Professional Documents
Culture Documents
-3.5.1-
C. W. Brice
1
August 2002
PCB
1
2
3
I1
C1
V3
V2
I2
C2
I3
TRIPPING
DIRECTION
C3
P23
P31
P12
V1
V23
V23
TS
AC
T
N
CO
I2
I1
I1
V12
I3
NT
CO
V31
E
OS
L
C
N
PE
O
TS
AC
Figure 3.5.1. AC elementary diagram for a phase directional overcurrent relay. Phase sequence
is 1-2-3.
Ground directional overcurrent relays use the residual current and a polarizing quantity that can
be (a) zero-sequence voltage, (b) neutral current from a delta-wye (grounded) transformer (or
other grounding transformer), or (c) negative-sequence voltage and negative-sequence current.
Figure 3.5.2 shows a ground directional overcurrent relay with a broken-delta connection of PT's
-3.5.2-
Protective Relays
to obtain zero-sequence voltage (actually 3 Vo). Since Vo is zero for balanced loads, threephase faults, and phase-phase faults, and since any fault involving ground will produce a zerosequence voltage that is always in the same polarity, then the zero-sequence voltage is useful as
a polarizing quantity for ground relays.
1
PCB
1
2
3
I1
C1
I2
C2
I3
TRIPPING
DIRECTION
C3
Co
V3
V2
P23
P31
P12
V1
V3
V2
3 Vo
V1
-3.5.3-
Potential
C. W. Brice
1
August 2002
PCB
1
2
3
I1
I2
I3
PHASE
RELAYS
TRIP
DIRECTION
Co
3 Io
TO NEUTRAL OF
POWER TRANSFORMER
SCHEMATIC, FOR
ILLUSTRATION ONLY
Protective Relays
amperes may be substituted as a polarizing quantity since the negative sequence is virtually
unaffected by mutual coupling and ground faults always produce both negative and zero
sequence quantities.
3.5.2 Distance relays
Distance relays actually respond to impedance (or reactance) seen at the relay. Obviously, such
a relay needs both current and potential measurements. For a given line and for given CT and
PT ratios, the impedance seen by the relay is proportional to the distance to a short circuit, thus
the name. We are really interested in directional distance relays, since the directional
characteristic is necessary.
Zone 2
Zone 1
R
Figure 3.5.4. Distance relay characteristics. Trip zones are inside the figures indicated. In
sequence, the directional impedance relay, the mho distance relay, the mho distance relay with
reactance relays, and three zones of mho distance relay characteristics.
-3.5.5-
C. W. Brice
August 2002
The simplest idea is the impedance relay, which is not directional, but may be supervised or
controlled by a directional element to make a directional impedance relay. Its characteristic is
shown in Figure 3.5.4(a), which is an example of the use of the R-X diagram. The mho distance
relay characteristic is shown in Figure 3.5.4(b). Note that it has an inherently directional
characteristic. Reactance units may be combined with a mho distance relay to create several
zones, as in part (c) of the figure. Alternately, several mho units may be combined for several
zones, as in part (d) of the figure.
The reactance characteristic is better for short lines since the arc resistance may be significant
compared to the line impedance. Arc resistance will add vectorially to the line impedance, along
a line approximately parallel to the R axis. On a short line, the mho distance characteristic may
not enclose the resulting locus, so the relay may miss a fault. The reactance characteristic has a
wider zone 2 parallel to the R axis, and will catch most short line faults. The mho distance
characteristic is better for long lines, since it is less likely to trip for a heavy overload or transient
swing. Transient swings are discussed later. Heavy loads appear as a point far out the R axis
and somewhat up the X axis (depending on the power factor). It is very important not to set
distance relays to enclose heavy loads in trip zones, since important lines will be lost
unnecessarily. The infamous 1965 Northeast power blackout was triggered by a distance relay
that tripped its zone 3 element on a heavy load. Zone 3 is sometimes used as a backup that trips
only after a long time delay, but this practice should be examined and the settings should be
checked periodically to ensure that this zone does not enclose load loci.
Electromechanical distance relays develop an operating torque Top and a restraining torque Tre,
and the net torque is given by:
Tnet = Top - Tre - Tspring
If the spring torque is neglected, the operating torque will equal the restraining torque at the
balance point, which is the relay's reach. The mho distance relay has an operating torque:
Top = k1 V I cos ( - )
where = relay angle of maximum torque and = line impedance angle. The restraining torque
is:
Tre = k2 V2
so at the balance point:
k2 V2 = k1 V I cos ( - )
Z = V / I = (k1 / k2) cos ( - )
Y cos ( - ) = k2 / k1 = K
-3.5.6-
Protective Relays
If Y cos ( - ) = K is plotted versus for constant values of K and , the result is the mho circle
that passes through the origin of the R-X diagram and has a maximum reach of Z = 1/K when
= .
Static distance relays perform the same function by checking the phase angle of an operating
quantity such as (I Zr - V) compared to a polarizing quantity such as V = I Zf, where Zr is the
relay reach and Zf is the impedance to the fault. If the phase angle of this operating quantity
leads the phase angle of the polarizing quantity by 90 degrees or less, then the relay impedance
falls within the mho circle and the relay trips. Otherwise, the relay sees an impedance outside
the mho circle and restrains. Note that the timer that is used to detect the phase difference may
be set to other angles less than or greater than 90 degrees. These settings would produce trip
zones that are not circles, but shaped like a lens if the setting is less than 90 degrees or a tomato
if greater than 90 degrees. Figure 3.5.5 shows these shapes on a phasor diagram.
Figure 3.5.5. Lens and tomato shaped trip zones may be obtained with static distance relay.
Since a distance relay cannot be set for 100% of the line, a scheme involving several zones of
protection is used to cover the line. Figure 3.5.6 shows a mho distance relay with three zones of
protection. The first zone trips the breaker with no intentional time delay, and is set to reach
about 80-90% of the line impedance (it underreaches the line end). Since this would leave 1020% of the far end of the line unprotected, the second zone, which trips after a time delay of
about 0.2 to 0.5 seconds, overreaches the line (110-120% of the line impedance). The time delay
is needed to allow the next breaker to clear faults beyond the next station. The third zone, which
may be set to trip after a time delay of about 0.5 to 1 second, can be used for backup protection
of the lines out of the next station. This has some danger, since the trip circle may become so
large that the third zone trips on heavy loads. Some schemes use the third zone reversed to look
behind the relay. This idea will be examined in a later section on carrier pilot protection.
Note that the mho characteristic passes through the origin, so one might wonder what would
happen if the fault were right at the relay. In this case, the voltage would be very small, so the
relay could fail to trip, except for the fact that some "memory" action is always included on the
-3.5.7-
C. W. Brice
August 2002
zone 1 relay. In essence, the memory action is just a long time constant (or reset time) in the
potential circuit, just enough to make sure the relay trips on close-in faults.
Also, the relay is set based on steady-state calculations of the line impedance at 60 Hz. One
might wonder what effects the transient currents, especially the DC transient in the short-circuit
current, have on the relay. The DC components decay fairly rapidly, so only the zone 1 element
is affected, and it sees an impedance that is somewhat less than the actual impedance to the fault
(based on steady-state calculations). The zone 1 distance relay will often exhibit a transient
overreach for this reason. The relay manufacturers control this overreach so that if the relay is
set to see no more than about 90% of the line, no operational problems will result.
X
ZONE 3
ZONE 2
ZONE 1
-3.5.8-
Protective Relays
1
PHASE SEQUENCE
IS 1-2-3 OR 3-2-1
PCB
1
2
3
I1
I2
I3
C12
C23
C31
C31
C12
C23
TRIP
V3
V2
P23
P31
P12
V1
Figure 3.5.7.
protection.
Figure 3.5.8 shows a one-line diagram of a three-terminal line, which presents interesting
problems for distance relays. The relay at R sees an impedance:
Z = Vr / Ir = [(Ir + It) Zs + Ir Zr] / Ir = Zs (1 + It/Ir) + Zr
The impedance to the fault is Zr + Zs, so the infeed from terminal T increases the apparent
impedance seen by the relay. Consequently, the relay will underreach and may not adequately
protect the line. Note that the first zone must be set for no more than 90% of the smaller of (Zr +
Zt) and (Zr + Zs), which means that a very short tap line will cause problems. The second zone
would have to be set much greater than the first, and may reach way too far past the station on
the short tap.
-3.5.9-
C. W. Brice
August 2002
S
Zr
Zs
Ir
Ir + It
21
It
Zt
T
EQUIVALENT SOURCE
Figure 3.5.8. One-line diagram of a three-terminal line. Infeed from T increases the apparent
impedance seen by the relay, causing underreach.
Another problem is illustrated in Figure 3.5.9, where terminal S and T are shown connected by
an equivalent impedance that represents the rest of the network (not shown). For an external
fault at F compared to an internal fault on the line near T, it is not immediately clear which will
result in the greater impedance seen by the relay. Depending on the impedance to the fault and
on the relative values and directions of the line currents during the fault, the relay may see a
greater impedance for the internal fault (which is how its setting would normally be made) than
for the external fault. Thus the relay may overreach and trip (incorrectly) for external faults
under some conditions. This case is tricky and careful short-circuit calculations are needed to
check the settings.
In some cases, distance relays will not protect three-terminal lines without overreaching or
underreaching one of the terminals. Even a simple two-terminal line may be impossible to relay
with distance relays, particularly if a long line is followed by a short line. Zone two of the relay
at R in Figure 3.5.10 should overreach the station at S by at least 10% (that is, the reach should
be 110% of the line impedance). This may also overreach the short line from S to T, which
means that coordination is lost. In these cases, some form of pilot protection is needed.
-3.5.10-
Protective Relays
R
S
Zr
Zs
Ir
Ir + It
21
It
F
EQUIVALENT
SOURCE
Zt
Figure 3.5.9. Three-terminal line with an external fault. Detailed calculations are needed to
determine whether the internal fault at T presents a greater or lesser value of impedance to the
relay at R than does the external fault at F.
R
S
21
ZONE 1
RELAY R
ZONE 2
RELAY R
ZONE 1
RELAY S
Figure 3.5.10. Long line protected by distance relays that is followed by a short line. Since zone
2 of distance relay at R must overreach the terminal S, they may overreach the terminal at T
(incorrect coordination).
Ground distance relays are relatively more complicated than are phase distance relays. Some
utilities prefer to use directional overcurrent relays for ground protection because of this
complexity. Ground distance relays are available, and so we make a brief discussion of them.
The problem occurs because the voltage of the faulted phase is not directly comparable to the
current, since the zero-sequence reactance of the line is not equal to the positive-sequence
reactance. Thus, to obtain a relay that detects the impedance to the fault in a consistent way,
-3.5.11-
C. W. Brice
August 2002
some type of compensation is used. For example, the faulted phase (call it phase A) voltage for
a single phase to ground short circuit is related to the sequence currents by:
Va = V1 + V2 + V0 = Z1 (I1 + I2) + Z0 I0 = (2 Z1 + Z0) I0
where I1 = I2 = I0, but Z0 is not equal to Z1 in general. Thus, the voltage restraint could be set
for:
Vr = Va - 2 Z1 I1 = I0 Z0
at the balance point.
The application of relays using such compensation produces some extra complexity, which
explains the continued use of ground directional overcurrent relays. Note that this discussion is
meant as an introduction only.
3.5.3 Pilot protection
Pilot protection is intended here to be a general term that covers any type of relaying that uses
communication from the far end of the line as part of the relaying scheme. The channel may be
wire, microwave, or fiber; the signal may be a carrier turned on, an audio tone shifted in
frequency, or an analog signal. The signal itself may represent an indication to permit tripping,
to block tripping, to force tripping, or an indication of the current at the far end (for example,
pilot wire, a form of differential relaying).
Communication channels in common use include wire lines, power-line carrier, microwave, and
fiber optic. Wire lines may be divided into telephone circuits (leased or owned by the utility)
suitable for audio tones, and pilot wire channels in which a circulating current is used to transfer
current magnitude and phase information from line end to line end. Power-line carrier, a popular
system, uses the protected transmission line as the communication channel, carrying an RF
signal. The signal is coupled to the line via coupling capacitors (a series resonant LC circuit
tuned for low impedance at the carrier frequency and high impedance at 60 Hz). Line traps
(parallel resonant circuits tuned to pass 60 Hz and block the RF) confine the signal to the
protected line. This is illustrated in Figure 3.5.11.
Since the channel may be lost in case the line suffers a three-phase fault (although there is a
surprisingly high probability that the signal will be attenuated but still get through most faults),
the carrier channels usually carry a blocking signal instead of a tripping signal. Microwave
channels, on the other hand, are sometimes subject to momentary fading during conditions
unrelated to the presence or absence of a power line fault, and so often utilize a tripping signal
since the blocking signal would be momentarily lost. These issues are fully discussed below.
-3.5.12-
Protective Relays
LINE TRAP
PCB
COUPLING CAPACITOR
RF CHOKE
XMTR/RCVR
RELAY
LINE TRAP TUNED TO PARALLEL RESONANCE AT RF (Y=0)
BUT PASSES 60 HZ WITH NEGLIGIBLE ATTENUATION.
COUPLING CAPACITOR AND RF CHOKE TUNED TO SERIES
RESONANCE AT RF (Z = 0) BUT BLOCK 60 HZ.
Figure 3.5.11. Power-line carrier communication channel. Line traps keep RF carrier signal
confined to protected line but pass 60 Hz, and coupling circuit passes RF but blocks 60 Hz.
Pilot protection, in many cases, allows high-speed clearing of faults at any location on the line.
By clearing faults at both ends simultaneously, the protection is improved both with regard to
minimizing conductor burn-down and with regard to improving the transient stability margin of
the system. If it were not for high-speed relay and circuit breaker operation, transient stability
would have become an almost insurmountable problem since every other trend is in the wrong
direction: increased line loadings, difficulty in siting lines and generation, delays in licensing
plants, decreased generator inertia and increased generator reactance (less iron per MVA).
Pilot Wire Relays
Pilot wire relays are the most straightforward types of pilot protection, since they are a simple
extension of the idea of differential relays to protection of a line. The modern application of
pilot-wire relays is for protection of short lines. The system consists of a pair of pilot wires that
connect the two stations at the line ends. The pilot wire circuit normally carries a circulating
current as the voltages at the line ends have opposite polarity, as shown in Figure 3.5.12.
External faults will produce the same polarity and cause most of the current to circulate which
produces a restraint. Internal faults will cause the polarity at one end to reverse, which then
forces significant current through the operating coils, tripping both ends.
-3.5.13-
C. W. Brice
August 2002
PCB
PCB
ISOLATING
TRANSFORMER
OP
OP
ISOLATING
TRANSFORMER
SATURATING
TRANSFORMER
OP
OP
OP
INTERNAL FAULT
Figure 3.5.12. Pilot-wire relay, a form of current differential relaying, which is useful for
protection of relatively short transmission and subtransmission lines.
To allow the pilot-wire system to protect against any type of short circuit over a single pair of
wires, the three phase currents are combined by a sequence network or filter. This network
produces a voltage that is proportional to some linear combination of the positive and zero
sequence currents. This voltage is put into the saturating transformer that, along with other
components not shown, limits the voltage applied to the operating coil (or operating circuit for a
static relay).
Application of pilot-wire systems is reasonably simple, unless the current transformers at each
end of the line are of different types. If both CT's are bushing type, or if both are wound type,
and if they have the same ratios and if their burdens are not widely different, then no problems
are expected. If the CT's do not match, then special calculations, including the magnetizing
branch of the CT's, will be needed to ensure proper operation.
The length of the pilot wire is limited by economic factors (cost of the wire) but also by the
series resistance of the loop (less than 2000 ohms for the whole loop) and by the shunt
-3.5.14-
Protective Relays
capacitance (less than 1.5 microfarads). The relay will impose no more than 100 mA or 60 V,
rms, when a fault is present. This should not pose great problems for a metallic twisted-pair
cable. The twisted-pair cable is used to limit the transverse voltage caused by mutual inductance
to the power line.
Pilot-wire relay systems can be equipped with automatic pilot-wire circuit monitoring. This
monitoring may be accomplished by using a DC supervision current on the AC pilot wire
system. If the supervision current signal is above or below its normal limits, then an alarm is
sounded. Transfer tripping may also be included, if desired.
Phase Comparison Schemes
Phase comparison schemes use a similar idea to the pilot-wire schemes, except there is no need
for the pilot wires to carry the current signal directly. In essence, this system uses a form of
modulation, where transmitters at each end send trains of pulses on alternate half cycles of the
AC waveform. For an internal fault, the transmitters will key on the same half cycle. By means
of an electronic circuit, this is detected at each end, and each relay then trips its local breaker.
For an external fault, the transmitters key on alternate half cycles and no trip signals are
produced.
The phase comparison scheme is not widely used (at least in the Southeastern US), but it has
some applications. It is too early to completely write off this system, since it could make a
comeback (in a slightly different form) if fiber optic systems become prevalent. For now, we
will not concentrate on the phase comparison system.
Directional Comparison Schemes
Directional comparison schemes are popular, especially when used with power-line carrier
channels. The directional comparison is sometimes lumped with the transfer tripping scheme
since it is similar and both are suitable for carrier systems. Usually the directional comparison
blocking scheme is used with power line carrier, so that is the version that we will discuss in
detail.
Often, the directional comparison system is based on directional distance relays, or on phase
distance and ground directional overcurrent relays, set to overreach the far end of the line. In
either case, the idea is that the carrier is keyed when the local relay sees a fault behind it (an
external fault). Receipt of the carrier at the remote end blocks the relay there from tripping, thus
the name blocking scheme. Figure 3.5.13 shows phase distance relay settings for the relay
overreaching the line, and for the zone set to look backwards to start the carrier to block for
external faults.
If the transmitter, receiver, or channel fails, the blocking signal will not prevent the remote relay
from tripping on the external fault, since the relay at each end is set to overreach the far end of
the line. For this reason, the blocking scheme is very dependable for internal line faults (trips
when it should), but it also can overtrip (not secure for external faults if the communication
system is not reliable). Blocking schemes are not usually applied on systems that use microwave
channels that are subject to fading, since temporary loss of the channel could prevent the
blocking signal from reaching the remote end, resulting in false tripping. On the other hand,
-3.5.15-
C. W. Brice
August 2002
power-line carrier is a very reliable communication channel in most cases, except when a severe
fault is on the protected line. In that case, a blocking signal is not being transmitted, so the
system works correctly. The signal can be a simple on-off signal, such as transmit carrier or not
transmit carrier.
REVERSED
ZONE 3 (R)
ZONE 2 RELAY R
REVERSED
ZONE 3 (S)
ZONE 2 RELAY S
21
R
RCVR
RCVR 21
S
Figure 3.5.13. Directional comparison (blocking) scheme. Phase distance relays overreach the
remote end of the line. Reversed third zone transmits a blocking signal for external faults.
Transfer Tripping Schemes
Transfer tripping schemes are similar to the directional comparison (blocking) scheme, except
that the communication channel transmits a tripping signal that either permits or forces tripping
at the remote end. Since the tripping schemes must get a trip signal through, they often use
frequency-shift keying (FSK), which means that they continuously transmit a certain frequency
(guard signal), shifting to a different frequency (trip signal) if a transfer trip is indicated. This
way, loss of the channel is detected, and tripping is blocked. As a result, this scheme is secure
from false tripping but not as dependable as the blocking scheme. In the following discussion, it
is assumed that the fault is near one end of the line, called the local end, and distant from the
other end, called the remote end. In any of these schemes (including the blocking scheme) a
close-in fault will cause the local relay to immediately trip its breaker.
The simplest type of tripping scheme is the direct underreaching transfer trip (DUTT). Here, the
local relay underreaches the remote end, and for a fault in this underreaching zone, initiates a
direct transfer trip to the breaker at the remote end and (simultaneously) initiates a direct trip of
the local breaker. The remote end will trip if the trip signal is received and the guard signal is
not received. This scheme is simple, but loss of the channel means that a portion of the line is
not covered by high-speed tripping. Interference on the channel may produce a false trip.
In the permissive underreaching transfer trip (PUTT) scheme, the local relay first zone
(underreaching the remote end) initiates a trip of the local breaker and keys the transmitter to
send a trip signal. The remote breaker will trip if it receives a trip signal, does not receive a
-3.5.16-
Protective Relays
guard signal, and sees the fault in the tripping direction in zone 2 or zone 3 (overreaching the
other end). In essence, the zone 2 or 3 relay is used to supervise the transfer trip receiver, and
the transfer trip signal merely permits the tripping.
The permissive overreaching transfer trip (POTT) scheme uses only the zone 2 (or 3) relays to
overreach the remote end. The local relay sends a trip signal to the remote end. If the remote
end receives a trip signal (and no guard signal) and also sees the fault (with its overreaching
zone) then it trips. Note that on a multiterminal line, the relay must receive a trip signal from all
terminals to permit tripping. Figure 3.5.14 illustrates the POTT scheme.
ZONE 2 RELAY R
ZONE 1 RELAY R
21
R
21
S
ZONE 2 RELAY S
ZONE 1 RELAY S
21 ZONE 2
TRIP
GUARD RCVR
21
ZONE 1
52 a
52 TC
Figure 3.5.14. Permissive overreaching transfer tripping scheme. Relay must see a trip signal
from all terminals to trip.
The directional comparison unblocking scheme is really a tripping-type scheme, in which the
guard signal blocks tripping. Each terminal is allowed to trip if its zone 2 or 3 sees the fault and
the blocking signal is removed (either by the fault on a carrier system or by the transmitter at the
other end or both). This is really a POTT scheme, except that there is no need to transmit a
signal through a fault on a carrier system. The receiver has logic to compensate for attenuation
of the signal and for noise spikes. If excessive noise is detected, the trip will be blocked. If low
noise is detected, no blocking signal is detected, but a low tripping signal (or no tripping signal)
then tripping is permitted.
-3.5.17-
C. W. Brice
August 2002
-3.5.18-