POWER SYSTEM PROTECTION

LEARNER MODULE

Feeder differential
 Line Differential Protection

Differential Protection

Used where:
 Protection co-ordination is difficult / not possible
using time delayed elements
 Fast fault clearance is critical

Can be used on all items of plant in one form or another

(i.e. Generators, transformers, lines, cables, busbars,
motors) to form a defined zone of protection dictated by
current transformer location.

Page 1
 Biased Characteristic

Differential
Current 30% Bias
5.0

4.0

OPERATE

3.0

RESTRAIN
2.0

1.0

0.2

2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0

Through Current

Low Impedance Principle

 Does nott rely

D l on equall ratio
ti CT’
CT’s (ratio
( ti correction
ti can
be done via external interposing CT’s or internally to
the relay)
 Limit on allowable spill dictated by percentage of bias
used - practical limit to give sufficient sensitivity to
internal faults
 Amount of spill related to CT size - stability limits and
CT requirements are therefore dependant on each
other

Page 2
 Unit Protection Involving Distance
Between Circuit Breakers
A B

Relaying
R
Point
Trip
pB

Trip A

Simple Local Differential Protection

Unit Protection Involving Distance

Between Circuit Breakers

A B
Communication
Channel

Relaying Relaying
Point Point
R R

Trip A Trip B

Unit Protection Involving Distance

Between Circuits

Page 3
Basic Pilot Wire Schemes with
Bias

B B
I
V
OP OP

V
I

Circulating Current

Summation Transformers

Page 4
Summation Transformer Sensitivity
for Different Faults
IA
1
IB
1
IC Output for operation
= K

IN

Let output for operation = K

(1) Consider A-E fault

for relay operation : IA (1 + 1 + 3) > K
IA > 1/5K or 20%K

Summation Transformer Sensitivity

for Different Faults
(2) B-E fault
for relay operation : IB (1 + 3) > K
IB > 25%K

(3) C-E fault

for relay operation : IC x (3) > K
IC > 331/3%K

(4) AB fault
for relay operation : IAB x (1) > K
IAB > 100%K

(5) BC fault
for relay operation : IBC x (1) > K
IBC > 100%K

(6) AC fault
for relay operation : IAC (1 + 1) > K
IAC > 50%K

Page 5
Zero Summation Transformer Output for
2-1-1 Current Distribution

Fault Settings for Plain Feeders

Input transformer summation ratio is 1.25 : 1 : N
where N = 3 for normal use
and N = 6 to give low earth fault settings

Fault Settings
N = 3 N = 6
A-N 0.19 x Ks x In 0.12 x Ks x In
B-N 0.25 x Ks x In 0.14 x Ks x In
C-N 0.33 x Ks x In 0.17 x Ks x In

A-B 0.80 x Ks x In
B-C 1.00 x Ks x In
C-A 0.44 x Ks x In
A-B-C 0.51 x Ks x In
Ks is a setting multiplier, variable from 0.5 to 2.0
In is the relay rated current 1 Amp or 5 Amps

Page 6
 Line Charging Current

Line Charging Currents

Line charging currents flow in at one end of the feeder only

and is therefore potentially capable of unbalancing a
protective
t ti system.
t

Charging currents (or capacitance currents) of overhead lines

generally low.

Charging current levels of underground cables however can

be high enough to dictate minimum permissible operating
level of the protection.

Page 7
 Line Charging Current
ICA
1
ICB
1
ICC

Line charging
g g current produce
p certain amount of For stability ,
unbalancing AT’s under normal steady state conditions
when balanced 3Ø charging current flows. (2  N)  A  3 C

Unbalancing AT’s = 3 IC
3
A  C
AT’s of most sensitive fault setting = (2 + N) IA (2  N)

where IA = setting for A-E fault

For N  3,  A  0.35  C

Solidly Earthed System

Maximum unbalance amp-turns is obtained when an external

double phase to ground fault on phases B & C occurs.

Charging
Ch i currentt flows
fl in
i phase
h A only
l since
i B & C phase
h
capacitances are shorted out.

IC
1

Page 8
 Solid Earthed System

Unbalancing AT’s due to charging current = (2 + N) IC

AT’s
AT s of most sensitive fault setting = (2 + N) IA

where IA = setting for A-E fault

For stability ,

((2 + N)) IA > ((2 + N)) IC

IA > IC

Note: For EHV cables, one relay per phase is sometimes used to avoid
lack of sensitivity for phase faults.

Pilot Characteristics

Page 9
 Pilot Wire

Resistance and shunt capacitance of pilots introduce magnitude and phase differences
in pilot terminal currents.

Pilot Resistance
Attenuates the signal and affects effective minimum operating levels.

To maintain constant operating levels for wide range of pilot resistance, padding
resistor used.

R Rp/2 R

Rp/2

Padding resistance R set to ½ (1000 - Rp) ohms

Pilot Capacitance

Circulating current systems :

 Pilot capacitance effectively in parallel with relay

operating coil.

 Capacitance at centre of pilots has zero volts across

them.

Page 10
 Pilot Isolation

Electromagnetic Induction

Field of any adjacent conductor may induce a voltage in the pilot circuit.

Induced voltage can be severe when :

(1) Pilot wire laid in parallel to a power circuit.
(2) Pilot wire is long and in close proximity to power circuit.
(3) Fault Current is severe.

Induced voltage may amount to several thousand volts.

Danger to personnel
Danger to equipment

Difference in Station Earth Potentials

Can be a problem for applications above 33kV - even if feeder is short.

Formula for Induced Voltage

e = 0.232 I L Log10 De/S

where I = primary line E/F current

L = length of pilots in miles
De = Equiv. Depth of earth return in metres = 655 . e/f
e = soil resistivity in .m
f = frequency
s = separation between power line and pilot circuit in
metres

Effect of screening is not considered in the formula.

If the pilot is enclosed in lead sheath earthed at each end, screening is provided by
the current flowing in the sheath.
Sheath should be of low resistance.

0.3 V / A / Mile Unscreened Pilots

0.1 V / A / Mile Screened Pilots

Page 11
Pilot circuits and all directly connected equipment should be insulated to
earth and other circuits to an adequate voltage level.

T
Two levels
l l are recognised
i d as standard
t d d : 5kV & 15kV

Relay Case

5kV 15kV
Pilot
Terminal
Relay
Input Relay
Circuit
Pilot
Wire
2kV 5kV

Supervision & Intertripping

Page 12
 Supervision of Pilot Circuits

Pilot circuits are subject to a number of hazards, such as :

- Manual Interference
- Acts of Nature (storms, subsidence, etc.)
- Mechanical Damage (excavators, impacts)
Therefore supervision of the pilots is felt to be necessary.
Two types exist :
- Signal injection type
- Wheatstone Bridge type

Pilot Wire Supervision

Circulating
Current
Schemes

Pilot Wire Maloperate

Open Circuited
Pilot Wire Stable
Short Circuited
Pilot Wire Maloperate
Crossed

Maloperation occurs even under normal loading conditions if

3-phase setting < ILOAD.
Overcurrent check may be used to prevent maloperation.
Overcurrent element set above maximum load current.

Page 13
 Overcurrent Check Relays

50
A

PILOT
C
50C WIRE
RELAY
(87PW)

50
G

Overcurrent Check Relays

50A-1 87PW-1
+ TRIP CIRCUITS

50C-1

50G-1 Isoc > Ifl

0.9
Isef > 1.2
1 2 IZ
Isef < 0.8 x Ief

Page 14
 System Requiring Intertripping

Source

Feeder
Protection
Busbar
Protection

Transformer Feeder - Intertripping

PILOT WIRES

PILOT WIRE
TRIP TRIP
PROTECTION PILOT WIRE TRANSFORMER
PROTECTION PROTECTION

DESTABILISE AND
INTERTRIP

Page 15
 Numerical Feeder
Differential Protection

All Digital Design

0IIIIII0I0.....0I0IIIIII0
Digital messages
End A End B

A/D P

Digital communication interface

Page 16
Direct Optical Fibre Link

OPGW

Interfacing to PCM Multiplexers

interface
Multiplexer
unit G.703
850nm or V.35
multimode electrical
optical fibre
interface
ISDN
unit X.21
electrical

Page 17
 Multiplexed Optical Link

Earth wire optical fibre

Multiplexer Multiplexer
34 Mbit/s

64k Telephone
bits/s
Telecontrol
End A End B
Teleprotection

Multiplexed Microwave Link

PCM
Multiplexer

Telephone
64k
bits/s Telecontrol

End A End B
Teleprotection

Page 18
 Optical Budgets for Direct Optical
Connection Between Relays

850nm Multi 1300nm Multi 1300nm Single 1550nm Single

Mode Mode Mode Mode
Min. Transmit -19.8dBm -8.2dBm -8.2dBm -8.2dBm
Output Level
Receiver -25.4dBm -38.2dBm -38.2dBm -38.2dBm
Sensitivity
Optical Budget 5.6dB 30.0dB 30.0dB 30.0dB
Less Safety 2.6dB 27.0dB 27.0dB 27.0dB
Margin (3dB)*
Typical Cable 2.6dB/km 0.8dB/km 0.4dB/km 0.3dB/km
Loss
Max. 1km 33.7km 67.5km 90km
Transmission
Distance

Short Haul Long Haul

Key: * 3dB allowance for joint loss/ageing

Propagation Delay Problem

Relay A Relay B

Current at B

Current received from A

Propagation delay

Page 19
 Propagation Delay Compensation

 Synchronise sampling in both relays

 Direct comparison of samples
 IRIG-B a possibility, but not always
available (= protection out of service)

 Asynchronous
y sampling
p g
 Continual time difference measurement
 Vector transformation in software

MiCOM-P540-39

Capacitive Current Compensation

IL IR
ZL

IchL IchR
VL VR

 C
Conventional current differential
ff protection needs to be set
above Ich to avoid tripping for 50/60Hz line charging
current
 Modern relays can extract Ich to allow more sensitive
current settings to be applied

Page 20
Protection of Transformer Feeders

Power transformer

Ratio
correction

Vectorial
correction

Virtual interposing CT Virtual interposing CT

Direct Intertrip

Relay Relay
A B Transformer
Protection
DTT=1
Data
Message
- + - +

 The channel could alternatively be used along with

the distance elements and PSL to configure
permissive or blocking distance aided schemes

Page 21
 Permissive Intertrip

IB

Relay A Relay B
Busbar
Relay
PIT=1
Data Message

- + - +

 Example shows interlocked overcurrent protection

 Feeder fault seen within busbar zone
 Remote end trip after set delay for PIT & current >
Is1

Page 22
Chap10-152-169 21/06/02 8:42 Page 152

• 10 • Unit Protection of Feeders

Introduction 10.1
Convention of direction 10.2
Conditions for direction comparison 10.3
Circulating current system 10.4
Balanced voltage system 10.5
Summation arrangements 10.6
Examples of electromechanical
and static unit protection systems 10.7
Digital/Numerical current differential
protection systems 10.8
Carrier unit protection schemes 10.9
Current differential scheme
– analogue techniques 10.10
Phase comparison protection
scheme considerations 10.11
Examples 10.12
References 10.13
Chap10-152-169 21/06/02 8:43 Page 153

• 10 • Unit P rotection of Feeders

10 . 1 I N T R O D U C T I O N
The graded overcurrent systems described in Chapter 9,
though attractively simple in principle, do not meet all
the protection requirements of a power system.
Application difficulties are encountered for two reasons:
firstly, satisfactory grading cannot always be arranged
for a complex network, and secondly, the settings may
lead to maximum tripping times at points in the system
that are too long to prevent excessive disturbances
occurring.
These problems led to the concept of 'Unit Protection',
whereby sections of the power system are protected
individually as a complete unit without reference to
other sections. One form of ‘Unit Protection’ is also
known as ‘Differential Protection’, as the principle is to
sense the difference in currents between the incoming
and outgoing terminals of the unit being protected.
Other forms can be based on directional comparison, or
distance teleprotection schemes, which are covered in
Chapter 12, or phase comparison protection, which is
discussed later in this chapter. The configuration of the
power system may lend itself to unit protection; for
instance, a simple earth fault relay applied at the source
end of a transformer-feeder can be regarded as unit
protection provided that the transformer winding
associated with the feeder is not earthed. In this case
the protection coverage is restricted to the feeder and
transformer winding because the transformer cannot
transmit zero sequence current to an out-of-zone fault.
In most cases, however, a unit protection system
involves the measurement of fault currents (and possibly
voltages) at each end of the zone, and the transmission
of information between the equipment at zone
boundaries. It should be noted that a stand-alone
distance relay, although nominally responding only to
faults within their setting zone, does not satisfy the
conditions for a unit system because the zone is not
clearly defined; it is defined only within the accuracy
limits of the measurement. Also, to cater for some
conditions, the setting of a stand-alone distance relay
may also extend outside of the protected zone to cater
for some conditions.
Merz and Price [10.1] first established the principle of
current differential unit systems; their fundamental
differential systems have formed the basis of many

Network Protection & Automation Guide • 153 •

Chap10-152-169 21/06/02 8:43 Page 154

highly developed protection arrangements for feeders 10 . 2 C O N V E N T I O N O F D I R E C T I O N

and numerous other items of plant. In one arrangement,
It is useful to establish a convention of direction of
an auxiliary ‘pilot’ circuit interconnects similar current
current flow; for this purpose, the direction measured
transformers at each end of the protected zone, as
from a busbar outwards along a feeder is taken as
shown in Figure 10.1. Current transmitted through the
positive. Hence the notation of current flow shown in
zone causes secondary current to circulate round the
Figure 10.3; the section GH carries a through current
pilot circuit without producing any current in the relay.
which is counted positive at G but negative at H, while
For a fault within the protected zone the CT secondary
the infeeds to the faulted section HJ are both positive.
currents will not balance, compared with the through-
fault condition, and the difference between the currents
will flow in the relay.
An alternative arrangement is shown in Figure 10.2, in Source Source
+ _ + +
which the CT secondary windings are opposed for
through-fault conditions so that no current flows in the
series connected relays. The former system is known as
a ‘Circulating Current’ system, while the latter is known Fault
G H J
as a ‘Balanced Voltage’ system.
Figure 10.3: Convention of current direction
End G End H

Neglect of this rule has often led to anomalous

arrangements of equipment or difficulty in describing
Id>
the action of a complex system. When applied, the rule
Relay
will normally lead to the use of identical equipments at
U n i t P ro te c t i o n Fe e d e r s

the zone boundaries, and is equally suitable for extension

Figure 10.1: Circulating current system to multi-ended systems. It also conforms to the standard
methods of network analysis.
End G End H

10 . 3 C O N D I T I O N S
F O R D I R E C T I O N C O M PA R I S O N
The circulating current and balanced voltage systems of
Figures 10.1 and 10.2 perform full vectorial comparison
Id> Id> of the zone boundary currents. Such systems can be
Relay G Relay H treated as analogues of the protected zone of the power
system, in which CT secondary quantities represent
Figure 10.2: Balanced voltage system
primary currents and the relay operating current
corresponds to an in-zone fault current.
• 10 • These systems are simple in concept; they are
Most systems of unit protection function through the nevertheless applicable to zones having any number of
determination of the relative direction of the fault boundary connections and for any pattern of terminal
current. This direction can only be expressed on a currents.
comparative basis, and such a comparative measurement To define a current requires that both magnitude and
is the common factor of many systems, including phase be stated. Comparison in terms of both of these
directional comparison protection and distance quantities is performed in the Merz-Price systems, but it
teleprotection schemes with directional impedance is not always easy to transmit all this information
measurement. over some pilot channels. Chapter 8 provides a detailed
A major factor in consideration of unit protection is the description of modern methods that may be used.
method of communication between the relays. This is
covered in detail in Chapter 8 in respect of the latest
fibre-optic based digital techniques. For older ‘pilot wire’ 10 . 4 C I R C U L AT I N G C U R R E N T S Y S T E M
systems, only brief mention is made. For more detailed The principle of this system is shown in outline in
descriptions of ‘pilot wire’ techniques, see reference Figure 10.1. If the current transformers are ideal, the
[10.2] in Section 10.13. functioning of the system is straightforward. The

• 154 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:43 Page 155

transformers will, however, have errors arising from both unacceptable. One solution is to include a stabilising
Wattmetric and magnetising current losses that cause resistance in series with the relay. Details of how to
deviation from the ideal, and the interconnections calculate the value of the stabilising resistor are usually
between them may have unequal impedances. This can included in the instruction manuals of all relays that
give rise to a ‘spill’ current through the relay even require one.
without a fault being present, thus limiting the
When a stabilising resistor is used, the relay current
sensitivity that can be obtained. Figure 10.4 illustrates
setting can be reduced to any practical value, the relay
the equivalent circuit of the circulating current scheme.
now being a voltage-measuring device. There is
If a high impedance relay is used, then unless the relay is
obviously a lower limit, below which the relay element
located at point J in the circuit, a current will flow
does not have the sensitivity to pick up. Relay
through the relay even with currents IPg and IPh being
calibration can in fact be in terms of voltage. For more
identical. If a low impedance relay is used, voltage FF ’
details, see reference [10.2].
will be very small, but the CT exciting currents will be
unequal due to the unequal burdens and relay current IR
will still be non-zero. 10.4.2 Bias
The 'spill' current in the relay arising from these various
IPg
End G End H
IPh
sources of error is dependent on the magnitude of the
through current, being negligible at low values of
through-fault current but sometimes reaching a
RSh RLg RLh RSh disproportionately large value for more severe faults.
iSg Sh ieh Setting the operating threshold of the protection above
ieg
Id Relay the maximum level of spill current produces poor
Zeg R Zeh sensitivity. By making the differential setting
approximately proportional to the fault current, the low-

U n i t P ro te c t i o n Fe e d e r s
RR
level fault sensitivity is greatly improved. Figure 10.5
illustrates a typical bias characteristic for a modern relay
(a)
Subscripts:
p that overcomes the problem. At low currents, the bias is
G' S - CT Secondary small, thus enabling the relay to be made sensitive. At
F' L higher currents, such as would be obtained from inrush or
G''
G
h - end H through fault conditions, the bias used is higher, and thus
J H
G F the spill current required to cause operation is higher. The
relay is therefore more tolerant of spill current at higher
H' fault currents and therefore less likely to maloperate,
H
H'' while still being sensitive at lower current levels.
GG' '
GG'' '' Electro-motive forces with low impedance relay
(b)

I1 I2
Figure 10.4: Equivalent circuit
of circulating current scheme
• 10 •
I3

10.4.1 Transient Instability Idiff

It is shown in Section 6.4.10 that an asymmetrical = I1+I2+I3
current applied to a current transformer will induce a Percentage
flux that is greater than the peak flux corresponding to bias k2
Operate
the steady state alternating component of the current. It
may take the CT into saturation, with the result that the
dynamic exciting impedance is reduced and the exciting Percentage
current greatly increased. bias k1
Restrain
When the balancing current transformers of a unit Is1

protection system differ in excitation characteristics, or

have unequal burdens, the transient flux build-ups will
Is2 I1 + I2 + I3
differ and an increased 'spill' current will result. There is Ibias=
a consequent risk of relay operation on a healthy circuit 2
under transient conditions, which is clearly Figure 10.5: Typical bias characteristic of relay

Network Protection & Automation Guide • 155 •

Chap10-152-169 21/06/02 8:43 Page 156

10 . 5 B A L A N C E D V O LTA G E S Y S T E M exciting current, because the whole of the primary

current is expended as exciting current. In consequence,
This section is included for historical reasons, mainly
the secondary e.m.f. is an accurate measure of the
because of the number of such schemes still to be found
primary current within the linear range of the
in service – for new installations it has been almost
transformer. Provided the transformers are designed to
completely superseded by circulating current schemes. It
be linear up to the maximum value of fault current,
is the dual of the circulating current protection, and is
balance is limited only by the inherent limit of accuracy
summarised in Figure 10.2 as used in the ‘Translay H04’
of the transformers, and as a result of capacitance
scheme.
between the pilot cores. A broken line in the equivalent
With primary through current, the secondary e.m.f.’s of circuit shown in Figure 10.6 indicates such capacitance.
the current transformers are opposed, and provide no Under through-fault conditions the pilots are energised
current in the interconnecting pilot leads or the series to a proportionate voltage, the charging current flowing
connected relays. An in-zone fault leads to a circulating through the relays. The stability ratio that can be
current condition in the CT secondaries and hence to achieved with this system is only moderate and a bias
relay operation. technique is used to overcome the problem.
An immediate consequence of the arrangement is that
the current transformers are in effect open-circuited, as
10 . 6 S U M M AT I O N A R R A N G E M E N T S
no secondary current flows for any primary through-
current conditions. To avoid excessive saturation of the Schemes have so far been discussed as though they were
core and secondary waveform distortion, the core is applied to single-phase systems. A polyphase system
provided with non-magnetic gaps sufficient to absorb could be provided with independent protection for each
the whole primary m.m.f. at the maximum current level, phase. Modern digital or numerical relays
the flux density remaining within the linear range. The communicating via fibre-optic links operate on this
secondary winding therefore develops an e.m.f. and can basis, since the amount of data to be communicated is
U n i t P ro te c t i o n Fe e d e r s

be regarded as a voltage source. The shunt reactance of not a major constraint. For older relays, use of this
the transformer is relatively low, so the device acts as a technique over pilot wires may be possible for relatively
transformer loaded with a reactive shunt; hence the short distances, such as would be found with industrial
American name of transactor. The equivalent circuit of and urban power distribution systems. Clearly, each
the system is as shown in Figure 10.6. phase would require a separate set of pilot wires if the
protection was applied on a per phase basis. The cost of
The series connected relays are of relatively high
providing separate pilot-pairs and also separate relay
impedance; because of this the CT secondary winding
elements per phase is generally prohibitive. Summation
resistances are not of great significance and the pilot
techniques can be used to combine the separate phase
resistance can be moderately large without significantly
currents into a single relaying quantity for comparison
affecting the operation of the system. This is why the
over a single pair of pilot wires. For details of such
scheme was developed for feeder protection.
techniques, see reference [10.2].
End G End H

10 . 7 E X A M P L E S O F E L E C T R O M E C H A N I C A L
• 10 • RSg RLg RLh RSh A N D S TAT I C U N I T P R OT E C T I O N S Y S T E M S
As mentioned above, the basic balanced voltage principle
of protection evolved to biased protection systems.
Zeg Zeh
Pilot Several of these have been designed, some of which
Parameters
appear to be quite different from others. These
dissimilarities are, however, superficial. A number of
Id> Id> these systems that are still in common use are described
Relay G Relay H below.
Figure 10.6: Equivalent circuit
for balanced voltage system
10.7.1 ‘Translay’ Balanced Voltage
Electromechanical System
A typical biased, electromechanical balanced voltage
10.5.1 Stability Limit of the Voltage Balance System
system, trade name ‘Translay’, still giving useful service
Unlike normal current transformers, transactors are not on distribution systems is shown in Figure 10.7.
subject to errors caused by the progressive build-up of

• 156 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:43 Page 157

End G End H Bias is produced by a copper shading loop fitted to the

A pole of the upper magnet, thereby establishing a Ferraris
B
C motor action that gives a reverse or restraining torque
proportional to the square of the upper magnet flux value.
Typical settings achievable with such a relay are:
Summation
winding Least sensitive earth fault - 40% of rating
Secondaryy Least sensitive phase-phase fault - 90% of rating
winding
Pilot Three-phase fault - 52% of rating
Bias loop

Figure 10.7: Typical biased

electromechanical differential protection system. 10.7.2 Static Circulating Current
Unit Protection System – ‘Translay ‘S’ ’
The electromechanical design derives its balancing voltages A typical static modular pilot wire unit protection system
from the transactor incorporated in the measuring relay at operating on the circulating current principle is shown in
each line end. The latter are based on the induction-type Figure 10.8. This uses summation transformers with a
meter electromagnet as shown in Figure 10.7. neutral section that is tapped, to provide alternative
earth fault sensitivities. Phase comparators tuned to the
The upper magnet carries a summation winding to power frequency are used for measurement and a
receive the output of the current transformers, and a restraint circuit gives a high level of stability for through
secondary winding which delivers the reference e.m.f. faults and transient charging currents. High-speed
The secondary windings of the conjugate relays are operation is obtained with moderately sized current
interconnected as a balanced voltage system over the transformers and where space for current transformers is
pilot channel, the lower electromagnets of both relays limited and where the lowest possible operating time is
being included in this circuit. not essential, smaller current transformers may be used.

U n i t P ro te c t i o n Fe e d e r s
Through current in the power circuit produces a state of This is made possible by a special adjustment (Kt) by
balance in the pilot circuit and zero current in the lower which the operating time of the differential protection
electromagnet coils. In this condition, no operating can be selectively increased if necessary, thereby
torque is produced. enabling the use of current transformers having a
correspondingly decreased knee-point voltage, whilst
An in-zone fault causing an inflow of current from each
ensuring that through-fault stability is maintained to
end of the line produces circulating current in the pilot
greater than 50 times the rated current.
circuit and the energisation of the lower electromagnets.
These co-operate with the flux of the upper Internal faults give simultaneous tripping of relays at
electromagnets to produce an operating torque in the both ends of the line, providing rapid fault clearance
discs of both relays. An infeed from one end only will irrespective of whether the fault current is fed from both
result in relay operation at the feeding end, but no line ends or from only one line end.
operation at the other, because of the absence of upper
magnet flux.

• 10 •
A
T1 - Summation transformer
B
T2 - Auxiliary transformer
C
RVO - Non linear resistor

Trip Trip To - Operating winding

T2 Pr Pr T2
Tr - Restraining winding
Rs T1 Tr Tr T1 Rs
Ro - Linear resistor
c c
O TO
Pilot wires Pr - Pilots padding resistor

RVO Ro Ro RVO c - Phase comparator

V V

Figure 10.8: Typical static circulating current feeder unit protection circuit diagram

Network Protection & Automation Guide • 157 •

Chap10-152-169 21/06/02 8:43 Page 158

10 . 8 D I G I TA L / N U M E R I C A L C U R R E N T operate as a result. In older protection schemes, the

D I F F E R E N T I A L P R OT E C T I O N S Y S T E M S problem was eliminated by delta connection of the CT
secondary windings. For a digital or numerical relay, a
A digital or numerical unit protection relay may typically
selectable software zero sequence filter is typically
provide phase-segregated current differential protection.
employed.
This means that the comparison of the currents at each
relay is done on a per phase basis. For digital data The problem remains of compensating for the time
communication between relays, it is usual that a direct difference between the current measurements made at
optical connection is used (for short distances) or a the ends of the feeder, since small differences can upset
multiplexed link. Link speeds of up to 64kbit/s (56kbit/s the stability of the scheme, even when using fast direct
in N. America) are normal. Through current bias is fibre-optic links. The problem is overcome by either time
typically applied to provide through fault stability in the synchronisation of the measurements taken by the
event of CT saturation. A dual slope bias technique relays, or calculation of the propagation delay of the link
(Figure 10.5) is used to enhance stability for through continuously.
faults. A typical trip criterion is as follows:
For |Ibias| < Is2
10.8.1 Time Synchronisation of Relays
|Idiff | < k1 |Ibias| + Is1
Fibre-optic media allow direct transmission of the
For |Ibias| < Is2 signals between relays for distances of up to several km
without the need for repeaters. For longer distances
|Idiff | < k2 |Ibias| - (k2 - k1) Is2 + Is1
repeaters will be required. Where a dedicated fibre pair
Once the relay at one end of the protected section has is not available, multiplexing techniques can be used. As
determined that a trip condition exists, an intertrip phase comparison techniques are used on a per phase
signal is transmitted to the relay at the other end. Relays basis, time synchronisation of the measurements is
that are supplied with information on line currents at all vitally important. This requires knowledge of the
U n i t P ro te c t i o n Fe e d e r s

ends of the line may not need to implement intertripping transmission delay between the relays. Four techniques
facilities. However, it is usual to provide intertripping in are possible for this:
any case to ensure the protection operates in the event
a. assume a value
of any of the relays detecting a fault.
b. measurement during commissioning only
A facility for vector/ratio compensation of the measured c. continuous online measurement
currents, so that transformer feeders can be included in d. GPS time signal
the unit protection scheme without the use of
Method (a) is not used, as the error between the assumed
interposing CT’s or defining the transformer as a separate
and actual value will be too great.
zone increases versatility. Any interposing CT’s required
are implemented in software. Maloperation on Method (b) provides reliable data if direct
transformer inrush is prevented by second harmonic communication between relays is used. As signal
detection. Care must be taken if the transformer has a propagation delays may change over a period of years,
wide-ratio on-load tap changer, as this results in the repeat measurements may be required at intervals and
current ratio departing from nominal and may cause relays re-programmed accordingly. There is some risk of
• 10 • maloperation, depending on the sensitivity of the relays. maloperation due to changes in signal propagation time
The initial bias slope should be set taking this into causing incorrect time synchronisation between
consideration. measurement intervals. The technique is less suitable if
rented fibre-optic pilots are used, since the owner may
Tuned measurement of power frequency currents
perform circuit re-routing for operational reasons
provides a high level of stability with capacitance inrush
without warning, resulting in the propagation delay
currents during line energisation. The normal steady-
being outside of limits and leading to scheme
state capacitive charging current can be allowed for if a
maloperation. Where re-routing is limited to a few
voltage signal can be made available and the
routes, it may be possible to measure the delay on all
susceptance of the protected zone is known.
routes and pre-program the relays accordingly, with the
Where an earthed transformer winding or earthing relay digital inputs and ladder logic being used to detect
transformer is included within the zone of protection, changes in route and select the appropriate delay
some form of zero sequence current filtering is required. accordingly.
This is because there will be an in-zone source of zero
Method (c), continuous sensing of the signal propagation
sequence current for an external earth fault. The
delay, is a robust technique. One method of achieving
differential protection will see zero sequence differential
this is shown in Figure 10.9.
current for an external fault and it could incorrectly

• 158 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:43 Page 159

Digital communications link

A B
End A End B

Measured sampling time Propagation delay time TA1'TA2' - sampling instants of relay A
TB3*=(TA*-Tp2) Tp1=Tp2=1/2(TA*-TA1-Td) TB1'TB2' - sampling instants of relay B
Tp1 - propagation delay time
Current from relay A to B
TA1 vectors
TB1 Tp2 - propagation delay time
Tp1 TA1 from relay B to A
TA2
TB2 Td - time between the arrival
TB* of message TA1 at relay B
TA3 Td and despatch of message TB3
TB3* TB3 TA1* - arrival time of message TB3
TA4 and relay A
Tp2 TB4
vectors TB* - arrival time of message TA1
Current
TA5 Td and relay B
TA*
TB3 TA1 TB5 TB3* - the measured sampling
time of TB3 by relay A

Figure 10.9: Signal propagation delay measurement

Relays A and B sample signals at time TA1,TA2 …and communication channels under well-defined conditions.

U n i t P ro te c t i o n Fe e d e r s
TB1,TB2 …respectively. The times will not be coincident,
The technique can also be used with all types of pilots,
even if they start coincidentally, due to slight differences
subject to provision of appropriate interfacing devices.
in sampling frequencies. At time TA1 relay A transmits
its data to relay B, containing a time tag and other data. Method (d) is also a robust technique. It involves both
Relay B receives it at time TA1 +Tp1 where Tp1 is the relays being capable of receiving a time signal from
propagation time from relay A to relay B. Relay B a GPS satellite. The propagation delay on each
records this time as time TB*. Relay B also sends communication channel is no longer required to be
messages of identical format to relay A. It transmits such known or calculated as both relays are synchronised to a
a message at time TB3, received by relay A at time common time signal. For the protection scheme to meet
TB3 +Tp2 (say time TA*), where Tp2 is the propagation the required performance in respect of availability and
time from relay B to relay A. The message from relay B maloperation, the GPS signal must be capable of reliable
to relay A includes the time TB3, the last received time receipt under all atmospheric conditions. There is extra
tag from relay A (TA1) and the delay time between the satellite signal receiving equipment required at both
arrival time of the message from A (TB*) and TB3 – call ends of the line, which implies extra cost.
this the delay time Td. The total elapsed time is The minimum setting that can be achieved with such
therefore: • 10 •
techniques while ensuring good stability is 20% of CT
(TA* - TA1) = (Td + Tp1 + Tp2) primary current.
If it is assumed that Tp1 = Tp2, then the value of Tp1 and
Tp2 can be calculated, and hence also TB3. The relay B 10.8.2 Application to Mesh Corner
measured data as received at relay A can then be and 1 1/2 Breaker Switched Substations
adjusted to enable data comparison to be performed.
Relay B performs similar computations in respect of the These substation arrangements are quite common, and
data received from relay A (which also contains similar the arrangement for the latter is shown in Figure 10.10.
time information). Therefore, continuous measurement Problems exist in protecting the feeders due to the
of the propagation delay is made, thus reducing the location of the line CT’s, as either Bus 1 or Bus 2 or both
possibility of maloperation due to this cause to a can supply the feeder. Two alternatives are used to
minimum. Comparison is carried out on a per-phase basis, overcome the problem, and they are illustrated in the
so signal transmission and the calculations are required Figure. The first is to common the line CT inputs (as
for each phase. A variation of this technique is available shown for Feeder A) and the alternative is to use a second
that can cope with unequal propagation delays in the two set of CT inputs to the relay (as shown for Feeder B).

Network Protection & Automation Guide • 159 •

Chap10-152-169 21/06/02 8:46 Page 160

Bus 1 Bus 2
power system currents since the systems are designed to
B2 IF B1 operate at much higher frequencies, but each medium
may be subjected to noise at the carrier frequencies that
may interfere with its correct operation. Variations of
F signal level, restrictions of the bandwidth available for
Id> relaying and other characteristics unique to each
Id>
Stub medium influence the choice of the most appropriate
bus type of scheme. Methods and media for communication
inputs are discussed in Chapter 8.
A B

Figure 10.10: Breaker and 10 . 10 C U R R E N T D I F F E R E N T I A L S C H E M E –

a half switched substation
ANALOGUE TECHNIQUES
In the case of a through fault as shown, the relay The carrier channel is used in this type of scheme to
connected to Feeder A theoretically sees no unbalance convey both the phase and magnitude of the current at
current, and hence will be stable. However, with the line one relaying point to another for comparison with the
disconnect switch open, no bias is produced in the relay, phase and magnitude of the current at that point.
so CT’s need to be well matched and equally loaded if Transmission techniques may use either voice frequency
maloperation is to be avoided. channels using FM modulation or A/D converters and
For Feeder B, the relay also theoretically sees no digital transmission. Signal propagation delays still need
differential current, but it will see a large bias current even to be taken into consideration by introducing a
with the line disconnect switch open. This provides a high deliberate delay in the locally derived signal before a
degree of stability, in the event of transient asymmetric CT comparison with the remote signal is made.
saturation. Therefore, this technique is preferred. A further problem that may occur concerns the dynamic
U n i t P ro te c t i o n Fe e d e r s

Sensing of the state of the line isolator through auxiliary range of the scheme. As the fault current may be up to
contacts enables the current values transmitted to and 30 times the rated current, a scheme with linear
received from remote relays to be set to zero when the characteristics requires a wide dynamic range, which
isolator is open. Hence, stub-bus protection for the implies a wide signal transmission bandwidth. In
energised part of the bus is then possible, with any fault practice, bandwidth is limited, so either a non-linear
resulting in tripping of the relevant CB. modulation characteristic must be used or detection of
fault currents close to the setpoint will be difficult.

10 . 9 C A R R I E R U N I T P R OT E C T I O N S C H E M E S
10.10.1 Phase Comparison Scheme
In earlier sections, the pilot links between relays have
been treated as an auxiliary wire circuit that The carrier channel is used to convey the phase angle of the
interconnects relays at the boundaries of the protected current at one relaying point to another for comparison
zone. In many circumstances, such as the protection of with the phase angle of the current at that point.
longer line sections or where the route involves The principles of phase comparison are illustrated in
• 10 • installation difficulties, it is too expensive to provide an Figure 10.11. The carrier channel transfers a logic or
auxiliary cable circuit for this purpose, and other means 'on/off' signal that switches at the zero crossing points
are sought. of the power frequency waveform. Comparison of a local
In all cases (apart from private pilots and some short logic signal with the corresponding signal from the
rented pilots) power system frequencies cannot be remote end provides the basis for the measurement of
transmitted directly on the communication medium. phase shift between power system currents at the two
Instead a relaying quantity may be used to vary the ends and hence discrimination between internal and
higher frequency associated with each medium (or the through faults.
light intensity for fibre-optic systems), and this process Current flowing above the set threshold results in turn-
is normally referred to as modulation of a carrier wave. off of the carrier signal. The protection operates if gaps
Demodulation or detection of the variation at a remote in the carrier signal are greater than a set duration – the
receiver permits the relaying quantity to be reconstituted phase angle setting of the protection.
for use in conjunction with the relaying quantities
derived locally, and forms the basis for all carrier systems Load or through fault currents at the two ends of a
of unit protection. protected feeder are in antiphase (using the normal relay
convention for direction), whilst during an internal fault
Carrier systems are generally insensitive to induced the (conventional) currents tend towards the in-phase

• 160 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:46 Page 161

condition. Hence, if the phase relationship of through binary information, the techniques associated with
fault currents is taken as a reference condition, internal sending teleprotection commands. Blocking or
faults cause a phase shift of approximately 180° with permissive trip modes of operation are possible, however
respect to the reference condition. Figure 10.11 illustrates the more usual blocking mode,
since the comparator provides an output when neither
Phase comparison schemes respond to any phase shift
squarer is at logic '1'. A permissive trip scheme can be
from the reference conditions, but tripping is usually
realised if the comparator is arranged to give an output
permitted only when the phase shift exceeds an angle of
when both squarers are at logic '1'. Performance of the
typically 30 to 90 degrees, determined by the time delay
scheme during failure or disturbance of the carrier
setting of the measurement circuit, and this angle is
channel and its ability to clear single-end-fed faults
usually referred to as the Stability Angle. Figure 10.12 is
depends on the mode of operation, the type and function
a polar diagram that illustrates the discrimination
of fault detectors or starting units, and the use of any
characteristics that result from the measurement
additional signals or codes for channel monitoring and
techniques used in phase comparison schemes.
transfer tripping.
Since the carrier channel is required to transfer only

End G End H

Signalling equipment and

A communication channel C
Summation B D
Squarer Transmitter
network
D'
Receiver

U n i t P ro te c t i o n Fe e d e r s
Phase
E comparator

Pulse length
discrimination
Load or through fault Internal fault
F G IG IH H G IG IH H

A. Summation voltage at end G

1 1
B. Squarer output at end G
0 0

• 10 •
C. Summation voltage at end H

1 1
D. Squarer output at end H 0 0
(Received at end G via
ideal carrier system as D'

1 1
E. Comparator output at end G
0 0
E=B+D'

1 1
F. Discriminator output at end G
0 0

Stability setting

Figure 10.11: Principles of phase comparison protection.

Network Protection & Automation Guide • 161 •

Chap10-152-169 21/06/02 8:46 Page 162

both ends are nominally equal, so the receiver responds

θ=90°
equally to blocks of carrier from either end. Through-
fault current results in transmission of blocks of carrier
from both ends, each lasting for half a cycle, but with a
phase displacement of half a cycle, so that the composite
signal is continuously above the threshold level and the
θ=180°-Tripping O Stability R detector output logic is continuously '1'. Any phase shift
θ=0°
relative to the through fault condition produces a gap in
the composite carrier signal and hence a corresponding
'0' logic level from the detector. The duration of the logic
'0' provides the basis for discrimination between internal
and external faults, tripping being permitted only when a
θ=270°
time delay setting is exceeded. This delay is usually
θ System differential phase shift referred to through fault reference
condition
expressed in terms of the corresponding phase shift in
OR Through fault IG IH IG=-IH
degrees at system frequency ϕs in Figure 10.12.
reference condition G H
The advantages generally associated with the use of the
(IG' IH conventional relay currents at ends of protected feeder)
power line as the communication medium apply namely,
Discriminator stability angle setting. that a power line provides a robust, reliable, and low-loss
interconnection between the relaying points. In addition
Figure 10.12: Polar diagram for phase
comparison scheme dedicated 'on/off' signalling is particularly suited for use
in phase comparison blocking mode schemes, as signal
attenuation is not a problem. This is in contrast to
permissive or direct tripping schemes, where high power
Signal transmission is usually performed by voice output or boosting is required to overcome the extra
U n i t P ro te c t i o n Fe e d e r s

frequency channels using frequency shift keying (FSK) or attenuation due to the fault.
PLC techniques. The noise immunity is also very good, making the scheme
Voice frequency channels involving FSK use two discrete very reliable. Signal propagation delay is easily allowed
frequencies either side of the middle of the voice band. for in the stability angle setting, making the scheme very
This arrangement is less sensitive to variations in delay or sensitive as well.
frequency response than if the full bandwidth was used.
Blocking or permissive trip modes of operation may be
implemented. In addition to the two frequencies used 10 . 11 P H A S E C O M PA R I S I O N P R OT E C T I O N
for conveying the squarer information, a third tone is S C H E M E C O N S I D E R AT I O N S
often used, either for channel monitoring or transfer One type of unit protection that uses carrier techniques
tripping dependent on the scheme. for communication between relays is phase comparison
For a sensitive phase comparison scheme, accurate protection. Communication between relays commonly
compensation for channel delay is required. However, uses PLCC or frequency modulated carrier modem
since both the local and remote signals are logic pulses, techniques. There are a number of considerations that
• 10 • simple time delay circuits can be used, in contrast to the apply only to phase comparison protection systems,
analogue delay circuitry usually required for current which are discussed in this section.
differential schemes.
The principles of the Power Line Carrier channel 10.11.1 Lines with Shunt Capacitance
technique are illustrated in Figure 10.13. The scheme
operates in the blocking mode. The 'squarer' logic is used A problem can occur with the shunt capacitance current
directly to turn a transmitter 'on' or 'off' at one end, and that flows from an energising source. Since this current is
the resultant burst (or block) of carrier is coupled to and in addition to the load current that flows out of the line,
propagates along the power line which is being protected and typically leads it by more than 90°, significant
to a receiver at the other end. Carrier signals above a differential phase shifts between the currents at the ends
threshold are detected by the receiver, and hence produce of the line can occur, particularly when load current is low.
a logic signal corresponding to the block of carrier. In The system differential phase shift may encroach into the
contrast to Figure 10.11, the signalling system is a 2-wire tripping region of the simple discriminator characteristic,
rather than 4-wire arrangement, in which the local regardless of how large the stability angle setting may
transmission is fed directly to the local receiver along be. Figure 10.14 illustrates the effect and indicates
with any received signal. The transmitter frequencies at techniques that are commonly used to ensure stability.

• 162 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:46 Page 163

End G End H
Line trap
Line trapp

Coupling
Summation filter
network
A
Squarer Transmitter B
Identical
relay
C to end G
Receiver

Pulse length
discriminator
D Trip
Trip Load or through fault Internal fault
1 1
A. Squarer output at end G 0 0

Blocks of carrier transmitted from end G

Squarer output at end H 1 1

0 0

Blocks of carrier transmitted from end H

U n i t P ro te c t i o n Fe e d e r s
B. Composite carrier signal at end G

C. Carrier detector output

1 1
0 0

1 1
D. Discriminator output 0 0

Stability setting

Figure 10.13: Principles of power line carrier phase comparison

Operation of the discriminator can be permitted only

A
when current is above some threshold, so that • 10 •
measurement of the large differential phase shifts which
θc
IC
occur near the origin of the polar diagram is avoided. By
ϕs Through Fault
O
IL Reference choice of a suitable threshold and stability angle, a
'keyhole' characteristic can be provided such that the
capacitive current characteristic falls within the
resultant stability region. Fast resetting of the fault
Squarer Threshold
Starter Threshold detector is required to ensure stability following the
Limits of differential phase shift due to capacitive current IC clearance of a through fault when the currents tend to
Encroachment into tripping region for discriminator
with stability angle setting ϕs
fall towards the origin of the polar diagram.
`Keyhole' characteristic
Minimum starter threshold =
capacitive current The mark-space ratio of the squarer (or modulating)
IC sin ϕs
where ϕs = tan -1 waveform can be made dependent on the current
IL
Characteristic of system with amplitude dependent amplitude. Any decrease in the mark-space ratio will
compensation ϕs = angular compensation for current of magnitude OA
 IC  permit a corresponding differential phase shift to occur
 2sin-1 for squarer threshold IC 
 OA  between the currents before any output is given from the
IL = load current comparator for measurement in the discriminator. A
Figure 10.14: Capacitive current in phase comparison
squarer circuit with an offset or bias can provide a
schemes and techniques used to avoid instability

Network Protection & Automation Guide • 163 •

Chap10-152-169 21/06/02 8:46 Page 164

decreasing mark-space ratio at low currents, and with a In the absence of pre-fault load current, the voltages at
suitable threshold level the extra phase shift θc which is the two ends of a line are in phase. Internal faults are
permitted can be arranged to equal or exceed the phase fed from both ends with fault contributions whose
shift due to capacitive current. At high current levels the magnitudes and angles are determined by the position of
capacitive current compensation falls towards zero and the fault and the system source impedances. Although
the resultant stability region on the polar diagram is the magnitudes may be markedly different, the angles
usually smaller than on the keyhole characteristic, giving (line plus source) are similar and seldom differ by more
improvements in sensitivity and/or dependability of the than about 20°.
scheme. Since the stability region encompasses all
Hence |θG - θH| ≤ 20° and the requirements of Equation
through-fault currents, the resetting speed of any fault
10.3 are very easily satisfied. The addition of arc or fault
detectors or starter (which may still be required for other
resistance makes no difference to the reasoning above, so
purposes, such as the control of a normally quiescent
the scheme is inherently capable of clearing such faults.
scheme) is much less critical than with the keyhole
characteristic.
10.11.3 Effect of Load Current
10.11.2 System Tripping Angles When a line is heavily loaded prior to a fault the e.m.f.'s
For the protection scheme to trip correctly on internal of the sources which cause the fault current to flow may
faults the change in differential phase shift, θ0, from the be displaced by up to about 50°, that is, the power system
through-fault condition taken as reference, must exceed stability limit. To this the differential line and source
the effective stability angle of the scheme. Hence: angles of up to 20° mentioned above need to be added.
θ0 = ϕs + θc …Equation 10.1 So |θG - θH| ≤ 70° and the requirements of Equation 10.3
are still easily satisfied.
where
For three phase faults, or solid earth faults on phase-by-
ϕs = stability angle setting
U n i t P ro te c t i o n Fe e d e r s

phase comparison schemes, through load current falls to

θc = capacitive current compensation zero during the fault and so need not be considered. For
all other faults, load current continues to flow in the
(when applicable)
healthy phases and may therefore tend to increase
The currents at the ends of a transmission line IG and IH |θG - θH| towards the through fault reference value. For
may be expressed in terms of magnitude and phase low resistance faults the fault current usually far exceeds
shift θ with respect a common system voltage. the load current and so has little effect. High resistance
IG = |IG| ∠ θG faults or the presence of a weak source at one end can
prove more difficult, but high performance is still
IH = |IH| ∠ θH possible if the modulating quantity is chosen with care
Using the relay convention described in Section 10.2, the and/or fault detectors are added.
reference through-fault condition is
IG = -IH 10.11.4 Modulating Quantity
∴ IG ∠ θG = -IH ∠ θH = IH ∠ θH ± 180°
• 10 • Phase-by-phase comparison schemes usually use phase
∴ |θG - θH| =180° current for modulation of the carrier. Load and fault
currents are almost in antiphase at an end with a weak
During internal faults, the system tripping angle θ0 is the source. Correct performance is possible only when fault
differential phase shift relative to the reference current exceeds load current, or
condition.
for IF < IL’ |θG - θH| ≈ 180°
∴ θ0 =180° - |θG - θH| for IF > IL’ |θG - θH| ≈ 180° …Equation 10.4
Substituting θ0 in Equation 10.1, the conditions for where IF = fault current contribution from weak source
tripping are: IL = load current flowing towards weak source
180 - |θG - θH| ≥ ϕS + θc To avoid any risk of failure to operate, fault detectors
∴ |θG - θH| ≤ 180 - (ϕS + θc) …Equation 10.2 with a setting greater than the maximum load current
may be applied, but they may limit the sensitivity of
The term (ϕs + θc) is the effective stability angle setting
scheme. When the fault detector is not operated at one
of the scheme. Substituting a typical value of 60° in
end, fault clearance invariably involves sequential
Equation 10.2. gives the tripping condition as
tripping of the circuit breakers.
|θG - θH| ≤ 120° …Equation 10.3

• 164 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:46 Page 165

Most phase comparison schemes use summation The fault current in Equation 10.6 is the effective earth
techniques to produce a single modulating quantity, fault sensitivity IE of the scheme. For the typical values of
responsive to faults on any of the three phases. Phase
M = 6 and N = -1
sequence components are often used and a typical
modulating quantity is: M
= −6
IM = MI2 + NI1 …Equation 10.5 N
where 3
∴ IE =− IL
I1 = Positive phase sequence component 5
I2 = Negative phase sequence component Comparing this with Equation 10.4, a scheme using
M,N = constants summation is potentially 1.667 times more sensitive
With the exception of three phase faults all internal than one using phase current for modulation.
faults give rise to negative phase sequence (NPS) Even though the use of a negative value of M gives a
currents, I2, which are approximately in phase at the lower value of IE than if it were positive, it is usually
ends of the line and therefore could form an ideal preferred since the limiting condition of Im = 0 then
modulating quantity. In order to provide a modulating applies at the load infeed end. Load and fault
signal during three phase faults, which give rise to components are additive at the outfeed end so that a
positive phase sequence (PPS) currents, I1, only, a correct modulating quantity occurs there, even with the
practical modulating quantity must include some lowest fault levels. For operation of the scheme it is
response to I1 in addition to I2. sufficient therefore that the fault current contribution
Typical values of the ratio M: N exceed 5:1, so that the from the load infeed end exceeds the effective setting.
modulating quantity is weighted heavily in favour of For faults on B or C phases, the NPS components are
NPS, and any PPS associated with load current tends to displaced by 120° or 240° with respect to the PPS
be swamped out on all but the highest resistance faults.

U n i t P ro te c t i o n Fe e d e r s
components. No simple cancellation can occur, but
For a high resistance phase-earth fault, the system instead a phase displacement is introduced. For tripping
remains well balanced so that load current IL is entirely to occur, Equation 10.2 must be satisfied, and to achieve
positive sequence. The fault contribution IF provides high dependability under these marginal conditions, a
equal parts of positive, negative and zero sequence small effective stability angle is essential. Figure 10.15
components IF /3. Assuming the fault is on 'A' phase and illustrates operation near to the limits of earth fault
the load is resistive, all sequence components are in sensitivity.
phase at the infeed end G: Very sensitive schemes may be implemented by using
MI FG NI FG high values of M _ but the scheme then becomes more
∴ I mG = NI L + + N
3 3 sensitive to differential errors in NPS currents such as
the unbalanced components of capacitive current or spill
and from partially saturated CT's.
θG ≈ 0 Techniques such as capacitive current compensation and
reduction of M _ at high fault levels may be required to
At the outfeed end load current is negative, N • 10 •
ensure stability of the scheme.
MI FH NI FH
∴ I mH = − NI L + +
3 3
10.11.5 Fault Detection and Starting
Now, for
For a scheme using a carrier system that continuously
ImH > 0,θH = 0, and |θG - θH| = 0°
transmits the modulating quantity, protecting an ideal
and for line (capacitive current=0) in an interconnected
ImH < 0,θH = 180°, and |θG - θH| = 180° transmission system, measurement of current magnitude
Hence for correct operation ImH ≥ 0 might be unnecessary. In practice, fault detector or
starting elements are invariably provided and the scheme
Let ImH = 0
then becomes a permissive tripping scheme in which
Then both the fault detector and the discriminator must
operate to provide a trip output, and the fault detector
3I L
I FH = = IE may limit the sensitivity of the scheme. Requirements
M 
 +1 for the fault detectors vary according to the type of
 N  …Equation 10.6 carrier channel used, mode of operation used in the

Network Protection & Automation Guide • 165 •

Chap10-152-169 21/06/02 8:46 Page 166

ensure that during through faults, a High Set is never

System voltage reference operated when a Low Set has reset and potential race
conditions are often avoided by the transmitting of an
MIE
MIE 1.1
3 unmodulated (and therefore blocking) carrier for a short
0.9 3 ImH time following the reset of low set; this feature is often
NILH
NILH NIE θH=0
1.1 referred to as 'Marginal Guard.'
NI 3
0.9 E
3
ImG
ImG θG=0
θG=180° MIE 10.11.7 Scheme without Capacitive
MIE 1.1
0.9 NILG 3 NILG Current Compensation
3 NIE
0.9 NIE
3 1.1 3 The 'keyhole' discrimination characteristic of depends on
(a) A phase to earth fault IF = 0.9 IE (b) A phase to earth fault IF = 1.1 IE
|θG- θH |=180° |θG- θH |=0°
the inclusion of a fault detector to ensure that no
measurements of phase angle can occur at low current
NIE levels, when the capacitive current might cause large
3 NILH NIE NILH
3 MIE phase shifts. Resetting must be very fast to ensure
MIE
θH 3 stability following the shedding of through load.
3
ImH θG ImH
θH θG

NILG
NILG 10.11.8 Scheme with Capacitive
120° NI
E NIE 120° Current Compensation (Blocking Mode)
3 3 ImG
ImG
MIE MIE When the magnitude of the modulating quantity is less
3 3 than the threshold of the squarer, transmission if it
(c) B phase to earth fault IF = IE (d) C phase to earth fault IF = IE
occurred, would be a continuous blocking signal. This
|θG- θH |=70°
might occur at an end with a weak source, remote from
U n i t P ro te c t i o n Fe e d e r s

Assumptions for examples: a fault close to a strong source. A fault detector is

Infeed of load IL at end G
Outfeed of load IL at end G
required to permit transmission only when the current
M =-6 therefore I = 6I - I and from Equation 10.6
m 2 2
exceeds the modulator threshold by some multiple
N 3
effective earth fault sensitivity IE =- IL
5
(typically about 2 times) so that the effective stability
IF
also IF1 =
3
angle is not excessive. For PLCC schemes, the low set
element referred to in Section 10.11.6 is usually used for
Figure 10.15: Effect of load current on differential
phase shift |θg - θH| for resistive earth faults this purpose. If the fault current is insufficient to
at the effective earth fault sensitivity IE operate the fault detector, circuit breaker tripping will
normally occur sequentially.
phase angle measurement, that is, blocking or
permissive, and the features used to provide tolerance to
capacitive current. 10.11.9 Fault Detector Operating Quantities
Most faults cause an increase in the corresponding phase
current(s) so measurement of current increase could form
10.11.6 Normally Quiescent Power Line Carrier
• 10 • (Blocking Mode)
the basis for fault detection. However, when a line is
heavily loaded and has a low fault level at the outfeed
To ensure stability of through faults, it is essential that
end, some faults can be accompanied by a fall in current,
carrier transmission starts before any measurement of
which would lead to failure of such fault detection,
the width of the gap is permitted. To allow for
resulting in sequential tripping (for blocking mode
equipment tolerances and the difference in magnitude of
schemes) or no tripping (for permissive schemes).
the two currents due to capacitive current, two starting
Although fault detectors can be designed to respond to
elements are used, usually referred to as 'Low Set' and
any disturbance (increase or decrease of current), it is
'High Set' respectively. Low Set controls the start-up of
more usual to use phase sequence components. All
transmission whilst High Set, having a setting typically
unbalanced faults produce a rise in the NPS components
1.5 to 2 times that of the Low Set element, permits the
from the zero level associated with balanced load current,
phase angle measurement to proceed.
whilst balanced faults produce an increase in the PPS
The use of impulse starters that respond to the change in components from the load level (except at ends with very
current level enables sensitivities of less than rated low fault level) so that the use of NPS and PPS fault
current to be achieved. Resetting of the starters occurs detectors make the scheme sensitive to all faults. For
naturally after a swell time or at the clearance of the schemes using summation of NPS and PPS components
fault. Dwell times and resetting characteristics must for the modulating quantity, the use of NPS and PPS fault

• 166 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:46 Page 167

detectors is particularly appropriate since, in addition to Parameter Setting Range

any reductions in hardware, the scheme may be Differential Current Setting, Is1 0.2 -2.0 In
characterized entirely in terms of sequence components. Bias Current Threshold Setting, Is2 1-30 In
Fault sensitivities IF for PPS and NPS impulse starter Lower Percentage Bias Setting, k1 0.3-1.5
settings I1S and I2S respectively are as follows: Higher Precentage Bias Setting, k2 0.3-1.5
In - CT rated secondary current
Three phase fault IF = I1S Table 10.1: Relay Setting Ranges
Phase-phase fault IF = √3I2S
Phase-earth fault IF = 3I2S
Is2 = 2.0pu
10 . 1 2 E X A M P L E S k1 = 30%
k2 = 150%
This section gives examples of setting calculations for
simple unit protection schemes. It cannot and is not To provide immunity from the effects of line charging
intended to replace a proper setting calculation for a current, the setting of IS1 must be at least 2.5 times the
particular application. It is intended to illustrate the steady-state charging current, i.e. 4.1A or 0.01p.u., after
principles of the calculations required. The examples use taking into consideration the CT ratio of 400/1. The nearest
the ALSTOM MiCOM P541 Current Differential relay, available setting above this is 0.20p.u. This gives the points
which has the setting ranges given in Table 10.1 for on the relay characteristic as shown in Figure 10.17.
differential protection. The relay also has backup The minimum operating current Idmin is related to the
distance, high-set instantaneous, and earth-fault value of Is1 by the formula
protection included in the basic model to provide a Idmin = (k1IL + Is1)/(1-0.5k1)
complete ‘one-box’ solution of main and backup
protection. for Ibias <Is2
and

U n i t P ro te c t i o n Fe e d e r s
Idmin = (k2IL -(k2-k1)Is2 + Is1)/(1-0.5k2)
10.12.1 Unit Protection of a Plain Feeder for Ibias >Is2
The circuit to be protected is shown in Figure 10.16. It where IL = load current
consists of a plain feeder circuit formed of an overhead and hence the minimum operating current at no load is
line 25km long. The relevant properties of the line are: 0.235p.u. or 94A.
Line voltage: 33kV
Z = 0.157 + j0.337Ω/km In cases where the capacitive charging current is very
Shunt charging current = 0.065A/km large and hence the minimum tripping current needs to
be set to an unacceptably high value, some relays offer
To arrive at the correct settings, the characteristics of the
the facility of subtracting the charging current from the
relays to be applied must be considered.
measured value. Use of this facility depends on having a
The recommended settings for three of the adjustable suitable VT input and knowledge of the shunt
values (taken from the relay manual) are: capacitance of the circuit.

• 10 •

33kV 25km 33kV

400/1 400/1

Digital communications link

Id> Id>

Steady state charging current = 0.065A/km

Figure 10.16: Typical plain feeder circuit

Network Protection & Automation Guide • 167 •

Chap10-152-169 21/06/02 8:46 Page 168

8
The delta/star transformer connection requires phase
shift correction of CT secondary currents across the
transformer, and in this case software equivalents of
7 interposing CT’s are used.
Since the LV side quantities lag the HV side quantities by
6 30°, it is necessary to correct this phase shift by using
software CT settings that produce a 30° phase shift.
There are two obvious possibilities:
5
a. HV side: Yd1
Idiff

LV side: Yy0
4
b. HV side: Yy0
LV side: Yd11
3
Only the second combination is satisfactory, since only
this one provides the necessary zero-sequence current
2 trap to avoid maloperation of the protection scheme for
earth faults on the LV side of the transformer outside of
the protected zone.
1
Ratio correction must also be applied, in order to ensure
that the relays see currents from the primary and
0
0 1 2 3 4 5 6 secondary sides of the transformer feeder that are well
Ibias balanced under full load conditions. This is not always
Figure 10.17: Relay characteristic; inherently the case, due to selection of the main CT
plain feeder example ratios. For the example of Figure 10.18,
U n i t P ro te c t i o n Fe e d e r s

10.12.2 Unit Protection of a Transformer Feeder Transformer turns ratio at nominal tap
Figure 10.18 shows unit protection applied to a 11
= = 0.3333
transformer feeder. The feeder is assumed to be a 100m 33
length of cable, such as might be found in some
industrial plants or where a short distance separates the Required turns ratio according to the CT ratios used
33kV and 11kV substations. While 11kV cable 400
capacitance will exist, it can be regarded as negligible for = 1 = 0.32
1250
the purposes of this example. 1

20 MVA
33kV 33/11kV
11kV
400/1 Dyn1 Cable 1250/1
100m
350A 1050A
• 10 •
0° -30°

0.875A 0.84A

Digital communication
channel

Id> Id>

Ratio correction: 1.14 Ratio correction: 1.19

software CT: Yy0 software CT: Yd11

Figure 10.18: Unit protection of a transformer feeder

• 168 • Network Protection & Automation Guide

Chap10-152-169 21/06/02 8:46 Page 169

Spill current that will arise due to the incompatibility of

the CT ratios used with the power transformer turns ratio
may cause relay maloperation. This has to be eliminated
by using the facility in the relay for CT ratio correction
factors. For this particular relay, the correction factors
are chosen such that the full load current seen by the
relay software is equal to 1A.
The appropriate correction factors are:
HV: 400/350 = 1.14
LV: 1250/1050 = 1.19
where:
transformer rated primary current = 350A
transformer rated secondary current = 1050A
With the line charging current being negligible, the
following relay settings are then suitable, and allow for
transformer efficiency and mismatch due to tap-
changing:
IS1 = 20% (minimum possible)
IS1 = 20%
k1 = 30%

U n i t P ro te c t i o n Fe e d e r s
k2 = 150%

10 . 1 3 R E F E R E N C E S
10.1 Merz-Price Protective Gear. K. Faye-Hansen and
G. Harlow. IEE Proceedings, 1911.
10.2 Protective Relays Application Guide – 3rd
Edition. ALSTOM Transmission and Distribution
Protection and Control, 1987.

• 10 •

Network Protection & Automation Guide • 169 •

FEEDER DIFFERENTIAL PROTECTION

Differential, or 'Unit' Protection

Feeder current differential protection is one of the family of protection elements that are used in unit
protection. Unit protection is a type of protection that is sensitive to faults within a protected zone
with clearly defined boundaries, and insensitive to faults outside this zone. These boundaries are
defined as the Current Transformers on either end of the protected object. These current
transformers feed current information to the protection scheme.

Examples of unit protection are:

1. Buszone protection;
2. Restricted earth fault protection;
3. Transformer differential protection;
4. Line differential protection;
5. Impedance protection aided by teleprotection (I.e. permissive schemes);
6. Directional current comparison schemes aided by teleprotection.

Examples of non-unit protection are:

1. Time Delayed Overcurrent Protection;
2. Instanstaneous Overcurrent Protection;
3. Earth fault protection;
4. Impedance protection;
5. Directional overcurrent and earth fault protection

Merz-Price Principle of circulating current – Circulating Current Differential Scheme.

The most basic of current differential protection scheme is the simple single line representation of a
current differential scheme shown in Figure 1. This uses the Merz-Price principle of circulating
current.

There are two basic fault scenarios that may occur in a current differential scheme. For an external,
or “through” fault, it is desired that the scheme remain stable, i.e. that it doesn't trip. For an internal
fault i.e. a fault that is within the unit zone of protection, it is desired that the protection scheme trip
as quickly as possible.

For the “through” fault analysis, consider the simplified diagram of a unit protection scheme in
Figure 1. A fault is applied to the system at position F1. It can be seen that this fault is outside of the
unit protection zone. Since there is a source at either end of the feeder, currents I1 and I2 flow toward
the fault from each end of the feeder. It can be seen that only I1 flows in the protected object.
The current transformers at each end of the feeder generate identical currents i1 in the secondary
circuit that flow in the same direction. This current circulates in the secondary circuit and does not
flow through the High Impedance Relay, 87. Thus it can be seen that the scheme is stable, and that
the relay will not operate.

I1 I2

i1 i1
F1
87 High
Impedance
Relay

Figure 1: External Fault in a circulating current differential protection scheme

For the analysis of the internal fault, consider the simplified diagram of a unit protection scheme in
Figure 2. A fault is applied to the system at position F1. It can be seen that this fault is now inside
the unit protected zone. The sources at either end of the feeder cause currents I1 and I2 to flow
toward the fault from each end of the feeder. It can be seen now that I1 flows through the CT on the
left hand side and I2 flows through the CT on the right hand side of the protected object.
The current transformers at each end of the feeder generate currents i1 and i2 in the secondary circuit
that flow in opposite directions. The only path for these currents to flow is through the High
Impedance Relay, 87. The currents sum together and i1 + i2 flow through the relay and thus cause the
relay to operate, thus tripping the circuit breakers and removing the fault from the system.

I1 I2
F1
i1 i2
High Impedance
87
Relay
i1+i2

Figure 2: Internal Fault in a circulating current differential protection scheme

In practice, because of the high impedance coil of the relay, the CT's generate a voltage because it is
not possible to drive currents of fault magnitude into a high impedance coil. The coil is actually
voltage sensing, and it is set to activate if a certain voltage appears across it. Alternatively, a low
impedance current sensing coil can be used, which is placed in series with a stabilising resistor.

It is obvious that the scheme described above is a simplified. In real life, there are three phases and
thus a more complex system is required.

Opposing Voltage Current Differential Scheme

I1 I2

v1 v1

F1

Figure 3: External Fault in an opposing voltage current differential protection scheme

Consider the diagram of a simplified opposing voltage current differential scheme in Figure 3. The
external or “through” fault current of I1 flows in the primary windings of the current transformers at
each end of the feeder. The CT's develop a voltage v1 at each end of the secondary circuit. Because
the connections are cross-wired, the voltages oppose one another and there is no potential difference
developed across the coils of the relays at each end of the feeder. Thus no current flows and the
relays do not operate.
 I1 I2

v1 v2
`
F1
i

Figure 4: Internal Fault in an opposing voltage current differential protection scheme

Consider the diagram in Figure 4. The in-zone fault F1 on the protected object is fed by currents I1
and I2, flowing in opposite directions from either end of the feeder. The Current transformers at each
end of the feeder produce voltages v1 and v2, which are in opposite directions. Because the
connections between the CT's are crosswired, these voltages add together and cause a current I to
flow in the coils of the relays at each end of the scheme, causing the scheme to trip.

Phase segregated High Impedance Differential Scheme

Consider the diagram in Figure 3. This diagram shows a scheme that is phase segregated. In this
deployment, each element will operate for a fault that involves the phases involved. For example, an
in-zone phase to earth fault involving the 'a' phase will operate the 87-a element, a phase-to-phase
fault involving the 'b' and 'c' phases will operate the 87-b and 87-c elements, and a 3 phase fault will
operate all 3 elements.

A practical high impedance buszone protection scheme will be very similar to this scheme.

87-a 87-b 87-c

Figure 5: Phase Segregated High Impedance Differential Scheme such as that used in Bus-Zone
protection.
 I1
i1 I1 i1 I1

i1 i1

87-a 87-b 87-c

i1 i1
i1 i1

Figure 6: Phase Segregated High Impedance Differential Scheme with external fault between the 'b'
and 'c' phases.

Consider the diagram in Figure 6 showing an external phase-phase fault between the 'b' and 'c'
phases. It can be seen that the current i1 supplied by the 'b' and 'c' phase current transformers at each
end of the secondary wiring scheme flows in the same direction, i.e. it circulates in the same way as
the simplified single line diagram and therefore does not flow through the elements 87-b and 87-c.
Thus the scheme is stable for an external fault and does not trip.

I1 I1 I2 I2
I1 i1 I1 I1+I2 I2 i2 I
2

End A End B
i1 i1+i2 i2
i1+i2 i1+i2
87-a 87-b 87-c
i1 i2

i1 i2

Figure 7: Phase Segregated High Impedance Differential Scheme with internal fault between the 'b'
and 'c' phases.

Consider the diagram in Figure 7 showing an internal, or 'in-zone' phase-phase fault between the 'b'
and 'c' phases. On the primary circuit, current I1 flows toward the fault from the source at End 'A',
and I2 flows in the opposite direction from the source at End 'B'. In the secondary circuit , current i1
supplied by the 'b' and 'c' phase current transformers at End 'A', flows in the opposite direction to
the current i2 supplied by the 'b' and 'c' phase current transformers at End 'B'. The currents i1 and i2
add to make a total current of i1+i2 which flows through the operating coils 87-b and 87-c, thus
tripping both relays 87-b and 87-c. Thus this scheme will indicate which phases were involved in
the fault if the relays have indication flags or LEDs.
Biased Differential Principle
The Protection schemes shown in the examples listed earlier are all examples of unrestrained
protection. They have high impedance coils that operate when enough current is driven through
them. Because of the lack of restraint current (i.e. a current that is stabilising the scheme to refrain
from tripping), high impedance schemes are prone to maloperation and other problems such as:
Current Transformer Saturation;
High Voltages during fault conditions that may cause arcing;
Tendency to maloperate due to CT mismatching.

Therefore, the idea of biasing comes into play.

IBIAS IDIFF

IDIFF
Op
No Trip Trip
Bias Bias

Figure 8: Basic Concept of Biasing

Consider the diagram in Figure 8. This shows the typical coil arrangement and principle of the
biased differential protection scheme. The balance beam concept illustrates that the principle is that
the relay evaluates two opposing currents and based on the relative dominance of one over the
other, takes a measured decision to restrain or to trip. The relay calculates IBIAS = Average current
∣I RHS∣+∣I RHS∣
magnitude flowing through each bias coil, i.e. I BIAS = and IDIFF = Magnitude of the
2
current flowing in the differential circuit, i.e. I DIFF =∣I LHS +I RHS∣ . If there is more bias current than
differential current, the balance beam tips to the 'No Trip' side and the relay does not trip. If there is
more differential current than bias current, then the balance beam tips to the 'Trip' side, and the relay
trips. So it is important to note that the relay is always evaluating Bias against Differential current,
and basing a tripping decision based on which is greater. It is important to note that an
electromechanical relay will do this based on opposing magnetic fields, where a numerical relay
will do this based on evaluating calculated values within its Digital Signal Processing algorithms,
but essentially, they are doing the same thing.

IDIFF

IBIAS
IDIFF
Op
No Trip Trip
Bias Bias

Figure 9: Biased differential circuit during normal load or through fault conditions

Referring to the diagram in Figure 9, it can be seen that during normal operating conditions or
during an external fault, current circulates in the Bias coils only, and no differential current flows
through the Operate coil. Thus referring this to the balance beam concept, it is clear to see that the
Bias current is 'heavier' than the diff current, and therefore the balance beam tips to the 'No Trip'
side.
 IBIAS
IDIFF
IDIFF
Op
No Trip Trip
Bias Bias

Figure 10: Biased differential circuit during in-zone fault conditions

Referring to the diagram in Figure 9, it can be seen that during an internal, or in-zone fault, the
differential current is twice the bias current (since the bias current is the average current in each bias
coil and the differential current is both of these currents added together), and therefore the
differential current will be 'heavier' than the bias current and the balance beam will tip toward the
'Trip' side.

Operate
Differential Region
Current Slope = k1
IDIFF Restaint
Region

Bias Current IBIAS

Figure 11: Basic Single Slope Biased differential Operating Characteristic

The simple biased differential scheme described above using the 'balance beam' analogy, operates
on a single slope characteristic, as illustrated by the graph in Figure 11. If the magnitudes of both
Differential and Bias current are compared, and a higher differential current causes tripping, the
characteristic is as above with a slope of k = 1. In more sophisticated relays, dual and triple slope
characteristics are used. The diagram in Figure 12 shows a dual slope bias characteristic. There are
settings, the two slope settings k1 and k2 which determine the slopes of the first and second curve,
the initial Bias setting Is1 which determines the initial differential current required for the relay to
trip with zero bias, and the second Bias setting Is2 which determines the bias current above which
the higher slope k2 becomes effective.

Operate Slope 2
Region = k2
Differential
Current Restaint
IDIFF Slope 1 = k1 Region

Is1

Bias Current IBIAS Is2

Figure 12: Dual Slope Biased Differential Operating Characteristic

Creating a Practical Pilot Wire Current Differential Scheme for Feeders

Referring back to the phase segregated high impedance scheme shown in Figure 5, it can be argued
that this is a good starting point to create a Feeder Differential Scheme. However, there are some
practical considerations and constraints that need to be taken into account:

1. The scheme in Figure 5 only has one relay. In a practical feeder situation, there are two
Stations at either end of the feeder, with circuit breakers attached. It is therefore necessary
to have a relay at each end of the feeder that can trip the breaker directly.
2. The phase segregated scheme needs 4 wires in order to function. It would be better to use 2
pilot wires that will be able to carry a composite signal from one end of the feeder to the
other.
3. Because the pilot wires will be travelling along with power cables, substantial voltages and
currents may be induced in the pilot wires.
4. The earth mat potentials of the Substations at each end of the feeder may be different by an
order of magnitude of kilovolts, and this may cause problems since the pilot wire will
effectively join the two earth mats together.
5. Because of the length of pilot wires, it is not practical to send the secondary currents along
them because substantial volt drops will occur.

In practice, a pilot wire scheme incorporates a summation transformer, with a pilot wire relay at
each end of the feeder. The summation transformer performs two important functions:

• It combines all of the 3 phase currents into 1 composite signal that only requires 2 pilot
wires;
• It provides isolation between the pilot wires and the Current Transformer secondary circuits
at each end of the feeder.
• It brings the signals down to a lower level that can travel a long distance and tolerate a large
range of pilot wire resistances.

STATION A Protected Feeder STATION B

Summation Summation
Transformer Transformer

Bias Bias
Diff Diff

Figure 13: Simplified Pilot wire Feeder Differential Scheme

In Figure 13, a Simplified Basic Pilot wire Feeder Differential Scheme is shown. In practice, the
summation transformer and the Bias and Operate coils are all integrated to form part of the same
Pilot wire relay.

As can be seen, the summation transformer integrates the current signal from all three phases into a
composite signal using only two wires, but in order to do this some compromises have to be made.
This is because if the currents of a balanced system were added together in equal proportions, the
resultant signal would be zero. This would mean that during in-zone three phase faults, the output of
the summation transformer would be zero, and the scheme would not trip.

Thus the summation transformer adds the currents together in different proportions. This means that
there is always a composite output signal, but the disadvantage of this arrangement is that because
of the unequal proportions, the sensitivity is different depending on the faulted phase.

IA
IB 1T
1T Output > K
IC
3T

Figure 14: Typical Summation transformer

In Figure 14, the summation transformer has a turns ratio of [Link]. Since currents from each phases
flow through a different number of windings, the sensitivity is different for each phase and fault
type.

Summation Transformer Effects: Sensitivities of Pilot Wire for Different Fault Types

Referring to Figure 14, let us assume that for operation of the relay, the ampere turns output of the
Summation transformer > K (must be greater than a pick up value, K).

Earth Faults

1. For a Phase A to Earth Fault, the current IA will flow in the top windings, and out of the
bottom neutral point. Thus the number of ampere turns generated by IA will be 5*IA. For
operation of the relay, 5*IA > K, and therefore IA > 0.2K
2. For a Phase B to Earth Fault, the current IB will only flow through 4 windings. Thus the
number of ampere turns generated by IB will be 4*IB. For operation of the relay, 4*IB > K,
and therefore IB > 0.25K
3. For a Phase C to Earth Fault, the current IC only flows through 3 windings. Similarly, for
operation of the relay, 3*IC > K, and therefore IC > 0.33K

Phase-Phase Faults

1. For a Phase A to Phase B Fault, the current IAB will flow through only one turn. For
operation of the relay, IAB > K, or IAB > 1.0K
2. For a Phase B to Phase C Fault, the current IBC will flow through a single winding. Thus for
operation of the relay, IBC > 1.0K
3. For a Phase A to Phase C Fault, the current ICA will flow through 2 windings. Thus for
operation, 2*ICA > K, and therefore IBC > 0.5K

3-Phase Faults

1. For a 3 phase fault, all currents are the same in magnitude and the B and C phase currents
can be described in terms of the A phase current as follows: IB = a2IA, IC = aIA. The total
ampere turns generated by IA, IB and IC is (IA*1 – IC*1) > K (since they are flowing in
opposite directions in the coils). Thus (IA – aIA) > K, and therefore IA (1-a) > K. From this,
1.73 IA > K, and IA > 0.577K

The sensitivities of a pilot wire scheme using a [Link]N summation transformer are shown in the table
below, if nominal current In with settings multiplier K will cause operation. Note that for systems
which have high impedance or resistance earthing, (for example earthing transformers and Neutral
earthing resistors) increased earth fault sensitivity is required, and therefore a sensitive tapping
(with more neutral windings) is used.

Fault Type Sensitivity

N = 3 (Normal E/F Settings) N = 6 (Sensitive E/F Settings)
A-N 0.2 x K x In 0.125 x K x In
B-N 0.25 x K x In 0.143 x K x In
C-N 0.33 x K x In 0.167 x K x In
A-B 1.0 x K x In
B-C 1.0 x K x In
C-A 0.5 x K x In
A-B-C 0.577 x K x In

What does this mean in practice? If we set K to 1, N = 3, and we are operating at a nominal current
of 1A, then 200mA injected between the A and Neutral wires will cause operation. Similarly,
250mA in the B phase and 330mA in the C phase will cause operation.

For Phase-Phase faults, 1A injected between the A-B and B-C phases will cause operation, but
500mA between the C-A phases will cause operation. For a 3 Phase fault, a balanced current of
577mA injected into each phase will cause operation.

There are certain network conditions where the summation transformer as listed above will not
produce an output, for example in Figure 15, which shows a network situation where a phase –
phase fault can create a 2-1-1 current distribution. The currents in the A and C phase produce
opposing fields which cancel each other out. Thus the output of the Summation transformer is zero.

Thus a more practical ratio to deal with this situation is 1.[Link]N.

STATION A Protected Feeder STATION B

Summation Summation
Transformer Transformer

Bias Bias
Diff Diff

Figure 15: Phase-phase fault on LV side of delta-start transformer causes 2-1-1 current
distribution and opposing ampere turns.
Effect of Line Charging Currents

In any feeder, a certain component of the current flowing is used to charge the line, which offers
capacitance to ground on each phase. This charging current only flows in one end of the feeder (the
end with the strongest source), and not the other. The charging current is therefore an unbalancing
current, and will manifest itself as a spill, or differential current. If the charging current is large
enough it can destabilise the pilot wire scheme, and cause it to trip.

Overhead lines have a small charging capacitance, and so the charging currents are generally low
and can usually be ignored.

Underground cables have substantial charging capacitance. The resulting charging currents can be
high enough to dictate the minimum settings of the scheme.

ICA
ICB 1T
1T
ICC
N

Figure 16: Summation transformer with line charging currents

Referring to the diagram in Figure 16, the Line charging currents ICA, ICB and ICC can be rewrittten
as follows: ICB = a2 ICA and ICC = a ICA.

The total unbalancing ampere turns produced by the line Charging Current Ampere Turns is:

AT CH =I CA×1 Turn− I CC ×1Turn

AT CH = I CA−a I CA
AT CH = I CA (1−a)
AT CH = √3 I CA

if one takes only magnitudes into account (since phase angles are not important here)

AT CH =√ 3 I ch …..... Equation 1

The fault that operate the scheme for the least amount of fault current is a phase-earth fault on the A
phase.

This produces ampere turns of AT Ia =I A×(2+N ) …..... Equation 2

For stability, Ampere turns produced by Equation 2 must be greater than ampere turns produced
by Equation 1, so
I A (2+N )>I ch

and so
√ 3 I ch
I A>
2+N
Effect Of Phase-Phase-Earth Fault On A Solidly Earthed System

Protected Feeder STATION B

ICH STATION A
ICH
0
0
Summation Summation
Transformer Transformer

Bias Bias
Diff Diff

Figure 17: Pilot wire scheme with external phase-phase-earth fault involving the B and C phases

As can be seen from the diagram in Figure 17, an out-of-zone fault between the B and C phases and
earth shorts out the faulted phases and the charging capacitance to ground. Therefore the only
charging current that will flow will be in the unfaulted 'A' phase.

Thus the ampere turns produced by the charging current is

AT CH =(2+N ) I ch and for stability for an 'A' phase fault

AT ( I a )=(2+N ) I a> AT (I CH )= I CH (2+ N )

and so I a>I CH , which means that the minimum current sensitivity setting must be set above the
line charging current.

Pilot Wire Resistance

The resistance of the pilot wire varies greatly and is determined by the feeder length and the length
of the pilot wire. In order to make sure that the pilot wire scheme functions optimally, padding
resistors are used to 'make up' the total loop resistance to a specified value so that the scheme can be
applied over a wide range of feeder lengths. A typical designed value for the total loop resistance is
1000Ω. The diagram in Figure 18 shows how the padding resistors Ra can be adjusted to cater for
the variable pilot wire loop resistance Rp. In order to ensure a total designed loop resistance of
1000Ω, 2Ra+ Rp = 1000Ω. Therefore the padding resistance on each relay needs to be set to:

Ra = (1000 – Rp)/2
Pilot Wire
Relay Cable Relay
Ra Rp/2 Ra

Rp/2

Figure 18: Variable padding resistor Ra ensures that the total pilot wire resistance is within the
designed value.
 The pilot wire loop resistance can be measured by disconnecting the pilot wires at both ends,
shorting the remote end pilots together and measuring the resistance from the local end using a
multimeter.

Effects of Pilot Wire Faults on Pilot Wire Feeder Differential Schemes.

The cable containing the pilot wire is routed in the outside environment and therefore it is
vulnerable to risks and hazards. These may compromise the integrity of the pilot wire scheme:
Some examples are:
• Human interference such as cable theft;
• Acts of nature such as storms, floods, soil subsidence and lightning damage;
• Mechanical damage excavation, impact damage.

Therefore some type of supervision is necessary to guard against the effects of pilot wire damage,
and to detect whether the pilot wire has been compromised. There are two types of supervision
mechanisms:
• Signal injection type, where a small magnitude signal is injected onto the pilot wire similar
to a carrier signal on a transmission line. Absence or distortion of the signal indicates that
the pilot wire has been compromised.
• Wheatstone bridge type, where a Wheatstone bridge is used: a voltage across the
Wheatstone bridge would indicate that the pilot wire has been compromised.

Some examples of pilot wire faults are:

1. Short circuited pilot wire, where a short circuit develops somewhere in the pilot wire cable
between the two relaying points;
2. Open circuited pilot wires, where there is a break in the pilot wire circuit.
3. Pilot wires cross-wired.

Short circuit pilot wire fault:

R R

Figure 19: Short circuited pilots with an in-zone fault on a Circulating Current Pilot Wire Scheme

As shown in Figure 19, the short circuited pilot bypasses operating current through the short circuit,
and thus the relays can never operate.

R R

Figure 20: External fault with short circuited pilots on an Opposing voltage pilot wire scheme.

As can be seen from the diagram in Figure 20, the short circuited pilot routes the driving voltage
from each summation transformer directly to the relay operating coil, causing the relay to operate
incorrectly during external faults and load currents.
Open circuit pilot wire fault:

R R

Figure 21: Open circuited pilots with load current or an external fault on a Circulating Current
Pilot Wire Scheme

As can be seen in Figure 21, the open circuit does not allow the current to circulate, and therefore
the only route for the circulating current to follow is through the relay, thus causing the relay to
operate incorrectly. This can occur both during faults and when the load current is high enough.

R R

Figure 22: Open circuited pilots on an Opposing Voltage Pilot Wire Scheme

As can be seen from the diagram in Figure 22, the open circuit does not allow a voltage difference
to be developed across the relay coil for an internal fault, and thus the relay can never operate.

Cross-wired Pilot Wires:

In both circulating current and opposing voltage schemes, cross wiring reverses the unit protection
zone: the scheme will block for internal faults and trip for external faults.

The table below summarises the effects of pilot wire faults on the two different pilot wire schemes.

Failure Mode Circulating Current Scheme Opposing Voltage Scheme

Short Circuit Pilots Stable (will not operate for faults) Unstable, Maloperate
Open Circuit Pilots Unstable, Maloperate Stable (will not operate for faults)
Reversed Pilots Unstable, Maloperate Unstable, Maloperate
DIGITAL CURRENT DIFFERENTIAL PROTECTION

In many ways, digital current differential protection is similar to pilot wire protection.

A pilot wire is a crude communications medium: it feeds information about the situation at one end
of a protected object to the other end. Each end, passively, is evaluating the information they
provide against the local situation, and then determines whether to trip or not from them.

With the advent of the microprocessor and improved communications, the relays at each end of the
protected object have intelligence, and actively arrive at the decision as to whether to trip the relay
or not. The communications link between either end of the line carries information in the form of a
digital stream instead of a composite analogue signal, as in the case of the pilot wire differential
relay.

IED Internal Components

The term commonly used for a modern numerical protection relay is IED, short for Intelligent
Electronic Device. It consists of:

ANALOG INPUTS

ANALOG CONDITIONING
CIRCUITRY

ANALOG TO DIGITAL
CONVERTER

1. Analogue conditioning circuitry and A/D converters, where the analogue inputs (I.e.
currents and voltages) are converted to digital signals;
2. Digital Signal Processing circuitry, which manipulate the signals in the digital domain. Any
type of processing can be done like scaling, phase shifting, extraction of fundamental
frequency, extraction of sequence components etc. Software, in the form of algorithms, is
used to manipulate the digital signals to build complex protection functions. A modern IED
is so sophisticated, that most of the value in the functionality of the IED rests in the
software, and not in the hardware. Many manufacturers use a standardised hardware
platform and change the firmware to determine what protection function the IED will
perform.
3. Digital Communication Interfaces or ports – These are for:
 1. Engineering access - to set the parameters and logic functions of the IED, extract
Sequence of Events records, Disturbance records etc.
2. SCADA or telecontrol functionality, where alarms and analogue values are reported to a
control centre, and where controls (like tripping and closing of the circuit breaker) are
accepted from a control centre and executed by the IED.
3. Remote end communications, where current information is sent and received from an
IED at the remote end of a protected object. This is particular interest if we are studying
a current differential scheme, because the communication link is crucial for the entire
scheme to function.
4. Digital inputs and outputs, where plant status information (like the status of a breaker) is fed
to the IED and control signals (like tripping and closing of the circuit breaker) are sent out to
other equipment.
5. Other devices, like displays, keypads, RAM, non-volatile RAM, real time clocks etc. In
many ways the IED is similar to a PC.

End-to-End Communications
Of crucial importance to the proper functioning of a digital current differential scheme is the
communication link. It must be reliable, and available at all times. The communication link carries
information about current magnitudes and angles from the remote end to the local end and vice
versa. The data is encoded in a particular format, or protocol, that the IED's at each end of the
protected object understand. This is usually a proprietary format, so generally for current
differential schemes to work the relays at each end must be of the same type and manufacturer. For
example a Siemens relay will not be compatible with an Areva relay! With some manufacturer's
products not only must the relays be of the same type, but the firmware they are running must be
the same version.

Types of Communications
There are a number of communication interfaces that can be used for the remote end
communications.
1. Optical Interface – Dark fibre.
A pair of optical fibres connect the relays directly without any intermediate equipment. The
IED's manage the communication link. The term commonly used for this type of
communications is 'dark fibre', so called because the fibre is 'lit' by the IED's and not with
bandwidth management equipment. This is the first choice for end-to-end communications.
Dark fibre has proven itself to be the most reliable for protection current differential
communications. Most of the applications of dark fibre is via OPGW, Optical Ground Wire,
where a fibre cable is embedded in the centre of the earth wire on an overhead feeder.
However, on a cable feeder, the optical cable could be buried in the same (or a different)
trench as the power cable.
2. Optical Interface – connecting to Managed channel.
In this application of optical connection, the IED is connected to a bandwith management
multiplexer via an optic cable. The bandwidth management equiment may carry the signal
over a number of mediums to a BME at the remote end. Not all relays function well when
connected to a multiplexing system: the Siemens 7SD511 does not perform well in this
application since the IED prefers to talk to a remote end without an intermediary.
3. Electrical Interface – connecting to an optical multiplexer.
The advantage for the utility is that more information can be passed through a single optic
fibre by multiplexing (teleprotection, telecontrol, telecommunications and relay end-to-end
communications) – the trade-off is that, as above, not all IED's will function in this mode,
and that a single mode failure will lead to the interruption of all of the services being carried
by the optical fibre pair.
4. Electrical Interface – connecting through a multiplexer to a micowave radio system.
One of the major disadvantages of radio systems is that they are susceptible to failure or
poor performance in bad weather, and this is the time when it is most likely that a fault may
 develop and when it is crucial that the differential scheme works properly.

Types of Optical Interfaces

1. Multimode fibre interfaces 850nm. A multimode interface is inexpensive but not designed to
carry large distances. Generally this type of interface is used to connect (patch) a relay to
another device in the same relay room. 850Nm Multimode fibre is unsuitable for use in a
dark fibre application.
2. Multimode fibre interfaces 1300nm. This multimode interface is inexpensive and can carry a
signal an intermediate distance. Generally this type of interface is used for short line
applications and dark fibre application.
3. Single Mode interfaces. These are usually driven by laser diodes and are suitable for long
distances and dark fibre application. Different optical frequencies will travel different
distances as shown in the optical budget table.
Optical Budgets

Depending on the length of the feeder under consideration, optical budgeting will determine what
type of line driver is required. The table lists various types of fibre and from this the maximum
distance can be determined.

Interface Type 850nm 1300nm 1300nm Single 1550nm Single

Multimode Multimode Mode Mode
Min. Transmit Output Level -19.8 dBm -8.2 dBm -8.2 dBm -8.2 dBm
Receiver Sensitivity -25.4 dBm -38.2 dBm -38.2 dBm -38.2 dBm
Optical Budget 5.6 dB 30dB 30dB 30dB
Safety Margin (3db) 2.6 dB 27dB 27dB 27dB
Typical Cable Loss 2.6 dB / km 0.8 dB / km 0.4 dB / km 0.3 dB / km
Max Transmission Distance 1km 33.7km 67.5km 90km
Distance Between Stations Short Medium Long Very Long

Let's examine some of the parameters in the specification and explain what they mean.

Very briefly, an IED or an optical interface has a connection for an optical transmitter (usually a
powerful LED, or a laser Diode) and a connection for an optical receiver. The Transmitter is
guaranteed to transmit a signal at a certain power level, measured in dBm. This is the minimum
Transmit Output level. The receiver has a certain minimum sensitivity. A signal with a strength
below this sensitivity will generate errors and the received signal will be of a poor quality. The
datastream will have intermittent errors and if there are too many errors, the IED's will fail the link
and revert to other forms of protection other than differential. The optical cable used has a certain
cable loss expressed in dB/kilometre. A safety margin of 3dB is also used to cater for cable splice
joints and deterioration of the fibre during its lifetime.

Thus for a good quality optical link:

Minimum Transmit Signal – Cable loss – Safety Margin >= Receiver Sensitivity
where Cable loss = Cable Length x dB loss / km

Example
Question
Calculate the Optical Budget and then the maximum cable length that can be used for the following
optical equipment:

Interface Type 1300nm Multimode

Min. Transmit Output Level -8.2 dBm
Receiver Sensitivity -34.2 dBm
Safety Margin 3dB
Typical Cable Loss 1.2 dB / km
Answer

Optical Budget=[Link] Level−Receiver Sensitivity=−8.2−(−34.2) dBm=26 dBm

Optical Budget−Safety Margin (26−3) dBm 23

MaximumCable Length= = = =19.16 km
Cable Loss 1.2 dBm/ km 1.2

Propagation Delay
A problem with digital current differential systems is that the current signal takes time to travel
from the remote end. Thus when the currents from the local and remote end are compared, they do
not align, i.e. there is a phase difference between the currents, and thus a 'phantom differential
current'. Some type of technique has to be employed in order to compensate for this.

Two techniques exist:

1. Synchronise sampling: this can be done by using a GPS clock signal to ensure that the IEDs
sample the current waveform at exactly the same time. The problem with this is that a GPS
signal is not always available – so synchonising samples is not a feasible option.
2. Asynchonous sampling: the IED's at each end run 'free', and then the received samples are
time shifted by the local IED because the propagation delay is measured or is a known
quantity. In some relays such as the 7SD511, the propagation delay is a fixed setting; in
others the propagation delay is measured by a special timing message, that is reflected or
'bounced back' by the remote end IED. These type of systems can cope with managed links
where the bandwidth management system may reroute the communications, thus changing
the propagation delay. The LFCB is an IED such as this.

Intertrips

Most digital current differential protection systems have direct intertripping functionality built in.
Thus a trip signal can be relayed from one end of a line to the other. This is used where there is no
breaker at one end of the line and a seperate protection scheme requires a remote end breaker to trip
to isolate a fault (I.e. a transformer-feeder scheme). This can also be used in interconnected systems
where a breaker fail condition requires a local bus strip operation and tripping of a remote end
breaker.
Primary Injection Tests: Pilot wire protection
The sensitivity check can be performed by injecting with the single-phase test set through each CT
primary in all phases. This checks the earth fault sensitivity of each phase as well as determining
that the correct phases have been connected to the pilot wire relay summation CT shown in Figure
32.
Single phase primary
injection Test Set

PILOTS TO REMOTE
END RELAY

Figure 4: Earth Fault Sensitivity check of pilot wire relay

Phase A injection should require the least primary current to cause operation, phase C the most. This
is because of the unequal turns for each phase in the summation transformer: phase A is applied
across all of the turns of the summation transformer.
 A

Single phase primary

injection Test Set

PILOTS TO REMOTE
END RELAY

Figure 5: Phase Fault Sensitivity check of pilot wire relay