay. The comparator operates when Sbc leads Sab; this can
be mathematically expressed as
0' 5 (AS,, - LS, b) 5 180'
(3)
It is possible to have a polyphase amplitude comparator
which has the same characteristic as the phase comparator
described by Equations 1 to 3. One possible design of such a
comparator would use the following inputs.
s O=s b c +j s u b (4)
Sr =Sb c - j Su b ( 5)
The operating condition of the polyphase amplitude
comparator is
ISBN 0-7803-4147-3AEEE CAT. NO. 97CH36117 Q 1997 IEEE 1991 Conference on Communications. Power and Computing
WESCANEX'97 Proceedings; Winnipeg, MB; May 22-23, 1997; pp. 179-184
180
when plotted on R-X diagram and are independent of the
source impedance but depend on the relay settling
impedance.
C. Phase-to-Ground Design
The most frequent type of fault experienced on a power
system is the phase-to-ground fault. It can occur between
any of the three phases and the ground. Utilities have been
using single phase units to protect their lines during phase-
to-ground faults [3]. Three single phase units are needed for
zone one protection, while one polyphase unit can provide
protection during any single phase to ground fault, thereby
saving considerable space and cost.. A more attractive
technique, that uses positive-, negative- and zero-sequence
components of voltages and currents for phase-to-ground
polyphase design, is introduced in this section.
The input signals to an amplitude comparator for correct
measurement of a-ground, b-ground, and c-ground
short-circuits are:
so= vxo+ vx2 (12)
Sr=V.l (13)
The operating signal, So, is made up of negative and zero
sequence components, VX2 and Vxo. The restraining signal, Sr,
consists of a positive sequence component denoted by V,,.
The sequence components can be expressed in terms of
compensated voltages, V,, V,and V,, at the relay location as:
V,=( u3)( vx+ vy+ V,) (14)
The characteristic equation describing the behavior of the
phase and amplitude comparators can be derived and they
are identical. The characteristic plotted on R-X plane is
shown in Figure 1. The characteristic is in the shape of a
circle which expands and shrinks depending on the source
impedance. If source impedance is zero, the characteristic
circle passes through the origin of the R-X diagram. For
faults behind the relay, current reversal occurs and the relay
restrains. Memory action is not needed because sufficient
torque exists for close-in faults.
Fig. 1. Characteristic of the phase-to-phase relay on the R-X
diagram.
B. Three-phase Design
Westinghouse Electric Corporation [3] are:
The inputs to the phase comparator introduced by the
The comparator operates when V,leads VZy i.e.
0 S( LV, . . -LVz, , ) S 180
(9)
Three-phase faults can also be detected by using an
amplitude comparator which has the same characteristics as
the phase comparator described by Equations 7 to 9. The
operating and restraining signals of such a comparator
would be as follows:
The operating condition of this three-phase amplitude
comparator is given in Equation 6.
It can be shown that the characteristic equations of both
comparators are identical. The characteristics are circular
V,~=(I/~)(V,+ a2 Vy+a vz) (16)
The voltages V,, V, and V, are defined as
V,= Vc-(Ic+ KIo)Zn
where:
K=(ZO-Z&ZI
is the complex constant made of the
positive and zero sequence impedance of
the transmission line.
is the zero sequence current measured at
the relay location.
IO
It is possible to have a polyphase phase comparator which
has the same characteristic as the amplitude comparator
defined by Equations 12 and 13. One possible design of
such a comparator is given as:
The signals SI and Sz can be calculated from Equations 2.0
and 21. The comparator will operate when S, leads Sz i.e.
181
0' 5 ( LS, - LS,) 5180' (22)
The characteristic equations derived for the amplitude and
phase comparators are identical. The characteristic of the
comparators is circular whose center and radius can be
computed. Also, the operating characteristics of the phase-
to-ground relays are the same as that of the phase-to-phase
relays.
HI. DETERMINATION OF FUNDAMENTAL
FREQUENCY COMPONENTS
Signals formed from the voltage and current phasors of the
fundamental frequency are used for making decisions. It is,
therefore, essential that the fundamental frequency
components of voltages and currents be determined.
Microprocessor based relays sample voltages and currents
and compute phasors representing their fundamental
frequency components [4,5].
Voltage and current signals contain transient components
such as, exponentially decaying dc, harmonics and other
high frequencies. This is especially true for currents and
voltages experienced during a fault. Since phasors of the
fundamental frequency components are to be used in the
designed comparator, decaying dc and harmonics must be
eliminated from the inputs. In the work reported i n this
paper the real and imaginary parts of the fundamental
frequency phasors were estimated using Least Error Square
(LES) digital filters [6]. During the design of the filters, it
was assumed that each input signal is composed of an
exponentially decaying dc and fundamental, second, third,
fourth and the fifth harmonics. A data window of 13 samples
taken at 720 Hz was used.
IV. TEST RESULTS
Four software modules were written to test the performance
of the phase and amplitude comparators. Module A was
written to simulate, on the EMTDC, a sample power shown
in Figure 2. The voltages and currents provided by the
simulation performed by Module A were processed in
Module B which included 4th order anti-aliasing low pass
filters. Outputs from Module B were then provided to
Module C which estimated the real and imaginary parts of
the fundamental frequency phasors. The estimates of the
phasors were presented to Module D which made the
'trip'/'do not trip' decisions.
Polyphase relays were set to protect 80% of the 200-km long
line. Several phase-to-phase, three-phase and phase-to-
ground faults at different locations on the line were
simulated. The data were processed by Module D which used
phase and amplitude comparison for making decisions.
Some results are presented in this section.
Module D
TRIP
S1& s2
P
away from the relay. Figure 3 shows the difference between
the phase angles of sbc and Sab, the signals used for phase
comparison. Originally, the difference between the phase
angles is out of the operating range. After the fault, the
difference changes and moves into the operating range and
the relay operates. The comparator took about 9.7 ms (7
samples) to produce the trip command. Figure 4 shows the
operating and the restraining signals of the amplitude
comparator. Immediately after the inception of the fault, the
operating signal starts to increase and the restraining signal
starts to decrease. A trip command is issued 9.7 ms (7
samples) after the inception of the fault. A phase-to-phase
fault (c-a) was also applied on the line at 190 km away from
the relay. Both comparators remained stable and did not
produce any trip signals as shown in Figures 5 and 6.
Figures 7 and 8 show the decision-making signals for
phase and amplitude comparators when b-c fault is applied
at the relay. Both comparators produce trip signals in 8.3
ms (6 samples).
transmission line
SAMPLES
Module A
va, vb, vc ia, ib, ic
4t h ORDER DIGITAL
Low PASS FILTER
Module B
is, ib, ic va, vb, vc
13 SAMPLE
LEAST ERROR SQU-
DIGITAL FILTER
Module c
Fig. 2. Flow Chart for testing the comparators.
A. Results of Phase-to-Phase Design
Figures 3 and 4 show the performance of the comparators for
a line side phase-to-phase (a-b) fault at a location 50 km
182
B. Results of Three-phase Design
A three-phase fault (a-b-c) was simulated at 100 km away
from the relay. The phase comparator issued a trip
command in 13.9 ms (10 samples) as shown in Figure 9.
The amplitude comparator also issued a trip signal in 13.9
ms; the operating and restraining signals are shown in
Figure 10. The phase and amplitude comparators did not
issue any trip commands for a three-phase fault at 190 km
away from the relay, as is shown in Figures 11 and 12.
C. Results of Phase-to-Ground Design
Figures 13 and 14 show the performance of the comparatalrs
for a line-side b-g fault 100 km from the relay. Both
comparators produced trip commands 15.3 ms (1 1 samples)
after the inception of the fault. Faults behind the relay
location (bus-side faults) were also simulated. Figures 15
and 16 illustrate the results for a phase-a-g fault. It can Ibe
seen from the figure that the comparators do not operate.
90
9
0
3
-90
-180 -
180 T
..........................
I
.I
..
I
..
.. I.
u i i i i m i m m i i i i m i . ' 1
. _
180
90
2
9
a 0
-90
-180
0 -
?
I!
3 :: .................
1.
Fig. 3. Angle difference between sbc and Sab for a a-b fault
50 km away from relay.
-
.........................
.I..
I
..
I
..
I
I
I
.. I
...............
*
450000
N
- 300000
8
8
4
4 150000
0
-180 1
.............. 450000
1.
"I 0.081 0.096 0.111 0.127 0.142
......................... , - Time (4
8
I 2
4
6 150000
........................
0 ,
J
0.081 0.096 0.111 0.l27 0.142
Time 6)
................
\
............. ..........
0
0.081 0.096 0.111 0.127 0.142
Time (s)
Fig. 4. Magnitude square of So and S, for a a-b fault 50 km
away from relay.
Fig. 6. Magnitude square of So and S, for a c-a fault 190 km
away from relay.
Fig. 5. Angle difference between sbc and Sab for a C-afault
190 km away from relay.
Fig. 7. Angle difference between sbc and sab for a b-c fault
at the relay.
183
-90 I
Fig. 11. Angle difference between V, and V,, for a-b-c
fault 190 km away from relay. Fig. 8. Magnitude square of SO and S, for a b-c fault at the
relay.
.ll.ll..lllllllll
I
180 T II
I
I
. m m i
300000 T
_...-_
0
0.082 0.097 0.113 0.128 0.143
Time (a)
Fig. 9. Angle difference between V, and V, for a-b-c fault
100 km away from relay.
Fig. 12. Magnitude square of SO and S, for a-b-c fault 190
km away from relay.
-
0
0.082
0.097 0.113 0.128 0.143
Time (9)
Fig. 13. Angle difference between SI and SZ for b-g fault
Fig. 10. Magnitude square sf SO and S, for a-b-c fault 100
kmaway from relay.
km away from relay.
500007
- H
8
;j 25OOo
a
I
O +
Fig. 15. Angle difference between SI and S2 for a-g fault MS. Advancements in
behind the relay.
Microprocessor Based Protection and Communication, IEEE
Tutorial Course Text, Publication No. 97TP120-0, 1997.
phase comparator, has been proposed for phase-to-phase and
_ . . . - _ three-phase designs. A unique amplitude and phase
comparison approach has been introduced for the polyphiise
phase-to-ground design. Characteristics of the proposed
-- amplitude comparators are identical to that of ithe
conventional phase comparator. Both comparators benefit
from the polarizing signals in detecting close-in faullts.
Extensive testing showed that the tripping times for the
faults on the line-side are similar for both techniques.
. O- . - - - . - . - - . .
-.
.-
.-__._--__.---.-.-_
. -
0 -
-
$
3 -90
3
-180 7
[2] A.R. van C. Warrington, Protective Relays Their Theory
and Practice Volume I & I1 Chapman and Hall 1974.
..
" ' 8
I.............' .. ........................ [3] Applied Protective Relaying, Westinghouse Electric
Corporation, Relay-Instrument Division, Newark, New
Jersey, U.S.A., 1964.
[4] M.S. Sachdev (coordinator), Microprocessor Relays and
-
w-
3oooo
PI
2
Q 2oooo
I
0 U2
loo00
0 -
_ - _ _ _ Apparatus and Systems, Vol. PAS-98, No., Nov./Dec. 1979,
pp. 2232-2240
..- ---..___ _ _ .
\ - *
. . '.
' c -
* , \
.-
- __._-_.-____.__.______
--
.-
. _