Anda di halaman 1dari 461




Enter the world of the AREVA T&D Training web site:





Overview Of Protection Fundamentals
Notes Overcurrent Protection
Directional Overcurrnt
Transformer Protec:tion Notes
Transformer Setting Tutorials
Generator and Generator Transf - Protection
Generators Setting Criteria
Distance Protection Notes
Distance Protectiorr Schemes
Busbar Protection
Motor Protection
A C Motor Protection
Motor Setting Criteria
Notes 1 C T S
Notes Additional Analysis
Notes Unbalanced Faults
Tutorial Balanced Faults
Tutorial Grading Examples
Tutorials Generator Protection
Tutorial C T Selection
Tutorial Busbar Protection

Overview Of Protection

APPS- Cornbined course

Overview Of Protection Fundamentals
Page 2 of 0




Relays are compact devices that are connected throughout the power
system to detect intolerable or unwanted conditions within an assigned
area. They are in effect, a form of active insurance designed to maintain
a high degree of service continuity and limit equipment damage. They
are "Silent Sentinels". While protective relays will be the main emphasis of
this chapter, other types of relays, applied on a more limited basis or used
as part df a total protective relays system will also be covered.


Relays can be divided into five functional categories:



a. P r o t e c t i v e Relays, which detect defective lines, defective apparatus, or

other dangerous or intolerable conditions. These relays can either
initiate or permit switching or simply provide an alarm.


b. M o n i t o r i n g R e l a y s , which verify conditions on the power system or in the

protection system. These relays include fault detectors, alarm units,
channel-monitoring relays, synchronism verification, and network
phasing. Power system conditions that do not involve opening circuit
breakers during faults can be monitored by these relays.

c. P r o g r a m m i n g R e l a y s , which establish or detect electrical sequences.

Programming relays are used for reclosing and synct-~ronising.

d. R e g u l a t i n g R e l a y s , which are activated when an operating parameter

deviates from predetermined limits. Regulating relays function through
supplementary equipment to restore the quantity to the prescribed

e. Auxiliary Relays, which operate in response to the opening or closing of

the operating circuit to supplement another relay or device. These
include timers, contact-multiplier relays, sealing units, receiver relays,
lock-out relays, closing relays and trip relays.

OALSTOM Limited, Energy Automation & Information

APPS- Combined cc.?~~rse

Overview Of ProtectionFundamentals
Page 3 of O-

In addition to these functional categories, relays may be classified by

input, operating principle or structure and performance characteristic:


Operating Principle of Structure

Solid state

and definitions are based on the ANSI Standard

The above c~assifi~ation
37.90 (IEEE 313).



Technically, most relays are small systems within themselves. Throughout

this chapter, however, the term systems will be used to indicate a
combination of relays of the same or different types. Properly speaking,
the protective relaying system includes circuit breakers as well as relays.
Relays and circuit breakers must function together; there i s little or no
value in applying one without the other.
Protective relays or systenls are not required to function during normal
power system operation, but must be immediately availa,ble to handle
intolerable system conditions and avoid serious outages and damage.
Thus,. the true operating life of these relays can be on the order of a few
seconds, even though they are connected in a system for many years. In
practice, the relays operate far more during t.esting and maintenance
than in response to'adverse service conditions.


OALSTOM Limited, Energy Automation & Information


APPS- Cornbined course



.. .

Overview Of Protection Fundamentals

Page 4 of 0

In theory, a relay system should be able to respond to the infinity of

abnormalities that can possibly occur within the power system. In
practice, the relay engineer must arrive at a compromise based on the
four factors that influence a n y re!oy rrpp!icatisn:
a. Economics - Initial, operating and maintenance.
b. Available measure of fault or trouble - Fault magnitudes and location of
current transformers and voltage transformers.
c. Operating practices - Conformity to standard and accepted practices;
ensuring efficient system operation.
d. Previous experience - History and anticipation perhaps better expressed
of trouble likely to be encountered within-thesystem-.
The third and fourth considerations are perhaps better expressed as the
"personality of the system and the relay engineer".
Since it is simply not feasible to design a protective relaying system
capable of handling any potential problem, compromises must be made.
In general, only those problems, which according to past experience are
likely to occur, receive primary consideration. Naturally, this makes
relaying somewhat of an art. Different relay engineers will, using sound
logic, design significantly different proteclive systems for essentially the
same power system. As a result there is little standardisation in protective
relaying. Not only may the type of relaying system vary, but also will the
extent of the protective coverage. Too much protection i s almost as bad
as little.
Nonetheless, protective relaying i s a highly specialised technology
requiring an in-depth understanding of the power system as a whole. The
relay engineer must know, not only the technology of the abnormal, but
have a basic understanding of all the system components and their
operation in the system. Relaying, then, i s a "Vertical" specialty requiring
a "horizontal" viewpoint. This horizontal, or total system, concept of
relaying includes fault protection and the performance of the protection
system during abnormal system operation such as severe overloads,
generation deficiency, out-of-step conditions, and so forth. Although
these areas are vitally important to the relay engineer, his concern has not
always been fully appreciated or shared by his colleagues. For this
reason, close and continued communication between the planning, relay
design, and operation systems should be mandatory, since power systems
grow and operating conditions change.

OALSTOM Limited, Energy Automation & Information

APPS- Combined course

Overview Of Protection Fundamentals
Page 5 of 0

A complex relaying system may result from poor system design or the
economic need to use fewer circuit breakers. Considerable savings can
be realized by using fewer circuit breakers and a more complex relay
system. Such systems usually involve design compromises requiring careful
evaluation, if acceptable protection is to be maintained.



The application logic of protective relays divides the power system into
several zones, each requiring its own group of relays. In all cases, the five
design criteria listed below are common to any well-designed and
efficient protective system or system segment:
a. Reliability - the ability of the relay p r relay system to perform correctly
when needed (dependability) and to avoid unnecessary operalion
b. Speed - minimum fault time and equipment damage:
c. Selectivity - maximum service continuity with minimum system
d. Economics - maximum protection at minimum cost.
e. Simplicity - minimum equipment and circuitry.
Since it is impractical to fully satisfy all these design criteria simultaneously
the necessary compromhes must be evaluated on the basis of
comparative risks.

System reliability consists of two elements - dependability and security.
Dependability is the certainty of correct operation in response to system
trouble, while security i s the ability of the system to avoid mis-operation
between faults. Unfortunately, these aspects of reliability tend to counter
one another: increasing security tends to decrease dependability and
vice versa. In general, however, modern relaying systems are highly
reliable and provide practical compromise between security and
Protective relay system must perform correctly under adverse sysfem and
environmental conditions. Regardless of whether other systems are
momentarily blinded during this period, the relays must perform
accurately and dependably. They must either operate in response to
trouble in their assigned area or block correctly i f the trouble is outside
their designated area.

OALSTOM Limited, Energy Automalion & informalion


APPS- Combined course

Overview Of Protection Fundamentals
Page 6 of 0

Dependability can be checked relatively easily in the laboratory or during

installation by simulated tests or staged faults. Security on the other hand is
much more difficult to check. A true test of system security would have to
measure response to an almost infinite variety of potential transients and
counterfeit trouble indicalions in the power system and its environment. A
secure system is usually the result of a good background in design
combined with extensive miniature power system testing and can only be
confirmed in the power system itself and its environment.
Relays that could anticipate a fa~lltwo!~ldbe utopian. But, even if
'available, they would doubtlessly raise the question of whether or not the
fault gr trouble really required a trip-out. The development of faster relays
must always be measured against the increased probability of more
unwanted or unexplained operations. Time, no matter how short, is still the
best method of distinguishing between real and counterfeit trouble.

Applied to a relay, high speed indicates that the operating time usually
does not exceed 50 ms (3 cycles on a 60-hertz base). The term
instantaneous indicates that no delay is purposely introduced in the
operation. In practice, the terms high speed and instantaneous are
frequently used interchangeably.
Selectivity versus Economics
High speed relays provide greater service continuity by reducing fault
damage and hazards to personnel. These relays generally have a higher
initial cost, which, however, cannot always be justified. Consequently,
both low and high-speed relays are used to protect power systems. Both
types have high reliability records. Records on protective relay operations
consistently show 99.5% and better relay performance.

As in any other engineering discipline, simplicity in a protective relay
system is always the hallmark of a good design. The simplest relay system,
however, is not always the most economical. As previously indicated,
major economies are possible with a complex relay system that uses a
minimum number of circuit breakers. Other factors being equal, simplicity
of design improves system reliability - if only because there are fewer
elements that can malfunction.

OALSTOM Limited, Energy Automation & Information

APPS- Combined course

Overview Of Protection Fundamentals
Page 7 of 0




Relay performance i s generally classed as:

( 1 ) Correct
(2) No conclusion
( 3 ) lncorrect
lncorrect operation may be either failure to trip or false tripping. The
cause of incorrect operation may be, a) Wrong application, b) lncorrect
settings, c ) A personnel error or 4) Equipment mal-function. Equipment
that can cause an incorrect operation includes current transformers,
voltage transformers, circuit breakers, cable and wiring, relays, channels
or station batteries.
lncorrect tripping of circuit breakers not associated with the trouble area
is often as disastrous as c failure to trip. Hence, special care must be
taken in both appiication and installation to ensure against the possibility
of incorrect tripping.
" No conclusion" is the last resort when no evidence is -available for a
correct or incorrect operation. Quite often this is a personnel involvement.
Zones of Protection
The general philosophy of relay application is to divide the power system
into protective zones that can be protected adequately with the
mininwm amount of the system disconnected. The power system is
divided into protective zones for:


Transmission and distribution circuits

A typical power system and its zones of protection are shown in Figl. The
purpose of the protective system is to provide the first line of protection,
within the guide-lines outlined above. Since failures .do occur, however
some form of backup protection is provided to trip out the adjace13f
breakers or zones surrounding the trouble area. Protection in each zone is
overlapped to avoid the possibility of unprotected areas

OALSTOM Limited, Energy kutomation & Informalion

APPS- Combined course

Overview Of Protection Fundamentals
Page 10 of 0

The device switching equipment are referred to by numbers, with

appropriate suffix letters when necessary, according to the functions they
These numbers are based on a system adopted as standard for
automatic switchgear by IEEE and incorporated in American Standard
C37.2 - 1970. 'This system is used in connection diagrams, in instruction
books and in specifications.

Device Numb'erina
Device Number
Master Elemenl

It is an initiating device, such as a
control switch, voltage relay, float
switch, etc., which serves either
directly or through such permissive
devices as protective and time
delay relays.
place an equipment in or out of
Delay It i s a device which functions to give
or a desired amount of time delay
before or after any point of
Closing Relay
operation in a switching sequence
protective relaying
except as specifically provided by
device function 48, 62 and 79
1 described later.
or It is a device which operates in
response to the position of a number
of other devices (or to a number of
predetermined conditions), in an
equipment, to allow an operating
sequence to proceed, to stop, or to
provide a check of the position of
these devices or of these conditions
i for any purpose.

- --


OALSTOM Limited, Energy Automation & Information

ibined course

I Device Number

Overview Of Protection Fundamentals

Page 1 1 of 0

It is a device, generally controlled by
the device No.1 or equivalent, and
perrr~issive and
protective devices, that serve to
make and break the necessary
control circuits to
equipment into operation under the
desired conditions and to take it out
of operation under other or
abnormal conditions.
Stopping Device It is a control device used primarily
to shut down an equipment and
hold it out of operation This device
may be manually or Electrically
actuated, but excludes the function
of electrical lockout (see device
function 86) on abnormal conditions. 1
Starting Circuit It is a device whose principal)
function is to connect a
its source of staitina voltaae.
Circuit It is one used in-theanode circuits of
a power rectifier for the primary
purpose of interrupting the rectifier
circuit if an arc back should occur.
Control Power It is a disconnecting device - such
as a knife switch, circuit breaker or
pullout fuse block, used for the
disconnecting the source of control
power to and from the control bus
or equipment.
Note: Control power is considered to
include auxiliary power, which
supplies such apparatus as sn~all
motors and heaters.
It is used for the purpose of reversing
a machine field or for performing 1
/ any other reversina functions.
( Unit Sequence ( It is used to change the sequence in 1
which units may be placed in and



OALSTOM Limited, Energy Automation & Information

APPS- Cornbined course



Overview Of Protection Fundamentals

Page 10 of 0

The device switching equipment are referred to by numbers, with

appropriate suffix letters when necessary, according to the functions they
These numbers are based on a system adopted as standard for
automatic switchgear by IEEE and. incorporated in American Standard
C37.2 - 1970. This system is used in connection diagrams, in instruction
books and in specifications.

Device Numb'erina
Device Number -Definition

Master Element It is an initiating device, such as a
control switch, voltage relay, float
switch, etc., which serves either
directly or through such permissive
devices as protective and time
delay relays.
place an equipment in or out of
Delay It is a device which functions to give
or a desired amount of time delay
before or after any point of
Closing Relay
operation in a switching sequence
protective relaying system,
except as specifically provided b y
device function 48, 62 and 79
described later.
It is a device which operates in
response to the position of a number
( Interlocking
of other devices (or to a number of
predetermined conditions), in an
equipment, to allow an operating
sequence to proceed, to stop, or to
provide a check of the position of
these devices or of these conditions




/ for any purpose.

OALSTOM Limited, Energy Automation & Information

pps- Combined course

Overview Or Protection ~undamental~
page 1 1 of 0

Device Nurr~

---=I :


It is a device, generally controlled b y
the device No.1 or equivalent. and
and I
protective devices, that sr=rL/rs to
make and break the rlecessarY
equipment into operation vrde! 'the I
desired ~onditiof?s
end to toke i i out
of operation
~ n d m TJ~I-!F-:
abnormal conditic:;~.
____---. Stopping Device It is a coritiol d,-\:jCe usee ;,:irr~arily I
to shut down c;;equipn.5~1:
hold it out of oge:2-i&n- -,;rli:
C;e/ice i
may be mancc:ii. or I:5c.::i~~lly ,
actuated, but exc,,zss ;r,5 f;,..ctior~
1 of elect!-ical I C ~ ~ - , - 1-, ~ :;edice
I function 86)on c ~ - - ~ .~
~ = : :.~__
; ,-T ! C!J T~
I ~ . ~i ,
lt i s a device .....--zs5
5:ir!.cipaI /
function is to c o ~ - - ~~ ~; ~- ~ vto! ! ~ r
its source of stariir:.~-.G ; T C ~ ~ .
- .
( .
2i:i of
Circuit It is one used in i.-e
a power- recjifie- :-r,gy,gr'j ;
- ,:- 4:;'l t l ~ e r
purpose oi inte---, ---2
circuit if ar: arc
F7 :T;~,.J~.
Power It i s a discanne;- -;
, - ;:,C~I
as a knife switc:
- L-3r;/er or
pullout fuse b!.zqzq - ---the
disconnecting ii-5 5 z . - - z-l -' -_,<-,irC~l
power t2 ~ : ; df: I-- -- - - -- - * / r~1.J:
or equipr:ie!~i.
Note: C ~ r ] i l .PO\.*.
~ / 51 1 -;:;*5?
include zct..rilic-.
1 ... :- ,,.,< ~f-r
supplies sctci, r - - - - - -- -- --- *'5~\1
nec:~.--,... - - )
( Device
a mactlifie fiei& -- - :-- ---./-,s - -<,!irtr;
an other re\,ersir,z -- -.y
.T; :
Unit Sequence It is used :p , - ~ ? C ~-. -- . Z--.- _ - s - - - :=- 'c. irl
which ur;i;- _ -,?c.,v - _ -- -- - -,-'"; '


- ---

- /

: A

p L

' -



-- -

b - - -


-- ..


OALSTOM Limited, Energy Automation & Info:;:-z'ior.


J .

APPS- Combined course

Overview Of Protection Fundamentals
Page 14ofO

Device Number

Synchronising or It is a device that operates when
Synchronismtwo ac circuits are within the desired
limits of frequency, phase angle and
Check Device
voltage, to permit' or to cause the
paralleling of these two circuits.
It is a device, which functions when
Thermal Device the temperature of the shunt field or
the armature winding of a machine,
or that of a load limiting or load
shifting resistor or of a liquid or other
medium exceeds a p.redefermined
value ; or if the temperature of the
protected appa~.atus, such as a
power rectifier, or of any medium
decreases below a predetermined
I Under Voltage It is a device, which functions on a
given value 'of undervoltage.
Flanie detector
It is a device that monitors the
presence of the pilot or main flame
-In such apparatus as a gas turbine
or a steam boiler.
1 It is a device used for disconnecting
one circuit from another for the
( Contcctor
purposes of emergency operation,
maintenance, or test.
It is a non-automatically reset
device that gives a number of
separate visual indications upon the
functioning of protective devices
and which may also be arranged to
perform a lockout function.
It connects a circuit such as the
field of a synchronous
converter, to a source of separate
sequence ; or one which energises
the excitation and ignition circuits of
a nower rectifier.

1 .







I Direc:'anal

It is a device which functions on a

OALSTOM Limited, Energy Automation & Information

~ p p sCombined
Overview Of Protection Fundamentals
Page 15 of 0


e Number






desired value of power fiow in
given direction, or upon reverse
power, like resulting from arc back in
the anode or cathode circuits of a I
power rectifier.
Position Switch
It makes or breaks contact when the
1 main device or piece of apparatus, 1
which has no device function
number, reaches a given position.
It is a device such as a motorMaster
operated multi-conjact switch, or
the equiv'alent, or a programming
device, such 0 s a computer, that
establishes or
operating sequence of the major
devices in an equipment during
starting and stopping or during other
sequential operations.
It is used for raising, lowering, or
Brushor ] shifting, the brushes of a machine, or I
for short circuiting its slip rings, or for
disengaging the
contacts of a mechanical rectifier.
or It operates or permits the operation
Voltage Device predetermined polarity only
verifies the presence of a polarising
voltage in an equipment.
Undercurrent or It functions when the current or
power power flow decreases below a
predetermined value.
1 Relav
It functions on excessive bearing
temperature, or on other abnormal
mechanical conditions, such as
undue wear, which may eventually
Power Relay



OALSTOM Limited, Energy Automation & Information


APPS- Combined course

Overview Of Pr~?xt!on
Page 1 6 of 0

I Device Number I


associated with bearings as covered
under device function 38), such as
excessive vibration, eccentricity,
expansion, shock, tilting, or seal
It functions on a given or abnormally
Field Relay
low value or failure of machine field
current, or on an excessive value of
armature current in an ac machine
indicating abnormally low field
Circuit It is a device, which functions to
apply, or to remove the field
excitation of a machine.
function is to connect a machine to
its source of running or operating
voltage. This function may also be
used for a device, such as a
contactor, that is used in series with
a circuit breaker or other fault
protecting means, primarily for
frequent opening and closing of the
Manual Transfer It transfers the control circuits so as
Selector to modify the plan of operation of
Device transfers the switching equipment or of some
of the devices.
Unit Sequence It is a device, whichfunctions to start
the next available unit in a multipleStarting Relay
unit equipment on the failure or on.
( the non-availability of the normally 1
preceding unit.
It is a device that functions upon the
I Atmospheric
I Condition
such as
damaging fumes, ex'plosive mixture,
smoke or fire.
I Reverse-Phase. It is a relav which functions when the 1
. .



OALSTOM Limited, Energy Automation & Information

Overview Of Protection Fundamentals

Page 1 7 of 0

Device Number


Current Relay

poly-phase currents are of reverse
phase sequence, or when the polyphase cu:cer?,?s fire vnbolonced zl;
contain negative phase-sequence
- It functions, upon a predetermined
1 Sequence
- 1 value of poly phase voltage in the ]
Voltage Relay
desired phase sequence.
It is a relay that generally returns the
Sequence Relay equipment to the normal, or off,
position and locks it out if the normal
starting, operating or stopping
sequence is not properly completed
within a predetermined time. If the
device is used for alarm purpose
only, it should preferably be
designated as 48A (alarm).
or It is a relay that functions w.hen the
temperature of
armature, or other -load carrying
Thermal Relay
winding or element of a machine, or
the temperature of a power rectifier
or power transformer (including a
power rectifier transformer) exceeds
an medetermined value.
[ I t is a relay that
overcurrent, or instantaneously on an excessive
rise value of current, or on an excessive
current rise, thus indicating a fault in
the apparatus or circuit being
Time It is a relay with either a definile or
an inverse time characteristic lhat
functions when the current in an ac
circuit exceeds a predetermined
Circuit It i s a device that is used to close
I Breaker
and interrupt an ac power circuit
under normal conditions or to



OALSTOM Limited, Energy Automation & Information

APPS- Cornbined course


Overview Of Pr~?w-t!on

Page 16 of 0

Device Number



associated with bearings as covered
under device function 38), such as
excessive vibration, eccentricity,
expansion, shock, tilting, or seal
It functions on a given or abnormally
Field Relay
low value or failure of machine field
current, or on an excessive value of
armature current in an ac machine
indicating abnormally low field
Circuit It i s :a device, which functions to
apply, or to remove the field


Circuit It is a device whose principal

function i s to connect a machine to
its source of running or operating
voltage. This function may also be
used for a device, such as- a
contactor, that is used in series with
a circuit breaker or other fault
protecting means, primarily
frequent opening and closing of the
Manual Transfer It transfers the control circuits so as
Selector to modify the plan of operation of
Device transfers the switching equipment or of some
of the devices.
Unit Sequence It is a device, which functions to start
the next available unit in a multipleStarting Relay
unit equipment on the failure or on
1 the non-availability of the normally 1




I Monitor



atmospheric condition, such

damaging fumes, explosive mixture,
smoke or fire.
lt is a reiay which functions when the

OALSTOM Limited, Energy Automation & Information

APPS- CcmSined CGiiiSs

Overview Of Protection Fundamentals
Page 1 7 of 0

Device Number


poly-phase currents are of reverse
phase sequence, or when the polyphclse currents me unbalaficzd o i
contain negative phase-sequence
It functions upon a predetermined
value of poly phase voltage in the
desired phase sequence.
Volta e Rela
It is a relay that generally returns the
Sequence Relay equipment to the normal, .or off,
position and locks it out i f the normal
starting, operating or stopping
sequence is not properly completed
within a predetermined time. If the
device is used for alarm purpose
I only, it should preferably be /
designated as 48A {alarm).
or It i s a relay that functions when the
armature, or other load carrying
Thermal Relay
winding or element of a machine, or
the temperature of a power rectifier
or power transformer (including a
power rectifier transformer) exceeds
an predetermined value.
It is a relay that functions
overcurrent, o i instantaneously on an excessive
rise value of current, or on an excessive
current rise, thus indicating a fault in
, Relay
the apparatus or circuit being
Time / It i s a relay with either a definile or
an inverse time characterisiic l l m t
functions when the current in an ac
( circuit exceeds a predetermined
Circuit It is a device that is used to c:lose
1 and interrupt an ac power circuit
under normal condilions or to
Current Relay


OALSTOM Limited, Energy Automation & Information

APPS- Combined course

Overview Of Protec?bnFundamentals
Page 20 of 0

Device Number




mechanical positioning.
directional It is a relay that functions on a
desired value of ac overcurrent
I flowing in a predetermined I
Blocking Relay
It is a relay that initiates a pilot signal
for blocking of tripping on external
faults in a transmission line or in other
apparatus under predetermined
conditions, or co-ordinates with
other devices to block tripping or to
block re-closing on an out-of-step
condition or on powerswing:.
It is - generally a two position,
! Control ~ e v i c e ' manudlly operated switch that in
one position permits the closing of a
circuit breaker, or the placing of an
equipment into operation, and in
the other posilion prevents the
circuit breaker or the equipment
from being operated.
It is a variable resistance device
I Rheostat
used in an electric circuit, which is
electrically operated or has other
electrical accessories, such as
auxiliary position or limit switches.
It is a switch which operates on
1 Level switch
given values, or on a given rate of
change, of level.
circuit It is used to close and interrupt a dc
Breaker conditions or to interrupt this circuit
i Load - Resistor It is used to shunt or insert a step of
load limiting, shifting, or indicating
' Contactor
resistance in a power circuit, or to
switch a space heater in circuit, or
to switch a light, or regenerative
load resistor of a power reclifier or





OALSTOM Limited, inergy Automation & Informalion

APPS- Combined course

Overview Of Protection Fundamentals
Page 21 of 0

Device Number


other machine in and out of circuit.
It i s a device other than the
annunciator, as covered under
device No.30, which is used to
operate in connection with a visual

Alarm ~ e b y

that isused for

moving a muin device from one
position to another i n a n equipment
;as . for
removable circuit breaker unit to
disconnected, and test positions.
DC Overcurrent It is a relay that functions when the
current in a dc circuit exceeds a
aiven value.
Pulse Transmitter It i s used to generate and transmit
I pulses over a telemetering or pilotwire circuit to the remote indicating
or receiving device.
Angle It is a relay that funclions at a
or predetermined
between two voltages or between
two currents or between voltage







0Flow Switch




that controls
automatic reclosing and locking out

It is a switch, which operates on

on a given rate of I
change, of flow.
It is a relay that functions on a
predetermined value of frequency,
either underlover on normal system
frequency or rate of , change of
1 automatic closing and reclosing of a (
dc circuit interrupter, generally in
resoonse to load circuit conditions.


1 given values, or

OALSTOM Limited, Energy Automation & Information

APPS- Combined course

Overview Of Protec?i@r!
Page 20 of 0

Device Number


mechanical positioning.
directional It is a relay that functions on a
desired value of ac overcurrent
( Relay
Blocking Relay
It is a relay that initiates a pilot signal
for blocking of tripping on external
faults in a transmission line or in other
apparatus under predetermined
conditions, or co-ordinates with
other devices to block tripping or to
block re-closing on an out-of-step
condition or on po\der swings:
1 Permissive
It is generally a two position,
i Control Device
manually operated switch that in
one position permits the closing of a
circuit breaker, or the placing of an
equipment into operation, and in
the other position prevents the
circuit breaker or the equipment
from being operated.
It is a variable resistance device
I Rheostat
used in an electric circuit, which is
electrically operated or has other
electrical accessories, such a:
auxiliary position or limit switches.
It is a switch which operates or
i Level Switch
given values, or on a given rate of
change, of level.
circuit It is used to close
1 Breaker I
conditions or to interrupt this circuit
: Load - Resistor It is used to shunt or insert a step of
load limiting, shifting, or indicating
resistance in a power circuit, or to
switch a space heater in circuit, or
to switch a light, or regenerative
load resistor of a power rectifier or









OALSTOM Limited, inergy Automation & Information

APPS- Combined course

Device Number













Overview Of Protection Fundamentals

Page 21 of 0

other machine in and out of circuit.
It is a device other than the
Alarm ~ e l a y
annunciator, as covered under
device No.30, which i s used i o
operate in connection with a visual
or audible alarm.
It is a mechanism that i s used for
moving a main device from one
position to another in an equipment
removable circuit breaker unit to
disconnected, and test posilions.
DC 0ve:current It is a relay that functions when the
l current in a dc circuit exceeds a
given value.
Pulse Transmitter It is used to generate and transmit
pulses over a telemetering or pilotwire circuit to the remote indicating
or receiving device.
I Phase
Angle It is a relay that functions at a
or predetermined
between two voltages or between
two currents or between voltage
and current.
Re-closing It is a relay that controls the
automatic reclosing and locking out
of an ac circuit interrupter.
It is a switch, which operates on
Flow Switch
given values, or on a given rate of
change, of flow.
It is a relay that functions or-) a
predetermined value of frequency,
e~therunderJover on normal system
frequency or rate of change of
Re-closing It is a relay that controls
automatic closing and reclosing of a
dc circuit inter~upter,generally in
response to load circuit conditions.

OALSTOM Limited, Energy Automation & information

APPS- Combined course

Overview Of Protection Fundamentals
Page 24 oi 6

Device Number

to 94 i s suitable.



Devices Performing
Than One Function
If one device performs two relatively important functions in an equipment
so that it is desirable to identify both of these functions, this may be done
by using a double function number and name such as:
50/51 - Instantaneous and Time Overcurrent Relay.
If h.10 or more devices with the same function number and suffix letter (if
used) are present in the same equipment, they rlay be distinguished by
numbered suffixes as for example, 52X-1, 52X-2 and 52X-3, when

Suffix Letters
Suffix letters are used with device function numbers for va~~ious
purposes. In
order to prevent possible conflict each suffix letter should have only one
meaning in an individual equipment. All other words should use the
abbreviations as contained in ANSI Y 1.1 latest revision, or should use some
other distinctive abbreviation, or be written out in full each time they are
used. The meaning of each single suffix letter, or combination of letters,
should b e clearly designated in the legend on the drawings or
publications applying to the equipment.
Lower case (small) suffix letters are used in practically all instances on
electrical diagrams for the auxiliary, position, and limit switches. Capital
letters are generally used for all other suffix letters. Th,e letters should
generally form part of the device function designation, are usually written
directly after the device function number, as for example, 52CS. 71 W, or
49D. When it is necessary to use two types of suffix letters in connection
with one function number, it is often desirable for clarity to separate them
by a slanted line or'dash, as for example, 20D/CS or 2OD-CS. .
The suffix letters which denote parts of the main device, and those which
cannot or need not form part of the device function designation, are
generally written directly below the device function number on drawings,
as for example, 52/CC or 43/A.


Standard reference positions of some typical devices

OALSTOM Limited, Energy Automation & Information

APPS- Combined course

Overview Of Protection Fundamentals
Page 250f 0

Power Circuit Breaker
Disconnecting Switch
Load-break swiich Valve

Adjusting Means
Relay (2)
Contactor (21
Contactor (latched-in-type)
Temperature Relay (3)
Level Detector (3)
Flow Detector (3)
Speed Switch (3)
Vibration Detector (3)
1 Pressure Switch 131
) Vacuum Switch (3)

Standard Reference Position

Main Contacts Open
Main Contacts Open
Main Contacts Open
Closed Posilion )
Closed Posilion
Disengaged Position
Disengaged Position
Maximum Gap Position
Maximum resistance Position
De-energised Position
Mairi Contacts Open
Lowest Temperature
Lowest Level
Lowest Flow
Lowest Speed
Minimum Vibration
I Lowest Pressure
1 Lowest Pressure i.e., Highest Vacuum

Note : If several similar auxiliary switches are present on the same device, they should be
designated numerically 1.2.3etc, when necessary.
( 1 ) 'These may be speed, voltage, current, load, or similar adjusting devices comprising
rheostats, springs. levers, or other components for the purpose.

( 2 ) These electrically operated devices are of the non-latched-in type, whose contact
p,osition is dependent only upon the degree of energisation of the operating or
restraining or holding coil or coils which may or may not be suitable for continuous
energisation. The de-energised position of the device i s that with all coils deenergised.
(3) The energising influences for these devices are considered to be, respectively, rising
temperature, rising level, increasing flow, rising speed, increasing vibration, and
increasing pressure.

The simple designation "a" or "b" is used in all cases where there is no
need to adjust the contacts to change position at any particular point in
the travel of the main device or where the part of the travel, where the
contacts change position is of no significance in the control or operating
scheme. Hence fhe "a" or "b" designations usually are sufficient for
circuit breaker auxiliary switches.
OALSTOM Limited, Energy .4utomation & Information

APPS- Combined course

Device Number

Overview Of Protection Fundamentals

Page 24 oi O

to 94 is suitable.


Devices Performing ~ 6 r Than
One Funciion
If one device performs two relatively important functions in an equipment
so that it i s desirable to identify both of these functions, this may be done
by using a double function number and name such as:

50151 - Instantaneous and Time Overcurrent Relay.

Suffix Numbers
If two or more devices with the same-function number and suffix letter (if
used) are present in the same equipment, they may be distinguished by
numbered suffixes as for example, 52X-1, 52X-2 and 52X-3, when
Suffix Letters
Suffix letters are used with device function numbers for various purposes. In
order to prevent possible conflict each suffix letter should have only one
meaning in an individual equipment. All other words should use the
abbreviations as contained in ANSI Y 1 . l latest revision, or should use some
other distinctive abbreviation, or be written out in full each time they are
used. The meaning of each single suffix letter, or combination of letters,
should be clearly designated in the legend on the drawings or
publicalions applying to the equipment.

Lower case (small) suffix letters are used in practically all instances on
electrical diagrams for the auxiliary, position, and limit switches. Capital
letters are generally used for all other suffix letters. The letters should
generally form part of the device function designation, are usually written
directly after the device function number, as for example, 52CS, 71 W, or
490. When it is necessary to use two types of suffix letters in connection
with one function number, it is often desirable for clarity to separate them
by a slanted line or'das!?,as for example, 20DJCS or 20D-CS.
The suffix letters which denote parts of the main device, and those which
cannot or need not form part of the device function designation, are
generally written direcily below the device function number on drawings,
as for example, 52lCC or 43lA.


Standard reference positions of some typical devices


OALSTOM Limited, Energy ~utomatidn& Information

APPS- Combined course

Overview Of Protection Fundamentals
Page 25 of 0

. ..-


Standard Reference Position

Main Contacts Open
Main Contacts Open
Main Contacts Open
Closed Position
Closed Position
I Disengaged Position
Disengaged 'Position
Turning Gear
Power Electrodes
Maximum Posilion
Maximum resistance Posilion
Low or Down Position
Adjusting Means ( 1 )
De-energised position
Relay (2)
Contactor (2)
De-energised Position .
Contactor (latched-in-type) Main Contacts Open
Lowest Temperature
Temperature Relay ( 3 )
I Level Detector 131
! Lowest Level
1 Lowest Flow
Flow Detector (3)
Lowest Speed
Speed Switch (3)
Vibration Detector (3)
Lowest Pressure
Pressure Switch (3)
1 Lowest Pressure i.e., Highest Vacuum
Vacuum Switch (3)


Note : If several similar auxiliary switches are present on the same device, they should be
designated numerically 1,2,3 etc, when necessary.

( 1 ) These may be speed, voltage, current, load. or similar adjusting devices conlprising
rheostats, springs, levers, or other components for the purpose.

(2) These electrically operated devices are of the non-latched-in type, whose contact
position is dependent only upon the degree of energisation of the operating or
restraining or holding coil or coils which may or may not be suitable for continuous
energisation. The de-energised position of the device is that with all coils deenergised.

(3) The energising influences for these devices are considered to be, respectively, rising
temperature, rising level, increasing flow, rising speed, increasing vibration, and
increasing pressure.

The simple designation "a" or "b" is used in all cases where there is no
need to adjust the contacts to change position at any particular point in
the travel of the main device or where the part of the travel, where the
contacts change position is of no significance in the control or operating
scheme. Hence fhe "a" or "b" designations usually are sufficient for
circuit breaker auxiliary switches.
OALSTOM Limited, Energy .4utomation & Informalion

APPS- Cornbined course

Overview Of Pr~:ec:Icn Fundamentals
Page 28 of 0

The following Chart gives a birds-eye view of the relay classifications

based on technology.






The simplest form of overcurrent protection is the fuse. The fuse is capable of operating
in less than 10ms for very large values of current, thus considerably limiting fault energy.
However, it does have a number of disadvantages, namely;
Can be difficult to co-ordinate
Its characteristic is fixed
Needs replacing ioiiowing iauit ciearance
Has limited sensitivity to earthfaults since it is rated above the full load current of the
Operation of single fuse results in a condition refereed to as single phasing. Single
phasing .can be disastrous for rotating plant such as motors.
The fuse characteristic is split into two sections, the 'Pre-arcing Time' and the 'Arcing
Time'. The addition of these times is referred to as the 'Total Operating Time'.



The purpose of overcurrent protection, as with other forms of protection, is to detect
faults on a power system and as a result, initiate the opening of switchgear in order to
isolate the faulty part of the system. The protection must thus be discriminative, that is
to say it shall, as far as possible, select and isolate only the faulty part of the system
leaving all other parts in normal operation.
Discrimination can be achieved by overcurrent, or by time, or by a combination of
overcurrent and time.

Discrimination by current relies upon the fact that the fault curren't varies with the
position of the fault. This variation is due to the impedance of various items of plant,
such as cables and transformers, between the source and the fault. Relays throughout
the system are set to operate at suitable values such that only the relay nearest to the
fault operates.
of operation are generally termed Instantaneous
Relays which adopt this
overcurrent relays.
(Where the fault level does not vary greatly between two relay location then the use of
i n s t a n t a n e ~ sovercurrent relays is not possible).
Page 1


If the fault level over a system is reasonably constant then discrimination by current will
not be possible. An alternatlile Is tc use time discrimination in which each overcurrent
relay is given a fixed ?irr?edelay with the relay farthest away from the source having the
shortest time delay. Operating time is thus substantially,independent of fault level but
the main disadvantage is that the relay nearest the source will have the longest time
delay and this is the point with the highest fault level.
Relays which adopt this principle of operation are generally termed definite
(independent) time overcurrent relays.
NOTE : When applying definite time overcurrent relays care must be taken to ensure
that the thermal rating of the current measuring element is not exceeded.

(Relay Current Setting)


Due to the limitat~onsimposed by the independent use of either t ~ m eor current, the
inverse time overcurrent characteristic has been developed. With this character~st~c
time of operation is inversely proportional to the current applied, i.e.; basically the higher
the current applied, the faster the relay operates. Thus, the actual characteristic IS a
function of both t ~ m eand current settings, thereby gaining the advantages of the
previous mentioned methods and eliminating some the disadvantages.


(Relay Current Setting)

Applied Current'

The principle of co-ordination refers to the procedure of setting overcurrent relays to
ensure that the relay nearest the fault operates first and all other relays have adequate
additional time to prevent them from operating. If the relay nearest fo the fault fails to
clear the fault, and the co-ordination is correct, then the next up-stream relay should
operate and so on towards the source, thus isolating the minimum amount of plant.
The principle of co-ordination is often referred to as 'grading'.
When performing any co-ordination exercise the following need to be considered:
Relay Characteristics
Relay Current Setting
Grading Margin
Time Multiplier Setting

Relay Characteristics
There are numerous characteristics, however they all confirm to either BS142lIEC or
ANSIIIEEE standards. The BS142lIEC standard incorporates the following
Standard Inverse
Very lnverse
Extremely Irlverse
Long Time Inverse
The ANSIIIEEE standard incorporates the following characteristics:
Moderately InverseVery lnverse
Extremely lnverse
Short Time lnverse
The BS142lIEC standard curves are mainly adopted in the LIK and the most commonly
used ones are explained in more detail below:
Standard lnverse - This characteristic is commonly known as the 3110 characteristic,
i.e. at ten times setting current and TMS of 1 the relay will operate in 3 secs.
The characteristic curve can be defined by the mathematical expression :

where I


applied current
setting current
multiple of setting current

The standard inverse time characteristic is widely applied at all system voltages - a s
back up protection on EHV systems and as the main protection on HV and MV
distribution systems.


In general, the standard inverse characteristics are used when

There are no co-ordination requirements with other types of protective equipm2nt further
out on the system, e.g. Fuses, thermal characteristics of transformers, motors etc.
The fault levels at the near and far ends of the system do not vary significantly.
Page 3

There is minimal inrush on cold load pick up. Cold load inrush is that c u i e n t which
occurs when a feeder is energised after a prolonged outage. In general the relay cannot
be set above this value but the current should decrease below the relay setting before
the relay operates.

Very lnverse Time This type of characteristic is normally used to obtain greater time
selectivity when the limiting overall time factor is very low, and the fault current at any
point does not vary tno :vlde!y with system conditions. It is particularly suitable, if there is
a substantial reduction of fault current as the distance from the power source increases.
The steeper inverse curve gives tonger time grading intervals. Its operating time is
approximately doubled for a reduction in setting from figures 7 to 4 times the relay
current setting. This permits the same time multiplier setting for several relays in series.
The characteristic curve can be defined by the mathematical expression :
t =


{i: I]

Extremely lnverse Time With this characteristic the operating time is approximately
inversely proportional to the square of the current. The long operating time of the relay
at peak values of load current make the relay particularly suitable for grading with fuses
and also for protection of feeders which are subject to peak currents on switching in,
such as feeders supplying refrigerators, pumps, water heaters etc., which remain
connected even after a prolonged interruption of supply.
For cases where the generation is practically constant and discrimination with low
tripping times is difficult to obtain, because of the low impedance per line section, an
extremely inverse relay can be very useful since only a small difference of current is
necessary to obtain an adequate time difference.

Another application for this relay is with auto reclosers in low vo'ltage distribution circuits.
As the majority of faults are of a transient nature, the relay is set to operate before the
normal operating time of the fuse: thus preventing perhaps unnecessary blowing of the
Upon reclosure, if the fault persists, the recloser locks itself in the closed posjtion and
allows the fuse to blow to clear the fault.
This characteristic is also widely used for protecting plant against overheating since
overheating is usually an I,t function.

Page 4

This characteristic curve can be defined by the mathematical expression :

t =



b o n g Time Inverse This type of characteristic has a long time characteristic and may
be used for protection of neutral earthing resistors (which normally have a 30 sec rating).
The relay operating time at 5 times current setting is 30 secs at TMS of 1.
This can be defined by :

Current Setting
The current setting of a relay is typically described aS either a percentage or multiple of
the current transformer primary or secondary rating.
If the CT primary rating is equal to the normal full load current of the circuit then the
percentage setting will refer directly to the primary system. This is an important point as
if, for example, the normal primary full load current was, say, 400 amp but the CT ratio
was 50015 then a relay with setting range 50-200% of 5 amp set at 100% would not
represent a "full load" setting;-the actual setting would in fact be 125% of full load
The choice of current setting thus depends on the load current and the CT ratio and is
normally close to but above the maximum load current (typically'lO%) - assuming of
course the circuit is capable of carrying the maximum foreseeable load. It should be
stressed at this point, that the relay is neither designed nor intended to be used as an
overload relay but as a protective relay to protect the system under fault conditions.
It is also important to consider the resetting of the relay. The relay will reset when the
current is reduced to 90%-95% of the setting (Depending on relay design) and if the
normal load current is above this value the relay will not reset after starting to operate
under through fault conditions which are cleared by other switchgear.
The setting for a typical overcurrent relay with a reset ratio of 95% can be determined
using the following:


Is = Setting
IF^ = Full Load Current

Grading Margin
As previously mentioned, to obtain correct discrimination it is necessary to have a time
interval between the operation of two adjacent relays. This time interval or grading
margin depends upon a number of factors :
The circuit breaker fault interrupting time
The overshoot time of the relay
Final margin on completion of operation (safety margin)
The discriminating relay can only be de-energised when the circuit breaker has
completely interrupted the fault current. It is now normal practice to use a value of 50 Page !

100 ms for circuit breaker overall interrupting time but obviously if it is known that the
switchgear is slower than this time, this must be taken..injo account.Operating of the relay may continue for a short time after the relay is de-energised until
a n i stored energy is dissipated. For example, an induction disc 'element will have stored
kinetic energy (or inertia) and a numerical relay may have stored energy in capacitors.
Although these factors are minimised by design, some allowance is usually necessary.
It is common to use a figure of 50 ms.


The overshoot time is not the actual time during which some forward
operation takes plan but is the time that the relay would have taken to travel
the same distance had the relay remained energised.



t l = relay de-energised
t3 - t l = actual overshoot time
t2 - t l = overshoot time used in the
calculation of margin




All measuring dev~cessuch as rejays and current transformers are subject to some
degree of error The t ~ m echaracteristic of either or both of the relays involved may have
positive or negat~veerrors. Current transformer errors are mainly due to the
rnagnetis~ngcharacteristic. It should be noted the CT errors do not affect definite time
overcurrent relays.
A safety margin of 100 ms is normally added to the final calculated margin to ensure
correct discrirn~nation.This additional time ensures a satisfactory contact gap (or
equivalent) is maintained.
In the past, a fixed margin of 0-4 secs was considered adequate for correct
discrimination. With faster modern switchgear and lower overshoot times a figure of
0.35 secs is quite reasonable and under the best possible conditions 0-3secs may be
However, rather than using a fixed margin it is better to adopt a fixed time for circuit
breaker operation and relay overshoot and add to this a variable time value which takes
into account relay and CT errors and the safety margin. This is particularly so when
grading at low values multiples of setting current where the relay operating time is
longer and a fixed total margin may be of the same order as the relay timing error.
A fixed value 0-25 secs is chosen which is made up of 0.1 secs for circuit breaker
operating time. 0.05 secs for relay overshoot time and 0.1 sec for safety margin.

In considering the variable time value, it is assumed that each IDMT relay complies with
basic assigned error class 7.5 according to British practice in BS 142. The error for a
class 7.5 relay IS 5 7.5%, but allowance should be made for the effects of temperature,
frequency and departure from the reference conditions as laid down in the BS. A more
practical approximation is to assume a total effective error of 2 x 7-5 i.e. 15% and
is to apbly to the relay nearest the fault which'is considered slow. To this total effective
relay error a further 10% is added to allow for overall CT error.
Page 6

Thus it is proposed to adopt the following equstior: ,:t determine the grading margin
between IDMT relays :

0.25 + 0-25 secs

where t = 'normal operating time of

relay nearest the fault

As far as definite time overcurrent relays are concerned, the fixed value will remain the
same but the relays are assumed to comply with error class 10 i.e. 10%. For the
reasons stated .previously, a practical approximation is to assume a total effective error
of 20% with the relay nearest the fault considered slow. As previously stated, CT errors
will have little effect of the operating time, thus it is proposed to adopt the equation :


0.25 + 0.25 secs

For the majority of systems an overcurrent grading exercise can be performed quite
adequately using a fixed margin of 0.4 secs. It is only when a number of stages are
involved and difficulties are being encountered that it may become necessary to
invmtigate margin times in more detail. To summarise, each system is different and
should be treated as so, it is not possible to lay down rigid rules regarding grading
margins and every grading exercise will ultimately be a compromise of some form.
Grading Overcurrent Relays With Downstream Fuse
For some applications ~twill be necessary to grade overcurrent relays with fuses. When
the fuse is downstream of the relay the following formula can be used to calculate the
grading margin.


Grading Margin = 0.4Tf + 0.1 5s over the whole characteristic.

The above formula assumes a minimum fuse operating time of 0.01 seconds
Generally for this type of application a Extremely Inverse characteristic should be
chosen to grade with the fuse and the current setting of the relay should be 3 - 4 x rating
of fuse to ensure co-ordination.

Time Multiplier Setting

The time multiplier setting is a means of adjusting the operating time of an inverse type
characteristic. It is not a time setting but a multiplier.
In order to calculate the required TMS (Treq), calculate the operating time of the'nearest
downstream protection device at the maximum fault level seen by both devices, add to
this the grading margin, calculate the operating time of the upstream device at this fault
level with a TMS equal to one (TI) and then use the following for formula:

TMS = Treq 1 T I
Plotting Of Characteristic
It is convenient to show the standard inverse time characteristic on logllog graph paper
with the 'y' axis scaled in seconds and the"x' axis in terms of "multiples of current
setting". By doing this the characteristic can be applied to any relay, irrespective of
setting range and nominal rating.





Where the source impedance is small in comparison with the protected circ~lit.i.mped.ance,. .
the use of high set instantaneous overcurrent units can be advantageous (for example on
long transmission lines or transformer feeders).


The application of an instantaneous unit makes possible a reduction in the tripping time at
high fault levels and also allows the discriminating curves behind the high set unit to be
lowered thereby improving overall system grading.
i i is important io note iiiai when grading with the relay immediately behind the high set
units, the grading interval should be established at the current setting of the high set unit
.and not at the maximum fault level that would normally be used for grading IDMT relays.
When using high set units it is important to ensure that the relay does not operate for
faults outside the protected section. The relays are normally set at 1.2 - 1.3 times the
maximum fault level at the remote end of the protected section.
.This particularly applies when using instantaneous units on the HV side of a transformer
when the instantaneous unit should not operate for faults on the LV side.
The 1-2- 1.3 factor allows for transient overreach, CT errors and slight errors in
transformer impedance and line length.


Transient overreach occurs when the current wave contains a dc component. Although a
relay may have a setting above the rms value of current, the initial peak value of current
due to the dc offset may be sufficient to operate the relay, if it has high transient
Percentage transient overreach is defined as



Where :


relay pick-up current in steady state rms amps

rms value of current which when fully offset will just pick up the relay

Modern Relays have integral instantaneous elements which have low transient overreach.
The degree of transient overreach is normally affected by the time constant of the
measured fault current. For example, a typical transient overreacn of a numerical
overcurrent relay is less than 5% for time constants up to 30 ms and less than 10% for
t ~ m econstant up to 100 ms. This allows the instantaneous elements to be used as h ~ g h
set un~tsfor application to transformers and long feeders. The low'transient overreach
allows settings to be just above the maximum fault current at which discrimination IS
required. The instantaneous elements are also suitable for use as low set elements in
conlunction with auto-reclose on distribution systems


Earth faults, which are by far the most frequent type of fault, will be detected by phase
overcurrent units as previously described but it is possible to obtain more sensitive
protection by utilising a relay which responds only to the residual current in a system.
Residual (or zero sequence) current only exists when a current flows to earth.
The residual current can be detected either by connecting a CT in an available neutral to
earth connection or by connecting Ilne CT's in parallel. By using this parallel connection
the earth fault relay is completely unaffected by load currents whether balanced or
unbalanced. The parallel connection can be extended to include either two or three
Page 8


overcurrent units without any effect on the earth fault relay. Two elements are often
considered sufficient 2s any interphase fault must affect at least one of the relays,
however, consideration must be given to the possibility of 2-1-1 current distribution i n the
system (refer deltalstar transformer protection).
It should be noted that on an LV 4 wire distribution system, 4 CT's will be required to
ensure stability under all load conditions, the 4th CT being placed in the neutral
connection. This fourth CT can be omitted if the earth fault relay setting is above the
maximum spill current caused byunbalanced loads, but as the degree of unbalance is not
riorii~aiiykiiowii (accilrately) the inclusion of the 4th CT is recommended.

Time Grading
The procedure for grading is similar to that for phase fault relays.
It is important to appreciate that fuses cannot discriminate between phase faults and earth
faults and therefore grading of earth fault relays (which have relatively sensitive settings)
with fuses is not possible.
When the system contains some neutral earthing impedance, the earth fault level is
practically constant over the whole system and grading is carried out at !his fault level. As
the fault level is consJant there is no particulai advantage is using IDMT earth fault relays
over definite time earth fault relays.
Sensitive Earth Fault Relays
Where the earth path resistivity is high which may be the case on systems that do not
utilise earth conductors, the earth fault current may be limited to such an extent that
normal earth fault protection may not be sensitive enough. To overcome these problems
a very sensitive relay is requirgd, but the relay must have a very low burden in order that
the effective setting is not increased. This very sensitive protection cannot be graded with
other conventional systems and it is normal to apply this protection with a definite time
delay of up to 10 or 15 secs. This time delay will prevent unwanted operation due to
transient unbalance under phase fault conditions. Care must be taken to ensure that the
relay setting is above any residual current that may be present under normal load
conditions. This may be due to slight d~fferencesin CT characteristics or unbalanced
leakage (capacitive) currents in the primary system. In order to ensure that the relay will
reset after the transient operation of the current measuring unit, the dolpu ratio should be
high, i.e.. approximately 99%.

The foregoing has basically looked at grading procedure as applied to radial feeders. If
the system is interconnected and involves parallel paths and rings, the grading can
become increasingly more complex.
For example, the operation of a particular circuit breaker may not itself result in the
isolation of the faulty plant, but may affect the fault current distribution in the other circuits.
The affect of this may be to start other relays operating or to change the operating
parameters of relays that havealready started. On such interconnected systems the fault
level does not tend to vary very much and it may be found impossible to obtain correct
discrimination for all faults. The system must be looked at in detail under maximum and
minimum fault conditions and the best compromise reached. Very often directional
overcurrent relaying can help to overcome the problems slightly.

Page 9

Directional Overcurrent


If fault current can flow in both directions through the relay location it is necessary to
add directional properties to the overcurrent relays in order to obtain correct
discrimination. Directional protection is commonly applied in two areas, namely,
parallel feeders(transf0rmers) and ring mains.


The more usual application of directional relays is to ring mains. In the case of a ring
system, fed at one point only the relays at the generafion end and at the mid-point
substation, where the setting of both overcurrent relays are identical, the relays can
be made non-directional, provided that in the latter case the relays are located on the
same feeder, one at each substation. In this respect it is interesting to note that
when the numbers of feeders in the rings is an even number, the two relays with the
shme operating time are at the same substation and.will have to be directional
whereas when the number of feeders is odd, the two relays.with the same operating
time are at different substations and therefore, do not need'to be directional. Also at
intermediate substations it will be noted that whenever the times of the two relays at
a substation are different, the difference in operating time is never less than the
grading interval of 0-4 seconds and consequently it is permissible for the relay with
the larger operating time to be non-directional.
Grading Ring Mains
The usual practice for grading relays in an interconnected system is to open the ring
at the supply point and to grade the relays first clockwise and then anti-clockwise.
Thus, the relays looking in a clockwise direction around the ring are arranged to trip
in the sequence 1 - 2 - 3 - 4 - 5 -3and the relays looking in the anti-clockwise
direction are arranged to trip in the sequence 1' - 2' - 3' - 4' - 5' - 6'. The arrows
indicate the direction in which the power must flow in order that the directional units
will close their contacts and prepare the overcurrent elements for operation. The
double headed arrows on each of the two ieeders at the generating station indicate
non-directional relays, directional features being unnecessary at these points,
because power can flow in one direction only, that is out of the generating station. At
all other points s~ngleheaded arrows are shown. These ind~catedirect~onalrelays
connected so as to operate with power flow in the direction of the arrow which is in
every case from the substat~onbus bars and into the protected line. See Figure 1.
This rule is invariable and applies to all forms of directional relays. Selection of the
faulty section is by time and fault power direction. Fault power has two phases x and
y. It divides between the two paths in the inverse ratio of the impedances an.d
passes through all the substations in the ring. Thus, at every substation one set of
relays will be inoperative because the power flow is against the arrow and other set
operative because the flow is with the arrow. In every case it will be found that the
time settings of the relays that are inoperative are shorter than those of the operative
relays, except in the case of substation C where the settings happened to coincide.
In this way, all relays with short time on sections between the fault one and the
generating station are prevented from operation. The others, which are operative are
graded downwards towards the fault and the last to be traversed by the fault current,
namely that on the faulty feeder section, has the shortest time and operates first.
This applies to both paths to the fault. Consequently, the faulty sectioh is the only
one to be isolated and supply is maintained to all substations.

Page 1

Grading Ring Mains With More Than One Source

When grading ring systems with more than one infeed (say two sources of supply)
the best method of approach is to either :


Open the ring at one the supply points by means of a suitable high set
instantaneous overcurrent relays and then proceed to grade the ring as in the
case of a single infeed.
Treat the inter-connector between the two sources of supply as a continuous
bus, separate from the ring and protect it by means of a unit system of
protection such as pilot wire relays. Then proceed to grade the ring as in the
case of a single infeed.


If non-directional overcurrent relays are applied to parallel feeders any faults

occurring on any one line will inevitably, irrespective of the relay setting chosen,
isolate both lines and completely disrupt the supply. To ensure discriminative
operation of the relays during line faults, it is usual with this type of system to design
and connect relays Rq' and R2' such that they will only operate for faults occurring on
the protected line in the direction indicated by the arrows. See Figure 2. With
parallel feeders to ensure correct discrimination during line faults, it is important that
the correct direct~onalrelay R1' or R2' operates before the non-d~rectionalrelays Rq
and R2. For this reason relays R1' and R2' are given lower time settings than relays
R1 and R2 and also lower current settings. The usual practice is to set relays Rq'
and R2' to 50% of the normal full load of the circuit (ensure that the relays are
capable of carrying without damage, twice their setting current continuously),
operating with an IOMT characteristic with a TMS 4.0
Care should be taken when using definite time relays. For such applications the
directional relays should be set above full load current to prevent them operating due
to load current reversal as a result of a phase to phase fault on the other side of the


The direction of alternating current can only be determined with respect to a common
reference. In relay terms, the reference is commonly referred to as the polarising
quantity. The most convenient reference quantity is polarising voltage taken from the
power system voltages.
The relay compares the power system current against this fixed polarising reference
to determine direction of operation.


This is a setting on the relay and is defined as the angle by which the current applied
to the relay must be displaced from the voltage applied to the relay to produce
maximum sensitivity.

Page 2

This is the angle by which the current applied to the relay is displaced from the
voltage applied to the relay at unity power factor.
The 90" connection (quadrature connection) is now used for all overcurrent relays.
30" and 60" connections were used in the past, but no longer, as the 90" conneciion
gives better performance. The 90" connection is achieved by using IA and VBC.
Hence, for an A phase fault the polarising voltage does not collapse. Without a
polaring voltage most relays are unable to make a directional decision. Modern
numerical relays are able to use prefault data to make a decision, a technique
referred to as memory .polarising.

90" Connection - 45" RCA

The 'a' phase relay is supplied with la current and Vbc volts displaced 45" in an anticlockwise direction. ln-this case the relay maximum sensitivity is produced when the
current lags the system phase to neutral voltage by 45". This connection gives a
correct directional tripping zone over the range of current 45" leading to 135" lagging
See Figure 3.


A relay designed for quadrature connection and having an RCA of 30" is
recommended when the relay is being used for the protection of plain feeders with
zero sequence source behind the relaying point.
In the case of transformer feeders or feeders which have a zero sequence source in
front of the relay, a quadrature connected relay is recommended but it is preferable
when protecting this type of feeder that the directional relay is designed to have an
RCA 45".
An RCA 45" is necessary in transformers and transformer feeders, to ensure correct
relay operation for faults beyond the starldelta transformer.
'Three fault conditions may theoretically cause mal-operation of the directional relay.
They are phase to phase to ground on a plain feeder; phase to ground fault on -a
transformer feTdmwlththezero sequence s o u r c e ~ l n f r o n t o t t h e l a ~ p h W t t o
phase fault on a transformer with the relay looking into the delta winding of the


These relays are similar in construction to the overcurrent relays but are polarised by
residual voltage or current. The polarising voltage is obtained from the secondary of
a three phase voltage transformer connected in broken delta. It is essential to
ensure that the correct voltage is fed to the relay that the voltage transformer primary
neutral is earthed and that it be a three phase, five limb type or consist of three single
phase units. Current polarisation is normally obtained by connecting a current
transformer in a local jransformer neutral. If voltage polarisation is used a 45" RCA
is normally used for solidly earthed systems and 0" for resistance earthed systems.

Page 3

Voltage Polarised Earth Fault Relays

Some care is necessary when using voltage polarised relays on solidly earthed
systems, as the residual voltage under single phase to earth fault conditions will be
equal to the phase to neutral voltage at the fault location or a sol~dearth fault only.
Any line impedance between the fault point and the relay, or resistance in the fault
itself will tend to reduce the value of the voltage and it can be very small if the line
impedance between the fault point and the relaying point is large compared with the
source impedance behind the relay. With modern directional relays however, which
will operate down to 1OO/ of normal voltage, no trouble should be experienced.

Current Polarised Earth Fault Relays

As already mentioned, current polarised relays may be polarised by a current
transformer connected in the power transformer neutral Only certa~ntypes of power
transformers however, are suitable as sources of polarising current, as in some the
direction of the current in the neutral can reverse*depending upon the fault position
and the ratio of system zero sequence impedances.
A staristar power transformer is not suitable for polarising relays even if both star
points are earthed. A current transformer in one neutral would not be suitable as the
current would reverse depending upon which side of the transformer the fault is on.
Paralleling two current transformers, one in each neutral connection, will not be
satisfactory as the resultant current would zero.
Three winding or two winding power transformers with one winding delta connected
are suitable for relay polarisation. Provided the star point is earthed, then a current
transformer in t h ~ sneutral can be used to supply the relay. In the case of three
winding transformers, if two star connected windings have the star point earthed,
then current transformers in each neutral connected in parallel must be used having
ratios inversely proport~onalto the power transformers voltage ratio. An alternative to
this is to use one current transfomer within the delta winding provided that no load is
taken from the delta. If load is taken from the delta winding it is necessary to use a
current transformer in each leg of the delta to prevent unbalanced load or fault
current producing incorrect polarising current.
Dual Polarised Earth Fault Relays
As the polarising current for current polarised earth fault relays is taken from a
current transformer in a local power transformer neutral, this may be lost if the
particular transformer is switched out of service and for this reason voltage
polarisation is in general more reliable. However, as pointed out, in solidly earthed
systems where the zero sequence source impedance is small the value of the
residual voltage can be very low and dual polarised relays, with both current and
voltage are used. It should be noted, however, that with modern relays the possibility
of voltage polarised relays failing to operate is very remote and that for all practical
conditions this possibility can in general be ignored.

Page 4

The operation of earth fault indication relays on systems earthed through a Petersen
Coil or totally insulated system is dependent on the capacitive current flowing in the
healthy feeders and when a Petersen Coil is used on the current due to the
suppression coil flowing in the faulty phase.
In the case of overhead lines the majority of earth faults are of a transient nature and
it is preferred that these faults shall not lead to automatic isolation of the faulty line. It
is desirable, however, that an indication should be given of sustained system faults in
order that the system may be supervised continuously and so that the faulty section
of the network is indicated.
For detection of a system earth fault, a sensitive directional relay or wattmetric relay
is used (Petersen Cod Systems)
Petersen Coil Earthed System
The diagram in Figure 4 shows asystem 0-f radial feeders, with a phase to ground
fault on the 'C' phase of one of the feeders. No current will flow in the 'C' phase of
the healthy feeders as they will be at earth potential. Capacative current will flow in
the healthy phases of all feeders to earth and back to the source via the fault. The
vector sum of the currents-in the current coil of the relay on the faulty feeder Is is


proportional to :

Where :

The vector diagram of the currents in the sound phases'shows that the total wattage
component of the currents in the restraining quadrant, hence the relays on the
healthy feeders will not operate. However, the current in the faulty feeder show that
the wattage component of the currents'is in the operating quadrant and hence, the
relay in the faulty feeder will operate.


The current transformers are of a special design, class 0.2, having an exceptionally
low phase angle error and because of this cannot be balanced accurately for currents
greatly in excess of rated current. The relay is provided with 0" MTA.
Insulated System
The diagram in Figure 5 shows a system of radial feeders, with a phase to ground
fault on the 'C' phase of one of the feeders. The residual current flowing in the
current coil of the relay on the faulty feeder, neglecting the effect of magnetising
current, is proportional to the 2 lc where lcis the vector sum of the currents in the
healthy phases Ica and Icb. Since the system is an insulated one, the fault has the
effect of raising the neutral point of the system by a voltage equivalent to the phase
voltage and the voltages'of the healthy feeders by A .
The relay is provided with a 90" leading MTA.

Page 5


Page 6

. :.




- -



. .

Page 8

a b c
















-\ - - ---+ - I- ,








- - 1-1c


Location of CT's



Healthy Feeders

Faulty Feeder



a b c

~ o c a t i i nCT'S
Faulty Feeder








- = -21,

Restrain 4




ction Notes

Power transformer is one of the most important links in a power system. Its development stems
from the early days of electromagnetic induction, when it was discovered that varying magnetic
flux in an iron core linking two coils produces an inducted voltage. From the basic discovery
has evolved the power transformer we know today using advanced insulation materials and
having complex windings on a laminated core using special magnetic steels cold rolled to
ensure grain orientation for low loss and high operating density.
With transformers of large capacity, a single transformer fault can cause large interruption to
power supplies. If faulted transformer is not isolated quickly, this can cause serious damage
and also power system stability problems. Protective systems applied to transformers thus play
a vital role in the economics and operation of a power system.
In common with other electrical plants, choice of suitable protection is governed by economic
considerations brought more into prominence by the range of size of transformers which is
wider than for most items of electrical plant. Transformers used in distribution and transmission
range from a few KVA to several hundred MVA.
Fo( transformers of the lower ;stings, only the simplest protection such as fuses can be justified
and for large rating transformers; comprehensive protection scheme should be applied.


Transformer faults are generally classified into four categories :


Winding and terminal faults

Core faults
Abnormal operating conditions such as overvoltage, overfluxing and overload
Sustained or uncleared external faults.

With tlie development of polyphase systems with more complex transformer winding
connections and also possible phase displacement between primary and secondary windings,
standardisation was necessary to ensure universal compatibility. (BS171 : 1970)
There are a number of possible transformer connections but the more common connections are
divided into four main groups :
Group 1


Phase displacement

e.g. Yyo
Phase displacement
e.g. Yd6
Group 2
Group 3
lag Phase displacement
e.g. Y d l
D Y ~
Group 4
lead Phase displacement
e.g. Y d l l
Yzl 1
High voltage windings are indicated by capital letters and low voitage windings by small letters
(reference to high and low is relative). The numbers refer to positions on a clock face and
Page 1

indicate the phase displacement of the low voltage phase to neutral vector with respect to the
high voltage phase to neutral vector, eg Y d l indicates that the low voltage phase vectors lag the
high voltage phase vectors by 30" (-30"phase shift).
Individual phases are indicated by the letters A, B and C, again capital letters for the high
voltage winding and small letters for the low voltage winding. All windings on the same limb of a
core are given the same letter. A further numerical subscript serves to differentiate between
each end of the winding.

Determination of Transformer Connections



This is best illustrated by considering a particular example. The following points should be

The line connections are normally made to the end of the winding which carries the
subscript 2, ie : A2, 62,C2 and a2, b2, c2.


The line terminal designation (both letter and subscript) are the same as those of the
phase windins to w h ~ c hthc line terminal is connected.

Consider the connection Yd1


Draw the primary and secondary phase to neutral vectors showing the required phase
displacemed :

Phase rotation

Phase rotation



ii) .

Complete the delta winding connection on the secondary side and indicate the respective
vector directions :



A 4

It is now possible to indicate the winding subscript numbers bearing in mind t-hh if
the direction of induced voltage in the high voltage winding at a given instant is Crom
A1 to A2 (or vice verse) then the direction of the induced voltage in the low voliage
winding at the same instant will also be from a1 to a2.
Page 2


It can now be seen'that the delta connection should be made by connxting a2

to c l , b2 to a1 and c2 to b l :


Small distribution transformers are commonly protected only by fuses. In many cases no circu~t
breaker is provided, making fuse protection the only available means of automatic isolation.
Fuses are overcurrent devices, and must have ratings well above the maximum transformer
load current in order to carry, without blowing, the short duration overloads that may occur
because of such as motor starting. Also the fuse must withstand the magnetising inrush of the
transformer. It follows that fuses will do little to protect the transformer, serving only to protect
the system by disconnecting a faulty transformer after the fault has reached an advanced stage
Overcurrent Relays
Overcurrent relays are often the only form of protection applied to small transformers. They are
used for backup protection for larger transformers and both instantaneous and time delayed
overcurrent can be applied.
Inverse time relays on the HV side of a transformer must grade with those on the LV
side which in turn must grade with the LV outgoing circuits. Due to this, the HV
overcurrent relays could have operating times which might cause operation of relays at
other substations. To overco-me this problem, high set instantaneous overcurrent relays
with low transient overreach are sometimes used. The settings of these relays should
be 120-13O0/0 of the through fault level of the transformer to ensure that the relays are
Page 3


stable for through faults. Care must also be taken to ensure that the relays d o not
operate under magnetising inrush conditions.

The function of differential protection is to provide faster and more discriminative phase
fault protection than that obtainable from simple overcurrent relays. Overall differential
protection may only be justified for larger transformers( Typically >5MVA).
CTs on the HV side are balanced against'CTs on the LV side. There are a number of
different connections but there are some important points that are applicable to all
Transformer Connection
Consider a deltalstar transformer. An external earth fault on the star side will result in
zero sequence current flowing in the line but due to the effect of the delta winding there
will be no zero sequence current in the line associated with the delta winding. In order to
ensure stability of the protection this zero sequence current must be eliminated from the
secondary connections on the star side of the transformer, ie the CTs on the star side of
the transformer should be connected in delta. With the CTs on the delta side of the
transformer connected in star, the 30" phase shift across the transformer is also catered
Since the majorlty of faults are caused by flashovers at the transformer bushings, it is
advantageous to locate the CTs in the adjacent switchgear.
Interposing CT (ICT)
Where it is not possible to correct for zero sequence current and the phase shift across
the transformer by using delta connected line CT's on the star side of the transformer, or
were CT ratio mismatch exists between primary and secondary CT's, then interposing
CT's are used. Tranditional ICT's were external devices, however modern numerical
relays are able to account for ratio error, phase shift and zero sequence current within
the relay. This eliminates the use of external ICT's and allows the protection to be set up
and installed more easily.
General Rules for CT Connections
CT connections opposite to main-transformer :

star CTs on delta side

delta CTs on star side

If similar primary terminals ie PI or P2 are towards the transformer, then delta and star
connection for the CTs should be the same as the transformer (or 180" opposite).
It is usual to assume that if current flows from P i
flow from S2 -+ SI.


P2 then the secondary current will

Note :If the transformer induced voltage is A1 --+ A2 then the secondary induced voltage
will be a1 -+ a2. Therefor?, current flo'w will be A1 --+ A2 and
a2 -+ a1

Page 4

Tap Changers
Any differential scheme can only be balanced at one point and it is usual to choose CT ratios
that match at the mid point of the tap range. Note that this might not necessarily be the normal
rated voltage. For example, if the tapping range is +1O0h, -20% then the CT ratio should be
based on a current corresponding to the -5% tap. The theoretical maximum out of balance in
the differential circuit is then +_ 15%.
Three Winding Transformers
Differential protection of three winding transformers is essentially similar to that of two winding
transformers. The same rules regarding CT connections still apply but the CT ratios used
shpuld be based on the MVA rating of one of the windings (usually the highest rated winding)
and not on the ratings of each individual winding.
For example, consider a 13213311'I kV transformer with windings rated for 100/60/40 MVA
respectively, then the current transformer ratios at all voltage levels should be based on 100
MVA, ie 44011. 176011 and 528011 respectively (these effective ratios are normally obtained by
the use of interposing CTs which means that, for example, all the main CTs associated with the
11 kV system can be made equal to 200011 - rated current).
If there is a source associated with only one of the transformer windings, then a relay with only
two bias coils can be used.- the CTs associated with the other two windings being connected in
parallel. If there is more than one source of supply then it is necessary to use a relay with three
bias windings in order to ensure that bias is available under all external fault conditions.

Page 5

Combined Differential and Restricted Earth Fault Protection

Although it is preferable to use separate CTs for restricted earth fault protection, it can be
combined with differential protection using the same current transformers, together with
interposing current transformers. A CT is required in the neutral connection and should be the
same ratio as the line current transformers.




Page 6

Magnetising Inrush

When a transformer is first energised, a transient magnetising currqnt flows, which may reach
instantaneous peaks of 8 to 30 times those of full load. The factors controlling the dirration and
magnitude of the magnetising inrush are :


Size of the transformer bank

Size of the power system
Resistance in the power system from the source to the transformer bank
Residual flux level
Type of iron used for the core and its saturation level.

There are three conditions which can produce a magnetising inrush effect :


First energisation


Voltage recovery following external fault clearance


Sympathetic inrush.due to a parallel transformer being energised.

Under normal steaay state cond'itions the flux in the core changes from maximum negative
value to maximum positive value duriqg one half of the voltage cycle, ie a change of 2 0
maximum. Since flux cannot instantly be created or destroyed this transformers are normally
designed and run at values of flux approaching the saturation value, an increase of flux to
double this value corresponds to relationship must always be true. Thus, if the transformer is
energised at a voltage zero when the flux would normally be at its maximum negative value, the
flux would rise to twice its normal value over the first half cycle of voltage. This initial rise could
be further increased if there was any residual flux in the core at the moment the transformer
was energised.
Since extreme saturation which requires an extremely high value of magnetising current.
As the flux enters the highly saturated portion of the magnetising characteristic, the inductance
falls and the current rises rapidly. Magnetising impedance is of the order of 2000% but under
heavily saturated conditions this can reduce to around 40% ie an increase in magnetising
current of 50 times normal, This figure can represent 5 or 6 times normal full load current.
Analysis of a typical magnitude inrush current wave shows (fundamental = 100%) :



2nd H

3rd H

4th H

5th H

6th H

7th H








The offset in the wave is only restored to normal by the circuit losses. The time constant of the
transient can be quite long, typ~cally0.1 second for a 100 KVA transformer and up to 1 second
for larger units. Initial rate of decay is high due to the low value of air core reactance. When
below saturation level rate of decay is much slower.
The magnitude of the inrush current is limited by the air core inductance of the windings under
extreme saturation conditions. A transformer with concentric windings will draw a higher
magnetising current when energised from the LV side, since this winding is usually on the inside
and has a lower air core inductance. Sandwich windings have approximately equal magnitude
currents for both LV and HV.
Resistance in the source will reduce the magnitude current and increase the'iate of decay.

Page 7

. ..

Effect on Differential Relays

Since magnetising inrush occurs on only one side of the transformer, the effect is similar to a
fault condition as far as differential protection is concerned. The following methods are used to
stabilise the relay during magnetising inrush period.


Time delayed - acceptable for small transformers or where high speed operation is not so
important. (Note : necessary time delay when associated with parallel transformers could be

Harmonic restraint - usual to use 2nd H restraint since magnitude inrush current contains
pronounced 2nd harmonics.
Note : 3rd H restraint should not be used for two reasons :

Due to-delta connections in the main transformer and in the CT circuits (which provide a
closed path for third harmonic currents), no third harmonic current would reach the relay.


CT saturation under internal fault conditions'also produces harmonics of which the 3rd is
the most predominant. Second harmonics are also produced under these conditions
(combination of dc offset and fundamental) so excessive saturation of CTs should be

The problem of any restraining tendency due to 2nd H currents produced by CTs saturating
under heavy internal fault conditions is usually overcome by using high set instantaneous un~ts
set at 8-10 x rated current.
While the second harmonic produces a useful restraint during external faults, it can produce
unwanted restraint for Internal faults, due to dc saturation of CTs. Extremely large CTs are
required such that they do not saturate and affect the operating times of the differential relay.
Gap Detection - If the various current waveforms that occur during magnetising inrush are
analysed, it can be found that the magnetising currents have a significant period in each cycle
where the current is substantially zero. Fault current, on the other hand, passes through zero
very quickly. Detection of this zero is considered a suitable criteria.

Thus, a transformer differential relay can be made to restrain if zero is detected In a cycle for
more than a certain period (typically 114f seconds). With the above principle of detection of
magnetising inrush, fast operation of the relay can be achieved for internal faults and
economically designed CTs can be used, without affecting the speed of operation.

Page 8



. .


An earthfault is the most common type of fault that occurs in a transformer.

For an earthfault current to flow, the following conditions must be satisfied :


a path exists for the current to flow into and out of the windings, ie a zero sequence path

the ampere turns balance is maintained between the windings.

The magnitude of earthfault current is dependent on the method of earthing solid or resistance
and the transformer connection.

Star Winding

- Resistance Earthed

An eadhfault on such a winding will give rise to a current which is dependent on the value of
earthing impedance and is also proportional to the distance of the fault from the neutral point,
since the fault voltage will be directly proportional to this distance.
The ratio of transformation between the primary winding and the short circuited turns also varies
with the position of the fault, so that the current which flows through the transformer terminals
will be proportional to square of the fraction of the winding which is short circuited.

If the earthing resistor is rated to pass full load current, then

Assuming V, = V,. then T, = 43 TI

Page 9

For a fault at x p.u. distance from the neutrgl,

Effective turns ratio = T2 I x TI

Primary C.T. ratio is based on lF.L. for differential protection.


C.T. secondary current (on prin~aryside of transformer)


= --F

I f differential setting = 20%

For relay operation


> 20/c

thus x > 59% i.e. 59% of winding is unprotected.

Differential relay setting


of winding protected

If as

of ~ F . L .

Pase 10

Star Winding

Solidly Earthed

In this case, the fault current is limited only by the leakage reactance of the winding, which
varies in a complex manner with the position of the fault. For the majority of the winding the fault
current is approximately 3 x Iflc, reaching a maximum of 5 x Iflc.

From a study of the various current distributions shown for earth faults, ~tis evident that
overcurrent relays do not provide aaequate earth fault protection. If the system is solidly
earthed, some differential relays adequately cover the majority of faults, but in general separate
earth fault protection is necessary.


Balanced earth fault for a delta (or unearthed star) winding can be provlded by connecting three
line CTs in parallel (residual connection). The relay will only operate for internal earth faults
since the transformer itself cannot supply zero sequence current to the system. The
transformer must obviously be connected to an earth source.

It is usual to provide instantaneous earth faultprotection to transformers since it is relatively

easy to restrict the operation of the protection to transformer faults only, ie the protection
remains-stable for external faults. This protection is called balanced Gr restricted earth fault and
the high impedance principle is utilised. However, modern numerical relays provide do provide
both high and low impedance restricted earthfault protection.


Earth Fault

For an earthed star winding, the residual connection of line CTs are further connected in parallel
with a CT located in the transformer neutral. Under external earth fault conditions the current in
the line CTs is balanced by the current in the neutral CT. Under internal fault conditions. current
only flows in the neutral CT and since there is no balancing current from the line CTs, the relay
will operate.
On four wire systems in order to negate the effect of the neutral return current a further CT
placed in the neutral and wired in parallel with the existing CT's. On a four wire system with the
transformer earthed at the neutral point 5 CT's are required. However, if the transformer is
earthed at the LV switch board only 4 CT's are required. If no neutral CT is used then therelay
will have to be set above the maximum expected unbalance current in the neutral return.

' I

Page 11

A relay, insensitive to the dc component of fault current is normally used for this type of
protection. If a "current operated" relay is used, an external stabilising resistor is placed in
series with the relay to ensure protection stability under through fault conditions. The protection
setting voltage is calculated by conventional methods. To reduce the setting voltage it is often
useful to run three cores from the neutral CT in order that the relay is connected across

.. .. .


Typical Settings for REF Protection (From ESI 48-3 1977)

10-60% of winding rated cur-rent

Solidly earthed
Resistance earthed

10-25% minimum earth fault current for fault at

transformer terminals

Unrestricted Earthfault Protection

Unrestricted earth fault or backup earth fault protection can be provided by utilising a single CT
mounted on an available earth connection eg transformer neutral, or (on an earthed systern) by
using a residual connection of three line CTs. In this case, the relay should be of the inverse or
definite time type in order to ensure correct discrimination.



On resistance earthed system, unrestricted earth fault protection is referred to as standby earth ,..;
fault protection. An inverse time relay is used which matches the thermal characteristic of the .::
earthing resistor. Earthing resistors normally have a 30 second rating and are designed to limit :
the earth fault current to transformer full load current.

: :,


Under fault conditions, currents are distributed in different ways according to winding
connections. Understanding of the various fault current distribution is essential for the design of
differential protection. performance of directional relays and settings of overcurrent relays.
Fault current distribution on a delta-star transformer, star-star transformer with unloaded tertiary
and star-delta transformer with earthing transformer for phase and earthfaults are shown in the
diagrams below :



PH-E Fault



PH-PH Fault

Fault Current Distribut~onon a Star - Star Transformer with Unloaded Tertiary

All types o i fault wlth~na transformer w~llproduce heat which will cause decomposition of the
transformer oil The resulting gases that are formed rise to the top of the tank and then to the
conservator. A buchholz relay connected between the tank and conservator collects the gas
and glves an alarm when a certain volume of gas has been collected. A severe fault causes so
much gas to be produced that pressure is built up in the tank and causes a surge of oil. The
buchholz relay will also detect these oil surges and under these conditions is arranged to trip the
transformer circu~tbreakers.
The maln advantage of the buchholz re4ay is that it will detect incipient faults which would not
oiherw~sebe detected by conventional protection arrangements. The relay is often the only way
of detect~nginterturn faults which cause a large current to flow in the shorted turns but due to
the large ratlo between the shorted turns and the rest of the winding, the change in terminal
currents IS very small

Parallel transformers are typically protected by directional overcurrent and earthfault protection
on the LV side set to look back into the transformers. Where an LV bussection exists the
directional relays can be replaced by non-directional relays, with the addition of a non-directional
overcurrent and earthfault relay at the bus-section.

If a transformer is connected in parallel with another transformer which is already energised,

magnetising inrush will occur in both transformers. The dc component of the inrush associated
with the switched transformer creates a voltage drop across the line resistance between the
source and the transformer. This voltage causes an inrush in the opposite direction in the
transformer that was already connected. After a time the two currents become substantially
equal and since they flow in opposite directions in the transmission line they cancel and produe
no more voltage drop in the line resistance. The two currents then become a single circulating
current flowing around the loop circuit made up of the two transformers in series -the rate of
decay being determined by the R/L ratio of the transformer.

Page 14


Overloads can be sustained for long periods with the limiting factor being the allowable
temperature rise in the windings and the cooling medium. Excessive overloading will result in
deterioration of insulation and subsequent failure.

p s far as protection IS concerned, non-harmonic restraint should not be used due to the long
time delay required. A harmonic restrained relay should be used for each transformer since if a
common relay were used the 2nd harmonic resGaint could be lost due to cancellation as
described above.

Overloads can be split into two categories :

Overloads which do not reduce the normal expectation of life of the transformers. Overloads in
this category are possible because the thermal time constant of the transformer means that
there is a con,siderable time lag before the maximum temperature correspond to a particular
load is reached. Quite high overloads can therefore be carried for short period.

Overloads in which an allowance is made for a rapid use of life than normal.
The length of life of insulation is not easily determined but it is generally agreed that the rate of
using life is doubled for every 6C temperature increase over the range 80-140C (below 80C
the use of life can be considered negltgible).

A hot spot temperature of 98C gives what may be considered the normal rate of using life, ie a
normal life of some tens of years. This temperature corresponds to a hot spot temperature.rise
of 78C above an ambient temperature of 20C. The graph below indicates the relative'rate of
using life against hot spot temperature.

rate of


80 90 100 110 120 130 140 "C

Hot Spot T e m p

Protection f o r Overloads
Since overloads cause heating of the transformer above the normal recommended
temperatures, protection against overloads is normally based on winding temperature

Page 15

Transformer Setting Tutorials

Advanced Industrial Power System


Transformer setting Criteria &

Page 1 of 33


Power Transformer- is one of the most important links in a power

system. W~thTransformers of larger capacity , a single transformer
fault can cause large interruption to power supplies. If faulted
transformer is not isolated quickly, this can cause serious damage
and also power system stability problems. Protective system applied
to transformers thus play a vital role in the economics and operation
of a power system.
In common with other electrical plants, choice of suitable protection
is governed by economic considerations brought more into
prominence Ly the range of size of transformers which is wider than
for most items of eiectrical plant.
For transformers of the lower ratings , only the simplest protection
such as fuses can be justified and for large rating transformers ,
comprehensive protection scheme should be applied.

Transformer faults are generally classified into four categories:

1 ) Windlng terminal faults'
2) Core faults
3 ) Abnormal operating conditions , such as overvoltage, Overfluxing
and overload
4 ) Sustained or uncleared external faults

With the development of poly phase systems with more complex
transformer connections and also poss~ble phase displacement
between primary and secondary windings, standardisation was
necessary to ensure universal compatability( BS 171: 1970)
There are a number of possible transformer connections but the more
common connections are divided into four groups.

Odegree phase displacement


E.g YyO

OALSTOM Limited, Energy Automation & Information


Advanced Industrial Power System


Transformer setting Criteria &

Page 2 of 33


180degree phase displacement


30degree Lagphase displacement

E.g Yy6
E.g Ydl


30degree Leadphase dispiacement E.g Yd 1 1

Dyl 1
Yzl 1

High voltage windings are indicated by capital letters and low

voltage windings by snmll letters (reference to high and low is
relative). The numbers refers to positions on a clock face and
indicate the phase displacement of the low voltage phase to neutral
vecior , e.9,Yd 1 indicates that the low voltage phase vectors lag the
high voltags phase vectors by 30 degree (-30 degree phase shift]
Individual phases are ind~catedby the letters A,B &C , again capital
letters for the low voltage winding. All windings on the same limb of a
core are given the same letter. A further numerical subscript serves to
differentiate between each end of the winding.


Over current and earth fault protection(Unrestricted)

Plain overcurrent and earth fault protection utilising IDMTL relays are
used primarily to protect the transformer against the effects of
exiernal short circuits and excess overloads. The current settings of
the protection must be above the permitted sustained over load
allowance and below the minimum short circuit current. The ideal
characteristic i s the extremely inverse (CDG14)as it is closely
approximates to the thermal curve of the transformer.
The protection is located on the supply side of the transformer and is
arranged to trip both the H V and LV circuit breakers. In many cases
the requirements for protecting the transformer and maintaining
discrimination with similar relays in the remainder of the power syslern
are not corilpatibile. In these circumstances , negative sequence
filter protecrion or under voltage blocking may be used to obtain the
desired ser?sii~vity..

:3ALSTOM Limited, Energy Automation & Information



Advanced Industrial Power System


Transformer setting Criteria &

Page 3 of 33

1 . High set overcurrent cut-Off:

On small transformers where the main protection is provided with
overcurrent devices and where the transformer i s fed from one end
only, a high set instantaneous relay i s utilised to provide protection
against terminal and internal winding faults.
The ,relay is set to be above the short circuit level on the
secondary(load ) -side of the transformer and below: that for a
terminal fault on the primary (supply)sideof the transformer.
On choosing the type and setting of the high set relay, it i s important
to consider the magnetising inrush currents under normal switching ,
offset fault currents and starting currents of motors.The first two
problems can be overcome by using a relay sensitive only to
fundamental frequency currents, while the third is overcome by
setting the relay above the max. starting current level.

2. Stand-by earth fault protection

Where transformers are earthed via an earthing resistance which is
short time rated , stundby earth fault protection is applied to protect
the resistor from damage when an earth fault persists for a time
longer than the rating of the resistors. The relay is energised from a CT
in the neutral connection and its time of operation is made to match
the thermal rating of the resistor. It is arranged to completely isolate
the transformer.
Some times a two stage relay is employed, each stage set to operate
at a different time. The first staqe arranged to trip the LV breaker and
if still the fault is persisting, Ihe second stage relay trips the HV side
breaker thus isolating the transiormer completely.


The funciion of differential protection is to provide faster and more

discriminative phase fault protection than that obtainable from
simple over current relays.CTs on the primary and the secondpry sides

OALSTOM Limited, Energy Automation 8 Information

Advanced Industrial Power System


Transformer setting Criteria &

Page 4 of 33

are connected to form a circulating current system. The following

figure illustrates the principle.





Basic Considerations for transformer differential protection

. A;.




1. Line current transformer primary ratings

. ..
.$ .

The rated currents of the primary and the secondary sides of a two
winding transformer will depend on the MVA rating of thetransformer
and will be in inverse ratio to the corresponding voltages. For three
winding transformers the rated current will depend on the MVA rating
of the relevar-rt winding. Line current transformers should therefore
have primary ratings equal to or above the rated currents of the
transiormer windings to which they are applied.

2. Current transformer connections

The CT connections should be arranged , where necessary to

compensate for phase difference between line currents on each
side of the power transformer. If the transformer is connected in
delta/star as shown in figure, balanced three phase through current
suffers a phase angle of 30 degree which must be corrected in the
CT secondary leads by appropriate connection of the CT secondary
Further more , zero sequence current flowing on the star side of the
power transformer .will not produce current outside the delta on the
other side. The zerosequence therefore be eliminated from the star
side by connecting the CTs in delta, from which i t follows that the CTs
on the delta side of the transformers must be connected in star, in
order to give 30 degree phase shift. This is a general rule ; if the

OALSTOM limited, Energy Automation & Information


. . ,....
.. .






. r....,.


. ,h..

Transformer setting criteria &

Pcge 5 of 33

Advanced Industrial Power System

transformer were connected starlstar
need to be connected in delta.

the CTs on both sides would

- .


< .

- ---

- . .-



. A , . -

. . . . .--.







, .



. . . . . .-



. . . .

.. . ._ ._

. ,

... -.


. ' .


.--,.-. :'.., !-.: ,

. .-. .- . . . . -. .
.. . . . .


. . .. . .. . .: .




> 3 . -


: ,





. .---,i
. . . .:. - I,
.:.-?: , .,, .. . . . . ... .,.







--- ---..

-. ..:

V ' ,


: . .;

.. ,


<>-. ;

. -.
.: .

--.:., *. .:,. -,..!



.... .-,.


. -.


0 ;


c ;

.. ,







.. -


,. . :
, i


_ I




. .. . ... . .
. ,, ,: .->-:
., .. :. a : .






, , 8 .

- -,


.. . .



. -

. . . . . .,. . .,i . . .




-. .
- . ..
, . .






- ...




. ._

< .





.: .




T a 0.4d


WIWQ wu ~ o r urv r to b C-OUCW

S l udZ.mdt-lharwdhrnbp.1

T h

em.rpo..64 ff. r e W-QDmCYd.


m -1

m h ull



- ........




d CT.

W ~ C O a t l * * S

r. MU-d

v. . I

IN n(n- a
1d 3 k*WchCLbrdlon-rtuCI'srn-ang.2.~3-vb.conaa.d

~ . r * c ar o c . ~ e a r o..(rrrr cap?rrd w lhor .prl(..d

I<C 76: 1967 rrd




.- .--

I* a u r u r l IN, C? c-m.

. -.. ;

. . . . . .


. .
. ,



.. , .. ..
. .


- -,


. . -:! !



m n d .rn' u
. ormlod




as 171-relo

OALSTOM Limited, Energy Automation & Information

Advanced Industrial.Power System


Transformer setting Criteria &

Page 6 of 33

Figure 1

When Cis are connected I n delta, their secondary rating must be

1 / 43 times the secondary rating of star connected
reduced to
Cis , inorder that the currents outside the-delta may balance with the
secondary currents of the star connected CTs.
When line CT ratios proiide adequate matching between currents
supplied to the differential relay under through load and through
fault conditions , the necessry phase shift can be obtained by
suitable connection of the' line CTs . Figure 1 above shows the
required connections for various power transformer winding
When delta connected CTs are required it is a common practice to
use star conr;ecled line CTs and to obtain fhe delta connection by
means of stal-/delta interposing CTs.

3. Bias to cover Tap-Changing facility and CT mismatch

If the transfor-n~er
hcs a tapping range enabling its ratio to be varied ,
this must be allowed for in the differential system. This is because the
CTs selecled to balance, for the mean ratio of the power transformer, ,
a variation in ratio from the mean will create an unbalance
proportional to the ratio change. At maximum through fault current ,
the spill oputput produced by the small percentage unbalance may
be substantial.
Differential protection should be provided with a proportional bias of
at-\ amount which exceeds in effect the maximum ratio deviation. This
stabilises ihe protection under through fault conditions while still
permitting the system to have good system sensitivity.
The bias characteristic for a typical differential protection is shown in
figure2, iron-\ which it can be seen that the cursent required to
operate the relay increases as the through fault increases.




OALSTOM Limited, Energy Automation & Information

Advanced Industrial Power System


Transformer setting Criteria &

Page 7 of 33

Figure 2



.. ..

- -



- - . - . --



- . --.






When applying a differential relay, care should be taken that its

characteristic will prevzr-it operation due to the combination of tap
change variation and CT mismatch . To mininiise the effect of tap
change variations , current inpuis lo the differeniial relay are usually
matched at the mid poini of the tap range..

Figure 3 below shows percentage biased differential protection for a

two winding transformer. The two bias windings per phase are
conimonly provided on the same electromagnet or auxiliary
core. .

OALSTOM Limited, Energy Automation & Information


Transformer setting Criteria &

Page 8 of 33

Advanced Industrial Power System


The Merz-price principle remains valid for a system having more than
two connections, so a transformer with three or more windings can
still be protected by the application of above principles.
When the power transforn~erhas only one of i t s three windings
connected to a source of supply, with the other two windings
feeding loads at differ-er~tvoltages, a relay of the same design. as
that used for two winding transformer can be employed.



. . .8 . .



. .
. .


. .

W j ,- >g - j

.... -, .

i . .


- - J > ' :

, ... ..i




.. ..

- .



.; .







1 I

d l . . ti::



,-. ,




. ,. .





; ! i


... .-



Biased differeniiol pro:ec;ion for two winding DeJtalStar transformer

Figure 3

The separate load cv!rents are summated in the CT secondar;/

circuits, and will balar~ce\with the infeed current on the supply side.

When more tho!] one winding is connected to a source , the

disiributiol~of CLII rer~tcz11,1ot readily be predicted and there is u
danger in the s c h e r ~ ~sflown
(a) in Ihe figure 4 of current circulaling
between the two paralleled sets of CTs with out producing any bias. It
i s therefore impor-tani i:lat bias be obtained seperately from the
current flowing in each set of line connections. In this case a ~eloy
used with separate bios \vindings , arranged so that their mechanical
or' electrical effects a l ~ ~ c add
y s numerically , that is not vectoria'ily . lo
give the total bias eiiezr. This is shown as (b)in the figure 4.

OALSTOM Limited, Enersy Automation & Information

Advanced Industrial Power System


Transformer setting Criteria &

Page 9 of 33

These considerations do not apply when the third winding consists of

a delta connected tertiary with no connections brought out. Such an
arrangement may be regarded GS t\.w \vvir,dlr,gtrafisformer --for
protection purposes and may
be protected as (c)in the figure 4.

3. Inter posing CTs to compensate for mismatch of Line CTs

Besides their use for phase ccjmpensation, interposing CTs may be

used to match up currents supplied to the differential protection from
the line CTs for each winding. The amount of CT mism.atch which a
relay can tolerate with out maloperation under through fault
conditions will depend on i t s bias characteristic and the range over
which the tap changer can operate. If the combined mismatch due
to CTs and tap changer is above the accepted level , then
interposing CTs may be used to achieve current matching at the mid
point of the tap changer range.

OALSTOM Limited, Energy Automation & Information





2 :

. ..


Transformer setting Criteria &

Page 10 of 33

Advanced Industrial Power System

Figure 4

. .



. .. - . , .

. - ..

(~1.: THjlEE WlNDitllG W S m R M E R Icne poviar rource)






!. -.-

@ALS:OivZ ?irnit+?d,Energ\/ Automation 8 Inforn;aiion


Transformer setting Criteria &

Page 1 1 of 33

Advanced Industrial Power System


For the protection of two winding transformers interposing CTs should

ideaiiy t-1-~aichihe---relaycurrents under through load conditions
corresponding to the maximum MVA rating of the tran~fo~mer.
example of this for an 1 1 KV/132KV 30MVA, DELTA /STAR transformer is
shown in the figure 5.

First the primary ratings of 1600A and200A chosen for the main CTs
should not be less than the max. full load currents in each winding ,
which are ,

30 x 100
= 1575 A
For 1 1 KV winding
3 3 i~i x 103
30 x 10.
= 164 ,A
For 132KV winding
\/3x 132 x 103
r the 1 1 KV winding this is also the nominal full load current , but for 1172
132KV winding , with -5% tap , the latter is:
30 x 10"
43 x 0.95 x 132 x 103

= 138A

For 1 1 KV winding

Equivaienf secondary currents in the line CTs are 0.984 A and 0.69 A.
Thus the ratio of the star /delta interposing CTs to achieve ideal
n7atching is given by:
0.69 /


= 0.70 / L

A OR 0.70/0.577 A

.' >

OALSTOM Limited, Energy Automation & Information


Transformer setting Criteria

Tutorials ..g
Page 12cf 33- $

Advanced Industrial Power System









. _


. . .
. A


_- .














-- - -- - i







.. *-.,





,. ,,

i L:




.; .. .: ,' .
, .








- 2


Figure 5
The pro!eciiori of three winding transfor-nier-sis complicated by the
fact that line
CTs for each winding ar-e riormally based on different MVA levels and
will 1701 ttienlselves achieve balance under ttirough current
ronditior~s.To achieve correct balarice , it is necessary to use
inlerposir~gCTs wllicti hlill provide the relay with raied currerli when
the rating of the highest rated winding is applied to all windings.

An exaniple for a 500KV/138KV/13.45KV, 120MVA/90MVA/30MVA,

star/slar/delta transfornler is shown in the figure 6.
Load cur rent a i 599

KV =

120 x 10"
- ~ x500x

= 138.6A

OALSTOM Limiled, Energy Automaiion & Iriforniation



,- ,-4
.? ~p-.-_ll
.- --*


Advanced Industrial Power System


Transformer setling Criteria &

Page 13 of 33


Load current at 138 KV = 90 x 106

~ 3 138x
Load current at 13.45 KV = 30 x 106
i 3 x 13.45 x 103

= 376.5 A

= 1288 A

Line CT ralio at 500KV = 20015 A

I-ine CT ratio at 138KV = 40015 A .
Line CT ratio at 13.45KV = 150015 A
Current at 138 KV corresponding to 120 MVA = 120 x 1 O6
3 x 138x lo3

Current at 13.45 KV corresponding to 120 MVA = 120 x 1 O6

\I3 x 13.45 x lo"

= 502 A

~ 5 1 5 1A

Secondary current from 500KV line CTs corresponding to 120 MVA

=138.6 x 5
=3.46 A

OALSTOM Limited, Energy Automation & Information

Transformer setting Criteria &

Advanced Industrial Power System



W k V

Page 14 of 33

1 W V


Figure 6


lnele iore ratio of required starldelta interposing CTs
= 3.461 5 A OR 3.461 2.89 A

Secondary current from 138 KV line CTs corresponding io 120 M V A

OALSTOM Limited, Enerhy Automation t Information


Advanced Industrial Power System


Transformer setting Criteria &

Page 15 of 33

There fore ratio of required starldelta interposing CTs

Secondary current from 13.45 KV line CTs corresponding to 120 MVA

= 5151 x 5
= 17.17 A
1 500
Therefore ratio of required star/star interposing CTs
= 17.1715 A




Under full load conditions of 30MVA, for the 13.45 KV delta winding ,
the current appearing in the primary of the 17.17/5 A inter posing CT
will only be 4.29 A , the corresponding secondary current being 1.25
A . However the ratings of the primary and the secondary windings
should ideally be 17.1 7 A and 5 A respectively to minimise winding


Tl~en~agnetisinginrush phenomenon produces current input to the

energised winding which has no equivalent on the other sides of the

transformer. The whole of the inrush current appears therefore as
unbalance and is no1 distingushable from internal fault current. The
normal bias is not , iherefore effective and increase of protection
setting to a value which would avoid the operation would make the
protection of little value.
Harmonic Restraint.

The inrush current, clthough y!~.nerallyresembling an inzone fault

current, differs greaily when the waveforms are compared. The
distinctive difference in the4 woveforms can be used to distinguish

OALSTOM Limited, Energy Automation & Information


between the condition;. The inrush contains all orders of harmonics,

but these are not all equally suitable for providing bias. The study of
this svbject is complex, as the wave form depends on the degree of
saturation and on the grade of iron in the core.

a) D.C Offset component (Zero harmonic)

A uni-directional component will usually be present in the inrush
current of the single phase transformer and in the principal inrush
currents of a three phase unit. However if at the instant of switching
the residual flux for any phase is equal to the flux which would exist in
the steady staie at that point on the voltage wave , then no transient
disturbance should take place on that phase.


- :


3.; !


'! :

Large inrush c ~ r ~ - e nwill

i s flow in the other two phases , corresponding
to high peak lux values established in these phase cores. The high
flux circulates through the yokes , the saturation of which affects the
iir-s: phase , L-:iiich. would have had no inrush effect, causing c
substantial transient current to flow in this phase as well. This latter
current , however will not be off set from the zero axis , althougt-1the
current waveform will be distorted..
If the uni-directional current component were used to stabilise a

differential sysiem, some sort of cross phase biasing would be

of this effect.
required becc~lse

b) Second Harmonic component

This connpone!3t is present in all inrush wave forms . It is typical of
wave forms in which successive half period portions do not repeat
with reversal oi polarity but-in which the mirror image symmetry can
be found abo:lt certain ordinates.
The portion of second harmonics varies some what with the degrce
of saturation of core , but is always present as long as the
unidireciior~ai~ormponentof flux exists. It has been shown to have a
minimum value of about 20% of the amount by which the inrush
current exceecjs the steady state magnetising current.




Transformer setting Criteria & .$

Tutorials :,
Page 16 of 33 !(

Advanced Industrial Power System




OALSTOM Lirni:-?d, Energy Automation & Information

Advanced Industrial. Power System



Transformer setting Criteria &

Page 17 of 33

Normal fault current do not contain second or other even harmonics,

nor do distorted currents flowing in saturated iron cored coils under
steady state conditions.

The output current of a current transformer which is energized into

steady state saturation will also contain odd harmonics but i ~ oeven
harmonics. However, should the current transformer be saturated by
the transient component of the fault current, the resulting saturation i s
not symmetrical and even harmonics are introduced into the output
current. This can have the advantage of improving the through fault
stability performance of a differential relay, but it also has the
adverse effect of increasing the operatioh time for internal faults.
The second harmonic is therefore an attractive basis for a stabilizing
bias against inrush effects, but care must be taken to ensure that the
current transformers are sufficiently large so that the harmonics
produced by transient saturation do not delay normal operation of
the relay.

The differential current is passed through a filter which extracts the

second harmonic; this component is then applied to produce a
restraining quantity sufficient to overcome theoperating tendency
due to the whole of the inrush current which flows in the operating
By this means a sensitive and high speed system can be obtained.
With the type DTH relay, a static design, a setting of 15% is obtained
with an operating time of 45 milliseconds for all fault currents of twice
or more times the current rating.
The relay will restrain when the second harmonic component
exceeds 20% of the current.
Third harmonic
The third harmonic is also present in the inrush current in roughly
comparable proportion to the second harmonic. The separate
phase inrush currents are still related in phase to the primary applied
electromotive forces and the harmonics have a similar time spacing,
which brings the third harmonic waves in the three windings into
phase. If the windings are connected in delta, the line currents are
each the difference of two phase currents.
As the inrush
components vary during the progress of the transient condition it is
possible for this qifference to pass,through zero, so that the third





f; .,:.



OALSTOM Limited, Energy Automation & Information


Advanced Industrial Power System


Transformer setting Criteria &

. Tutorials
Page 18 of 33

harmonic component in the line current vanishes; this component,

therefore, be regarded as a reliable source of bias.
To this must be added the further consideration that a sustained third
harmonic component is quite likely to be produced by CT saturation
under heavy in-zone fault .conditions.
All this means that the third harmonic is not a desirable means of
stabilizing a protective system against inrush effects.

Higher harmonics ,
All other harmonics are theoretically present in inrush current but the
relative magnitude diminishes rapidly as the order of harmonic
increases; there may be 5% of fourth harmonic in a given inrush
'current. This component would be similar in response to the second
harmonic but the small magnitude hardly justifies the provision of an
extra filter circuit.

A still smaller proportion of fifth harmonic will be present. This

component is not subject to cancellation as is the third harmonic,
and can be present in the output of a CT in an advanced state of
saturation, therefore offering no benefit. Still hlgher harmonics are of
magnitude too small to be worth consideration.
The percentage of fifth harmonic in the transformer magnetizing
current increases significantly when the transformer is subjected to a
temporary overvoltage condition. Some manufacturers apply a
measure of fifth harmonic bias to the relay to restrain operation for
this condition. Typically such relays are restrained if the magnetizing
current contains 30% fifth harmonic.


A simple overcurrent and earth fault system will not give good

protection cover for a star-connected primary winding, part~cularlyil

the neutral is eaithed through an impedance. The degree of
protection is very much improved by the application of a ur-lil
differential earth fault system or restricted earth fault protection, as
shown in Figure 7.
The residual current of three line current

OALSTOM Limited, Energy Automation & Information

Advanced Industrial Power System


Transformer setting Criteria &

Page 19 cf 33

transformers is balanced against the output of a current transformer

in the neutral conductor. The relay is of the high impedance type.

The system is operalive for faults within the region between current
transformers, t h a t Is, fs: fai;lts on the star winding in question. The
system will remain stable for all faults outside this zone


Restric ted earth fault protection for a star winding. Figure 7

The gain in protection performance comes not only from using an
instantaneous relay with a low setting, but also because the whole
fault current i s measured, not merely the transformed component in
the HV primary winding. Hence, although the prospective current
level decreases as fault positions progressively nearer the neutral end
of the winding are considered, the square law which controls the
primary line current is notapplicable, and with a low effective setting,
a good percentage of the winding can be covered.
Restricted earth fault protection is often applied even when the
neutral is solidly earthed. Since fault current then remains at a high
value even to the last turn of the winding , virtually complete cover

OALSTOM Limited, Energy Automation & Information

Advanced Industrial Power System


Transformer setting Criteria &

for earth faults if obtained, which is a gain compared with the
performance of systems which do not measure the neutral conductor


Earth fault protection applied to a delta-connected or unearthed

star winding in inherently restricted, since no zero sequence
component can be transmitted through the transformer to the
secondary system. A high impedance relay can therefore be used,
giving fast operation and phase fault stability.

. ;+?$






Both windings of a transformer can be protected separately with

restricted earth fault protection, thereby p~oviding high speed
protection against earth faults for the whole transformer with
relatively simple equipment.








This protection is based on high impedance differential principle,

offering stability for any type of fault occuring outside the protected
zone and satisfactdry operation for faults with in the zone.
A high impedance relay is defined as a relay or a relay circuit whose
voltage setting is not less tnan the calculated rnax. voltage which.
can appear across its terminal under the assigned-max.through fault
current condition.






. .,f

,... ,



. '.





- -2


It can be seen frorn the figure that during an external fault current
should circulate between the current transformer secondaries. The
only current that can flow through the relay circuit i s due to any
difference in CT output for the same primary current. Magnetic
saturation will reduce the output of a CT and the most extreme case
of stability will be if one CT is completely saturated and the other
unaffected. At one end of the CT can be considered fully saturated
with i t s magnetising impedance , while the CT at the other end being
unaffected , delivers its full current output which which will then
divide between the relay and the saturated CT. This division will be in
the inverse ratio of R relay circuit and Rct +2RL and obviously if R relay
circuit is high compared with Rct +2RL, the relay will be prevented
from undesirable operation.


To achieve the stability for external faults, the.stability voltage for the
protection Vs must be determined by the formula,
Vs = If (Rct +2RL )
Where Rct = CT secondary winding resistance
RL= max. lead resistance from the CT to the common point

OALSTOM Limited, Energy Automation & Information

. .._.
. .
. ..-..


Advanced Industrial Power System


Transformer setting criteria.&

of 33

To ensure satisfactory operation of the relay under internal fault

conditions the CT Knee point voltage should not be less than twice
the relay setting voltage. i.e

The setting of the -stabilising resistor must be calculated in the

following manner, where the setting is a function of the relay ohmic
impedance at setting Rr , the required stability voltage setting Vs and
the relay current setting Ir.

Rst =



The ohrriic impedance can be calculated using the relay VA burden

at current setting and the current setting Ir,



When the max. through fault current is limited by the protected
circuit impedance: such as in the case of power transformer REF
protection , it i s generally found unnecessary to use non -linear
voltage limiting resistors or Metrosils. How ever when the max. through
fault current is high , it is always advisable to use a non linear resistor
across the relay circuit. Metrsils are used to limit the peak voltage
developed by the CTs under internal fault conditions, to a value
below the insulation level of the CT s, relay and the connecting leads,
wh~chare normally withstand 3000V peak.

The following formulae should be used to estimate the peak transient

voltage that could be produced for an internal fault. The peak
voltage produced during an internal fault will be a function of the CT
Knee point voltage and the prospective voltage that would be
produced for an internal fault if CT saturation did not occur. This
prospective voltage will be a function of max. internal fault
secondary current , the CT ratio , the CT secondary winding
resistance, the CT lead resistance to the common point , the relay
lead resistance, the stabilising resistor value and the relay burden at
relay operating current.


OALSTOM Limited, Energy Automation 8, Information


Advanced Industrial Power System


Transformer setting Criteria &

. .

Page 22 of 33

Vp= 2 d 2 Vk (Vf - Vk)

Vf = If (Rct +2rl+Rst + Rr )




. .: >

Where Vp= Peak voltage developed by the CT under internal fault

Vk = CT knee point voltage
Vf = max. voltage that would be produced if CT saturation did
not occur
Max. internal fault sec. Current
Rct = CT secondary winding resistance
RL = max. lead burden from CT to relay
Rst = stabilising resistor
Rr = relay ohmic impedance .at setting

When the value given by t h e formula is greater than 3000 V peak,

non-linear resistors (metrosils) should be applied. These non-linear
resistor s
(metrosils)are effectively connected across the relay circuit, or phase
to neutral of the ac bus wires, and serve the purpose of shunting the
secondary current output of the current transformer from the relay
~norderto prevent very high secondary voltages.
These non-hear resistors (metrosils) are externally mounted and take
of 152 mm diameter and approximately 10
tne form of a n n ~ ~ l discs,
mm thickness. The operating characteristics follow the expression:


l-10 2 5

V= Instantaneous voltage applied to the non-linear
C= Constant of the non-linear resistor (metrosil)
I = lnstantaneous current through the non-linear resistoi
For satisfactory application of a metrosil, its characteristic should be
sucli that i t requires the following requirement.
At the relay voltage setting , the non linear resistor, current should be

as low as possible but no greater than approximately 30mA r.m.s for

1 A CT and approx. 100mA for 5A CTs.

OALSTOM Limited, Energy Automation & Information









. .


Advanced Industrial Power System


Transformer setting Criteria &

Total impedance = 14 p.u


inere foi-e ! I = 1/:4 = Z.C714 p.u

Base current = 80 x 10 6

= 1 1 1 296 Amps
There fore fault current = If = 3x 0.071 4 ~ 1 1 1 2 9 6
= 23840 dmps ( Primary)
= 14.9 Amps (Secondary)
Setting voltage Vs = If ( R c t + 2 R L )
Assuming Earth CT saturates,
Rct = 4.8 ohms
2RL = 2x 100 x 7 . 4 1 ~
= 1.482 ohms

. ..


.. ..


. .. . . ,.


. .

Therefore setting voltage = Vs = 14.9 x ( 4.4+ 1.482)

= 93.6 V

Stabilising Resistor

Rst = Vs /IS - VA/ls

Where VA is the burden of the relay
Is = relay setting current

Adopt the relay setting as lo%,

Rst = 93.610.1 - 1 / (0.112

= 836 ohms

Effective setting = Ip = CT ratio x (Is + nle)

Where n= the no. of CT in parellel
le = magnefisingcurrent of each transformer

OALSTOM Limited, Energy Automation & Information

Transformer setting Criteria &

Advanced Industrial Power System



Page 25 of 33

From the CT characteristics -& the Table

i i n e side CTs:

Flux density at 93.6V, = 93.6'1158 = 0.592Tesla

Magnetising force at 0.592T = 0.015 AT/mm
Therefore magnetising current = 0.015 x 0.341 = 0.0051Amps

Line side CT
Earth CT

1 58

1 0.341

1 0.273

GALSTOM Limited, Energy Automation & Information

OALSTOM Limited, Energy Automalion & Information

Advanced Industrial Power System


Transformer selting Criteria &

Page 27 of 33

Flux density at 93.6 V = 93.6 / 236 = 0.396 Tesla

Therefore mag current = 0.012 x 0.275 = 0.0033 Amps

Thus effective setting = 1600 x [0.1 + (3x 0.0051 + 0.0033) ]

= 190 Amps

Transformer full load current = 1391 Amps

Peak Voltage developed across the relay circuit = Vp= 2

4 2 Vk


Vf= 14.9xVs/\s= 14.9x936= 13946volts

For earth CT froni the graph,
Vk = 1.4~236= 330 V

Therefore Vp = 2


2 x330 ( 13946 - 330)

Since this value is more than 3 KV

Metrosil voltage limiter will be


Power frequency overvoltage causes both an increase in stress on

the insulation and a proportionate increase in the working flux. The




increase ip



loss . and

Advanced industriai Power System


Transformer setting Criteria &

Page 28 of 33

disproportionately great increase in magnetizing current. In addition,

flux is directed from-the laminated core
structure into steel structural parts. In particular, under conditions of
over-excitation of the core, the core bolts, which normally carry little
flux, may be subjected to a large component of flux diverted from
the highly saturated and constricted region of core alongside.
Ui~der such conditions, the bolts may be rapidly heated to a
temperature which destroys their own insulation and will damage the
coil insulation if the condition cantinues.
Reduction of frequency has an effect with regard to flux density,
similar to that of overvoltage.
It follows that a rransformer can operate with some degree of
overvottage with a corresponding increase In frequency, but
with a high voltage input at a low
operation must not be continued

Operation cannot be sustained when the ratio of voltage to

frequency, with these quantities given values per unit of their rated
values exceeds un~tyby more than a small amount, for instance if V/f
> 1 .l. The base of 'unit voltage' should be taken as the highest
voltage for which the transformer is designed if a substantial rtse in
system voltage has been catered for in the design.
The condition does no1 call for high speed tripping; instantaneous
operation is undesirable as this would cause tripping on momentary
system disturbar~ces which can be borne safely, but normal
conditions must b e restored or the transformer must be isolated within
one or iwo minutes at most. The fundamental equation for the
generat~onof e.m.f. in a transformer
can be arranged to give:

OALSTOM Limited, Energy Automation & Information


Advanced Industrial Power System


Transformer setiing Criteria &

Page 29 of 33

ii is necessary to detect a ratio of E l f exceeding unity, E and f being

expressed in per unit values of rated quantities.The system voltage, as
measured by a voltage transformer, is applied to a resistance to
produce a proportionate current; this current, on being passed
through a capacitor, produced a voltage drop which is proportional
to the function in question, Elf, and hence to the flux in the power
Feedback techniques are used in the type GTT relay to make the
measured ratio accurate over a wide range of frequency and
voltage.Two time delay outputs are given by auxiliary elements, each
with multiple contacts. One element, the contact of which are used
to effect a control operation to reclify the abnormal condition,
operates after a pre-selected fixed,time delay between 0.5s and 1.0s
or between 2s and 5 s.
The second element is arranged to trip the supplies to the transformer
after a pre-set time delay of 5s to 30s or 12s to 120s if the abncrmal
condition persists.
Overfluxing protection is mostly confined to generator-transformers
for which the risks appear to be greatest, although overfluxing trouble
has been known to occur for other transformers as

& Information
OALSTOM Limited, Energy Automation

Advanced Industrial Power System


Page 30 of 33







. ., .<is

...?$ 5


The use of instantaneous relays for the primary side of the transformer
is recommended inorder to improve fault clearance time and enable
a lower time multiplier setting on relays elsewhere on the system. The
relay should have low transient overreach and be set to
approximately 125% of the maximum through fault level of the
transformer, in order to prevent operation for faults on the secondary

:* .
..:. -.




. ..








,' '$5

...?, :


OALSTOM Limited, Energy Automation & Information



Generator and Generator

Transf - Protection

Generator and



'The core of an electric power system is the generation. $.': - ~..'&.'.:<.'.: ~

With the exception of emerging fuel cell and solar-cell $ , : ~ ; ~ ~ y : ~ ~ ~ - ~ ~ technology for power systems, the conversion o f the
fundamental energy into its electrical equivalent
normally reqdires a 'prime mover' to develop mechafiical
power as an intermediate stage.


The nature of this machine depends upon the source of

energy and i n turn this has some bearing on the design
o f the generator. Generators based on steam, gas, water
or wind turbines, and reciprocating combustion engines
are all i n use. Electrical ratings extend from a f e w
hundred kVA (or even less] for reciprocating engine and
renewable energy sets, up t o steam turbine sets
exceeding 1200MVA.
Small and medium sized sets may be directly connected
to a power distribution system. A larger set may be
associated with an individual transformer, through
which it is coupled to the EHV primary transmission
Switchgear may or may not be provided between t h e
generator and transformer. In some cases, operational
and economic advantages can be attained by providing
a generator circuit breaker i n addition to a high voltage
circuit breaker, but special demands will be placed on
the generator circuit breaker for interruption o f
generator fault current waveforms that do not have an
early zero crossing.
A unit transformer may be tapped o f f t h e
interconnection between generator and transformer for
the supply of power to auxiliary plant, as shown i n
Figure 17.1. The unit transformer could be of the order
of 10% of the unit rating for a large fossil-fuelled steam
set with additions( flue-gas desulphurisation plant. b u t
it may only be of the order o f 1% o f unit rating for a
hydro set.










., ... .
.;:, ;:..:
,. :y.;:..:,
. . . . .:::,..:

required. The amount of protection applied w i l l be

governed b y economic considerations, t a k i n g i n t o
account the value of the machine, and t h e value o f its
output t o the plant owner.

Main ttansformcr


HV busbars
Unit transforrncr

i :
I :


The following problems require consideration f r o m the

point o f view o f applying protection:
a. stator electrical faults

-.-supplics switchboard

b. overload

d. unbalanced loading
Industrial or commercial plants w i t h a requirement for
steamlhot water n o w often'include generating plant
utilising or producing steam t o improve overall
economics, as a Combined Heat and Power (CHP)
scheme. The plant w i l l typically have a connection t o the
public Utility distribution system, and such generation is
referred t o as 'embedded' generation. The generating
plant may b e capable of export nf w r p l u s power, or
simply reduce the i m p o r t o f power from the Utility. This
is shown i n Figure 17.2.

e. overfluxing

f. inadvertent energisation
e. rotor electrical faults

f. loss o f excitat~on
g. loss of synchron~sm
h. failure o f prlme mover

j. lubrication oil failure

I. overspeed~ng

m. rotor distortion
n. difference i n expansion between rotating an
stationan/ parts

o. excessive v i b r a t ~ o n


p. core lamination faults

The neutral point of a generator is usually earthed

facilitate protection of the stator winding and associat
Earthing also prevents damaging transic
overvoltages i n the event of an arcing earth fault

f'l;inl l r c d r r ,

- tola1

d c n ? : t n d : xtbtlV

PCC- Poirll

G ! Common C o u l > l ~ n ~

W h c n plan1 o r n r r r l < l r I \


II y>r, PI:,": n A y c r f l o r : l o U1111:y;:to,>

I! x>y, Plrn; - 3 . orrn;l.>: om I::.Ib:v


~ r ~ l ~ c c d

For HV generators, impedance is usually inserted i n

stator earthing connection t o limit the magnitude
earth fault current. There is a wide variation i n the e:
fault current chosen, common values being:
1. rated current

2 . 200A-400A (low impedance earthing]

3. IOA-20A (high impedance earthing)

A modern yencrating u n i t is a complex system

comprising the generator stator winding, associated
transformer and u n i t transformer (if present), the rotor
w i t h its field winding and excitation system, and the
prime mover w i t h its associated auxiliaries. Faults of
many kinds can occur wi:hin this syslem for which
diverse forms of electrical and mechanical protection are

The main methods of impedance-earthing a genet

are shown i n Figure 17.3. Low values of earth '
current may limit the damage caused from a fault
they simultaneously make detection of a fault tov
the stator winding star point more difficult. Excel
special applications, such as marine, LV generator
normally solidly earthed t o comply w i t h s
requirements. Where a step-up transformer is ap

the generator and the lower voltage winding o f the

transformer can be treated as an isolated system that is
not influenced by the earthing requirements o f the
power system.

sufficient that the transformer be designed to have a

primary winding knee-point e m f . equal t o 1.3 times the
generator rated line voltage.

Failure of the stator windings or connection insulation

can result in severe damage to the windings and stator
core. The extent of the damage will depend on the
magnitude and duration of the fault current.


la1 Oirccr carlhing

Typical scrring
(%of carthing
rcsisror raring)

will7 n v C I C ~ f l c t l relay

i ? :.

:;:: :,.:.::. (.! .;:.



.. , .



An earthicg transformer or a series impedance can be

impedance. If an earthing transformer is
ntinuous rating is usually in the range 5250kVA. The secondary winding is loaded with a resistor
of a value which, when referred through the rransformrr
turns ratio, will pass the chosen short-time earth-fault
current. This is typically i n the range of 5-20A. The
resistor prevents the production o f high transient
overvoltages i n the event of an arcing earth fault, which
i t does by discharging the bound charge i n the circuit
Capacitance. For this reason, the resistive component of
fault current should not be Icss. than the residual
Capacitance current. This is the basis of the design, and
in practice values of between 3-5 I,,,
are used.
It is .important that the earthing transformer never


becomes saturated; otherwise a very undesirable

Condition of ferroresonance may occur. The normal rise
ofthe generated voltage above the rated value caused by
a sudden loss of load or by field forcing must be
as well as flux doubling in the transformer
point-on-wave o f voltage application. 11 is

p r . t r.....

? A


The most probable mode of insulation failure is phase to

earth. Use of an earthing impedance limits the earth
fault current and hence stator damage.
An earth fauit involving the stator core results in burning
of the iron at the point of fault and welds laminations
together. Replacement of the faulty conductor may not
be a very serious matter (dependent on set
rating/voltage/construction) but the damage to the core
cannot be ignored, since the welding o f laminations may
result in local overheating. The damaged area can
sometimes be repaired, but i f severe damage has
occurred, a partial core rebuild will be necessary. A
flashover is more likely to occur in the end-winding
region, where electrical stresses are highest.
resultant forces on the conductors would be very large
and they may result in extensive damage, requiring the
partial or total rewinding o f the generator. Apart from
burning the core. the greatest danger arising from failure
to quickly deal with a fault is fire. A large portion of the
insulating material is inflammable, and in the case of an
air-cooled machine, the forced ventilation can quickly
cause an arc flame to spread around the winding. Fire
will not occur in a hydrogen-cooled machine, provided
the stator system remains sealed. In any case, the length
of an outage may be considerable, resulting in major
financial impact from loss of generation revenue and/or
import o f additional energy.

Phase-phase faults clear o f earth are less common; they

may occur on the end portion of stator coils or in the
slots if the winding involves two coil sides i n the same
slot. In the latter case, the fault will involve earth i n a
very short time. Phase fault current is not limited by the
method of earthing the neutral point.

lnterturn faults are rare, but a significant fault-loop

current can arise where such a fault does occur.

Conventional generator protection systems would be

blind t o a n interturn fault, b u t the extra cost and
complication of providing detection of a purely interturn
fault is n o t usually justified. I n this case, an interturn
fault must develop into an earth fault before it can be
c!eared. An exception may be where a machine has an
multiple winding
arrangement, where the probqbility o f an interturn fault
might be increased.

calculation, after measurement of the individual q :

secondary currents. I n such relay designs, there is full.:
galvanic separation o f the neutral-tail and terminal Q.;
secondary circuits, as indicated i n Figure 17.5(a). This is
not the case for the application of high-impedance
differential protection. This difference can impose some
special relay design requirements t o achieve s t a b i l i t y f ~ ~ ,
biased differential protection i n some applications.









To respond quickly t o a phase fault w i t h damaging heavy

current, sensitive, high-speed differential protection is
normally applied t o generators rated in excess of 1MVA.
For large generating units, fast fault clearance will also
maintain stability of the main power system. The zone
of differential protection can be extended t o include an
associated step-up transformer. For smaller generators,
IDMT/instantaneous overcurrent protection is usually the
only phase fault protection applied. Sections 17.5-17.8
detail the various methods that are available for stator
winding protection.

The relay connections for this form of protection are

shown in Figure 17.5(a) and a typical bias characteristic
is shown i n Figure 17.5(b). The differential current
threshold setting I,, can be set as low as 5% o f rated :.
generator current, to provide protection for as much of:
the winding as possible. The bias slope break-point$&
threshold setting I;, would typically be set t o a value.?$,
above generator rated current, say 12O01o, to achieve::?.
external fault stability i'n the event of transient,{2
asymmetric CT saturation. Bias slope I;, setting would:?
- !
typically be set at 150%.
.: c:


i: ,.:.

The theory o f circulating current differential protection is

discussed fully in Section 10.4.


. - ..-.

High-speed phase fault protection is provided, by use of

the connections shown i n Figure 17.4. This depicts the
derivation of differential current through CT secondary
circuit connections. This protection may also offer earth
fault protection for some moderate impedance-earthed
Either biased differential or high
impedance differential techniques can be applied. A
subtle difference w i t h modern, biased. numerical
generator protection relays is that they usually derive the
differential currents and biasi'ng currents by algorithmic

This d~ffersfrom biased differential protection by

manner in which relay stability is achieved for eXt
faults and by the fact that the differential current
be attained through the electrical connections
secondary circuits. I f the impedance of each re
Figure 17.4 is high, the event oC one CT bec
saturated by the through fault cutrent (leadin!




1 0 ~ 4 4 page 285

latively low CT impedance), will allow the current from

unsaturated CT t o flow mainly through the saturated
rather than through the relay. This provides the
uired protection stability where a tuned relay element
is employed. I n practice, external resistance is added t o
the relay circuit i o prwidc the necessary high
impedance. The principle of high-impedance protection
application is illustrated i n Figure 17.6, together with a
summary of the calculations required t o determine ttie
value o f external stabilising resistance.

To calculate the primary operating current, the following

expression is used:

I,, = N X(is, + nl,)


= prima y operating current

N = CT ratio
= relay
l setting

n = number of CT's in parallel with relay element

I, = CT magnetising currerft at

Hcalthy CT

Saturated CT

L', = K V ,

wherc J . O < K r l . S
Stabilising resistor, R,, limits spill currcnt to <I, lrclay sctting)
whcn RE

5I - R R


I,, is typically set to 5% of generator rated secondary


It can be seen from the above that the calculations for

the application of high impedance differential protection
are more complex than for biased differential protection.
However. the ~rotectionscheme is actually cuite simple
and it bffers high level of stability for through f a k
and external switching events.

With the advent of multi-function numerical relays and

with a desire to dispense with external components; high
impedance differential protection is not as popular as
biased differential protection i n modern relaying
. .
. .

. .,

. .ii..
... . : ,..,.l
. a.
.. . . ,...",


i n some applications, protection may be required to limit

voltages across the CT secondary circuits when the
differential secondary current for an internal phasc fault
flows through the high impedance relay circuit(s1, but
this is not commonly a requirement for generator
differential applications unless very high impedance
relays .are applied. Where necessary, shunt-connected.
non-linear resistors, should be deployed, as shown in


... .i.Ji...


:., ;




- ..


..- . - . .






Many factors affect this, including the other protection

functions fed by the CT's and the knee-point
requirements of the particular relay concerned. Relay
manufacturers are able to provide detailed guidance on
this matter.

A common connection arrangement for large generators

is to operate the generator and associated step-up
transformer as a unit without any intervening circuit
breaker. The unit transformer supplying the generator
auxiliaries is tapped off the connection between
generator and step-up transformer.
protection can be arranged as follows.


The CT requirements for differential protection will vary

according t o the relay used. Modern numerical relays
may not require Ci's specifically designed for differential
protection to IEC 60044-1 class PX (or BS 3938 class X).
However, requirements i n respect of CT knee-point
voltage will still have to be checked for the specific
relays used. High impedance differential protection may
be more onerous i n this respect than biased differential

;: ..:::..


p. ..i:;.j:-:.':.?+.i
. . . . . ......
$;:. .i.,:'<i.:-.:
. :':








Page 206


- .

.. .....;,<:...: ,*,\>'
. .



..<. .T:y<?$

. .....

~ $ . j ~ ~17.6.1
. ; ~ ~ :Gencrato~.iSrcp-upTra~isiorrner
. . ... . .. .

> , v,,
. . .,...
. ..


.' : * ,.

transformer rating is extremely low in relation to th

Dlffc-rcntial Protection

generator rating, e.g. for some hydro applications. ~h

location of the third set o f current transformers
normally on the primary side of the unit transformer.
located on secondary side of the unit transformer, th(
wnuld have-to- be of an exceptionally high ratio, :
exceptionally high ratio interposing CT's would have.
be used. Thus, the use of secondary side CTs is not to I
recommended. Cne advantage is that unit transform
faults would be' within the zone of protection of


The generator stator and step-up transformer can be

. . . _. ~
~ ~ by
. ~
a single
i ~ ~
zone- of
~ overall
... k:z:4>$;t.
.., .,.. . % . protection (Figure 17.8).
This will be in addition to
:....:. :.. . differential protection applied to the generator only. The
. .,.- ..< ..
current transformers should be located in the generator
... .
neutral connections- and i n the -transformer HV
Alternatively, C r s within the HV
. .
.. ..c:~$~..~;:



switchyard may be employed if the distance is not

technically prohibitive. Even where there is a generator
circuit breaker, overall differential protection can still be
provided i f desired.



generator. However, the sensitivity of the generat

protection to unit transformer phase faults
considered inadequate, due to the relatively low rating
the transformer in relation to that of the generat
Thus, the unit transformer should have its 0,







a. .

-;<i y7-p-!~

s - - .





... . .. . . . . . . . .

. '


-. . .




. .


The current transformers should be rated according to

Section 16.8.2. Since a power transformer is included
.rwithin the zone o f protection, biased transformer
differential protection, with rnagnetising inrush restraint
should be applied, as discussed in Section 16.8.5.
Transient overfluxing of the generator transformer may
arise due to overvoltage following generator load
rejection. In some applications, this may threaten the
stability of the differential protection. In such cases.
consideration should be given to applying protection
. ...,.
with transient overfluxing restraintlblocking (e.g. based
. . ..z.'-..$' . on a 5th harmonic differential currentthreshold).
Protection against sustained overfluxing is covered i n
Section 17.14.


. . . .
..... .. ..;_,;




...-. ..
,.:. .,. -

-.*.:.: .
;& -T&*;;


i.j,,,l - , ;


(I;.:,, . I . i . .






c;c,,j .L:L:yrc!l:,

For larger gener
Overcurrent protection can be applied as remote ba
~rotection,to disconnect the unit from any uncl
external fault. Where there is only one set of differ
main protection, for a smaller generator, the OverC
protection will also provide local back-up protecti
the protected plant, i n the event that the
protection fails t o operate. The general princi~
setting overcurrent relays are given in Chapter 9.


..... .

-. --


time-delayed plain overcu

protection to generators. For generators rated less
lMVA, this will form the principal stator wi'

.! :..,!.
. ;,


It is


;,; r>,;2ir;


principle form of protection for small generaton,.

for larger ones where differer
protection. is used asthe primaty method of ge&i
stat0.i winding protection.
Voltage dep;nc
ovekcurrent protection may tie applied where differel
protection isnot justified on larger generators, or w
problems are met i n applying plain overcur

. 172..


Overcurrent protection of generators may take :

forms. Plain overcurrent protection may be used a;

- .

5qurc :7.5: O,.rr:?:i :<?:rr~l<;t-i::n~rcrzi~i.

:~.i!cr<,cr;:;; Ji<<::m:i<nn


t r - .

3 , .







s....... A.





.differential protection scheme. Protection for the u

transformer is covered in Chapter 16, including methc
for stabilising the protection against magnetising inn



.,4 ,! ,.!(.!;.

In the case of a single generator feeding an i:

system, current transformers at the neutral end
machine should energise the overcurrent protect
allow a response to winding fault conditions.
characteristics should be selected to take into ;
the fault current decrement behaviour of the ge

The current taken by the unit transformer must be

allowed for by arranging the generator differential
protection as a three-ended scheme. Unit transformer
current transformers are usually applied to balance the
generator differential protection and prevent the unit
transformer through current being seen as differential
current. An exception might be where the unit

with allowance for the performance of the ex







& ,f.t.m.ri.a


- .... : z m




+tern and its field-forcing capability. Without the

provision o f fault current compounding from generator
CT's, an excitation system that is powered from an
wcitation transformer a t the generator terminals will
whibit a pronounced fault current decrement for a
terminal fault. With failure t o consider this effect, the
potential exists for the initial high fault current to decay
to a value below the overcurrent protection pick-up
setting before a relay element can operate, unless a low
current setting and/or time setting is applied. The
protection would then fail to trip the generator. The
settings chosen must be the best compromise between
assured operation i n the foregoing circumstances and
discrimination with the system protection and passage
of normal load current, but this can be impossible with
plain overcurrent protection.
In the more usual case o f a generator that operates i n
parallel with others and which forms part o f an extensive
interconnected system, back-up phase fault protection
for a generator and its transformer will be provided by HV
overcurrent protection. This will respond to the higherlevel backfeed from the power system to a unit fault.
Other generators i n parallel would supply this current
and, being stabilised by the system impedance, it will not
suffer a major decrement. This protection is usually a
requirement of the power system operator. Settings must
be chosen to prevent operation for external faults fed by
.the generator. I t is common for the HV overcurrent
protection relay t o provide both time-delayed and
instantaneous high-set elements. The time-delayed
elements should be set t o ensure that the protected items
of plant cannot pass levels o f through fault current i n
excess o f their short-time withstand limits. The
instantaneous elements should be set above the
maximum possible fault current that the generator can
supply, but less than the system-supplied fault current in
the event of a generator winding fault. This back-up
protection will minimise plant damage in the event of
main protection failure for a generat~ngplant fault and
instantaneous tripping for an HV-side fault will aid the
recovery of the power system and parallel generation.

The choice depends upon the power system

characteristics and level of protection t o be provided.
Voltage-dependent overcurrent relays are often found
applied t o generators used on industrial systems as an
alternative to full differential protection.

Voltage controlled overcurrent protection has t w o

timelcurrent ciiaracteristics which are selected according
t o the status of a generator terminal voltage measuring
element. The voltage threshold setting for the switching
element is chosen according t o the following criteria.
1. during overloads,

when the system voltage is

sustained near normal, the overcurrent protection
should have a current setting above full load current
and an operating time characteristic that will prevent
the generating plant from passing current to a remote
external fault for a period i n excess o f the plant shorttime withstand limits

r r u . r k


P r . r r r r i , n

. . . . . . . - . .
W Awr.-,ri.m






' - '

an external circuit. fault that.kill

. assist
- ..



Typical characteristics are shown i n Figure 17.9.






- ....




Currcnr pick-up k v c l


The plain overcurrent protection setting difficulty

referred to i n the previous section arises because
allowance has to be made both for the decrement of the
generator fault current with time and for the passage of
full load current.
To overcome the difficulty of
discrimination. the generator terminal voltage can be
measured and used to dynamically modify the basic relay
currentltime overcurrent characteristic for faults close to
the generating plant. There are two basic alternatives
for the application of voltage-dependent overcurrent
Protection, which are discussed in the following sections.


2. under close-up fault conditions, the busbar voltage

.must fall below the voltage threshold so that the
second protection characteristic will be selected. This
characteristic ihould be set t o allow relay operation
with fault current decrement for a close-up fault a t
the generator terminals or- a t the,HV- b u s b a .~ T h e . ~ ~ , ~ , & ; :
,-.. .' . .
. . . . .sh&d
also ..time~jrade..'.with:~xteI??l~......
jy&-be.gdditiond i"fe&ds t o I ~ ~ ~ ~ ~ : ' - : , . . ~ '

f:,r.;,r ! ;

c ~ ~ n l ~ ~r dl cl ~r yd


,.,, ,<

!: !.. : ; . . ; . j / . f r ,

Voltagc lcvcl


.=. .<:

The alternative technique is to continuously vary the

relay element ,pickup setting with generator voltage
variation between upper and lower limits. The voltage is
said to restrain the operation o f the current element.
The effect is to. provide a dynamic I.D.M.T. protection
characteristic, according to the voltage a t the machine

- <?:r

terminals. Alternatively, the relay element may be

regarded as an impedance type with a long dependent
time delay. In consequence, for a given fault condition,
the relay continues t o operate mnre or less
independently of current decrement i n the machine. A
typical characteristic is shown i n Figure 17.10.


This method is used i n the following situations:

a. direct-connected generators operating i n parallel

b. generators with high-impedance neotral earthin,
the earth fault current being limited to a few ter
of amps
c. installations where the resistance of the grour
fault path is very high, due to the nature of t[
In these cases, conventional earth fault protection
described i n Section is of little use.

5 2

V ,

The principles o f sensitive earth fault protection ;

described i n Sections 9.17.1. 9.18 and 9.19. The ea:
fault (residual] current can be obtained from residl
connection of line CT's, a line-connected CBCT, or a 0
the generator. neutral. The latter is not possible
directional protection is used. The polarising voltagt
usually the neutral voltage displacement input to .
relay, or the residual o f the three phase voltages, s.
suitable VT must be used. For Petersen Coil earthin!
wattmetric technique {Section 9.19) can also be usec

Voltage level

Earth fault protection must be appl~edwhere impedance

earthing is employed that limits the earth-fault current
t o less than the pick-up threshold of the overcurrent
andlor differential protection for a fault located down to
the bottom 5% o f the stator winding from the starpoint. The type of protection required will depend on the
method of earth~ngand connection of the generator to
the power system

A single direct-connected generator operating on an

isolated system will normally be directly earthed.
However, i f several direct-connected generators are
' 17*. . operated in parallel, only one generator is normally
For the unearthed generators, a
...\!. ....
. . earthed a t a time.
. .:..i",':;,.simple measurement of the neutral current is not
possible, and other methods of protection must be used.
~. .
The following sections describe the methods available.

With this form of protection, a current transformer i n the

neutral-earth connection energises an overcurrent relay
This provides unrestricted earth-fault
protection and 'so i t must be graded with feeder
protection. The relay element will thereforfhave a zimedelayed operating characteristic.
Grading must be
carried out in accordance with the principles detailed in
Chapter 9. The setting should not be more than 33/~of
the maximum earth fault current of the generator, and a
lower setting would be preferably, depending on grading

. For 'direct '&nnected.


operating in. para

earth: fault :protection. may
necessary. This is t o ensure that a faulted ge&rat~r
be tripped. before there is any ossibility of the:+
~ " e r c u r r e n tprotection tripping a parallel he;
generator. When being driven by residually-conne
phase CT's, the protection must be stabilised ag;
incorrect tripping with transient spill current i n thee
of asymmetric CT saturation when phase faul
magnetising inrush current is being passed. Stabil
techniques include the addition o f relay ci
impedance and/or the application of a time delay. W
the required setting o f the protection is very lo
comparison t o the rated current of the phase C
would be necessary to employ a single CBCT for the
fault protection to ensure transient stability.

. directional'..&nsitive

Since any generator i n the paralleled group m:

earthed, all generators will require to be fitted wit1
neutral overcurrent protection and sensitive direc
earth fault protection.
The setting o f the sensitive directional earth
protection is chosen to co-ordinate with ger
differential protection andlor neutral v
displacement protection t o ensure that 95% of thc
winding is protected.
Figure 17.11 illustrat,
complete scheme, including optional blocking
where difficulties i n co-ordinating the generat
downstream feeder earth-fault protection occur.



Page 285


As the protection is still unrestricted, the voltage setting

.: 3

o f the relay must be greater than the effective setting o f

any downstream earth-fault protection. It must also be
rn-nrriinatr w i t h c ~ r r h n
,-- tn
'. r .n. t r r t i n n
Sometimes, a second high-set element with short time
delay is used t o provide fast-acting protection against
major winding earth-faults. Figure 17.12 illustrates the
possible connections that may be used.

. . . . .



figorc 77. ll: CC~??M):C.~S;;~

pr8>:wrifir;S C I I C V ~:CI:C , ; , : c c I - ~ - , ; , I : ; ~ ,c~ . ~ c

o~nrra!:lrs :),8eroi;.za . ir: 13;;ruiir.i


For cases (b) and (c) above, it is not necessary t o use a

directional facility. Care must be taken t o use the correct
RCA setting - for instance if the earthing impedance is
mainly resistive, this should be 0". On insulated or very
.,-.hig.h impedance earthed systems, an RCA o f -90" would
. :used, as the .earth fault current is -predominately

.-2: . -..

sensitive earth-fault protection can also be

used for detecting winding earth faults. I n this case, the
. relay element is applied t o the terminalsof the generator
and is set to respond t o faults only within the machine
windings. Hence earth faults o n the external system do
not result i n relay operation. However, current flowing
from the system into a winding earth fault causes relay
operation. It will not operate on the earthed machine, so
that other types of earth fault protection must also be
applied. All generators must be so fitted. since any can
be operated as the earthed machine.


@ Ocrivcd from phasc ncutral voltagcs

@ Measurcd from carth impcdancc
@ Mcasurcd from brokcn dclta VI

... .. . .. .



. . . . . .. . ... .... .. .... .. ....




?,w,r< 17. 1.:: ?:<!i,fr":


. . .u. . .

: = .

..- .




7 ,


. .



. . . . . . .. . . . .. . . . . .

In a balanced network, the addition of the three phaseearth voltages produces a nominally zero residual
. voltage, since there would be little zero sequence voltage
Present. Any earth fault will set up a zero sequence
system voltage, which will give rise t o a non-zero
residual voltage. This can be measured by a suitable
relay element. The voltage signal must be derived from
a VT that is suitable - i.e. it must be capable of
transforming zero-sequence voltage, so 3-limb types and
those without a primary earth connection are not
suitable. This unbalance provides a means of
detecting earth faults.
third harmonic voltages that may be
Present in the system
as these will



As noted in Section 17.2, a directly-earthed generatortransformer u n i t cannot interchange zero-sequence

current with the remainder o f the network, and hence an
earth fault protection grading problem does n o t exist
The following sections detail the protection methods for
the various forms of impedance earthing of generators.
. . . , . ... ..I. _ .. .. . .. . .





E : .


A current transformer mounted on the neutral-earth

conductor can drive an instantaneous and/or time
delayed overcurrent relay element, as shown i n Figure
17.13. I t is impossible to provide protection for the whole
of the winding, and Figure 17.13 also details how the
percentage of winding covered can be calculated. For a
relay element with an instantaneous setting, protection is
typically limited t o 90% of the winding. This is t o ensure
that the protection will not maloperate w i t h zero
sequence current during operation of a primary fuse for a
...... .
W earth fault or with any transient surge currents that
. , -. could flow through the interwinding capacitance of the . . . . . ..,,.,...
, ..
.: .; . .-"?.-.;: ,,step-up transformer for an HV system earth fault.

. - . . . .-:;;.;.\.'..

A time-delayed relay ismore secure i n this respect, and if ,.;

may have a ~ t t i n gto cover 9% of the stator winding.
Since the generating
units under consideration are usually
large, instantaneous
and timedelayed relay elements are


,. . '



. -,
. ,--



Page 290


. ... ...:,:. ........
* ...
. . . . . ..+,..,


. .<...:'.:....':. .!.
........ .
.:.. .. -..
.:." ,.
\ . .. ,::,,>c .., :



.;. !.$.,. .,..> z , . .


:_ .>

, ..
,.+&. -:!,'.:.

often applied, with settings of 10% and 5% of maximum

current respectively; this is the optimum
...... : compromise i n performance. The portion of the winding
..& .z,..... :,.>' ..
:..-c<. .'-. :...v
left unprotected for an earth fault is a t the neutral end.
2 .
..:,.-.. the voltage t o earth a t this end of the winding is
.;> ..::<,.. ,.. '.;a-.
2%: . ,..
the probability of an eartn fauit occurring is also lo^.
. .. . .
~ .;.
............ : . . .
Hence additional protection is often not applied.
. ,. ' "

. x 2 :Y,,.

,.... .".. earth fau It






k .:








+ ..







. :



(a) Protection using a currcnt clcmcnt





- 7-71

I ( R






%co~rwd (I-omic1 x 1100%



. . . . . . .
F;q,,r< I?.?:;: <,::I,



... . . -. .



. . .;.j ,,,;.,>,;:
. . ,, '.., .

- ;:.

. . .,;,.;>:Gc. < . .

.. .... ...:.
. . ..


,.: ... . *

. ... .


.- .

In this arrangement, shown in Figure 17.14(a), the

generator is earthed via the primary winding of a
distribution transformer. The secondary winding is fitted
with a loading resistor to limit the earth fault current.
An overcurrent relay element energised from a current
transformer connected i n the resistor circuit is used to
measure secondary earth fault current. The relay should
have an effective setting equivalent to 5% of the
maximum earth fault current at rated generator voltage,
i n order to protect 95% of the stator winding. The relay
element response to third harmonic curre_nt should be
limited to prevent incorrect operation when a sensitive
setting is applied.

.. .. .. . . .


-. ....

.. . '2

. ..*$
.' .".>.'.


Earth fault protection can also be provided using a voltag

measuring element i n the secondary circuit instead. ;T.
setting considerations would be similar to those for:$
current operated ~ ~ o t e c t ~ obut
n * transposed to voltat
The circuit diagram is shown i n Figure 17.l4(b).
Application of both voltage and current opera.
elements to a generator with distribution transfon
earthing provides some advantages.
The curr
operated function will continue to operate i n the ev
of a short-circuited loading resistor and the volt
protection still functions in the event of an OF
circuited resistor. However, neither scheme will ope
i n the event of a flashover on the primary terminal
the transformer or of the neutral cable between
generator and the transformer during an earth faul
CT could be added in the neutral connection closet,
generator, to energise a high-set overcurrent
detect such a fault, but the fault current
high enough to operate the phase differ<

As discussed i n Section for neutral overcurrent

protection, the protection should be time delayed when
in order to prevent
a sensitive setting is
maloperation under transient conditions. It also must
grade with generator VT primary protection (for a, VT
primary earth fault). An operation time i n the range
0.5s-3s is usual. Less sensitive instantaneous protection
can also be applied to provide fast tripping for a heavier
earth fault condition.

. .


. .


. . .



z %-t;rrc,?r:!<.-:.:i

. .

.....- . .
. . .
...T. x
? I.?f;:
G:zni:fcr kl.!i::!:nc;
. . . r.afrh-bup.
.w%... .... :.
.-*.-3, :rc~$:c:?.?;~:&hh&

,>!o;?<l;cc c i l:!<h-,c5::!::7:?


ednhea $ c ? : : z i v z : d : ~ ~ r2...<::%?..?s
. : s : z-

<> ,



gcncrator stator winding using a currcnt clcmcnt

- ..





(bl Protection uring a voltagc clcmcnt






. .:.




This can be applied in the same manner as for c

connected generators (Section


N*tr*.l Pr.r<rli..




: :



Page 2 9 1


rence is that 'the are no grading problems as the

is inherently restricted. A sensitive setting
can therefore be used, enabling cover o f up t o 95% of
the stator winding t o be achieved.

E a r l ! i F:i~l: P~.nicc::ic~$

17.t3.3 Rcs:rir';ccl

This technique can beused on small generators not fitted

with differential protection to provide fast acting earth
fault protection within a defined zone that encompasses
the generator. It is cheaper than full differential
protection but only provides protection against earth
faults. The principle is that used for transformer REF
protection, as detailed in Section 16.7. However, in
contrast t o transformer REF protection, both biased lowimpedance and high-impedance techniques can be used.
!/ .,......... !'<;--:":
".":' ,I:;.:!.,L::+>::
: :.


:i .




This is shown i n Figure 17.15. The main advantage is

that the neutral CT can also be used in a modern relay t o
provide conventional earth-fault protection and no
external resistors are used. Relay 'bias is required, as
described i n Section 10.4.2, b u t the formula for
calculating the bias is slightly different and also shown
in Fiqure 17.15.


. .


Phasc CT ratio 100011

. .


,.-. 2-i.;:.a+,.

CT in the neutral connection. Settings of the order of 5010





o f maximum earth fault current at the generator

terminals are typical. The usual requirements i n respect
of stabilising resistor and non-linear resistor to guard
against excessive voltage across the relay must be taken,
where necessary.

-.-. .-.-'




g... , : ~ -. : ,
. . .!$
..... ..


wrtr .>.-

..i ;

. >


..;,\..: . : . - v.....

. . ;Q;


- . . . . . : . . ..,.
. . . . . . .

All of the methods for earth fault protection detailed so

far leave part of the winding unprotected. In most cases,
this is of no consequence as the probability of a fault
occurring in the 5010 of the winding nearest the neutral
connection is very low,due to the reduced phase to earth
voltage. However, a fault can occur anywhere along the
stator windings in the event of insulation failure due to
localised heating from a core fault. In cases where
protection for the entire winding is required, perhaps for
alarm only, there are various methods available.


.i .

.. : ..

.:, . . . . .


. ... ..

. . .. .. ...
. . ..




. .

r. 0






One method is to measure the internally generated third

harmonic voltage that appears across the earthing
impedance due to the flow o f third harmonic currents :.":c
through the shunt capacitance b f t h e
windings : .
. . .
etc. When a fa'ult occurs in.the p a r t o f the stator
winding nearest the neutral end, the third harmonic
voltage drops to near zero, and hence a relay element
that responds to third harmonic voltage can be used t o
detect the condition. As the fault location moves
progressively away from the neutral end, the drop in
third harmonic voltage from healthy conditions becomes
less, so that at around 20-30010 of the winding distance,
it no longer becomes possible to discriminate between a
healthy and a faulty winding. Hence, a conventional
earth-fault scheme should be used i n conjunction with a
third harmonic scheme, to provide overlapping cover
of the entire stator winding. The measurement o f third
harmonic voltage can be taken either from a star-point
VT or the generator line VT. In the latter case, the VT
must be capable of carrying residual flux, and this
prevents the use of 3-limb types. I f the third harmonic
voltage is measured at the generator star point, an
undewoltage characteristic is used. An ovewoltage
characteristic is used i f the measurement is taken from
the generator line VT. For effective application of this - .: .
form of protection, there should be a t least 1010 third :-.-.;..
harmonic voltage across the generator neutral earthing i....::::;
.. :..,:
impedance under all operating conditions.





(highrst of 1,. IS. I C J +(INx scoiina.foctorJ

u.hrrr scnlit~gfactor



T ratio



The initial bias slope is commonly set t o 0%to provide

maximum sensitivity, and applied up to the rated current
of the generator. I t may be increased to counter the
effects of CT mismatch. The bias slope above generator
rated current is typically set to 150% o f rated value. The
initial current setting is typically 5qo of the minimum
earth fault current for a fault a t the machine terminals.

:. .

p ;.....
: , , , : . ...., ,..... I , ,?..
. :,.

::: :.!.<::::,,::.:.#c.

The principle of high impedance differential protection is

given in Chapter 10 and also described further in Section
17-52. The same technique can be used for earth-fault





., .

A problem encountered is that the level o f third

harmonic voltage generated is related to the output of
the generator. The voltage is low when generator output






Page 292

i s low. I n order t o avoid maloperation when operdting at

low power output, the relay element can be inhibited
using an overcurrent or power element (kW, kvar or kVA)
and internal programmable logic.
; y,<?,,;..;






Another method for protecting the entire stator winding

of a generator is t o deploy signal injection equipment to
inject a low frequency voltage between the stator star
point and earth. An earth fault at any winding location
will result i n the flow of a measurable iniection current
t o cause protection operation. This form of protection
can provide earth fault protection when the generator is
a t standstill, prior t o run-up. It is also an appropriate
method t o apply t o variable speed synchronous
machines. Such machines may be employed for variable
speed motoring i n pumped-storage generation schemes
or for starting a large gas turbine prime mover.

. . ., . , . . . . .,. . .:.'.(,. >
, ........
LS2%j:;<:;>;. :,.>f.:j<>,:...,





mach~newith a healthy voltage regulator, but i t may be

caused by the fol!owing cont~hgencies.



a. defective operation of the automatic voltage

regulator when the machine is i n isolated operation



b. operation under manual control with the voltage

regulator out of service. A sudden variation of the
load, i n particular the reactive power component.
will give rise to a substantial change i n voltage
because of the large voltage regulatinn inherent in
a typical alternator



.. 17..


. . . . . . .

. ..




. .,.I"......
- .....
... .2,... . ..




,.: -.,:


., ..

,. ,

. .




... ...


c. sudden loss of load (due to tripping of outgoing

feeders, leaving the set isolated or feeding a very small
load) may cause a sudden rise in terminal voltage due
to the trapped field flux and/or overspeed
Sudden loss of load should only cause a transient
overvoltage w h ~ l ethe voltage regulator and governor act
to correct the situation. A maladjusted voltage regulator
may trlp to manual, ma~ntalnlngexcitation a t the value
prior to load loss while the generator supplies little or no
load. The terminal voltage will increase substantially,
and i n severe cases i t would be limited only by the
saturation characteristic of the generator. A rise in speed
simply compounds the problem. I f load that is sensitive
to overvoltages remains connected, the consequences i n
terms of equipment damage and lost revenue can be

. a -

, A,,.




. -A
I s ..



. A sustained overvoltage condition should not occur for a

Overvoltages on a generator may occur due t o transient

surges on the network, or prolonged power frequency
overvoltages may arise from a variety of conditions.
Surge arrestors may be required t o protect against
transient overvoltages, but relay protection may be used
to protect against power frequency overvoltages.



. I . .;
': ,.
: .


For these reasons, it is prudent to provide powerx

frequency overvoltage protection, i n the form of a timedelayed element, either IDMT or definite time. The time *; .
delay should be long enough t o prevent operation during .~.normal regulator action, and therefore should take ;V
account of the type of AVR fitted and its transient ,;:.
response. Sometimes a high-set element is provided as j
- .. .>.,
well, w i t h a very short definite-tirne delay
instantaneous setting t o provide a rapid trip in extreme<:$
circumstances. The usefulness of this is questionable
generators fitted with an excitation system other than
static type, because the excitation will decay in$
accordance with the open-circuit time constant of the;::
field winding. This decay can last several seconds. The-j
relay element is arranged to trip both the main circuit,
breaker (if not already open) and the excitation; trippingL!
the main circuit breaker alone is not sufficient.

Prolonged overvoltages may also occur on




Undervoltage protection i s rarely fitted to generators. l t

is sometimes used as an interlock element for anothe
protection function or scheme. such as field failu.6;
protection or inadvertent energisation protection,.+yheg
the abnormality t o be detected leads -directly$...
. -.
indirectly to an undervoltage condition.
. . .
. -..

A transmission system u n d e ~ o l t a g econdition may arix

when there is insufficient reactive power generation to
maintain the system voltage profile and the conditior
must be addressed to avoid the possible phenomenon
system voltage collapse.
However, i t should be addressed by the deployment o
'system protection' schemes. The generation should no
be tripped. The greatest case for undervoltage protectio
being required would be for a generator supplying a
isolated power system or to meet Utility demands fl
connection of embedded generation (see Section 17.21
In the case of generators feeding an isolated systei
undervoltage may occur for several reasons, typica
overloading or failure of the AVR. In some cases. t
performance of generator auxiliav plant fed via a u
transformer from the generator terminals could
adversely affected by prolonged undervoltage.
Where undervoltage protection is required, it sho
comprise an undervoltage element 'and an associa
time delay.
Settings must be chosen t o a\
maloperation during the inevitable voltage dips du,
power system fault clearance or associated with m1
starting. Transient reductions in voltage down to 80'
less may be encountered during motor starting.




Page 2Sf




:7Y - ~1


J ,

,Low forward power or reverse power protection may be

required for some generators to protect the prime mover.
Parts o f the prime mover may not be designed t o
experience reverse torque or they may become damaged
throuqh continued rotation after the prime mover has
suffered some form o f failure.

Low forward power protection is often used as an

interlocking function to enable opening of the main
circuit breaker for non-urgent trip< - e.g. for a stator
earth fault on a high-impedance earthed generator, or
when a norinal shutdown o f a set is taking place. This is
to minimise the risk o f plant overspeeding when the
electrical load is removed from a high-speed cylindrical
rotor generator. The rotor of this type o f generator is
highly stressed mechanically and cannot tolerate much
overspeed. While the governor should control overspeed
conditions, it is not good practice t o open the main
circuit breaker simultaneously with tripping of the prime
mover for non-urgent trips. For a steam turbine, for
example, there is a risk o f overspeeding due to-energy
-storage i n the trapped steam, after steam vaive tripping,
:.:.orliri the.everlt that the steam valve(s1 do not fully close
f o r some reason. For urgent trip conditions, such as
.:stator differential protection operation, the risk involved
in simultaneous prime mover and generator breaker
tripping must be accepted.


Dicxl Engine


(split shaft)
Gas Tuhinc ...........

firclcrplosion due
to unburnt fucl
Mechanical damaqc

. . . . . . . .


. (bladcs out>2 of watcrl


bladc and runncr


(bladcs in water)


Stcam Turbinc :


..- . . . . . . . . .

7lJh!< 17. 1 ;

turbinc bladc damaqc

gcarbox damayc
on gcarcd XIS


-. . . . . . . . .

Reverse power protection is applied to prevent damage

to mechanical plant items in the event o f failure of the
prime mover. Table 17.1 gives details o f the potential
problems for various prime mover types and the typical
settings for reverse power protection. For applications


. .-A


. -


:. :.. \!:..srr_,..,z:


A three-phase balanced load produces a reaction field

that, to a first approximation, is constant and;;;o
synchronously w i t h the rotor field system.
unbalanced corfdition can be resolved into positive,
negative and zero sequence components. The positive
sequence component is similar t o the normal balanced
load. The zero sequence component main

-- . . . . . . . . . . . - . . . .
. . . . .




armature reaction.

. . L


. ...

; / . ;


r . . ;:; :; ,:.






The negative sequence compdnent i j similar: to, the- v$<$$:<{2'

positive . sequence .system,.-.except )that<Jhe -.cesulti"q, ; 1*.,!<...&;.-:...>.
u : . f.
reaction field rotates in -the.@p+ite direction t d t h 6 . d ; ~field system, Hence, a flux, is produced which cuts the:. -.
rotor at twice the rotational velocity, thereby inducing


double frequency currents i n the field system and i n the

rotor body. The resulting eddy-currents are very large
and cause severe heating of the rotor.

A generator is assigned a continuous negative sequence

rating. For turbo-generators this rating is low; standard
values of 10% and 15% of the generator continuous
rating have been adopted. The lower rating applies when
the more intensive cooling techniques are applied, for
example hydrogen-cooling with gas ducts i n the rotor to
facilitate direct cooling of the winding.

gcarbox damage

(single *ah1

The reverse power protection should be provided with a

definite time delay on operation to prevent spurious
operation with transient power swings that may arise
following synchronisation or i n the event of a power
transmission system disturbance.

So severe is this effect that a single-phase load equal t o

the normal three-phase rated current can quickly heat
the rotor slot wedges to the softening point. They may
then be extruded under centrifugal force until they stand
above the rotor surface, when it is possible t h a t they may
strike the stator core.



where a protection sensitivity o f better than 3% is

required, a metering class CT should be employed t o
avoid incorrect protection behaviour due to CT phase
angle errors when the generator supplies a significant
level of reactive power a t close t o zero power factor.

Short time heating is of interest during system fault

conditions and it is usual i n determining the generator
negative sequence withstand capability to assume that
the heat dissipation during such periods is negligible.
Using this approximation i t is possible to express the
heating by the law:



fft'\;. ; 3 ,


sequence capacity and may not require protection.-,;@

Modern numerical relays derive the negative s e q u e n ~
current level by calculation, with no need for speIci
circuits to extract the negative sequence component. A.'.&
true thermal replica approach is often followed. to aliOw i.;$


IZR = negative sequence component

(per unit of MCR)
t = time (seconds)

= constant proportional

to the thennal capacity

of the generator rotor

a. standing levels of negative sequence current below

the continuous withstand capability. This has the
effect of shortening the time to reach the critial
temperature after an increase in negative sequence
current above the continuous withstand capability .'

For heating over a period of more than a few seconds, it

is necessary to allow for the heat dissipated. From a
combination of the continuous and short time ratings,
the overall heating characteristic can be deduced to be:

b. cooling effects when negative sequence current

levels are below the continuous withstand .
The advantage of this approach is thz: cooling effects are
modelled more accurately, but the disadvantage is that




the tripping characteristic m3Y not follow th; withstand

characteristic specified by the manufacturer accurately. .,

Ilegarive PIIaSe requetlce corrrilluousratillg ill

per unit of MCR

The typical relay element characteristic takes the form of ;:

The heating character;stics of various designs of

generator are shown in Figure 17.16.

time to trip


-. ................... >
coolcd (air)

... lndircctly coolcd (H2)

350MW dircct coolcd j

660MW dircct coolcd i

-1000MW d~rcctcooled,
Usmg I:t modcl
Usmg truc thcrmal




. . ..

0.01 . . . . . . . . . . . . . . .
Ncgativc scqucncc currcnl (p.u.1

: ...

. ..<:.:+.

.... >.......r i




; ,...


,, , , , , ,
- ,- .......

. ,,:.... .



m l o r qcr;r~irluri




1 I . iIj.7 N:~;J-.I,!~,I',;L:.V


I: a .

. . ~iun

'!:<q ?






. .:.,..;s.;.:; .,


negative sequence witltstand coeficient (Figure 1 7.1 6)

IZcmr = generator maximum coittinuous Iz withstand


generator rated primary curretlr

CTpriinary current


relay rated current

Figure 17.16 also shows. the thermal replica time



This protection is applied to prevent overheating due to

negative sequence currents.
Small -salient-pole
generators have a P
~ larger
P negative

characteristic described by Equation 17.1, frorn which it..

will be seen that a significant gain in capability is
achieved at low levels of negative sequence current'
Such a protection element will also respond to phase-:
earth and phase-phase faults where sufficient negative
sequence current arises. Grading with downstream.
power system protection relays is therefore required. A,
definite minimum time setting must be applied to the
negative sequence relay element to ensure
grading. A maximum trip time setting may also be used
to ensure correcttripping




Page 295

rrent level is only slightly i n excess of the continuous

thstand capability and hence the trip time from the
ay depart significantly from the rotor

Accidental energisation of a generator when it is not

running may cause severe damage to it. With the
generator a t standstill, closing the circuit breaker results
in the generator acting as an induction motor; the field
winding (if closed) and the rotor solid ironldamper
circuits acting as rotor circuits. Very high currents are
induced i n these rotor components, and also occur i n the
stator, with resultant rapid overheating and damage.
Protection against this condjtion is therefore desirable.
A combination of stator undervoltage and overcurrent
can be used t o detect this condition. An instantaneous
overcurrent element is used, and gated with a threephase undervoltage element (fed from a VT on the
generator side of the circuit breaker) t o provide the
protection. The overcurrent element can have a low
setting, as operation is blocked when the generator is
,;.;ioperating normally. The voltage setting should be low
;"enough t o ensure that operation cannot occur for
. .,transient fautts. - A setting of about 50% of rated voltage
. :-,. is -typical. VT failure can cause maloperation o f the
. .
protection, so the element should be inhibited under
these conditions.

These conditions are grouped together because these

problems often occur due to a departure from
synchronous speed.



Overfluxing occurs when the ratio o f voltage to

frequency is too high. The iron saturates owing to the
high flux density and results in stray flux occurring in
components not designed to carry it. Overheating can
then occur, resulting i n damage. The problem affects
both direct-and indirectly-connected generators. Either
excessive voltage, or low frequency, or a combination of
both can result i n overfluxing, a voltage to frequency
ratio i n excess of 1.05p.u. normally being indicative of
this condition. Excessive flux can arise transiently, which
is not a problem for the generator. For example, a
generator can be subjected t o a transiently high power
frequency voltage. at nominal frequency. immediately
after full load rejection. Since the condition would not
be sustained, it only presents a problem for the stability

L!. .

k-...- -..... -...........


v ~ ~
r . t rr , ri i , m


~ . r , - a r i a *

~ . i , r

of the transformer differential protection schemes

applied a t the power station (see Chapter 16 for
transformer protection). Sustained overfluxing can arise
during run up, i f excitation is applied too early with the
AVR i n service, or i f the generator is run down, w i t h the
excitation still applied. Other overfluxing instances have
occurred from loss of the AVR voltage feedback signal,
due to a reference VT problem. Such sustained
conditions must be detected by a dedicated overfluxing
protection function that will raise an alarm and possibly
force an immediate reduction i n excitation.
. .:' . ., , i . .

Most AVRs' have an overfluxing protection facility
- -.;$>.
< .........
- ..
.: -r.....
.-..,?. .
included. This may only be operative when the generator
. . ,.:
is on open circuit, and hence fail to detect overfluxing
. : ; s : ~ . : . : .,:... .
conditions due to abnormally low system frequency. " .
p..;~::*-.:. .' .
this facility is not engineered t o protection z....$
*,:,,. ,:- Q ::
relay standards, and should not be solely relied upon t o
.,. r
'Iprovide o v e k ~ u x i nprotection.
A separate relay element Y':>&
is therefore desirable and provided i n most modern


ow ever,




I t is usual t o provide a definite time-delayed alarm

setting and an instantaneous or inverse time-delayed
trip setting, t o match the withstand characteristics o f
the protected generator and transformer. It is very
important that the VT reference for overfluxing :
protection is not.the same as that.used.for the AVR. . . . :.

. . .

. . . . . .. . . . .. . .. .. . .. . . ? . . : > : .


-1:':- i



. . .

&, . : - .
W .




The governor fitted to the prime mover normally provides

protection against overfrequency. Underfrequency may
occur as a result of overload of generators operating on
an isolated system, or a serious fault on the power
system that results in a deficit of generation compared
to load. This may occur i f a grid system suffers a major
fault on transmission lines linking two parts o f the
system, and the system then splits into two. I t is likely
that one part will have an excess of generation over load,
and the other will have a corresponding deficit.
Frequency will fall fairly rapidly i n the latter part, and the
normal response is load shedding, either by load
shedding relays or operator action. However, prime
movers may have to be protected against excessively low
frequency by tripping of the generators concerned.


17 -

W ~ t hsome prime movers, operation i n narrow frequency

bands that lie close to normal running speed (either
above or below) may only be permitted for short periods, together with a cumulative lifetime duration of .. '
. . .
operation i n such frequency bands. This typically occurs ai,.:,..
....... :
due to the presence of rotor torsional frequencies in such ~ .... ~
frequency bands. In such cases, monitoring o f the period iij?.:::
of time spent i n these frequency bands is required. A :'.,...:. .
special relay is fitted i n such cases, arranged t o provide
alarm and trip facilities if either an inoividual or

I- ,>-.




Page 2 9 6

cumulative period exceeds a set time.

produce a balancing force on this axis. The result is an

unbalanced force t h a t in a large machine may be o f the
order o f 50-100 tons. A violent vibration is set up that
may damage bearing surfaces or even displace t h e rotor
by an amount sufficient t o cause it t o foul the stator.

The field circuit of a generator, comprising the field

winding of the generator and the armature o f the exciter,
together w i t h any associated field circuit breaker if it
exists, is an isolated d.c. c i r c u ~ twhich is not normally
earthed. I f an earth fault occurs, there will be no steadystate fault current and the need for action will not be








A = area

. . .!r:!.;,!>.!>,:
,,. .,,,,
. . . . . . . . .. ~

A 'blind spot' would exist a t t h e centre o f the field

winding. To avoid a fault at this location re,mainipg.,,.,.
'undetected., the tapping point d n the potentiometi$~~
could be varied b y a p u s h b u t t o ~o r switch:.. l~h%'i@l&j.*. . .
. setting is typically about 5% o f the exciter.vol-tage.- . +
:;.;. . . .



This is a scheme that was fitted t o older generators, and

it -is illustrated i n Figure 17.18. An earth fault on the
field winding w o u l produce a voltage across the relay,
the maximum voltage occurring for faults at the ends of
the winding.

F=- B ' A

: >;-::;j;jit


b. a.c. injection method

. .. . .. :. .. . . ;

More damage may be caused mechanically. If a large

portion o f the winding is short-circuited, the flux may
adopt a pattern such as that shown in Figure 17.17. The
attracting force at the surface o f the rotor is given by:

. .
; :<,<;;;<,!

a. potentiometer method


Two methods are available t o detect this type o f f a u l t

The first method is suitable for generators that
incorporate brushes in the main generator field windina:
The second method requires at least a slip-ring
connection t o the field circuit:

Danger arises if a second earth fault occurs at a separate

point i n the field system, t o cause the high field current
to be diverted, in part at least, from the intervening
turns. Serious- damage t o the conductors and .possibly
the rotor can occur very rapidly under these conditions.



flux density

. .,*--






Short Circtiil





[;?arc ;7. IS:

: . . . . . . .. .


.. .,. r,. . -

LI;;:: : U ! I I : p x : ~ c t t ~~fi;&f
,:ir8::,:: !~y?,):#:r!!;?;cc!,.r?crno,i

. .:..


F,qt,,c T ? . : 7: .F!"Y ,fr>lr,r>.,!~r-


Two methods are in common use. The first is based on i

low frequency signal injection, with series filtering, as ::
shown in Figure 17.19(a). It comprises an injection::'.
source that is connected between earth and one side
the field circuit, through capacitive couplir~gand the.:;
measurement circuit. The field circuit is subjected to a n j i
alternating potential at substantially the same level;l
throughout. An earth fault anywhere in thc ficld system':
will give rise to a Current that is detekted as an;?
equivalent voltage across the adjustable rcsistor by the.$
relay. The capacitive coupling blocks the normal d.~.fie143
voltage, preventing the discharge of a large d i r c ~ e
current through thc protection scheme. T h c c o m b i n a t i ~ ? ~ ~


wirh put:!~!+, thorl r,trv;l


.. ..

I t will be seen from Figure 17.17 that the flux is

concentrated on one pole but widely dispersed over the
other and intervening surfaces. .[he attracting force is in
consequence large on one pole but very weak on the
axis will
opposite one, while fluxon the

....c-. ;'. l:



... :,' . -!I:.:;;

P.,.,.. .:
..f.5. ...









~ 8 l w ~ ' rP kr , l r < , i . q




....... .',


,- <.b,









C w i l t





Page 297

of series capacitor and reactor forms a low-pass tuned

circuit. the intention being to filter higher frequency
rotor currents that may occur for a variety of reasons.

17.1 5.2 Rotor Earth Faul? Proreaior?

For 6rush;czs Gcr-erato:s

A brushless generator has an excitation system

consisting of:

Other schemes are based on power frequency signal

injection. An impedance relay element is used, a field
winding earth fault reducing the impedance seen by the
relay. These suffer the draw back of being susceptible to
static excitation system harmonic currents when there is
significant field winding and excitation system shunt

1. a main exciter with rotating zrmatzre and

stationary field windings

2. a rotating rectifier assembly, carried on the main
shaft line out

3. a controlled rectifier producing the d.c. field

voltage for the main exciter field from an a.c.
source (often a small 'pilot' exciter)

Greater immunity for such, systems is offered by

capacitively coupling the protection scheme to both ends
of the field winding, where brush or slip ring access is
~ossible(Figure 17.19(b)).

Hence, no brushes are required in the generator field

circuit. All control is carried out in the field circuit of the
main exciter. Detection of a rotor circuit earth fault is
still necessary, but this must be based on a dedicated . : - $ ~ ~ . ~ ~ $ ; ? ~
rotor-mounted system that has a telemetry link to
'a- .'
provide an alarmldata.


The low-frequency
injection scheme is also
advantageous in that the current flow through the field
winding shunt capacitance will be lower than for a
power frequency scheme. Such current would flow
through the machine bearings to cause erosion of the
bearing surface.
For power frequency schemes, a
solution is to insulate the bearings and provide an
earthing brush for the shaft.







.. j ..: : ,.. : ,. . ,,L..!.>.

. .





As detailed in Section 17.15 a shorted section of field
winding will result in an unsymmetrical rotor flux
pattern and in potentially damaging rotor vibration. ... b . . ..
& . -..: . ~etection
of such an electrical fault is possible using a -;.
probe consisting of a coil placed i n the airgap. The flux. -1: .-ct.. . . .
pittern i f t k positi"e and negative poles is me&ured". ,
2C .
and any significant difference in flux pattern between .".- PI
the poles is indicative of a shorted turn or turns.





Automated waveform comparison techniques can be

used to provide a protection scheme, or the waveform
can be inspected visually a t regular intervals. An
immediate shutdown is not normally required unless the
effects of the fault are severe. The fault can be kept
under observation until a suitable shutdown for repair
can be arranged. Repair will take some time, since it
means unthreading the rotor and dismantling the


L.F. injection

(a1 Low (rcqucn& a.c. voltage injection

- currcnt


Since short-circuited turns on the rotor may cause

damaging vibration and the detection of field faults for
all degrees of abnormality is difficult, the provision of a
vibration a detection scheme is desirable - this forms
part of the mechanical protection of the generator.




- 17-

A short-circuited diode will produce an a.c. ripple in the

exciter field circuit. This can be detected by a relay
monitoring the current in the exciter field circuit,
however such systems have proved to be unreliable. The
relay would need to be time delayed to prevent an alarm
being issued with normal field forcing during a power
system fault. A delay of 5-10 seconds may be necessary.



.. ,

, .,.,


Fuses t o disconnect the faulty diode after failure may be

fitted. The fuses are of the indicating type, and an
inspection window can be fitted over the diode wheel t o
enable diode health t o be monitored manually.

P. dinde that fails open-circuit occurs less often. I f there

is more than one diode i n parallel for each arm of the
diode bridge. the only impact is t o restrict the maximum
continuous excitation possible. If only a single diode per
bridge arm i s fitted, some ripple will be present on the
main field supply but the inductance of the circuit will
smooth this t o a degree and again the main effect is t o
restrict the maximum continuous excitation. The set can
be kept running until a convenient shutdown can be




17.I 5.5 i i c ' c



The need t o rapidly suppress the field o f Ti'machine in

which a fault has developed should be obvious, because
as long as the excitation is maintained, the machine will
feed its own fault even though isolated from the power
system. Any delay i n the decay o f rotor flux will extend
the fault damage. Braking the rotor i s n o solution,
because of its large kinetic energy.
The field - w i n d i n g current cannot b e interrupted
.;: - ~ n ' s t a n t a q e o u ias
l ~ i t fidws i n a highly inductive circuit.
: - Consequently, the flux energy must b e ' dissipated'.to
. ..i?
an excessive- inductive voltage rise i n the field
.: circuit. For machines of moderate size, it is satisfactory
to open the field circuit with a n air-break circuit breaker

without arc blow-out coils. Such a breaker permits only

a moderate arc voltage, which is nevertheless high
enough t o suppress the field 'current fairly rapidly. The
inductive energy is dissipated partly i n the arc and partly
in eddy-currents in the rotor core and damper windings.
With generators above about SMVA rating, i t is better t o
provide a more definite means o f absorbing the energy
without incurring damage. Connecting a 'field discharge
resistor' i n parallel with the rotor winding before opening
the field circuit breaker will achieve this objective. The
resistor, which may have a resistance value of
approximately five times the rotor winding resistance, is
connected by an auxiliary contact on the field circuit
breaker. The breaker duty is thereby reduced to that of
opening a circuit with a low L/R ratio. After the breaker
has opened, the field current flows through the discharge
resistance and dies down harmlessly. The use of a fairly
high value o f discharge resistance reduces the field time
constant t o an acceptably low value, though i t may still
be more than onc second. Alternatively. generators
fitted with static excitation systems may temporarily
invert the applied field voltage to reduce excitation
current rapidly to zero before the excitation system is

7.1 6 LcjsS i l F [_X::[';p,TigN


Loss of excitation may occur for a variety of reasons. i f . :

the generator was initially operating a t only 20%-3(&,,
o f rated power, it may settle t o run super-synchronously I
as an induction generator, at a low level of slip. I n doing !:
so, it will draw reactive current from the power system . ' 3,s
for rotor excitation. This form of response is particularlv ~ 2 z i
. .'(:I.,
true of salient pole generators. In these circumstances, ;$
the generator may be able to run for several minute- .<%without requiring to be tripped. There may be sufficient
time for remedial action to restore the excitation, but the '-:%.
reactive power demand o f the machine during the failure
may severely depress the power system voltage t o an {.$
unacceptable level. For operation at high initial power
output, the rotor speed may rise t o approximately 105% :$;
of rated speed, where there would be low power output
and where a high reactive current of up to 2.0p.u. may ,:$
..,,., .
be drawn from the supply.
Rapid .;automatic.'%
disconnection is then required t o protect the stator
windings from excessive current and t o protect the rotor :-'
from damage caused by induced slip frequency currents.





The protection used varies according to the s.ize of .i.2:
generator being protected.
. -. -+S+:. r!

On the smaller machines, protection again&.%
asynchronous running has tended t o be optional, but it$$
may now be available by default, where the functionality iz
,is available within a modern numerical generator.:;d!
protection package. I f fitted, it is arranged either to.:ii
provide an alarm or t o trip the generator. If the
generator field current can be measured, a relay element ;!;
can be arranged t o operate when this drops below a
preset value. However, depending on the generator
design and slze relative to the system, it may well be that
the machine would be required to operate synchronously
with little or no excitation under certain systemconditions.



The field undercurrent relay must have a setting below

the minimum exciting current, which may be 8% of that.
corresponding to the MCR of the machine. Time delay
relays are used t o stabilise the protection againsti
maloperation in response t o transient conditions and to!.
ensure that field current fluctuations due to pole slipping
do not cause the protection to reset.
I f the generator field current is not measurable. then the
technique detailed in the following section is utilised.

For generators above about SMVA rating, protection

against loss of excitation and pole slipping conditions is
normally applied.

Consider a generator connected to network, as shown i n

Figure 17.20. On loss of excitation, the terminal voltage
will begin to decrease and the stator current will increase,
resulting in a decrease of impedance viewed frEm the
generator terminals and also a change i n power factor.

The general case can be represented by a system of

circles with centres on the line CD; see Figure 17.21.
Also shown is a typical machine terminal impedance
locus during loss of excitation conditions.



A relay to detect loss of synchronism can be located at

point A. It can be shown that the impedance presented
to the relay under loss of synchronism conditions (phase
swingi~gor pole slipping) is given by:

!_. .

The special cases of EG=Esand EG=O result in a

straight-line locus that is the right-angled bisector of
CD, and in a circular locus that is shrunk t o point C,
When excitation is removed from a generator operating
synchronously the flux dies away slowly, during which
period the ratio of
is decreasing, and the rotor angle
of the machine is increasing. The operating condition
plotted on an impedance diagram therefore travels along
a locus that crosses the power swing circles. At the same
time, it progresses in the direction of increasing rotor
angle. After passing the anti-phase position, the locus
bends round as the internal e m f . collapses, condensing on
an impedance value equal to the machine reactance. 'The
locus is illustrated in Figure 17.21.



'The relay location is displaced from point C by the

. : ,... . .. .'. .
generator reactanceXG. One problem in determining the
position of these loci relative to the relay location is that
the value of machine impedance varies with the rate of :,:7:$,$$:i:G.:...... .-.., .
slip. At zero slip XG is equal to Xd, the synchronous ;>!:$@&?$$;I:~
reactance. and at 10090 slip XGis equal to X ' j , the sub- : ; : i ~ ~ t ~ ~ g ~ ; ~ .
transient reactance. The impedance in a typical case has
been shown to be equal to XId,the transient reactance, j:~?~$$+#~~:
at 50% slip, and to 2X; with a slip of 0.33%. The slip
likely to be experienced with asynchronous running is
. :&
.' &<$<:i

n i ~ g l cby 11~11icll
E(; 1cc~tl.cli,

If the generator and system voltages are equal, thcabove

expression becomes:


- (x, + x , 4 - ~ , ) ( 1 - , c o l O



' 9 9



.. .,..,.>?

.$;:f, 5;;;"..<-..:.:
.. ,.:,?.,';*.,:.':.-'."...
.. . ..,?....<,,,





f :'


low, perhaps 146, so that for the purpose o f assessing the

power swing locus it is sufficient to take the value



This consideration has assumed a single value for XG.

However, the reacranceXq on the quadrature axis differs
from the direct-axis value. the ratio of Xd/Xgbeing
known as the saliency factor. This factor varies with the
slip speed. The effect of this factor during asynchronous
operation is to cause XG to vary at slip speed. In
consequence, the loss of excitation impedance locus
does not settle at a single point, but it continues to
describe a small orbit about a mean point.-



A protection scheme for loss of excitation must operate


decisively for this condition, but its characteristic must

not inhibit stable operation of the generator. One limit
o f operation corresponds to the maximum cracticable
rotor angle, taken to be a t 120'. The locus of operation
can be represented as a circle on the impedance plane.
as shown in Figure 17.22. stable operation conditions
lying outside the circle.


I .













... ..


_ ._!.
. . . ..

.... . - ,. .

figvrc ;7.2;:




' '

s,.,:cc.,;,:<>"> ,?,:;<I:;,:,:


.,< .





. -.

On the~samediagram the full load impedance locus for

one per unit power can be drawn. Part of this circle
represents a condition that is not feasible, but the point
of intersection with the maximum rotor anqlc curve can
be taken as a limiting operating condition for setting
impedance-based loss o f excitation protection.

..:... .




ig+ ;







:$n .9

.. *+

1,;. -: [;.I;

: ,q. ,

.; !-\. - :
.. ..



. .

:. . ... ' '..


Figure 17.21 alludes t o the possibility that a protection

. - -....- -.. - --

Normal machinc opcraring impcdancc

Locus of conyant MVA



scheme for loss o f excitation could be based on

impedance measurement The impedance characteristic
must be appropriately set or shaped to ensure decisive
operation for loss of excitation whilst permitting stable
generator operation within allowable limits. One or two
offset mho under impedance elements (see Chapter. 11
for the principles of operation) are ideally suited for
providing loss of excitation protection as long as a
generator operating a t low power output (20-30%Pn)
does not settle down to operate as an induction
generator.The characteristics o f a typical two-stage loss
of excitation piotection scheme are illustrated i n Figure
17.23. The first stage, consisting of settings X,, and Xbl
can be applied to provide detection of loss of excitation
even where a generator initially operating at low power
might settle dowv to operate as an
output (20-30%P,]
induction generator.

... .

- ---- .-


Pick-up and drop-off time delays t d l and tdo, are

associated with this impedance element. Timer t d , is
used to prevent operation during stable power swings
that may cause the impedance locus o f the generator to
transiently enter the locus of operation set by Xbr
However, the value must short enough to prevent
damage as a result of loss of excitation occurring. If
pole-slipping protection is not required (see Section
17.17.2). timer tdo, can be set to give instantaneous
reset. The second field failure element, comprising
and associated timcrs IdI and tdo2can
settings X b l ,
be used to give instantaneous tripping follovring loss of
excitation under full load conditions.

1 /.I


~ : O : C f ' ~ ~ 2 : 1S~l:!r,v:

The typical setting values for the two elements vary

according to the excitation system and operating regimf
of the generator concerned, since :hew affect th(
generator impedance seen by the relay under normal an1
abnormal conditions. For a generator that is neve




page 301

*crated at leading power factor, or at load angles i n

e s s of 90" the typical settings are:

.! 7:!?.1

impedance element diameter Xbl = Xd

impedance element offset X,! = -0.5%

time delay on pick-up, t d r = 0.5s .- 10s

time delay on drop.-off, tdo, =


lf a fast excitation system is employed. allowing load

of up to 120" to be used, the impedance diameter
must be reduced to take account of the reduced
generator impedance seen under such conditions. The
offset also needs revising. In these circumstances,
typical settings would be:
pedance element diameter Xb, = 0.5Xd
impedance element offset X,
time delay on pick-up,



= 0.5s - 10s

time delay on drop-off, tdol = 0s

fietypical impedancesettings
for the second element, if
impedance element diameter
xb2 =

Protectio:~usinq Rc-VC~SC
Pov:r: Eic!-ynt

During pole-slipping, there will be periods where the

direction of active power flow will be i n the reverse
direction, so a reverse power relay element can be used
tn detect this, i f not used for other purposes. However,
since the reverse power conditions are cyclical, the
element will reset during the forward power part of the
cycle unless either a very short pick-up time delay andlor
a ,,itabledrop-off time delay is used to eliminate

The main advantage of this method is that a reverse

power element is often already present, so no additional
relay elements are required. The main disadvantages are
the time taken for tripping and the inability to control
the system angle at which the generator breaker trip
command would be issued, if i t is a requirement to limit
the breaker current interruption duty. There is also the
Determination of settings in the field, from a deliberate .
pole-slipping test is not possible and analytical studies
may not discover all conditions under which poleslipping will occur.

. : ~ . ~ < ~ ~ 2 . . ~ . y ; . . .


' j '.:+!(:.

, ,




. . . . .


:. : ,-,.!-!.,~~,~:::.,.;:? k ;:..,.:i





- . Q


. .

Protection can be provided using several methods. The

choice of method will depend on the probability of pole
occurring and on the consequences should it

.. .

. .

With reference to Figure 17.21. a !ass of excitation und&

impedanke characteristic may also be capable - o f
detecting loss of-synchronism, in applications where the
electrical centre of the power system and the generator
lies 'behind' the relaying point This would typically be
the case for a relatively small generator that is
connected to a power transmission system (XG >> (XT +
Xsll- With reference to Figure 17.23; i f pole-slipping
protection response is required. the drop-off timer tdorof
the larger d ~ ~ m e t eimpedance
measuring element
should be set to prevent its reset of in each slip cycle,
until the rdl trip time delay has expired.

The time delay settings i d 2 and tdO2 are Set to zero to give
instantaneous operation and reset.

A generator may pole-slip, or fall out of synchronism

with the power system for a number of reasons. The
principal causes are prolonged clearance of a heavy fault
on the power system, when the generator is operating at
a high load angle close to the stability limit, or partial or
Complete loss of excitation. Weak transmission links
between the generator 2nd the bulk of the Power system
aggravate the situation.
I t can also occur with
embedded generators running in parallel with a strong
Utility network if the time for a fault clearancean the
slow. perhaps because only lDMT relays
are provided. Pole slipping is characterised by large and
rapid oscillations in active and reactive power. Rapid
of the generator
the network is
required to ensure that damage to the generator is
avoidedand that loads supplied the
affected for very long.


- . =.-*
r : 2.:.






. pi







- 17.

with reverse power protection, this would be an

elementary form of pole-slipping protection. I t may not
be suitable for large machines where rapid tripping is
required during the first
slip cycleand where some
controlis required for the system angleat which ,the
generator circuit breaker trip command is given. Where
protection against pole-slipping must be guaranteed, a
more sophisticated method of protection
should be used.
A typical reset timer delay for pole-slipping protection

... .

might be 0.6~. . F generator

~ ~
transformer units, the
.. , '. ....
additional impedance infront of the relaying
point may
take the system impedance outside the under impedance ..::.: :~~:~~-.'~=!:;::'::.
. ..: .,
. _ ..
relay characteristic required for loss of excitation ~.~..:~;.-;:~<;:~.,
. . :.... *
protection. Therefore, the acceptability of this pole- . . .
... . . :. -,....
slipping protection scheme will be dependent on the
. .
. .. . . ,.. .. . . .
.: -:'.".:.:




.:" .........





.1. .::; .
.. ;,.. ......,.
-..::.: ....y8.s. .
." ~,':::~.:,~:,~~:.'.




_ _ -. .. .

. .

A more sophisticated approach is tb measure the

impedance of the generator and use a lenticular
impedance characteristic to determine i f a pole-slipping
condition e~ists.The lenticular characteristic is shown i n
Figure 17.25. The characteristic is divided into two haives
by a straight line, called the blinder.

Large generator-transformer units directly connected to

grid systems often require a dedicated pole-slipping
protection .scheme to ensure rapid tripping and with
system angle control. Historically, dedicated protection
schemes have usually been based on a n ohm-type
impedance measurement characteristic.





The inclination, 6, of the lens and blinder is

the angle of the total system impedance. The impedance ;::
of the system and generator-transformer determines the .:..
forward reach of the lens, ZA,and the transient reactance i :
o f the generator determines the reverse reach ZB.

Although a mho type element for detecting the change

i n impedance during pole-slipping can be used in some
applications, b u t with performance limits, a straight line
ohm characteristic is more suitable. The protection
principle is that o f detecting the passage of the
generator impedance through a zone defined by two
such impedance characteristics, as shown in Figure
17.24. The characteristic is divided into three zones, A,
B, and C. Normal operation o f the generator lies i n zone
A. When a pole-slip occurs. the impedance traverses
zones B and C, and tripping occurs when the impedance
characteristic enters zone C.



Ohm relay 1

Tripping only occurs i f all zones are traversed

sequentially. Power system faults should result i n the
zones not being fully traversed so that tripping will not
be initiated. The security of this type of protection
scheme is normally enhanced by the addition of a plain
under impedance control element (circle about the origin
of the impedance diagram) that isset t o prevent tripping



. .. . .. . .

. . .

for impedance trajectories for remote power system

faults. Setting of the ohm elements is such that they lie
parallel t o the total system impedance vector, and
enclose it, as shown in Figure 17.24.





Page 303



windings and t o issue an alarm or trip t o prevent

Although current-operated thermal replica protection
cannot take i n t o account the effects o f ambient
temperature or uneven heat distribution, it is often
applied as a back-up .to direct stator temperature
measuring devices t o prevent overheating due t o high
stator current. With some relays, the thermal replica
temperature estimate can be made more accurate
through the integration of direct measuring resistance
temperature devices.
Irrespective of whether current-operated thermal replica
protection is applied or not, it is a requirement t o
monitor the stator temperature of a large generator i n
order t o detect overheating from whatever cause.
..1 c j ~ ~ : ~
:., ; ~ ~ ;
c..,-. :.-:>-.:,
Tempe~aturesensitive elements, usually of the resistance
... 0 ,:;.
type, are embedded i n the stator winding a t hot-spot ..- +& :'3',
locations envisaged by the manufacturer, the number
.used being sufficient to cover all variations. The
- 2
elements are connected to a temperature sensing relay
element arranged to provide alarm and trip outputs. The
settings will depend on the type o f stator winding
insulation and on its permitted temperature rise.



I f the impedance locus lies above line PP', the swing lles
far out i n the power system - i.e. one part o f the power
system, including the protected generator. is swinging
against the rest of the network Tripping may still occur,
but only if swinging is prolonged - meaning that the
power system is i n danger of complete break-up Further
confidence checks are introduced by requiring t h a t the
~mpedancelocus spends a minimum tlme withln each
zone for the pole-slipping condition t o be valid. The trip
signal may also be delayed for a number o f slip cycles
even i f a generator pole-slip occurs - thls IS t o both
provide confirmation of a pole-slipping condition and
allow time for other relays to operate i f the cause of the
pole slip lles somewhere In the power system. Should
the impedance locus traverse the zones i n any other
sequence. trlpping IS blocked.


Various faults may occur on the mechanical -side o f a

generating set. The following sections detail the m i r e
important ones from an electrical point of view.

When a generator operating in parallel with others loses

its power input, it remains in synchronism with the
system and continues to run as a synchronous motor.
drawing sufficient power t o drive the prime mover. This
condition may not appear to be dangerous and i n some
circumstances will not be so. However, there is a danger
of further damage being caused. Table 17.1 lists some
typical problems that may occur.

Overheating of the stator may result from:

Protection is provided by a low forward powerlreverse

power relay, as detailed in Section 17.11

ii. failure o f the cooling system

iv. core faults

The speed of a turbo-generator set rises when the steam

input is in excess of that required to drive the load at
nominal frequency. The speed governor can normally
control the speed, and, i n any case, a set running i n
parallel with others i n an interconnected system cannot
accelerate much independently even i f synchronism is
lost. However. i f load is suddenly lost when the HV
circuit breaker is trippcd, thc set will begin t o accelerate

Accidental overloading might' occur through the

Combination o f full active load current component,
governed by the prime mover output and an abnormally
high reactive current component, governed by the level

a modern protection relay, it is ielatively simple to

Provide a current-operated thermal replica protection
element to estimate the thermal state of the stator

-. .................


:.- 2

.. f l

I ~hap17-280-315


10: 50

Page 304

rapidly. The speed governor is designed t o prevent a

dangerous speed rise even with a 100% load rejection,
but nevertheless an additional centrifugal overspeed trip
device is provided t o initiate an emergency mechanical
shutdown i f the overspeed exceeds 10%.
To minimise overspeed on load rejection and hence the
mechanical stresses on the rotor, the following sequence .
..:. . .:. .I . is used whenever electrical tripping is not urgently
- required:


i. trip prime mover or gradually reduce power input t o


ii. allow generated power to decay towards zero

event of loss of vacuum, as this would cause rapid

overheating of the low-pressure turbine blades.

GEFJERAioz pr<(jiEn!Q>:

17-20 CO>JP;flE

From the preceding sections, it is obvious

protection scheme for a generator has t o take account of
- -h
many possible faults and plant design variations. :;SEzJ
Determination of the types of protection used for a .,
particular generator will depend on the nature of the
plant and upon economic considerations, which i n turn :;$$$
is affected by set size. Fortunately, modern, multi- ,'-?$
function, numerical relays are sufficiently versatile to
include all of the commonly required protection
functions i n a single package,. thus simplifying the .{.$$
decisions to be made. The following sections provide
illustrations of typical'protection schemes for generators ,:j.$$:
'connected t o a grid network, but not all possibilities are ':$:
illustrated, due to the wide variation in generator sizes i$$:
and types.





iii. tripgenerator
circuitbreaker onlywhen


power is close t o zero*or when t h e power flow

starts t o reverse, t o drive the idle turbine -










A failure of the condenser vacuum i n a steam turbine

driven generator results i n heating of the tubes. This
then produces strain i n the tubes, anJ a rise i n
temperature o f the low-pressure end of the turbine.
Vacuum pressure devices initiate progressive unloading
of the set and, if eventually necessary, tripping of the
. turbine valves followed by the high voltage circuit
breaker. The set must not be allowed t o motor i n the



. .. . . .


.. . .. . .

. ... :


.... ..


A typical protection scheme for a direct-connected :;'$;>

generator is shown i n Figure 17.27. It comprises the:::@
.,: =:e
following protection functions:
- .~
... . .:
. .:__:
. =rv
. . . . . . . -:>;.

5 , .


Elcctr~caltrlp of govcmor

Emcrgcncv push b u ~ o n


Stator diffcrcntial (biascdlhigh




E/F (or ncutral voltagc

Back-up ovcrcurrcnt lor voltagc

dcpcndcnt OICI

- 0
Stator winding tcmpcraturc


Unbalanced loading

Low powcr



Mechanical fatrlts ( n o n - u r g c n t l ~

N.8. Alarms and I ~ m cdclays arnittcd for simplicity




Page 3 0 5


1. stator differential protection

2. overcurrent protection

conventional or voltage

3. stator earth fault protection

4. over;o!:agc

instantaneous electrical trip and which can be time

delayed until electrical power has been reduced t o a low
value. The faults that require tripping o f the prime mover
as well as the generator circuit breaker are also shown.


5. undervoltage protection
6. overloadllow forward power1 reverse.-power
protection (according t o prime mover type)

7. unbalanced loading
8. overheating
9. pole slipping
10. loss of excitation

'11. underfrequency
12. inadyertent energisation
13. overfluxing
14. mechanical faults

These units are generally of higher output than directconnected generators, and hence more comprehensive
protection is warranted. In addition, the generator
transformer also requires protection,. for which the
protection detailed in Chapter 16 is appropriate
Overall biased generatorlgenerator transformer
differential protection is commonly applied i n addition,
or instead of, differential protection for the transformer
alone. A single protection relay may incorporate all of
the required functions, or tbe protection o f the
transformer (including overall generatorlgenerator
transformer differential protection) may utilise a
separate relay.


. -.... ,.:


. .

frequency and voltage, or for other reasons.

From a Utility standpoint, the connection o f e
generation may cause problems w i t h voltage co
increased fault levels. The settings f o r protecti
in :he vicinity cf the ;!ant may require adjustment-with
the emergence of embedded generation. It must also be
ensured that the safety, security and quality o f supply of
the Utility distribution system is n o t compromised. The i
embedded generation must n o t be permitted t o supply .'
any Utility customers i n isolation, since the Utility supply
is normally the means of regulating the system voltage';
and frequency within the permitted limits. It aiso ;
normally provides the only system earth connection(s], to :.
ensure the correct performance o f system protection in
response to earth faults: If the Utility power infeed fails,
it is also important t o disconnect the embedded
generation before there is any risk of the Utility power':
supply returning on t o unsynchronised machines. In,
practice this generally requires the following protection
functions t o be applied a t the 'Point o f Common 'j
Coupling' (PCC) t o t r i p the coupling circuit breaker:

In recent years, through de-regulation o f the electricity

supply industry and t h e ensuing commercial
competition, many electricity users connected to MV
power distribution systems have installed generating st%
to operate i n parallel with the public supply. The
intention is either t o utilise sur.?lus energy from other
sources, or t o usewaste heat or steam from the prime
mover for other purposes. * Parallel connection o f
generators t o distribution systems did occur before deregulation, but only where there was a net power import
from the Utility. Power export t o Utility distribution
systems was a relatively new aspect. Since generation o f
this type can now be located within a Utility distribution
system, as opposed t o being centrally dispatched
generation connected t o a transmission system, the term
'Embedded Generation' is often applied.- Figure 17.2
illustrates such an arrangement. Depending on size, the
embedded generator(s) may be synchronous or
asynchronous types, and they may be connected a t any
voltage appropriate to the size o f plant being considered.
The impact o f connecting generation t o a Utility
distribution system that was originally engineered only
for downward power distribution must be considered,
particularly i n the area of protection requirements. I n
this respect, it is not important whether the embedded
generator is normally capable of export t o the Utility
distribution system or not. since there may exist fault
conditions when this occurs irrespective o f the design


a. overvoltage



;.;.,,. . , .

.?. .-.>,A:
. ,

..:. .
. ..-


.. >+.A i



..-., .. . .

Limits may be placed by the Utility on the amount of

powerlreactive power importlexport. These may demand
the use o f an in-plant Power Management System t o
control the embedded generation and plant loads
Some Utilities may insist on automatic
- .
tripping o f the interconnecting circuit breakers i f there is
a significant departure outside permissible levels of



b. undervoltage


e. loss o f ~ t i l i t ~ s .u.:, ~ .~ l ~. .

In addition: 'partichiar ci;cumstances

. - .:



..:-. ...i;g


may require:$

additional protection functions:

I f plant operation when disconnected from the Utility

supply is required, underfrequency protection (Section
17.4.2) will become an important feature of the in-plant
power system. During isolated operation, it may be
relatively easy to overload the available generation, such
that some form of load management system may be
required. Similarly, when running i n parallel with the
Utility, consideration needs t o be given t o the mode of
generator operation if reactive power import is to be
controlled. The impact on the control scheme of a
sudden break in the Utility connection t o the plant main
busbar also requires analysis.
Where the in-plant
generation is run using constant power factor or
constant reactive power control. automatic reversion t o
voltage control when the Utility connection is lost is
essential t o prevent plant loads being subjected to a
voltage outside acceptable limits.



f. neutral voltage displacement

g. reverse power
h. directional overcurrent
I n practice, it can be difficult t o meet the protection
settings or performance demanded by the Utility without
a high risk o f nuisance tripping caused by lack of COordination w i t h normal power system faults and
disturbances that do not necessitate tripping o f the
embedded generation. This is especially true when
applying protection specifically to detect loss of the
Utility supply (also called 'loss of mains') to cater for
operating conditions where there would be no
immediate excursion i n voltage or frequency to cause
operation o f conventional protection functions.

> I . : / : , I P ! o i r c ~ i o nI'.,(;;~inst Los5 ::I Il:ili.:v Sv;)~!y

If the normal power infeed to a distribution system, o r t a

the part of it containing embedded generation is lost, the
effects may be as follows:
a. embedded generation may be overloaded, leadin!
t o generator undervoltage/underfrequency




Page 307

embedded generation.may bc underloadcd, leading

to ove~oltage/overfrequency



Ee;ay Description

A ROCOF relay detects the rate o f change of frequency i n

excess of a defined setpoint. The signal is obtained from
a voltage transformer connected close t o the Point of
Common Coupling (PCC). The principal method used is
to measure the time period between successive zerocrossings to determine the average frequency for each
half-cycle and hence the rate of change of frequency.
The result is usually averaged over a number of cycles.

little change to the absolute levels of voltage or

frequency if there islittle resulting ,-hange to t h e
load flow through the PCC
rst t w o effects are covered by conventional voltage
equency protection. However, if condition (c)
conventional protection may not detect the loss
ty supply condition or it may be too slow t o do so
the shortest possible auto-reclose dead-times

. .



ay be applied i n association with Utility overhead

otection. Detection of condition (c) must be
d if the requirements of the Utility are to be met.
ossible methods have been suggested, but the
t often used is the Rate of Change of Frequency
relay. - I t s application is ba;ed on the fact that
of change o f small changes i n absolute
, in response to inevitable small load changes,
ter with the generation isolated than when the
is i n parallel with the public, interconnected
However, problems w i t h nuisance
tional power system events.




. %


~ ; : ! ; itj~

A voltage vector shift relay detects the drift i n voltage

phase angle beyond a defined setpoint as long as it takes
place within a set period. Again, the voltage signal is
a voltage transformer connected 'loseto
the Point o f Common Coupling (PCC). The principal
method used is to measure the' time period between
successive zero-crossings to determine the duration o f
each half-cycle, and then to
the durations
of earlier
half-cycles in
the memorised average duration
order to determine the phase angle






;. . . .:


following the loss of a large generator.or a .

r interconnector, have occurred.
ularly true for geographically islanded power
h as those of the British Isles. An alternative
otection is a technique sometimes referred
vector shifts protection.
lnthis technique
ase ,-hange between the directly
mpared with a memorised ax.

sho;ld loss
of the Utility supply occur, .it is
. - & - , .; c .
"nlikely that there will be an exact match between'the 5': o, : - - -...
:; .
output o f the. embedded generator(s) and the connected -.?,. o, -.:load. A small frequency change or voltage bhase angle
change will therefore occur, t o which can be added any
changes due t o the small natural variations i n loading o f
an isolated generator with time. Once the rate of change
of frequency exceeds the setting of the ROCOF relay for
a set time, or once the voltage phase angle drift exceeds
the set angle, tripping occurs t o open the connection
between the in-plant and Utility networks.

al form of earth fault protection

may also be demanded to prevent the backfeed of an
earth fault by embedded generation. The.only Way of
detecting an
neutral voltage displacement protection. The additional
requirement i s only likely t o arise for embedded
generation rated above 15OkVA. since the risk of the
m a l l embedded generators not being cleared b y other
means is negligible.

it is possible to estimate
the rate of change of

frequency from knowledge o f the generator set inertia

and MVA rating, this is not an accurate method for
setting a ROCOF relay because the rotational inertia of
the complete network being fed by the embedded
generation is required. For example, there may be other
~ generators
t o, consider. As a result, it is
invariably the case that the relay settings are determined
at site during commissioning. This is to ensure that the
Utility requirements are m e t while reducing the
possibility of a spurious trip under the various operating
scenarios envisaged. However, i t is very difficult t o .
determine whether a given rate of change o f frequency
will be due to a 'loss of mainsq incident o r a
load/frequeno/ change on the public power network, and
hence spurious trips are impossible to eliminate. Thus
the provision of Loss of Utility Supply protection t o meet
power distribution U t i l i t y interface protection




-. . . . . . . . . . . .




- J07



.: . . '
:'.;\ ,i<.......: ...,i:








.'..<.'.':.''. ,:.. ;.=;

. . .

-" ...
:... -r
. .:..
. .

...."....< + ..:'.





-. . . . . . . . . . . . . .

d generation are n o t normally

s a potential safety hazard. I n the
system earth fault, the Utility
rate t o remove the Utility power
Id also re&lt in removal of the
, through the action o f the
ency protection and dependable
n. H
~ in view
~ of safety
considerations (e.g. fallen overhead line conductors in


. .

. .. .% . = . .. . : .


;.-. . .



. ... : . . .
:+:.. . . .. . . .
:,.. . . :
. :

requirements, may actually conflict with the interests o f

the national power system operator. With the growing
contribution of non-dispatched embedded generation t o
the aggregate national power demand, the loss o f the
embedded generation following a transmission system
incident tnai may aireaiiy ciialleilge the security-of the
system can only aggravate the problem. There have been
claims that voltage vector shift protection might offer
better security. but it will have operation times that vary
with the rate o f change of frequency. As a result,
depending on the settings used, operation-times might
not comply with Utility requirements under all
circumstances. Reference 17.1 provides further details
of the operation of ROCOF relays and the problems that
may be encountered.

Salient details o f the generator, network and protection

required are given i n Table 17.2.
The example
calculations are based on a MiCOM P343 relay in respeq
o f setting ranges, etc.
. . .,


, ,


. . . . ... .. . .


Lw : PF

. -c--"-.,




;::: / 6250 1 ,5000 1 0.8 j
. . ..,-;
....... . .,..... !





.. .. . .



Gcncrator hlpc
hlicm ~


... ,......
.. ..-.


. 3. v. ,.

.,....,. ..,,..
. .,...












~t normal voltage, the current setting must be greater

Primc Movcr




: 1500 !Stcam Turbinc



a. . . ,



Maximum carth ! Minimum pharc

faull currcnl
fault currcnl

/ . 31.7.n >:I,';:

,, . . . .




Maximum downslrcam
p h a faull


, .2,,<e:Js:.!



The nearest settable value is 3 6 5 4 or 0.731,.


~.,,.:F,>,,,;g<:.: ?r' 2:
", ".'.,>
. i.
. .. -..

This protection is applied as remote backup to the

downstream overcurrent protection i n the event of
protection or breaker failure conditions. This ensures
that the generator will not continue to supply the fault
under thew conditions.

than the maximum generator load current of 328A. A '1

margin must be allowed for resetting of the relay at this ';j
current (reset ratio = 950101 and for the measurement 5
;~,+: :5~F.k35r..~,~r!
tolerances of the relay (Solo o f r, under reference :;
X; p . ~ .
CI Ratio
VT Ratio
: 11000,110 ; conditions), therefore the current setting is calculated as: ,.:



..-, ,. . . .



. ,~
. ... ...,.. :-.,. .....


This section gives examples of the calculations required

for generator protection. The first is for a typical small
generator installed o n an industrial system that runs in
parallel with the Utility supply. The second is for a larger
generator-transformer unit connected to a grid system.

. z






. . . . . . . . . . . . >.... :' 'i. ..:.


Biased differential protection involves the determinatio

of values for four setting values: I,,. ISz, K I and Kz in :;$
Figure 17.5. I;., can be set a t 5OIoof the generator rating, .@
in accordance with the recommendations for the relay, j:"
and similarly the values of I,, 1120010) and K2 (15Wo) of ;
generator rating. It remains for the value of K , to be
determined. The recommended value is generally @h, :;
but this only applies where CT's that conform t o IEC
60044-1 class PX (or the superseded BS 3938 Class X] /j:#$
are used - i.e. CT's specifically designed for use in
differential protection schemes. In this application, the
CT's are conventional class 5P CT's that meet the relay
requirements in respect of knee-point vcltage, e t c $3
Where neutral tail and terminal CT's can saturate at :!..:-&:
different times due to transiently offset magnetising
inrush or motor starting current waveforms with an r.m.s. :;(?':
level close t o rated current and where there is a high l j R .$&
time constant for the offset, the use of a 0% bias so
l pe;':&
may give rise to maloperation. Such waveforms can k :,$$
encountered when plant o f . similar ..rating t o . the~22~.~$
generator i s being energised or staited: Differen&;yX$
between CT de'sibns'ordiffering remanent fluilevels d":;I%
lead to asymmetric saturation and the producticin
differential spill current. here fore, it .is appropriate t o %
select a non-zero setting for K , , and a value of 5% isy@
usual i n these circumstances.

Nevertheless. because such protection is a common

requirement of some Utilities, the 'loss o f mains'
protyction may have to be provided and the possibility of
spurious trips will have t o be accepted in those cases.
Site measurements over a period of time of the typical
rates of frequency change occurring may assist in
negotiations of the settings with the Utility, and with the
fine-tuning of the protection that may already be

.. '



Eanh Faull Scllingr

j 20011 j"'


;144 1






- ' ..Z
<-.,s .... :.


1 T#rl~lcI?.?:

........ -

Oslo I:,,:.:r:::'.

~ : c ~ c ~ ' : l,?,,::r+.:*o~



: 48A

j 0.15

... .;


The minimum phase-phase voltage for a close-up single- ..

phase to earth fault is 57%. so the voltage setting V,
must be less than this. A value of 3010 is typically used, .giving V, = 33V. The current setting multiplying factor.

K must be chosen such that Kls is less than 50%

of the
generator steady-state current contribution to an
uncleared remote fault. This information is not available

an operation time of not less than 1.13s. At a TMS of 1.0,

the generator protection relay operating time will be:

protection such that:




L ..

Vrfl= eflective voltage setting



dorunstrearn earth-fault current setting

eaflhing resistarlce


Hence a setting of 48V is acceptable. Time grading is

required, with a minimum operating time of the NVD
protection of 1.13s at an earth fault current of 200A
Using the expression for the operation time of the NVD




Providing protection for 90% of the winding.


voltage seer1 by relay

= relay setting voltage

. .

?<,.+,. :...





:;.,71.7.5 loss


Page 3 1 0

(<;.r.;rc: lisi: s;-.;.;i.;ifir:

quantities (corresponding t o
voltage) is typically used, with
'10s t o allow for transients
offlrejection, overvoltages on
motor starting, etc.

Loss of excitation is detected by a mho impedance relay

element, as detailed i n Section 17.16.2. The standard
settings for the P340 series relay are:
X,= 0 . 5 ~x' (0
~ ratio/VT ratio)

The second element provides protection in the event of;

large overvoltage, by tripping excitation and th;
generator circuit breaker (if closed]. This must be set
below the maximum stator voltage possible, taking
account saturation. As the open circuit
the generator is not available, typical values must bec@
used. , Saturation will normally limit the maximum"'""
overvoltage on this type of generator to 130%, so a.
setting of 120% (132V secondary) is typically used:'
Instantaneous operation is required.
manufacturers are normally able t o provide'
recommendations for the relay settings. Far embedded
ger:erators, the requirements of the local Utility may a1
have to be taken into account. For both elemen&,
variety of voltage measurement modes are available t$@
take account of possible VT connections (single or three:@
phase, etc.], and conditions to be protected against I"
this example, a thee-phase VT connection is used, aid
overvoltages on any phase are t o be detected, so.'?
selection of 'Any' is used for this setting.

(in secondary quantities]

= -0.5 X 0.297 x 19.36 x 500/100


Xb = Xdx (CT r a t i o m ratio)


= 2.349fl x 19.36 x (500/100)

= 227Q

The nearest settings provided by the relay are X, = 14.552 Xb = 22752. The time delay t d l should be set t c
avoid relay element operation on power swings and a'
typical setting of 3s is used. This value may need t o be
modified i n the light of operating experience. To prevent
cyclical pick-up of the relay plement without tripping.
such as might occur during pole-slipping conditions, a
drop-off time delay td,, is provided and set t o 0.5s.
. ,,,.,,..... :.,:
:.:+., (<.
= < ,, .;:.::s:c
! < .:



This protection is required t o guard against excessive

heating from negative phase sequence currents. whatever
the cause. Thegentrator, i s o f salient-poledesign, so from
withstand is-8qo of rating
IEC '6&34-1;-.the'&ntinuok
and the ~ : r w l u e ii:20s. . ~sin<.G~uatTon
17.i, the
required relay settingsmn found as f2>>,=0.05 and K =
8.6s. The nearest available values are I,>> = 0.05 and
K = 8.6s. The relay also has a cooling time constant
K,,,,,that is normally set equal t o the value o f K. To coordinate with clearance o f heavy asymmetric system
faults, that might otherwise cause unnecessary operation
of this protection, a minimum operation time tmi, should
be applied. I t is recommended t o set this to a value of 1.
Similarly, a maximum time can be applied to ensure that
the thermal rating o f the generator is not exceeded (as
17,. ,:, this is uncertain, data not available) and to take account
of the fact that the P343 characteristic is not identical
6c;$&$$$i with that specified i n .lEC 60034. The recommended
.: ,..,".&.k%,~.
setting for t,,, is 600s.
. !..,. :.~,.'',
. . . .'.'.:f.:
- .. :. .,. <:.;. . . . . .
....,.. ."
: ;.


~ h i s ' i siequiied t o protect the generator from s;&i$$$

overloid conditions during periods df op'eration.isii~tdh
from the Utility supply. The generating set manufacturer#
will normally provide the details o f machine short-tim$$
capabilities. The example relay provides four stages
underfrequency protection. In this case, the first stageis!
used for alarm purposes and a second stage would b$,
applied to trip the set.



The alarm stage might typically be set to 49Hz, w i t h a

time delay of 20s, to avoid an alarm being raised unds
transient conditions, e.g. during plant motor starting;
The trip stage might be set to 48Hz. with a time delay of
0.5s, to avoid tripping for transient, but recoverable, dips
in frequency below this value.



1 .


.. ..


..; . - .





:j:..:. ,.:, ,


..:: :.

The relay setting i s 5% of rated power.

This is required to guard against various failure modes,

e.g. AVR failure, resulting in excessive stator voltage. A
two-stage protection is available, the first being a lowset time-delayed stage that should be set to grade with
transient overvoltages that can be tolerated following
load rejection. The second is a high-set stage used for'
instantaneous tripping in the event of an intolerable
overvoltage condition arising.





500 X 100



This value can be set in the relay. A time deb;;

to guard against power swings while genera$i
at low power levels, so use a time delay of 5s. NO ..::
timedelay isrequired.


en era tors can normally withstand 105% of rated

voltage continuously. SO the low-set stage should be set
higher than this value. A sctting of 117.7V in secondary


... ,
. .i. . . D
. . ;:,..
. . ,... ,:

. J10









: :.!





Page 3 1 1

Gcncrator MVA rating


_ ..


....... ,
i Gcncrator voltagc
( r..........,..........-....






~ ~ ~ ~ ! , e ~ F : ~ 2 ~ F ! ! ~ i ;

Oirccl-axis Uansicnt rcactancc

!8 ' .~ i n i G " & ;'@ti"i+li?&




I Gcncrator ncgatiw xqucncc capability

:.. ~ .

Loss of excitation

Id 1











. .






. ,.

. : : . a

Gcncntormotoring powcr

Gcncrator ovcwoltagc



Gcncrator third harmonic vol

i z ? ~ ' ::


.-._ ...........




Gcncrator ncgativc xqucncc factor, Kg



,I: . : ~ ~ ~ : ~ ~'1 ~;,k.i:": :-:,\, : , ~ ~ ~ " :


Gcnuator ~ n d c w o l t y c


Max polc slipping frcqucncy

Gcncntor trhsformcr rating





Ncgatiw phase xqucncc

. I




Gcncrator tnnsformcr lcaka~crcactancc




\ '


Gcrlcrator transformer ovcdux alarm


Nctwohrcristanct (rcfcncd to 18W


Nctwoh rcactancc (rcfcrrcd to 18kW


V, mcas modc

Gcncrator transfoicr o w d u x alarm

F<l tim; dclay

F<Z wtting


F<2 timc
- .-....
- . . . . . . . . . .0.5s
PI h a r t i o n
; lrvmc powcr
-.-.-.-. . . . . - .

PI tirnc dclay



PI DO limc



: I


- r


- 5








; lRWo/lZO
. .

. . i~


.... .!








,:I. w '::



: pu


Gcncrator VT ratio


1 .

Minimum load rcsistancc

j Numbcr o f gcncratas i n parallcl



Systcm impcdancc anglc (cstirnatcdl

Gmcrator CT'ratio


has t o be greater than the fullThe setting current

load current o f the generator (6019A). A suitable margin
must be allowed for operation a t reduced voltage, so use
a multiplying factor of 1.2. The nearest settable value is
72OUA. The factor K is calculated so that the operating
current is less than the current for a remote end three
phase fault. The steady-state current and voltage a t the
generator for a remote-end three-phase fault are given
by the expressions:




- 17-



The data for this unit are given i n Table 17.4. It is fitted
with t w o main protection systems t o ensure security of
$ : ' tripping i n the event of a fault. To econornise on space.
I.- .
the setting calculations for only one system, that using a
MiCOM P343 relay are given. Settings are given i n
P r h a r y quantities throughout.



VN =
The settings follow the guidelines previously stated. AS
1OOoIo stator winding earth-fault protection is provided.
,,.high sensitivity is n o t required and hence J,, can bc set
;::"' to 1O0Io of generator rated current. This equates t o 602A.
and the nearest settable value on the relay is M O A I=
0.08 of rated CT current]. The settings for K , . I,, and K,
f0ll0w the guidelines i n t h e relay manual.





110-load p h n r c - ~ r e u t r a gl o l c r a t o r volrngc

= golerator

d-aris synchronous rcoctance




~~:,';;,..-r:.?~:~. .;..,,, ,.-...- .......






A TMS value of 10 is selected, t o match the withstand


curve supplied by the manufacturer.

IJlr = 2893A

;;,72,7,6 ':()L?<b



I ' + I x , + ~ ~1')
---f i I ' ~+ I X~~ + X~, + j~2 ~ X ~



This is provided by a combination of neutral voltage

dispiaceiiicnt ai;d
:bird h8rmnnic undewoltagc
For the neutral voltage displace
protection to cover 90% o f the stator winding,
minimum voltage allowing for generator operation
minimum of 92% o f rated voltage is:












. . . . .



. . . . . .

.. ...

The generator has a maximum steady-state capability of

8010 of rating, and a value o f 1;9 of 10. Settings of I,,,,,,
= 0.06 (=480A) and I;, = 10 are therefore used.
Minimum and maximum timedelaysof 1s and 1300s are
used to co-ordinate with external protectiori and ensure
tripping at low levels of negative sequence current are


. ..

.., ..




. .. .

- c..........



The generator-transformer manufacturer supplied the

following characteristics:

. ... . .. , .


...... ;


> I .J

v/$ > J . ~ , ~ ~ Ili!i~c

, c ~cliaracfc"slic

+ .......


;. ; p y<;?


/ ,:


. .<":
.....I.-, .
'.,.?> . .
..' .3,

I :



h: .










.+. v.:


.... ,. .

Use a value of 935.3~, nearest setfable value
of the
iscovered. A 0.5s definih,
time delay is used t o prevent spurious trips. The third;
harmonic voltage
under normal conditions is 20h of rated
voltage, giving a value of:




. ..

This protection is a combination of overcurrent with

undervoltage, the voltage signa! being obtained from a
VT on :he generator side of the system. The current
setting used is that of rated generator current of 6019A.
in accordance with IEEE C37.102 as the generator-ii for .
installation i n the USA. Use 6000A nearest settablevalue. The voltage setting cannot be more than 85% of .
the generator rated voltage t o ensure operation does not
occur under normal operation. For this application, a
value of 50010 of rated voltage is chosen.

A suitable value of V,,,, is 120010 of V':. giving a value

of 1565V. The nearest settable value is 3000V, minimum
allowable setting. The value of V,,, is required
be above the minimum voltage seen by thegenerator for
fault. A value
voltage is used for V;,,,. 14400V.

0.92x18kVxO. 1

A suitable value of K is therefye 0-36x2=0.3



the alarm setting is J800xJ.05 = 3 1 5 V / H z .

A 0
A time delay of 5s is used to avoid alarms due to
transient conditions.
The tr,ip setting is J8000xJ-3/6O=360~/1fz

78kV xO.02




The setting- of,..the :.third har

protettion-hu; b e l o 6 this-"
beingacceptable: use-avslue 6f 166.3V. A tiiiiqde
of 0.5s is used. . lnhibiiion o f the e1ement:at.l
generator output requires determination dur~n
; ;....... / ,

;:,;:>.., :::



The client requires a two-stage loss

protection function. The first is alarm only. while th
second provides tripping under high load conditions.:
achieve this, the first impedance element of the P3
loss of excitation protection can be set i n a c c ~ r d a n . ~
with the guidelines of Section 17.16.3 for a generatok
operating at rotor angles up to 120", as follows:

Xbl = O.SXd


X a , = 0.75X',j = 0.245R
Use nearest settable values of 1.669fl and 0.25
time delay of 5s is used to prevent alarms
transient conditions. For the trip stage, settings fo
load as given in Section 17.16.3 are used:

k V 2 - 18'
MVA 187.65


The nearest settable value for Xb2 is 1.72512. A

delay of 0.5s is used.

Loss of cxcitalion

Alarm: 59.3H2, 0.5s t i m e delay

Voltagc conlrollcd ovcrcurrcnl

1st stage trip: 58.7Hz, 100s time delay

, time delay
tage trip: 5 8 . 2 ~ z Is

Alarm: 62Hz, 30s t i m e delay

Trip: 63.5Hz, 10s time delay
These characteristia can b e set in the relay directly.

The generator manufacturers' recommendation is:

lates i n t o the following relay settings:

Alarm: 19800V. 5s t i m e delay
Trip: 23400V, 0.1s t i m e delay


Reverse reach. Zn


~ W C ~

'The setting data, according t o the relay

Forward reach, Z,

+ Z,


0.02 + 0.22



Rrvcnc P

~ o Slipping
~ c

Rcrcnc Powcr

... F>2rime dclay


Reactance line. Zc = 0.9 x Z


0 . 9 x 0.22




Z,= gct~rroforr r a ~ ~ s f o n i ~leakage



rlerluork impcrior~ce


N ~ f w . t k P r a r a c r i s a ff A s f . - . r i . m

i 5 . R*r?

~ . .





- 0.5s


s sos
j,-..-.- --

Fc2 tirnc dclay




F<fxttifi$ "...'

. - :58.7Hz







blrgc ~ C I I C I O I ~ I I , ~ ~ O ; C C : ~ ~C~A U ~ P ! C


.. . ...
. .. .
. ... .
.. .:.. . .
. .
:,..... :!
.. ,




.? .7,:'


-- ,.?;:


.. .?. . . .
ky$z7L;. .. .:,
.*- . : ... .

. .,....--. .- -:... !.
,x7:.. ,.,.: ., ..>

G m i l c


F < 3 limc dclay


. DO
. tirnc
F<Isetting '.
Fcl i i c dclay
, ..
i . ,,F+ xnilig ,.::.


PI scrling
PItimc dclay



,.:.cP '.,.
,. ....

-. - .<<.s:
. >; .-... . .

:;,~ ,
, ~ h * " ' ~

. ..;-.
.. . . .,;....


{ ~ h a ~ 1 7 - 2 8 0 - 3 1 5 17/06/02




The nearest settable values are 0.2434 0.656Q, and

0.206Q respectively.
The lens angle setting; a, is found from the equation:

and, substituting values.

= 62.5'
Use the minimum settable value of 90". The blinder angle,
8,is estimated to be 80', and requires checking during
commissioning. Timers T, and T, are set to l5ms as
experience has shown that these settings are satisfactory
to detect pole slipping frequenries up to IOtlz.


This completes the settings required for the generator,

and the relay settings are given in Tabte 17.5. Of course,
additional protection is required for the generator
transformer, according t o the principles described in
Chapter 16.





2 - .


17.1 Survey o f Rote O f Change o f Frequency Relays

ond Voltoge Phase Shif? Relays for Loss o f Moins
Protection. ERA Report 95-0712R. 1995. ERA

Technology Ltd.

ng Criteria

APPS Combined course

Generator Protection -Setting
Criteria & Tutorials

Page 1 of 45

The action required f3llowing response of an electrical or mechanical
often catzgorised as follows:
. -

. ...



.:i p
. ..-.


. .: :.... ,.

* . ..

..d... .
a.-..;:- ;'.-.
.,.... ,,


Urgent shutdown
Non-urgent shutdo.:.
'i Alarm only

- A n urgent shutdown .:I 3uld .be required, for example, if a phase to phase
fault occurred within -vie generator electrical connection. A non-urgent
shutdown might be s ~ ~ u e n t i awhere
the prime mover may be shutdown
prior to electr-ically ur, zading the generator, in order to avoid over- spesd.
A non-urgent shutdc.-,,n may be initiated in the case of continued
unbalar~cedloading. n this case, it is desirable that an alarm should b e
given before shutdo\/. .- becomes necessary, in order to allow for operator
intervention to remed.. -iie situation.
For urgent tripping, - may be desirable to electrically niaintain The
shutdowr-7 condition :.!ith -latching protection output contacts, which
would requir-e manuc resetting. For a non-urgent shutdown, i?rnay be
required that ttie oc-zvt contacts are self-reset, so that production of
power can be re-stcr-53 as soon as possible.
Generator differential protection

winc--,gs,or connection isulalion, can result i ! i severe

Failure of stato~.
damage to ttie wine'.-3s and stator core. The extent of the dar-riage will
depend upon the f::ult current level and the duration of ihe fault.
Protection should b e zpplied to limit the degree of damage in order to
limit repair costs. For :::.iniary generating plant, high-speed disconnection
of the plant from t t i t - ::Jwer sysiem niay be necessary to niaintain sysieni


OALSTOM Limited, Eries-gy ~ u t o m a t i o n& Inforniation


APPS Combined course

Generator Protection -Setting
Criteria & Tutorials

Page 3 bf 45

High impedance differntial protection

The high impedance principle is best explained by considering a

differential scheme where one CT is saturated for an external fault, as
shown in Figure 3.
If the relay circuit i s . considered to be a very high impedance, the
secondary current produced by the healthy CT will flow through the
saturated CT. If 'the magnetising impedance of the saturated CT is
considered to be negligible, the maximum voltage across the relay circuit
will be equal to the secondary fault current multiplied by the connected
impedance , ( R w + R L +~ R.iz~12)
The relay can be made stable for this maximum applied voltage by
increasing the overall impedance of the relay circuit, such that the
resulting current through the relay is less than its current setting. As the
impedance of the relay input alone is relatively low, a series connected
external resistor is required. The value of this resistor, RST , is calculated by
the formula shown in Figure 3. An additional on linear resistor, Metrosil,
may be required to limit the peak secondary circuit voltage during
internal fault conditions.

OALSTOM Limited, Energy Automation & Information

APPS Combined course

Criteria & Tutori

Protected z o n e

V o i l a g e a c r o r s relay circuit

..~ . .

.. .

j where K .- i .S

Srab:li s i n g r e s i s t o r , R
R st


. limits

s p i I:

, .:"!


curren: t.= l c j r s l u y sertir;g.l



.. .


- R~

W'here R R = rdq burden


. ...

. =.:


To ensure that the protection will operate quickly during an internal fault
the CTs used to operate the protection must have a kneepoint voltage of
at least 4Vs.


Setting guidelines for high impedance differential protection










The differential current setting, should be set to a low setting to protect as

much of the machine winding as possible. A setting of 5-10 % of rated-$$
current of the machine is generally considered to be adequate.
setting may need to be increased where low accuracy class CTs are used ;;.:
to supply the protection. A check should be made to ensure that t h e.:$
primary operating current of the element is less than the minimum fault&
cyrrent for which the protection should operate.


OALSTOM Limited, Energy Automation & Information



APPS Combined course

Generator Protection -Setting
Criteria & Tutorials

Page 5 of 45

The primary operating current (lop) will be a function of the current

transformer ratio, the relay operating current , the number o f current
transformers is parallel with a relay element (n) and the magnetising
current of each current transformer (le) at the stability voltage(Vs). This
relationship can be expressed in three ways:
To determine the maxim~~m
current transformer mgnetising current
to achieve a specific primary operating current with a particular
relay operating current.


le < 1 /n (lop/CT ratio - I diff)


To determine the maximum relay current setting to achieve a

specific primary operating current with a given current transformer
magnetising current.


< ( Iop/CT ratio

nle )

To express the protection primary operating current for a particular

relay operaling current and with a particular level of magnetising

= (CT ratio) x (


+ nle)

In order to achieve the required primary operating current with the

current transformers that are used, a current setting (I diff) must be
selected for the high impedance element, as detailed in expression(ii)
above. The setting of the stabilising resistor (RST) must be calculated in the
following manner, where the setting is a function of the required stability
voltage setting (Vs) and the relay current setting (I diff).

(I diff)

Note: the above fo;mula assumes negligible relay burden.

OALSTOM Limited, Energy Automation & Informalion


APPS Carnbinecf course

Generator Protection -Setting

Criteria & Tutorials
Page 6 of 45


Metrosils are used to limit the peak voltage developed by the current
transformers under internal fault conditions, to a value below the insulation
level of the current transformers, relay and interconnecting leads, which
are normally able to withstand 3000V peak.
The following forn~ulaeshould be used to estimate the peak transienl
voltage that could be produced for an internal fault. The peak voltage
produced during an internal fault will be a function of the current
transformer kneepoint voltage and the prospective voltage that would be
produced for an internal fault if current transformer saturation did not
occur. This prospective voltage will be a function of maximum internal
. .
fault secondary current, the current transformer -ratio, . . t h e .current::$
transformer led resistance to the common point; the relay lead resistonc&-"j
. .
and the stabilising resistor value.

Vp = 2

< 2 Vk



Vp = peak voltage developed by the CT under internal fault condiiiol-1s.
Vk = current transformer knee-point voltage
Vr = Maximum voltage that would be produced if CT saturation did not


Setting guidelines for Stator earth fault protection function (51 N)

Current operated from a CT in the neutral earth path.

OALSTOM Limited, Energy Automation & Information

-- ----


APPS Combined course

Generator Protection -Setting
Criteria & Tutorials



Page 7 of 45

Two independent tripping stages.

First stage tripping can incorporate either a definite time or standard
inverse type IDMT delay
Second stage tripping can be instantaneous or definite time delayed. .
Immune to third harmonics.


Applied to directly connected generators.

The protection must be time graded with other earth fault protection.
The setting employed should be less than 33% of the earth fault level.
A setting of 5% of the earth fault level should be applied for applications

where the differential protection provides less than 95% coverage of the
stator winding.
Applied to in-directly connected generators.
(with the generator earthed via a distribution transformer)

Can be supplied from a CT in either the primary or secondary circuit of the

distribution transformer.
With a CT in the primary circuit, th.e protection has the advantage of
being able to detect an earth fault which causes flashover of the primary
winding of the distribution transformer. With the CT in the secondary circuit
the protection has the advantage of detecting a short circuit across the
loading resistor. A sensitive 5% setting can be applied to the first tripping
stage, a short time delay can be applied to stabilise the protection
against small earth currents due to VT failures or earth leakage during HV
system faults.
The second tripping stage can be utilised as a high set. A 10% setting and
instantaneous operation ensures fast clearance of generator earth faults.

In the case of direct generator connection, it is common that only one

generator of a parallel set is earthed at any one time, with the earth
connections of other machines left open. If the generating plant can also
be run directly in parallel with a medium voltage public supply, it i s a
common requirement that all generator earth connections are left open
during parallel operation. In such circumstances, the main earth fault

OALSTOM Limited, Energy Automation & Information

APPS Combined course


Generator Protection -Settins

Criteria 8 Tutorial


protection element (le>) will only be operational for an earthed machine

It will provide primary earth fault protection for the associated machine
backup earth fault protection for other machines and the rest of th
power system and thermal protection for the earthing resistor.

For indirectly connected applications, the time-delayed earth fat

protection function may be employed in one of two ways:

1. To measure earth fslult current indirectly, via .a CT in the secondc

circuit of a distribution transformer earthing arrangement.

2. To measure earth fault directly, via a CT in the generator winding ea

With the first mode of application, the current operated protecli
function (51 N) may be used in conjunction with voltage operat
protection function { 5 9 N ) , measuring the distribution transforr,
secondary voltage. This is a complementary arrangement, where .
voltage operated protection function (59N) is able to operate in
event of an open-circuited loading resistor and the current opera
protection function (51N) is able to operate in the event of a sh
circuited resistor.
The second mode of application would be used for cases of di
resistive earthing. For distribution transformer earthing, this mode offers
advantage of being able to respond to an earth fault condition
leads to a flashover of the distribution transformer primary connect
Such a primary short circuit would render protection on the secon
side of the transformer inoperative and it would also result in a very
and damaging primary earth fault current.
In either mode of application, the main stator earth fault current oper
protection element (le>) should be set to have a primary sensitiv'
around 5% of the maximum earth fault current as limited by the ear
impedance. Such a setting would provide protection for up to 95% c
generator stator windings. The probability of an earth fault occurring
lower 5% of the generator windings would be extremely low, due i
fact that the winding voltage with respect to earth is low in this regior

OALSTOM Limited, Energy Automation & Information

APPS Combined course

Generator Protection -Setting

Page 9 of 45

The time characteristic and setting of the main current operated

protection element (le>) should be set to prevent false sperufion during
HV system earth fault clearance, where a transient generator earth
connection current may appear as a result o: the inter-winding
capacitance of the generator step-up transformer. The protection
element should also co-ordinate with operation of generator VT primary
-fuses, for a VT primary earth fault, and with VT secondary fuses for a
secondary earth fault on a VT that has its primary windings earthed.
Depending on the VT fuse characteristics, and on HV system earth fault
protection clearance times, a definite time delay anywhere between 0.5s
and 3.0s would be appropriate.

In machiqes with complex winding connection arrangements, e.g. some

hydrogenerators, the probability of a fault occurring in the stator winding
star-end region (first 5% of-the winding) might be higher. For a highly rated,
expensive machine, such increased probability may prompt operators to
apply 100% stator earth fault protection. A suitable 100%stator earth fault
protection scheme can be3pplied in these cases.

100 % Stator Earth Fault Protection

The conventional unit type generator has the neutral earthed through a
resistance loaded distribution type transformer. For a single ground fault
near the neutral end of the winding , there will be proportionately less
voltage available l o drive the current through the ground, resulting in a
lower fault current and a lower neutral bus voltage.



OALSTOM Limited, Energy Automation & Information

- iiealtky Cor,dlt.ion F - Faulty Condition

Figure 4
If an earth fault occurs and remains undetected because o its location (
otherwise the probability of a second fault occurring is much greater. Ttsecond fault may result from insulation deterioration caused by transie
overvoltages due to erratic , low current , unstable arcing ut the first fat
point. This second fault may yields of larger magnitudes.

. A 100% stator earth fault protection is designed to detect earth fat

occuring in the regions of machine windings close to the neutral end
works on principle involving monitoring of the neutral side and line si
components of the third harmonic voltages produced by the


OALSTOM Limited, Energy Automation & Information


APPS Combined course

Generator protection -Setting

Criteria & Tutorials

Page 11 of 45
AC generators in service produce a certain magnitude of third harmonic
voltages in their windings. Under healthy conditions of working the third
harmonic voltage developed by the machine is shared between the
phase to ground capacitive impedance at the machine terminal and the
neutral to ground impedance at the machine neutral. In general, under
healthy conditions the line and neutral impedances are fixed. Thus
irrespective of the magnitudes of the generated third harmonic V3, the
third harmonic voltages at the machine line end VL3 and neutral end VN3
should bear a constant ratio.

APPS Combined course

Generator Protection -Setting
Criteria 8 Tutorials







Page 12 of 45-

Referring to figure 4 it may be noted that when fault occurs at a point

say F on the machine -winding, the voltage distribution VN3/VL3
undergoes a change from that during healthy running condition. In the
extreme case of a fault occuring on the machine neutral , VN3 becomes
zero and VL3 becomes equal to V3. Similarly when a fault occurs on the
phase terminal VN3 becomes equal to V3. Ffor all other fault positions,and
depending on the f a ~ ~resistance,
VL3 & VN3 magnitudes will vary.

f!.: .

[rom the figure 5 it is clear that in order to remain stable under healthy
conditions the relay should restrain with in the two lines . The slopes of the
two lifles namely n11 & m2 can be suitably set to ensure stability and the
same will vary from n~achineto machine.

Setting guidelines for Neutral voltage displacement protection function

Voltage operated
Single n7easuring element two time delay stages.
In~muneto third harmonics.

Applied to directly connected generators.

Supplied from a broken delta VT

The voltage setting should be greater than the effective setting of any
down-strean?earth fault protection.
A time delay sufficient to allow downstream earth fault protection to

operate first should be used.

Fast earth fault protection can be enabled when the generator is no1
connected to the rest of the system.

Application to a directly connected


OALSTOM Limited, Energy Automation & Information

APPS Combined course

Generator Protection -Setting
Criteria 8, Tutorials

Page 13 of 45

For .tl-~is
mode of application, the neutral voltage displacement protection
should be driven from a broken-delta-connected secondary

f a,n,-tin

winding of a generator terminal VT that has its primary winding star-point

earthed. 'This VT should be made up of three single-phase units or should
be a single-phase unit with a 5-limb core. If the VT is not provided with an
independent set of secondary windings for broken delta connection, a
set of three single-phase interposing VT's should be applied. The
interposing VT's should have their primary windings connected in star to
the main VT secondary winding terminals and star-point. Their secondary
windings should be connected in broken-delta ,;ormat, to drive the neutral
voltage displacement protection function. ~lternafively,this protection
function could be driven from a single-phase VT connected between the
generator winding star-point and earth.
The voltage setting of the neutral voltage displacement protection
function should be set higher than the effective setting of current
operated earth fault protection on any outgoing feeder from the
generator bus. The setting should also be higher than the effective setting
of the sensitive directional earth fault protection applied to any parallel
generator. The effeciive voltage setting of any current operated earth
fault proteclion may b e established by multiplying the primary operating
current of the protection b y the generator grounding impedance and
dividing by one-third of the VT winding ratio, in the case of a broken delta
VT arrangement, or by the actual VT winding ratio in the case of a singlephase star-point VT.

Applied to in-directly connected generators.

Supplied from the secondary winding of a distribution earthing transformer

or from a broken delta VT.
A sensitive setting can be applied.
A short time delay can be applied to stabilise the protection during
voltage fluctuations due to VT failures or earth linkage during H V system


OALSTOM Limited, Energy Automation & 1nf6rmation


APPS Combined cocrrse

Generator Protection -Setting

Criteria & Tutorials
Page 14 of 45

Application to an indirectly connected generator

For this type of application, the voltage operated stator earth fault
protection function should be driven from the secondary winding of a
distribution earthing transformer. In the case of direct resistive.earthing, or
of no deliberate earth connection, the protection should be driven from a
VT winding.
The voltage setting of the protection function should be set to 5% of the
voltage that would be applied to the relay in- the event of a solid fault
occurring on one of the generator terminals. This would offer
approximately 95% coverage of the generator winding. The voltage
operated protection function might be used to complement the current
operated protection fundion in the case of distribution transformer

Setting guidelines for Voltage-dependent overcurrent protection function

(51 V)

Provides back up protection for uncleared downstream faults.

The protection operating mode can be configured to be: a simple
overcurrent, a voltage controlled overcurrent or a voltage restrained
overcurrent function.
In any of the modes of operation, the associated time delay can be
either definite time or standard inverse IDMT.
-The voltage dependent overcurrent protection must be time graded
with down-stream overcurrent protection. Where overcurrent reluys with
start contacts are used on outgoing feeders, time grading con be
achieved by blocking the opera-tion of the voltage del-~cndent
overcurrent protection.
In the simple overcurrent mode the system voltage has no effecl on the
current setting of the protection.
At normal system voltage the current setting should be 5% above ful
load current.
.. ...
OALSTOM Limited, Energy Automation & Information

APPS Combined course

Generator Protection -Setting

Criteria (L Tutorials
Page 15 of 45

When a fault close to the generator will result in a fault current

decrement the system vgltage should be monitored to distinguish
between normal load current and a system fault. Here either the voltage
controlled or the voltage restrained modes of operation should be
se1ected.A step change in the current setting is initiated i f the system
voltage falls below a selected level.


Applied when t h e generator i s directly c o n n e c t e d to the system.

At normal system voltage the current setting should be 5% above full load

Under low voltage conditions, the current setting should be reduced to

less than 50%of the minimum steady state fault current

The voltage control threshold should be selected to ensure that a voltage

reduction due to a single phase to earth fault will not result in a change of the
current setting.

When negative phase sequence protection is also applied, the

calculation of the voltage threshold need only consider the effect of a
remote three phase fault.

The voltage-dependent overcurrent protection function is a three-phase

protection function that is driven by the general protection CT inputs and
which is intended to provide backup protection for an uncleared phase
fault on the generator busbar or on a feeder from the busbar.

In the case of a generator passing h~ghlyreactive current to a fault the

level of fault current can fall below the maximum possible machine load
current within 0.5s-1.Os unless a fast-acting automatic voltage regulator
(AVR) is available. This is because the AVR is able to boost the level of field
excitation during a fault. The problem of fault current decrement can be
most acute when the excitation supply is derived from a transformer
connected to the generator terminals. Where a fault current decrement

OALSTOM Limited, Energy Automation & Information

APPS Combined course


.' ':,A.

Generator Protection -Setting

Criteria 8, Tutorials
Page 16 of45


X ti

.., .

possible, voltage-dependent overcurrent protection provides limedelayed backup protection with adequate sensitivity for a multi-phase
busbar or feeder fault, whiist remaining siabie for ;he highest aniicipaieci
level of generatar load current. The generator terminal voltage is
monitored as a way of being able to distinguish between normal load
and system fault conditions.

In the voltage-controlled protection mode, a step-change in current

setting (I> to K.I>) is imposed when the monitored voltage signal drops
below an adjustable threshold setting (Vs).

The under voltage switching threshold setting (Vs) should be selected so

that switching does not take place with the minimum possible phase- 1.;
phase voltage for single phase to earth fault conditions. For a single phase :';
fault, the minimum possibje phase-phase voltage would be for a close-up :.!;
earth fault on a solidly earthed power system, where the voltage could fall iK::
to 57% of the nominal level. The voltage setting should also be set above..,$;.
the maximum phase-phase voltagefor anyelement required to operate-:$
for a remote-end feeder fault. If the negative phase sequence thermal::;:
protection function is set and enabled, a remote three-phase fault need *:.
only be considered when determining the voltage threshold setting (Vs). .:&.*..


. ....

Reverse power and low forward power protection functions (32R/32L)

.. :.:.
. ....,.:,,


Reverse power protection:,;


Detects active power flow into the generator.


., ..a

, \,


The level of power required to motor the generator will depend on the :;$
type of prime move[.



A high sensitivity current input is used to monitor the system power. his$
may be connected to the main system protection CT's or, for
which require a sensitive setting, the input can be driven from a high4
accuracy measuren7ent CT.


QALSTOM ~ i k i t e dEnergy
Automation & Information

Generator Protection -Setting

A compensation angle setting is provided to compensate for CT and VT

phase errors.
A time delay (typically 5s) should be used to prevent operation of the
protection during some system fault conditions and power system swings.

To detect fluctuating reverse power flow, which could result from failure of
a reciprocating prime mover, a delay on drop off timer is available, in
addition to .the delay on pick up timer.

Low forward power protection.


Operates when the fom/ard power falls below the set level.
Operation can be instanianeous or time delayed.

Usually interlocked with non-urgent protection to reduce over speeding of

the generator following breaker operation for a non-urgent fault.

~ypicallevels of motoring power and possible motoring damage that

could occur for various types of generating plant are given in table

Prime mover

Gas Turbines

power Possible damage

Risk of fire or explosion
Reverse torque on



Steam Turbines

The need for automatic disconnection is arguably less for plant that is
continuously supervised, but, in .the event of prime mover failure, the
of control
staff could be diverted by other aspects of the

OALSTOM Limited, Energy Automation & Information

A?PS Combined course

Generator Protecfioii -Seffing
Criteria 8 Tutorials
Page 18 of 45

mechanical failure. If motoring damage can occur rapidly, operator

action may be too slow to prevent the onset of damage, so there may be
a requirement for automatic generator disconnection or for an alarm to
be raised. For unattended generation plant, e.g. small hydro schemes
that are only periodically supervised, automatic generator disconnection
should occur even if immediate prime mover damage would not be
envisaged. If automatic disconnection did not occur in such cases,
motoring may be possible for hours, with plant damage being gradually
inflicted. Automatic disconnection would also prevent an unnecessary
power system loss-,

In many cases, prime mover failure can be detected by non-electrical

means; e.g. by a steam turbine differential pressure switch or by a
hydraulic flow device. If mechanical means of detecting prime mover
failure are provided, an electrical measurement method would not be
required or would only be used for backup detection.
- .--."?
: :.




,.A- .,*..


. ..
. .--4

The reverse power protection function needs to be time-delayed to'.$$ij

prevent false tripping or an alarm given during power system swings,
following power system disturbances or following synchronisation. In some .$!!
applications, the reverse power protection function should be disabled ;$' $
during certain modes of protected machine operation. One example of $@
such a situation is where, during dry seasons, a synchronous machine is
de-coupled from its hydraulic prime mover and operated as a
synchronous compensator for power system VAR control.
. ..!=;







Low forward power protection function (32L)

. .,.This protection function is offered for those users who wish to interlock non-

power measuring element confirms that the mechanical drive has bee
cut. Such an arrangement would ensure that there w o ~ ~be
l d no possibilif
of generator set over speed when any restraining electrical load is cut b
electrical tripping.
OALSTOM Limited, Energy Automation & Information

APPS Combined course




Generotor Protection -Setting

Criteria & Tutorials

Page 19 of 45

With any generator, tripping of the generator breaker and excitation

system should be accompanied by throttle or valve closure. There is
always a risk, however, that the throttle/valves may not close fully and
that machine over speed will result when electrical loading is removed.
With large high-speed steam turbo-alternator sets, an apparently small
over speed could result in machine damage or wreckage, as well as a
threat to human safety. Failure of a steam valve to fully close during a
shut-down is an obvious risk This over speed risk could be addressed by
using duplicate valves in series.

Even where valves, etc., do close fully, there will be some lag in dissipating
all the energy within a prime mover, especially in the event of a shutdown
from full-load. Some types of plant, are very prone to over speed following
rejection of full-load, but have a good over speed tolerance, e.g. slowspeed hydro generators. Large turbo-alternators, with slender,low-inertia
rotor designs, do not have a high over speed tolerance and trapped
steam in the turbine, downstream of a valve that has just closed, can
rapidly lead to over speed. To reduce the risk of over speed damage to
such sets, it i s sometimes chosen to interlock non-urgent tripping of the
generator breaker and the excitation system with a low forward power
check. The delay in electrical tripping, until prime mover energy has been
completely absorbed by the power system, may be deemed acceptable
for 'non-urgent' protection trips; e.g. stator earth fault protection for an
indirectly connected generator. For 'urgent' trips by instantaneous
electrical protection, e.g. stator winding current differential protection,
any potentially delaying interlock should not be imposed. With the low
probab~lity of 'urgent' trips, the risk of over speed and possible
consequences must be accepted.
With a large generator, even a very small percentage of rated power
could quickly accelerate an unloaded machine to a dangerous speed. A
typical under power setting requirement would be 0.5% of rated power.
The time delay associated with the low forward power protection function
(t) could be set to zero. However, some delay is desirable so that
permission for a non-urgent electrical trip is not given in the event of
power fluctuations arising from sudden steam valvelthrottle closure. A
typical time delay for this reason is 2s.

OALSTOM Limited, Energy Automation & Information

APPS Combined course


Generator Protection -Setting

Criteria 8 Tutorials
Page 20 of 45




Negative phase sequence thermal protection function (46)

Protects the rotor of a generator from damage resulting from the

heating effects of negative phase sequence currents.

Provides true negative phase sequence thermal protection and a

definite time alarm.

Accurate over a wide system frequency range.

The trip threshold should be set slightly higher than the constant negative
phase sequence current withstand of the generator.

The protection must be time graded to allow downstream protection to

clear an unbalance fault.


T o achieve easier grading with. down stream protection, during-:-f

clearance of a heavy asymmetric fault, a minimum operating time for the''$
negative phase sequence protection can be set.
For negative phase sequence currents slightly above setting, a
maximum trip time can be set.

Can provide back up protection for uncleared asymmetric faults.

Models the cooling characteristic of the generator, following exposure

to negative phase sequence currents.

The alarm element is commonly set to 70% of the trip setting with a time
delay well above the time taken to clear any system faults. The alarm element functions directly on the measured level of negative phase
sequence current.

The NPS protection function is provided for applications where a

generator (synchronous machine) is particularly susceptible to rotor
thermal damage, in the event of the current supplied to the power system
- unbalanced. The degree of susceptibility will depend on the
generator rotor design (cylindrical or salient construction), methods of

OALSTOM Limited, Energy Automation & Information

APPS Combined course

Generator Protection -Setting
Criteria & Tutorials

Page 21 of 45

forced cooling employed and the presence of any ancillary metallic rotor

Monitors the generators terminal irr~pedancein order to detect failures in

the excitation system.
Uses a circular, offset mho, irr~pedancecharacteristic.


The diameter of the impedance characteristic is based on the direct

synchro-nous reactance of the generator.
.The offset of the impedance characteristic based on the direct axis
transient reactance of the generator.


An associated definite time delay prevents operation of the protection

during stable power swings.
Can be interlocked with the under voltage protection element to
prevent operation during power swings.
A delay on drop off timer-can be used to detect cyclic operation of the
field failure protection. This could result during pole slipping.

This protection function measures the impedance at the terminals of a

generator that is run in parallel with another source to detect failure of the
generator excitation. The current used for single phase impedance
measurement is obtained from the 'general protection CT inputs and the
voltage is obtained from the main VT inputs. This protection function is
provided with an adjustable, offset circular impedance characteristic, see
Figure 6 , an adjustable tripping delay timer (t) and an adjustable
measuring element reset time delay (tDO).

OALSTOM Limited, Energy Automation

& Information

Figure 6
Complete loss of excitation-may arise as a result of accidental tripping of
the excitaiion system, an open circuit or short circuit occurring in l11e
excitation DC circuit, flashover of any slip rings or failure of the excilolion
power source. A pure open circuit in the excitation system i s unlikely to he
long-lasting in view of the high voltage that would be developed crc~oss
the open circuit with the machine running and connected to a po\ver
system. Such a fault is likely to evolve quickly into a short circuit fauli.

OALSTOM Limited, Energy Automalion & Information


APPS Combined course

Generator Protection -Setting

Criteria 8 Tutorials
Page 23 of 45

Where a generator stabilises at a high level of slip, following excitation

failure, the reverse inductive impedance seen at the generator terminals
will be highly reactive and will be less than the direct axis synchronous
reactance of the machine ( X d ). A typical minimum value for this
impedance is twice the direct-axis transient reactance of the generator
(2X d ' ) for a level of slip below 1%. Figure 6 shows a typical machine
terminal loss-of-field impedance locus, which illustrates the effect of rotor
flux decay, leading to gentle pole-slipping and eventual stabilisation as
an induction generator with a level of slip of around 1%.
To quickly detect a loss-of-field condhion where machine damage may
occur, the diameter of the relay field-failure impedance characteristic
(Xb) should be set as large as possible without conflicting with the
impedance that might be seen under normal stable conditions or during
stable power swing conditions. To meet this objective, it is recommended
that the diameter of the relay impedance characteristic is set equal to
the generator direct-axis synchronous reactance in secondary ohms. The
characteristic offset should be set equal to half the direct-axis transient
reactance (O.5X d ' ) in secondary ohms.

The above guidelines are suitable for applications where a generator i s

operated with a rotor angle of less than 90" and never at a leading power
factor. For generators that may be operated at slightly leading power
factors and which may be operated with rotor angles up to 1 20, by virtue
of high-speed voltage regulation equipment, the settings would need to
be different. The impedance characteristic diameter should be set to 50%
of the direct-axis synchronous reactance (O.5X d ) and the offset should
be set to 75% of the direct axis transient reactance (0.75X d ' ) .
The field failure protection time delay (t) should be set to minimise the risk
of operation of the protection function during stable power swings
following system disturbances or synchronisation. However, it should be
ensured that the time delay is not so long that stator winding or rotor
thermal damage will occur. The stator winding should be able to
withstand a current of 2.0 p.u. for the order of 15s. It is unlikely that rotor
damage would be incurred in much less time than this. It must also be
appreciated that it may take some seconds for the impedance seen at
the generator terminals to enter the selected characteristic of the
protection function. However, a time delay less than 10s would typically
be applied. The minimum permissible delay, to avoid potential problems

OALSTOM Limited, Energy Automation & Information

APPS Combined course

Generator Protection -Setting
Criteria & Tutorials

Page 24 of 45
of false tripping due to stable power swings with the above impedance
settings, would be of the order cf 0.5s.
Some operators have traditionally interlocked operation of impedancetype field failure protection with operation of under voltage detection
elements in order to allow a low field failure protection time delay without
the risk of unwanted tripping for stable power swings. This arrangement
may also have been used to prevent field failure protection operation for
hydrogenerators that may be run as synchronous compensator's, with the
turbine mechanically decoupled.

The field failure protection fucction is offered with an adjustable delay on

reset of the trip timer (tDO).This lime delay can be set to ovoid delayed
tripping that might arise as a result of cyclic operation of the impedance
measuring element during the period of pole-slipping following loss of
excitation. The delay on reset of the trip timer (tDO) might also be set to
allow the field failure protection funclion to be used for detecting pole
slipping of the generator when excitalion is not fully lost; e.g. following
time-delayed clearance of a nearby power system fault by delayed

Under voltage protection function (27)

Operates when the three phase voltages fall below the common set
point. An adjustable timer is available.
Can be interlocked with the field failure protection to prevent its
operation during stable power swings.
Can be used to initiate dead machine protection
Can detect failure of the AVR or system faults which have failed to be
cleared by other means.
Prevents damage to any connected loads which could occur- during
operation at less than rated voltage.

OALSTOM Limited, Energy Automation & Information

APPS Combined course

Generator Protection -Setting
Criteria &Tutorials
Page 25 of 45




The pick up level sho~lldbe set to less than the voltage seen for a three
phase fault at the remote end of any connected feeder.

.2-' 9..



The time delay should be set to allow the appropriate feeder protection
to operate first to clear the fault, and. also to prevent operation of the
protection during transient voltage dips.

A dedicated input is provided to block the operation of the under

voltage and under frequency protection during run-up or run-down of the

generator. This input can be driven from an auxiliary contact in the circuit

Under voltage protection can be used to detect abnormal operating

conditions or an uncleared power system fault that may not have been
detected by other generator protection.
- ...:......-. .
',: .



;.,..,. :..








.;, :..

For an isolated generator, or for an isolated set of generators, especially in

the case of standby generating plant, -a prolonged, under voltage
condition could arise for a number of reasons. One reason would be some
failure of automatic voltaae reaulation IAVRY eaui~ment.If such a
condition persists, automatic generator tripping should be initiated. to
prevent -possible damage to system loads. Another reason could be that
a fault exists somewhere on the power system that has not been cleared
by other means.







:.'. ,


. ...






,:.> :


In the case of generators feeding an industrial system, which is normally

fed from a public-power supply, system overcurrent protection settings
would have to be above maximum levels of system load current with the
normal supply available. If the public supply fails, the local generation
would be left feeding the entire system. Where the local generation is
unable to meet the entire system load, there would be a provision for the
automatic shedding of non-essential loads. If a fault subsequently
Occurred on the system, the relatively low fault current contribution of the
local generation and its decrement with time may result in the system
overcurrent protection failing to respond. In this case it would be
expected that the generator backup overcurrent should operate.

$:. .4;





Operation of generator overcurrent protection in the above

circumsfances can be assisted by employing voltage-dependent
protecfion.Wt-,erethere is a parallel set of generators,
-and where the fault

OALSTOM Limited. Enerav Automation & Information

APPS Combined course

Generator Protection -Setting
Criteria 8 Tutorials
Page 26 of 45




is relatively remote from the generators, even the generator voltagedependent proteciion r i i a y fail to respond to the fault. If the fault is
asymmetric, and if the negative phase sequence thermal protection
function has been set and enabled, the unbalanced fault current may be
sufficient to operate this form of generator protection. The worst situation
would be for an uncleared three-phase fault. Although such a fault would
be rare, it may be that the only form of protection that would reliably
detect the fault would be generator under voltage protection.

In the case of large thermal power plant generators, a prolonged under

voltage condition could adversely affect the performance of the auxiliary
plant such as boiler-feed pumps and air-blowers. This would ultimately
have an effect on the primary plant performance. If such a situation is
envisaged, the application of time-delayed under voltage protection to
trip the generator might be a consideration.
The under voltage protection function threshold (V<) should be set below
the steady-state phase-phase voltage 'for a. three-phase fault at the-'
remote end of any feeder connected to the generator bus or up to
selected locations within an industrial power network. Allowances should
be made for the fault curcent contribution of parallel generators, which
will tend to keep the generator voltage up. The time setting of the under
voltage protection function (t) should be set longer than the time required
for backup feeder protection to clear remote-end feeder faults. The delay
should preferably be longer than the time required for the generator
back-up overcurrent protection function to respond to such a fault.
~ddifionally,the delay should be long enough to prevent unwanted
operation of the under voltage protection function for transient voltage
dips during clearance of faults further into the power system or by starting
of local machines. The required time delay would typically be in excess ol
To prevent tripping of the under voltage protection function followinc
normal shutdown of a generator, a normally closed circuit breake
auxiliary contact should be used to energise the under voltage inhib
logic input.

Over voltage protection function


OALSTOM Limited, Energy Automation & Information



APPS Combined course

Generator Protection -Setting
Criteria 8, Tutorials

Page 27 of 45

* Operates when the three phase voltages are above the common s e t

Two tripping stages, each with an aajustable timer.


Protects against damage to the generator insulation and that of any

connected plant.

Recommended for


hydrogenerators which may suffer from load


*Time delayed protection should be set with a pick up voltage of 100120% of the nominal voltage and a time delay ~ufficientto overcome
operation during transient over voltages.


Instantaneous protection with a setting of 130% - 150% of the nominal

voltage can be implemented.



An unsynchronised generator terminal over voltage condition could arise

when the generator is running, but not connected to a power system, or
where a single generator is running and providing power to an isolated
power system. Such an over v ~ l t a g ecould arise in the event of a fault
with automatic voltage regulating equipment or if the voltage regulator is
set for manual control and an operator error is made. Over voltage
protection should be set to prevent possible damage to generator
insulation, prolonged over fluxing of the generating plant or damage to
isolated power system loads,


When a generator
i s synchronised to a power system with other sources, a
synchronised over voltage could only arise i f the generator was lightly
loaded and was requiced to supply a high level of power system
capacitive charging current. An over voltage condition might also be
possible following a system separation, where a generator might
experience full-load rejection whilst still being connected to part of the
original power system. The automatic voltage regulating equipment
should quickly respond to correct the over voltage condition, but over
voltage protection is advisable to cater for a possible failure of the
voltage 'regulator to correct the situation or for the possibility .of the
regulator having been set to manual control.

OALSTOM Limited, Energy Automation & Information

APPS Combined course

Gsnerctor Protection -Setting
Criteria & Tutorials
Page 28 of 45

The worst case of generating plant over voltage following a system
separation, which results in full-load rejection, could be experienced by
hydrogenerators. The response time of the speed governing equipment
can be so slow that transient over speeding up to 200% of nominal speed
could occur. Even with voltage regulator action, such over speeding can
result in a transient over voltage as high as 150%. Such a high voltage
could result in rapid insulation damage.
The time-delayed over voltage protection function threshold (V>) should
typically be set to ] 00%-120%of the nominal voltage . The time delay (t>)
silould be set to prevent unwanted tripping of the delayed over voltage
protection function due to transient over voltages that do not pose a risk
to the generating plant; e.g. following load rejection with non-hydro sets.
The typical delay to be applied would be 1s-3s.

Under frequency protection function (81 U)

Two under frequency stages each with an independent timer.

First stage can be used to initiate load shedding for industrial systems.
Time delayed to allow any down stream load shedding equipment to
operate first.
Second under frequency stage to trip more rapidly.
.A dedicated input is provided to block the operation of the under
voltage and under frequency protection during run-up or run-down of the
generator. This input can be driven from an auxiliary contact in the circuit

As well as being able to initiate generator tripping, the under frequency

protection can also be arranged to initiate local load-shedding, where
appropriate. Under frequency operation of a generator will occur when
the power system load exceeds the prime mover capability of an isolated
of generators. Where the system load exceeds
generator or of a group
- - I

OALSTOM Limited, Energy Automation & Information


APPS Combined course

- .

Generator Protection -Setting

Criteria & Tutorials
Page 29 of 45


the alternator rating, but not the prime mover rating, the alternator could
become overloaded without a frequency drop. !t would therefore be
important for the alternator manufacturer to provide stator winding
temperature measurem-entdevices, to give alarm or to automatically shut
down the generator before winding thermal damage results.


.. .

Power system overloading can arise when a power system becomes split,
with load left connected to a set of 'islanded' generators that is in excess
of their capacity. Such events should be allowed for by system planners
and automatic system load-shedding should be implemented so that the
load would rapidly be brought back within the generation capacity. In
this case, under frequency operation would be a transient condition; as
during power swings. The degree of load shedding would have to take
into account the fact that some generating plant, e.g. gas turbine plant,
may have a reduced power capability when running below nominal
frequency. In the event of under shedding of load, the generators should
be provided wit,h backup under frequency protection to shut down the
generating plant before plant damage or unprotected system load
.damage could' occur.


Under frequency running at nominal voltage will result in some over fluxing
of a generator, and its associated electrical plant, which needs to b e
borne in mind. However, the more critical considerations would b e in
relation to blade stresses being incurred with high-speed turbine
generators; especially steam-driven sets. When running away from
nominal frequency, abnormal blade resonance's can be set up which, if
prolonged, could lead to turbine disc component fractures. Such effects
can be accumulative and so operation at frequencies away from
nominal should be limited as much as possible, to avoid the need for early
plant inspections/overhaul. Under frequency running is most difficult to
contend with, since there is little action that can be taken at the
generating station in the event of load under shedding, other than to shut
the generator down.

To prevent under frequency protection tripping following normal

shutdown of a generator, a normally closed circuit breaker auxiliary
contact should be used to energise the under frequency protection
function inhibit logic.


DALSTOM Limited, Energy ge tom at ion & Information


8 . 7

A nnc r , , ~ : - ~
r r ~au r ~ i ~ A
~ course
r- i-.t .-f u-

Generator Proieciion -Seiiing

Criteria 8, Tutorials

Page 30 of 4


Over frequency protection function (810)

. .





Single over frequency stage with associated timer.





Should be set abdve the sustainable over frequency level with a time
delay sufficient to overcome transient over frequencies following load








..C i i


Over frequency running of a generating set arises only when the

mechanical power input to the alternator is in excess of the electrical
load and mechanical losses. The most common occurrence of over :I$
frequency is after substantial loss of electrical loading. When a rise in ;:;j\
running speed occurs, the governor should quickly respond to reduce the :$
mechanical input power so that normal running speed is quickly regained. ..,:!;,

r L

. ..




Over frequency protection may be required as a backup protection .@

function to cater for governor or thiottle control failure following l o i s o t . $
- . .- :.+?
load or dyring unsynchronised running.

. ..<?









. .-


Voltage balance protection function (60)

. ..,


~ . X

I .?




. ,...


Detects VT fuse failure.
-Supplied from the secondaries of two VTs or two separately fused'..
. ..
secondary circuits of a single VT.
Used to raise an alarm and block voltage sensitive protection if,::
-.... ..

Dead machine protection

For a multiple source power system, closure of a generator circuit breakel
must be controlled either by automatic synchronising equipment, or by
manual breaker closing carried out with the aid of synchronising
instruments, and supervised by a synchronism check relay.
.. ,

Whilst inadvertent closure of a generator circuit breaker should not b!

possible, a small risk does exist; especially when fault finding, carrying 0~


OALSTOM Limited, Energy Automation & Information



APPS Combined course

Generator Protedion -Setting
Criteria 8 Tutorials

Page 31 of 45


maintenance tests or testing control systems. The possible damage

caused by connecting a dead machine to a live power system, or
energlsing a steam turbo-alternator when on turning gear, could be
extremely costly if a method of quickly tripping the generator breaker is
not provided.
If a dead machine is energised from a live power system, rotor currents will
be induced and the machine will accelerate as an induction motor. The
induced currents in the rotor body and windings would be very high with
the machine initially at standstill and could rapidly result in thermal
damage unless the machine is designed for direct-on-line run-up as an
induction motor (possibly for starting a gas turbine prime m'over). 'The
unexpected shaft rotation could also result in rapid mechanical damage
if lubrication systems are not running or if a steam turbo-alternator is on
turning gear.

A number of

protection functions could. respond to the inadvertent

energisation of a dead .machine. The effective machine impedance.
during such enegisation would be similar to its sub-transient reactance
and so the current drawn from the power system would-be high. So under
voltage and overcurrent protection functions could respond to the
condition and can be interlocked with manual tripping Jogic to protect
the machine against the inadvertent energisation of a dead machine.

OALSTOM Limited, Energy Automation, & Information

APPS Combined course

Generator Protection -Setting
Criteria & Tutorials

Page 32 of 45

CTlG -

a a

dead tl~lachine
?rip ping


Pole-slipping protection
A generator might pole-slip, or fall out-of-step with other power system

sources, in the event of failed or abnormally weak excitation or as a result

of delayed system fault clearance; especially when there is a weak (high
reactance) transmission link between the generator and the rest of the
power system.
With large utility base-load generators, the requirement for pole-slipping f
protection will be dependent on the transmission system reactance. In the ;
case of generators connected to a dense, interconnected system, pole- ;'
slipping protection may not be required. In the case of remote generation :!
and a weak radial transmission link to the load centre, stability of :':
generation may be an issue. Pole-slipping protection is frequently .!;.a
requested for relatively small generators running in parallel with strong $#
public supplies. This might be where a co-generator runs in parallel with!$$
the distribution system of a public utility, which may be a relatively strong
source, but where high-speed protection for distribution system faults is not $
provided. The delayed clearance of system faults may pose a stability$
threat for the co-generation plant.
. ...


, u.

OALSTOM Limited, Energy Automation & Information




;-.: - .


APPS Combined course

Generator Profection -Setting.

Criteria & Tutorials
Page 33 of 45


The pole slipping relay ZTO has been designed to protect Synchronous
generators against the possibility of the machine running in the unstable
region of the power angle curve which would result in pole slip.

The relay consists of one directional relay and one blinder relay operating
in conjunction with 40-80 mSec timer . Both characteristics look into the
source and consequently ignore all condifions of load other than those
which produce a reversal of power flow such as would occur with a
condition of pole slip or power ;wing exceeding 90 degree.

The timer is incorporated so that the discrimination can be made

between a power swing and a pole slip -condition. A trip condition can
only occur if the timer has timed out before the fault moves into t h e


If the fault never reaches the operate regionof the b1inder:or moves
between the directional and blinder char.acteristics in a time less than
the timer setting , no operation will occur.




Generator Details
Terminal voltage
Synctironous Reactance Xd
Transient Raactance Xd'
Sub-Transient Reactance Xd' '
Continuous Negative withstand capability
12' t
Length of longest line emanating from the bus
Impedance of the line
Bus fault MVA

247 MVA
15.75 KV
205 %
23.4 %
17.9 %

2071-5 MVA

Full load current of the machine

( MVAx 1000/ 1.732~

Synchronous reactance of the machine Xd

(KVxKVx pu Xd / MVA)
Transient reactance of the machine Xd'
(KVxKVx pu X d ' / MVA)
Sub transient reactance of the machine Xd"
(KVxKVx pu Xd' ' / MVA)
1 OYdO
1.5 ohms
1.1 1 ohms

Generator line/neutral side CT primary current

Generator line/neutral side CT secondary current
CT sec resistance
Lead resistance
Generator PT primary voltage
Generator PT secondary voltage

OALSTOM Limited, Energy Automation & Information



k '.



APPS Combined course

Generator Protection -Setting

Criteria 6 Tutorials

Page 35 of 45

CT/ PT Ratio
CT primary for inter turn protection
CT secondary for in-terturn protection
CT sec resistance
Lead resistance

5000 A
0.75 ohms
1 ohm

Generator Transformer rating

Voltage ratio
240115.75 KV
Transformer lmpedance
Transformer lmpedance in ohms

250 MVA


Generator Differential Protection 87G - CAG 34

Maximum three phase fault current

(Full load current/sljb transient reactance)
Fault current referred to secondary
Voltage developed across the relay circuit
Fault current ref to sec ( CT resistance + 2 lead resistance)

93.68 V

Pick-up setting recommended

Voltage developed across the relay at pick up
(VA burden .of the relay / pick up current)
Stabilising resistance value


(Voltage developed across the relay-Voltage acro'ss rely at pick up/ Pick

Generator Inter-turn Protection 87G1- CAG 34

Maximum three phase fault current

O A L ~ ~ OLimited,
Energy Automation & Information

25183 A


APPS Crmbiiied course


Generator Protection -Setting

Criteria 8 Tutorials

Page 36 of 45

(Full load current/sub transient reactance)

Fault current referred to secondary
Voltage developed across the relay circuit
Fault current ref to sec ( CT resistance + 2 lead resistance)

69.25 V

Pick-up setting recommended

Voltage developed across the relay at pick up
(VA burden of the relay / pick up current)

Stabilising resistance value
(Voltage developed across the relay-Voltage across rely at pick up/ Pick
up current)

Generator Field Failure Protection 40G--YTGM15




Diameter setting
(0.5 x synchronous recatance)

1.03 ohms

Diameter setting ref to secondary

Offset setting of the relay
transient reactance)
Offset setting of the relay ref- to secondary

1.68 ohms

Timer settings:
Pick up timer
Drop off timer

OALSTOM Limited, Energy ~utomation& Information

10 Sec

2 Sec


Criteria 8, Tutorials

Page 37 of 45

0.1 6 ohms

0.14 ohms

3.07 ohms
75 degree

- .-

APPS Combined course

Generator Protection -Setting

Criteria 8, Tutorials

Page 38 of 45

Generaivr Irr-~pedzjiice
Generator transformer impedance
Source impedance angle (generally assumed to be 80")
Rate of slip ( should be provided by the manufacturer, otherwise
assumed to be .I 600 Elec. Deg per second)


.. .:>




I I.ti.,





Procedure :



; .>:

1. Select a suitable scale for the diagram. Draw the X and Y axes with
Origin as (0)






2. Plot .the Generator impedance along the negative "Y- axis" to get
point (G)

... .



, ...

3. Plot the Generator Transformer impedance at positive "Y - axis" to get

point (T)



4 . Draw source imped&nce at an angle of 80 (or source, impedance;::.<$

. .:angle i f available) from point (T) to getpoint (S)
. .-..:.,:2_ :...-.

. :,$<
. -.

5. Connect points (G) and (S) by a straight line.


6. The locus of Pole slip will be nearer to either (G) or (S) depending on the
ratio between emfs at (G) and (S). We assume this ratio to be equal to
1. Thus the pole slip locus is the bisector of line (G)---(S). Mark the point
( 1 ) on the line (G) ---(S) where the Pole slip locus cuts it (centre of the

7. With point ( 1 ) as center and (1)---(GIas radius draw a circle. Mark the
point where this circle cuts the locus as ( 2 ) .
8. Draw a line passing through the origin (0) at 75". Mark the point v,/here
this line cuts the pole slip locus as (3).This is the directional line.
9. Measure the obtuse angle at point (4) between the lines (G)---(4)and
This is named as (4.

1O.The obtuse angle at point (2) should be 270". Name this as (a,).

.. .
. .


OALSTOM Limited, Energy Automation & Information





APPS C o m b i n e d course'
Generator Protection -Setting
Criteria 8. Tutorials
Page 40 of 45


;*?,.;f '

OALSTOM Limited, Energy Automation & Information



. . . .%.


g::. :

APPS Combined course


Generator Protection -Settina

Criteria 8, ~utofiag
Page 41 of 45

g: .

Generator Back-up Impedance 21G YTGMl5

lmpedance required to cover the entire line under maximum generation


Where n is the No. of niachines in parallel




. ._..


~ : ~ o w areach
to be set in relay
Reverse reach
(25% of forward reach)

= 0.98 ohms

Timer setting

= I sec

Timer setting need to be co-ordinated with Zone 2 Timing

Negative Sequence Protection 46G- CTNM31

= 5%

Negative seq current 125

122t ( in terms of K1 & K3 )

QALSTOM Limited, Enerqy Automation & Information

= 8

APPS Combined course

- .


Criteria & Tutorials

12 Alarm setting

Alarm timer setting (fixed)


Generator Reverse Power Protection 32G


Pick-up setting
(depends on the type of prime mover)
Time delay

= 5 Sec

Generator 95 % Stator Earth Fault Protection 64s - VDGl4


Voltage setting

= 5.4 V
. .:Q

= 1 Sec

Time delay




Time setting need to be co-ordinated with down stream Earth Fault relays in;'!
case of Direct connected svstems.

Generator 100 % Stator Earth Fault Protection 64s - PVMM163

The required settings for this 95-100 % protection can be selected only on
measurement and studying the machine third harmonic behaviour a1 site.

Measurement of Generator
- - . - - - - -~
~ . ~ ~ Harmonic
. ..~


-~ -~

OALSTOM Limited, Energy Automation & Information


Generator Protection -Setting

Criteria 8, Tutorials

Page 43 of 45

a) Measure the filtered third harmonic voltages from neutral side (VN3) and
Line side (VL3) at the following TEST SOCKET pins provided on the front of
the reiay PVMM. Digital muitimeter in AC millivolts range can be used for
this purpose.
VN3 - Across 1 & 7
VL3 - Across 2 & 7
b) 'These measurements areto be made during voltage build up of
Generator before synchronization and after' synchronizing at different load
(MW) and excitation (MVAR j conditions.

Study of Generator Third Harmonic Voltages

The third harmoni; voltages measured above are plotted in a graph with
VN3 on X-axis and VL3 on Y-axis. Drawtwo lines enclosing.all,measured
values with some tolera-nee.
.. . v ' a l ~ a t e ~ i i o ~ e
s ' m l of these lines. The
slopes can b e calculbted by ielecting any pointalongthe line-andb y
computing its V N ~ / \ /ratio;
. .




Alternatively, calculate VL3/VN3 from each set of readings under different

load conditions. Select the maximum and minimum values of these ratio.
Then m l will be maximum VL3/VN3+5%and m2 will be VL3/VN3 - 5%.

The dead band setting K and the null setting potenliometer "a" can be
calculated as given below and set it accordingly in the relay.




OALSTOM Limited, Energy Automation & Information


APPS Combined course

Generator Protection-~eftj,$
Criteria 8, TvtorialSPage 44 o f z


The following table shows the actual values of the generator VL3 and VN3
obtained from site:

1. Before Synchronization:
Vn3(mvolts) Point: 187

V13(mvolts) Poi1

Volta e kV


49 1
-1 1 16
1 680



- 76.4
1 28
263 325 390.


2. At different load (MW) conditions

SI.No -




OALSTOM Limited, Energy Automation & Information

APPS. Combined course

Generafor Profecfion -Setting
criteria & Tutorials

Page 45 of 45








Generator 3rd Harmonics

Vn3(N eutra I Side)

From the table.

m 1 = 0.454




APPS Combined course

Generator Protection -Setting

Criteria 8, Tutorials Page 46 of 45 -.

Calculated values are

K = 3.1 and a = 0.409

Dead Machine Protection 6: B


Overcurrent setting
Under voltage relay setting

Timer setting

= 1 Sec


OALSTOM Limited, Energy Automation & Information

Distance Protection Notes




Zone 1 Setting


Zone 2 Setting


Zone 3 Setting


Settings for Reverse Reach and Other Zones


Amplitude and Phase Comparison

1.7.2 Plain lmpedance Characteristic


Self-Polarised Mho Relay


Offset MhoJLenticular Characteristics

Application of Lenticular Characteristic

1.7.5 Fully Cross-Polarised Mho Characteristic

1.7.6 Partially Cross-Polarised Nlho Characteristic

Protection Against Power Swings - Use of the Ohm Characteristic


1.7.9 Other Characteristics

1.8.1 Starters for Switched Distance Protection


Effect of Source lmpedance a_nd Earthing Methods

1.9.1 Phase Fault lmpedance Measurement
1.9.2 Earth Fault lmpedance Measurement

1.10.1 Minimum Voltage at Relay Terminals

1.10.2 Minimum Length of Line
1.10.3 Under- Reach - Effect Of Remote lnfeed
1.10.4 Qver-Reach
1.10.5 Forward Reach Limitations

4.10.6 Power Swing Blocking

1.10.7 Voltage Transf~merSupervision
Other Distance Relay Features
Distance !?e!ay Application Example
1.12.1 Line Impedance
1.12.2 Residual Compensation
1.12.3 Zone 1 Phase Reach

1.12.4 Zone 2 Phase Reach

1.12.5 Zone 3 Phase Reach
1.12.6 Zone Time Delay Settings
Reach Settings
1.12.7 Phase ~ a u lResistive
1.12.8 Earth Fault Impedance Reach Settings
1.12.9 Earth Fault Resistive Reach Settings

The problem of combining fast fault clearance with selective tripping of plant is a key aim
for the protection of power systems. To meet these requirements, high-speed protection
systems for transmission and primary distribution circuits that are suitable for use with the
automatic reclosure of circuit breakers are under continuous development and are very
widely applied.
Distance protection, in its basic form, is a non-unit system of protection offering
considerable economic and technical advantages. Unlike phase and neutral overcurrent
protection, the key advantage of distance protection is that its fault coverage of the
protected circuit is virtually independent of source impedance variations. This is illustrated
in Figure 1.1, where it can be seen that overcurrent protection cannot be applied
satisfactorily. Distance protection is comparatively simple to apply and it can be fast in
operation for faults located along most of a protected circuit. It can also provide both
primary and remote back-up functions in a single scheme. It can easily be adapted to
create a unit pr'otection scheme when applied with a signalling channel. In this form it is
eminently suitable for application with high-speed auto-reclosing, for the protection of
critical transmission lines.

Relay R, setting >7380fi


Therefore, for relay operation for line f a u l t s ,

Relay current setting <6640A a n d >7380A
This is impractical, overcurrent r e l a y not s u i t a b l e
Must use Distance or U n i t . P r o t e c t i o n

Figure I .1: Advantages of distance over overcurrent protection

. . . .
. :. . . . . . . .. .
. .. . . .. . . .
. .

Page 3


Principies of Distance Relays

Since the impedance of a transmission line is proportional to its length, for distana
measurement it is appropriate to-use a relay capable of measuring the impedance of :
line up to a predetermined point (the reach point). Such a relay is described as a distana
relay and is designed to operate only for faults occurring between the relay location an(
the selected reach point, thus giving discrimination for faults that may occur in differen
line sections.
The basic principle of distance protection involves the division of the voltage at th
relaying point by the measured current. The apparent impedance so calculated
compared with a predetermined impedance (normally the impedance of the circuit tieir
protected multiplied by some factor), known as. the reach point. If the measurc
impedance is less than the reach point impedance, it is assumed that a fault exists on-tt
line between the relay and the reach point.
The reach point of a relay is the point along the line impedance locus that is intersect
by the boundary characteristic of the relay. Since this is dependent on the ratio of volta
and current and the phase angle between them, it may be plotted on an RMdiagram. 7
loci of power system impedances as seen by the relay during' faults, 'power swings c
load variations may be plotted on the same diagram and in this manner the performar
of the relay in the preser,ce of system faults and disturbances may be studied.


Relay Performance
Distance relay performance is defined in terms of reach accuracy and operating t
Reach accuracy is a comparison of the actual ohmic reach of the relay under prac
conditions with the relay setting value in ohms. Reach accuracy particularly depend
the level of voltage presented to the relay under fault conditions. The imped:
measuring techniques employed in particular relay designs also have an impact.
Operating times can vary with fault current, with fault position relative to the relay sc
and with the point on the voltage wave at which the fault occurs. Depending o
measuring techniques employed in a particular relay design, measuring signal t r a ~
errors, such as those produced by Capacitor Voltage Transformers or saturating
can also adversely delay relay operation for faults close to the reach point. It is us1
electromechanical and static distance relays to claim both maximum and mir
operating times. However, for modern digital or numerical distance relays, the va
between these is small over a wide range of system operating conditions an1
1.3.1 ElectromechanicallStatic Distance Relays

With electromechanical and earlier static relay designs, the magnitude 1

quantities particularly influenced both reach accuracy and operating time
customary to present info-mation on relay performance by voltagelreach
as shown in Figure 7.2, and operating timelfault position curves for variou
of system impedance ratios (S.I.R.'s)as shown in Figure 1.3,where:

Page 4



ZS= system source impedance behind the relay location

ZL = line impedance equivalent to relay reach setting

% relay rated v o l t a g e



Phase-earth f a u l t s

% relay rated voltage

(b) Phase-phase f a u l t s


% relay rated voltage

Three-phase and three-phase-earth f a u l t s

Figure 1.2: Typical impedance reach accuracy characteristics for Zone I

10 20 30 40 50 60 70 80 90 100
Fault position (% relay setting)

(a) With system impedance ratio of I/?

Fault position (Oh relay setting)

( b ) With system impedance ratio of 3011

Figure 1.3: Typical operation time characteristics for Zone 1 phase-phase faults
Alternatively the above information was combined in a family of contour curves,
where the fault position expressed as a percentage of the relay setting is plotted
against the source to line impedance ratio, as illustrated in Figure 1.4.

: .



. . .


. ...
. .. .
. . . . . .. . .. .
. . . ..:...' . . . . . .:.
. . . . . . . . . .. . . ,
' : !
0.4 ............ J-! .


. ...

: : i


. .. .

. . ..

o o1

. . . - . .. .. . . .



. . . . . . . :....

. .


. . . .
. . . _ . . . . .

. . .:

02 -


. . . - . .


... . ..
. -. :
.:. , .. .
I :

... ...



/ - -

. .



Z p , or S i.R.
(a) Zone 1phase-phase fault: minlmum operation tlmeS


.I . . . . . ,


. . . . . . . .. .: .' ;.'




.'.I ................ I , .

, .

, -

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .. I.. . . . . . . . . .
i : ;;4-I.: ............. - !...-.i
.. ,
....... .-...
,. I

I : .

. .



. .

_ . . __,.__

. . . . . ..



. ..


I . .



Z,fZ, or S1.R.

Zone 1 phase-pnase fault: maximum operation times

Figure 1.4: Typical operation-time contours

. . .

1.3.2 DiqitalINumberical Distance Relays

DigitalINumerical distance relays tend to have more consistent operating times.

They are usually slightly slower than some of the older relay designs when
operating under the best conditions, but their maximum operating times are also
less under adverse waveform
conditions or for boundary fault conditions.

Relationship Between Re!ay Voltzge and ZS!ZL Rs;tio

A single, generic, equivalent circuit, as sbown in Figure 1.5(a), may represent any fault
condition in a three-phase power system. The voltage V applied to the impedance loop is
the open circuit voltage of the power system. Point R represents the relay location; iR and
VR are the current and voltage measured by the relay, respectively.
The impedances Zs and ZLare described as source and line impedances because of their
position with respect to the relay location. Source impedance Zs is a measure of the fault
level at the relaying point. For faults involving earth it is dependent OP the method of
system earthing behind the relaying point. Line impedance ZL is a measure of the
impedance of the protected section. The voltage VR applied to the relay is, therefore,
For a fault at the reach point, this may be alternatively expressed in terms of source to
line impedance ratio Zs/ZL are by means of the following expressions:

where :

. .

therefore :


(Zs / Z L )


... Equation 1.I

(a) P o w e r s y s t e m configuration

. .


. . . . . . .



012 0.3


i.. . . ! ..

. .
i ; ! !




,. . . . . . . . . . . . . .

: .


System ~ m p e d a n c eratio



.. . . .;.:. ....; !! . ./!

..:.. ....................

3 4 5




(b) Variation of relay voltage w ~ t hsystem source to line impedance r a

Figure 1.5: Relationship between source to line ratio and relay voltage

The above generic relationship between VR and Z d Z L , illustrated in Figure 1.5(b), i

for all types of short circuits provided a few simple rules are observed. These are:
For phase faults, V is the phase-phase source voltage and Z d Z L is the positive sec
source to line impedance ratio. VR is the phase-phase relay voltage and IR
is the
phase relay current, for the faulted phases

VR =

(Zs / Z L ) + 1

"P -P

... Equation 1.2

i. F o i earth faults, V is the phase-neutral source voltage and Zs/Zl

is a composite ratio involving the positive and zero sequence impedances'. V
phase-neutral relay voltage and IR
is the relay current for the faulted phase

Paoe 8


Voltaae Limit for Accurate Reach Point Measurement

The ability of a distance relay to measure accurately for a reach point fault depends on
the minimum voltage at the relay location under this condition being above a declared
value. This voltage, which depends on the relay design, can also be quoted in terms of an
equivalent maximum ZslZL or S.I.R.
Distance relays are designed so that, provided the reach'point voltage criterion is met,
any increased mea~uringerrors for faults closer to the relay will not prevent relay
operation. Most modern relays are provided with healthy phase voltage polarisation
andlor memory voltage polarisation. The prime purpose of the relay polarising voltage is
to ensure correct relay directional response for close-up faults, in the f o h a r d or reverse
direction, .where the fault-loop voltage measured by the relay.may be very small. . . .
Zones o f Protection
Careful selection of the reach settings and tripping times for the various zones of
measurement enables correct co-ordination between distance relays on a power system.
Basic distance protection will comprise instantaneous directional Zone 1 protection and
one or more time-delayed zones. Typical reach and time settings for a 3-zone distance
protection are shown in Figure 1.6. Digital and numerical distance relays may have up to
five zones, some set to measure in the reverse direction. Typical settings for three
forward-looking zones of basic distance protection are given in the following sub-sections.
To determine the settings for a particular relay design or for a particular distance
teleprotection scheme, involving end-to-end signalling, the relay manufacturer's
instructions should be referred to.

zone' 1 Settinq
Electromecha~nicallstaticrelays usually have a reach setting of up to 80% of the
protected line impedance for instantaneous Zone 1 protection. For
digitallnumerical distance relays, settings of up to 85% may be safe. The resulting
15-20% safety margin ensures that there is no risk of the Zone 1 protection overreaching the protected line due to errors in the current and voltage transformers,
inaccuracies in line impedance data provided for setting purposes and errors of
relay setting and measurement. Otherwise, there would be a loss of discrimination
with fast operating protection on the following line section. Zone 2 of the distance
protection must cover the remaining 15-20% of the line.

Page 9

1.6.2 _Zone2 Setting

To ensure full coverage of the line with allowance for the sources of error already
listed in the previous section, the reach setting of the Zone 2 protection should be
at least 120% of the protected line impedance. In many applications it is common
practice to set the Zone 2 reach to be equal to the protected line section +50% of
the shortest adjacent line. Where possible, this ensures that the resulting
maximum effective Zone 2 reach does n.ot extend beyond i i ~ erniniiiiiii-ii zffective
Zone 1 reach of the adjacent line protection. This avoids the need to grade the
-Zone 2 time . settings between upstream and downstream relays. In
electromechanical and static relays, Zone 2 , protection is provided either by
separate elements or by extending the reach of the Zone 1 elements after a time
delay that is initiated by a fault detector. In most digital and numerical relays, the
Zone 2 elements are implemented in software.
Zone 2 tripping must be time-delayed to ensure grading with the primary relayins
applied to adjacent circuits that fall within the Zone 2 reach. Thus completc
coverage of a line section is obtained, with fast clearance of faults in the first 80
85% of the line and somewhat slower clearance of faults in the remaining sectior
of the line.

A- <*




Zone 1 = 80435% of protected line impedance
Zone 2 (minimum)= 120% of protected line
Zone 2 (maximum) < Protected line + 50% of shortest second line
Zone 3F =-I .2 (protected line + longest second line)
Zone 3R = 20% of protected line

X = Circuit breaker tripping time

Y = Discriminating time

Figure 7.6: Typical time/distance characteristics for three zone distance protection

Remote back-up protection for all faults on adjacent lines can be provided t
third zone of protection that is time delayed to discriminate with Zone 2 protec
plus circuit breaker trip time for the adjacent line. Zone 3 reach should be set 1
least 1.2 times the impedance presented to the relay for a fault at the remote
of the second line section.

.. ..


Zorie 3 Settinq

Page 10


. ..

On interconnected power systems the effect of fault current infeed at the remote
busbars will cause the impedance presented to the relay to be much greater than
the actual impedance to the fault and this needs to be taken into amount wilen
setting Zone 3. In some systems, variations in the remote busbar infeed can
prevent the application of remote back-up Zone 3 protection. but on radial
distribution systems with single end infeed, no difficulties should arise.

1.6.4 Settinqs for Reverse Reach and Other Zones

Modern digital or numerical relays may have additional impedance zones that can
be utilised to provide additional protection functions. For example, where the first
three zones are set as above, Zone 4 might be used to provide back-up protection
for the local busbar, by applying a reverse reach setting of the order of 2.5% of the
Zone 1 reach. Alternatively, one of the forward-looking zones (typically zone-3)
could be set -with a small reverse offset reach from the origin of the- RIX diagram,
in addition to its forward reach setting. An offset impedance measurement
characteristic is non-directional. One advantage of a non-directional zone of
impedance measurement is that it is able to operate for a close-up, zeroimpedance fault, in situations where there may be no healthy phase voltage signal
or memory voltage signal available to allow operation of a directional impedance
zone. With the offset-zone time delay bypassed, there can be provision of 'SwitchOn-To-Fault' (SOTF) protection. This is required where there are line voltage
transfoimers, to provide fast tripping in the event of accidental line energisation
with maintenance earthing clamps left in position. Additional impedance zones
may be deployed as part of a distance protection scheme used in conjunction with
a teleprotection signalling channel.






1.7. Distance
Relay characteristics
. .

. . .


Some numerical relays measure the absolute fault impedanceand t h e n determine

whether operation is required according to impedance boundaries defined on the R/X
diagram. Traditional distance relays and numerical relays that emulate the impedance
elements of traditional relays do not measure absolute impedance. They cornpare the
measured fault voltage with a replica voltage derived from the fault current and the zone
impedance setting to determine whether then fault is within zone or out-of-zone. Distance
relay impedance comparators or algorithms which emulate traditional comparators are
classified according to their polar characteristics, the number of signal inputs they have,
and the method by which signal comparisons are made. The common types compare
either the relative amphtude or phase of two input quantities to obtain operating
characteristics that are either straight lines or circles when plotted on an R/X diagram. At
each stage of distance relay design evolution the development of impedance operating
characteristics shapes and sophistication has been governed by the technology available
and the acceptable cost. Since many traditional relays are still in service and since some
numerical relays emulate the techniques of the traditional relays, a brief review of
impedance comparators is justified.
1.7.1 Amplitude and Phase Comparison
Relay measuring elements whose functionality is based on the comparison of two
independent quantities are essentially either amplitude or phase comparators. For
the impedance elements of a distance relay, the quantities being compared are
the voltage and current measured by the relay. There are numerous techniques
available for performing the comparison, depending on the technology used. hey
vary from balanced-beam (amplitude comparison) and induction cup (phqse
comparison) electromagnetic relays, through diode and operational .amplifier

Page 11

. .


. .




comparators ir! static-type-distance relays, to digital sequence comparators in

digital relays and to algorithms used in numerical relays.

Any type of impedance characteristic obtainable with one comparator is also

obtainable with the other. The addition and subtraction of the signals for one type j
of comparator pioduces the required signals to obtain a similar characteristic using
the other type. For example, comparing V and I in an amplitude comparator results ~.
in a circular impedance ch3racteistlc-sentred at the ofiglr! of the--=diagram.
the sum and difference of V and I are applied to the phase comparator the result is
a similar characteristic.
Plain Impedance Characteristic

This characteristic takes no &ccount of the phase angle between the current an
the voltage applied to it; for this reason its impedance characteristic when plotted
on an R/X diagram is a circle with its centre at the origin of the co-ordinates and
radius equal to its setting in ohms. Operation occurs for all impedance values le
than the setting, that is, for all points within the circle. The relay characteris
shown in Figure 11.7, is therefore nondirectional, and in this form would oper
for all faults along the vector AL and also for all faults behind the busbars up to
impedance AM. It is to be noted that A is the relaying point and R A B is the a
by which the fault current lags the relay voltage for a fault on the line A 5 and
is the equivalent leading angle for a fault on line AC. Vector A 5 represents the
impedance in-front of the relay between the relaying point A and the end o
AB. Vector A C represents the impedance of line A C behind the relaying point. AL
represents the reach of instantaneous Zone 1 protection, set to cover 80% to 85%
of the protected line.
. .

Line GK

Line G H

Figure 1.7: Plain impedance relay characteristic

Page 12


Iirect i I
el erne

(a) C h a r a c t e r i s t i c of c o m b i n e d directional!
impedance r e l a y

2 e- ...a.


; Source


. - . . ..





-.._- ..\





. ;,


(b) Illustration of u s e of directional!impedance

circuit diagram


: ,R
d i s t a n c e element at A
directional element at A
( c ) Logic f o r directional and impedance
elements at A
Figure 1.8: Combined directional and impedance relays


.. r



Page 13

A relay using this characteristic has three important disadvantages:


it is non-directional; it will see faults both in front of and behin

point, and therefore requires a directional element to give it


it has non-uniform fault resistance coverage


it is susceptible to power swings and heavy loading of a long line, be

the large area covered by the impedance .circle

Directional control is an essential discrimination quality for a distance

make the relay non-responsive to faults outside the protected line. Thi
obtained by the addition of a separate directional control element. The i
characteristic of a directional control element is a straight line on the RI
so the combined characteristic of the directional and impedance re
semi-circle APLQ shown in Figure 1.8.
If a fault occurs at F close to C on the parallel line CD, the directio
will restrain due to current IF,. At the same time, the impedance unit i
from operating by the inhibiting output of unit RD.If this control is not
under impedance element could operate prior to circuit breaker C opening.
Reversal of current through the relay from IF,to IF2 when C opens could then result
in incorrect tripping of the healthy line if the directional unit Ro operat
impedance unit resets. This is an example of the need to consider t
ordination of multiple relay elements to attain reliable relay perfor
evolving fault conditions. ,. In older . relay designs, ,.the.type. of ..p
addressed was commonlyiefefred toas orie of 'coritact race'.



1.7.3 Self-Polarised Mho Relay


The mho impedance element is generally known as such

characteristic is a straight line on an admittance diagram. It cleverly combines the
discriminating qualities of both reach control and direction~lcontrol, thereby .X
eliminating the 'contact racet problems that may be encountered with separate
reach and directional control elements. This is achieved by the addition of a 2
polarising signal. 'Mho' impedance elements were particularly attractive for $
economic reasons where electromechanical relay elements were employed. As a ,!x
result, they have been widely deployed worldwide for many years and their -:;
advantages and limitations are now well understood. For this reason they are still jj
emulated in the algorithms of some modern numerical relays.

The characteristic of a 'mho' impedance element, when plotted on an R/X .:

diagram, is a circle whose circumference passes through the origin, as illustrated r,
in Figure 11.9(b). This demonstrates that the impedance element is inherently
directional and such that it will operate only for faults3n the forward direction along . .
line AB.







.. +



The impedance characteristic is adjusted by setting Z, the impedance reach, :'

along the diameter and cp, the angle of displacement of the diameter from the R::
axis. Angle cp is known as the Relay Characteristic Angle (RCA). The relay,
operates for values of fault impedance ZF within its characteristic.
. ..:

It will be noted that the impedance reach varies with fault angle. As the line to be
protected is
up of resistance and inductance,. its fault angle will be
dependent upon the relative values of R and X at the system operating frequency.
Under an arcing fault condition, or an earth fault involving additional resistance,
such as tower footing resistance or fault through vegetation, the value of the
resistive component of fault impedance will increase to change the impedance
angle. Thus relay having a characteristic angle equivalent to the line ang!e will
under-reach under resistive fault conditions.

( a ) P h a s e comparator i n p u t s



( b ) M h o impedance c h a r a c t e r i s t i c

Page 15

Relay i m p e d a n c e s e t t i n g
Relay c h a r a c t e r i s t i c angle setting
GL P r o t e c t e d l i n e
PQ Arc r e s i s t a n c e
6' L ~ n eangle


Figure 1.9: Mho relay characteristic

It is usual, therefore, to set the RCA less than the line angle, so that it is possi
to accept a small amount of fault resistance without causing under-rea
However, when setting the relay, the difference between the line ar
"EQUATION MISSING" and the relay characteristic angle cp must be known. resulting characteristic is shown in Figure 11,9(c) where AB corresponds to
length of the line to be protected. With cp set less than 0, the actual amount of
protected, AB, would be equal to the relay setting value AQ multiplied by CO!
(0-cp). Therefore the required relay setting AQ is given by:

Page 16

Distance Protection Schemes




. . . ,., .

Distance Protection

Conventional time-stepped distance protection is

illustrated i n Figure 12.1. One of the main disadvantages
o f this scheme is t h a t the instantaneous Zone 1
protection at each end of the protected line cannot be



in these zones are cleared in Zone 1 time by the

protection at one end o f the feeder and i n Zone 2 time
(typically 0.25 t o 0.4 seconds) by the protection at the
other end o f t h e feeder.



. . . .

l a ) Srcppcd tirncldisrancc



(b) Trip circuit (solid statc logic)



.. .

This situation cannot be tolerated in some applications.

for t w o main reasons:
a. faults remaining on the feeder for Zone 2 time may
cause the system t o become unstable
b. where high-speed auto-reclosing is used. the non-

time' during the auto-reclo;e

cycle for the fault to

cause permanent lockout o f the circuit breakers at

each end o f the line section


-- .. . . .. . - .











Even where instability does not occur, fhe increased

duration of the disturbance may give rise t o power
quality problems, and may result i n increased plant


. ... .. . ......


. ..

-... .-

Unit schemes of proieciiori tii%icompare the ccnditinns at

the two ends of the feeder simultaneously positively
identify nhether the fault is internal or external t o the
protected section and provide high-speed protection for
the whole feeder length. This advantage is balanced by the
fact-that the unit scheme does not provide the back up
protection for adjacent feeders given by a distance scheme.
The most desirable scheme is obviously a combination of
the best features o f both arrangements, that is,
instantaneous tripping over the whole feeder length plus
back-up protection t o adjacent feeders. This can be
achieved by interconnecting the distance protection
relays a t each end o f the protected feeder by a
communications channel. Communication techniques
are described i n detail i n Chapter 8.

Zone 2

Zonc 3
(b) Simplilicd logic

The purpose of the communications channel is t o

transmit informarion about the system conditions from
one end of the protected line t o the other. including
requests t o initiate or prevent tripping of the remote
circuit breaker. The former arrangement is generally
known as a 'transfer tripping scheme' while the latter is
generally known as a 'blocking scheme'. However, the
terminology o f the various schemes varies widely,
according t o local custom and practice.

is o u t of service: Reversion t o the
reach setting occurs only a t the end of the reclaimt
For interconnected lines, the Z1X scheme is establis

This scheme is intended for use with an auto-reclose

facility, or where no communications channel is
available, or the channel has failed. Thus it may be used
on radial distribution feeders, or on interconnected lines
as a fallback when no communications channel is
available, e.g. due to maintenance or temporary fault.
The scheme is shown i n Figure 1 2 . 2 .

The disadvantage of the Zone 1 extension scheme is th

external faults within the Z1X reach o f the relay result
tripping of circuit breakers external to the fault

The Zone 1 elements of the distance relay have t w o

settings. One is set t o cover 80010 of the protected line
length as in the basic distance scheme. The other, known
as 'Extended Zone l ' o r 'ZlX', is set to overreach the
protected line, a setting of 120010 of the protected line
being common. The Zone 1 reach is normally controlled
by the Z 1 X setting and is reset t o the basic Zone 1 setting
when a command from the auto-reclose relay is received.

circuit breakers operate.



. ?..
:-., .. . -


::;:e.: ....



-,< s-.



. .:..-. . . .
,,..::. .
.-*. -. .:. . . . .
2 s : \ - L > . : : . . ., )I.*

Fi =-





A contact operated by the Zone 1 relay element is

arranged to send a signal to the remote relay requesting a
trip. The scheme may be called a 'direct under-reach
transfer tripping scheme', 'transfer trip under-reaching
scheme', or 'intertripping under-reach distance protection
scheme', as the Zone 1 relay elements do not cover the
whole of the line.


p..:. .
. .. .-

2 5 : ; . ...


Fj:: F;J"rr


The direct under-reach transfer tr~ppingscheme described

above is made more secure by supelvising the received
signal with the operation of the Zone 2 relay element before
a ~ ~ b w i nang instantaneous trip; as shown in Figure 12.5. The
scheme is then known as a 'permissive under-reach transfer
tripping scheme' (sometimes abbreviated as PUP 22
scheme) or 'permissive under-reach distance protection', as
both relays must detect a fault before the remote end relay
is permitted to trip in Zone 1 time.

(bl Faul: within Zonc I cx:cnsion rcach o f distance rclays

(doublc circuit lincs)

72.:'. F:.?;,,,, :,,l-:<C






; :*i;.,:l:.i..c

..... !i. . . ,.- a:.:

:,;:,c . r L .::.,


r ' .~ . : ~ ,.


!. .

\;A number of these schemes are available, as described

below. Selection of an appropriate scheme depends on
i,the requirements'of the system being protected.

----c Signal scnd


LThe simplest way of reducing the fault clearance time at

Ithe terminal that clears an end zone fault in Zone 2 time
;Isto adopt a direct transfer trip or intertrip technique, the
'logic of which is shown in Figure 12.4.

Signal rcccivc -0

3 .

(a] Signal logic

7Sggnal scnd

Signalling cquipmcnl
-End A

(bl Signalltng arrangcmcnt


Signal r c c c w c




*; *

!?. $2":;.:,


..... .+-.
. -.<'..\;p
"t .'
%5. ,.fir"
' J

2 -:,,&:'

A fault F i n the end zone at end B i n Figure 12.l(a)

results i n operation of the Zone 1 relay and tripping o f
the circuit breaker a t end B. A request t o trip is also sent
' t o the relay at end A. -The receipt of a signal a t A
initiates tripping immediately because the receive relay
contact is connected directly t o the trip relay. The
disadvantage o f this scheme is the possibility o f
undesired tripping by accidental operation or
maloperation o f signalling equipment, or interference on
the communications channel. As a result, i t is not
c o m ~ [ a n l yused.





.%- ::

.r t.

. . ?-&C


_ :L -2..

. . )........
_ .....

.. - . .


IchaPl2 exe



Page 196


A variant o f this scheme, found on some relays, allows

tripping by Zone 3 element operation as well as Zone 2,
provided the fault is i n the forward direction. This is
sometimes called the PUP-Fwd scheme.
Time delayed resetting o f the 'signal received' element is
required t o ensure that the relays at both ends o f a
single-end fed faulted line o f a parallel feeder circuit
have time t o trip when the fault is close t o one end.
Consider a fault F i n a double circuit line, as shown i n
Figure 12.6. The fault is close to end A, so there is
negligible infeed from end B when the fault a t F occurs.
The protection at B detects a Zone 2 fault only after the
breaker at end A has tripped. I t is possible for the Zone 1
element a t A to reset, thus removing the permissive
signal t o B and causing th: 'signal received' element a t
B t o reset before the Zone 2 unit at end B operates. I t is
therefore.,necessary to delay the resetting sf the 'signal
received' element to ensure high speed tripping a t end B.

relays that share the same measuring elements for both'

Zone 1 and Zone 2. In these relays, the reach of thca
measuring elements is extended from Zone 1 t o Zone 2 S
by means o f a range change signal immediately, instead:
o f after Zone 2 time. It is also called a n 'accelerated:
underreach distance protection scheme'.

The under-reaching Zone 1 u n i t is arranged t o send a

signal to the remote end o f the feeder i n addition to
tripping the local circuit breaker. The receive relay:
contact is arranged to extend the reach o f the measurini
element from Zone 1 to Zone 2. This accelerates theA
fault clearance at the remote end for faults that lie i n the!
region between the Zone 1 and Zone 2 reaches. The
scheme is shown i n F~gure12.7. Modern distance relays,
do not employ switched measuring elements, so the,
scheme IS likely t o fall into disuse








. -

;. ..



- 2:,




(a1 Fault occurs-bus bar vollagc low so

ncgligiblc fault currcnl via cnd B






. "9

(b) Signal l o g ~ c

(bl End A r c l a v clcars fault and c u r r c n l

s t a r t s lccdnng from cnd R




The PUP schemes require only a single communications

channel for two-way signall~ngbetween the line ends, as the
channel is keyed by the under-reaching Zone 1 elements
When the circuit breaker a t one end is open, or there is
a weak infeed such that the relevant relay element does
not operate, instantaneous clearance cannot be achieved
for end-zone faults near the 'breaker open' terminal
unless special features are included, as detailed in
section 12.3.5.

This scheme


applicable only to zone sw~tchedd~stance

- :.+

. I..:keq

In this scheme, a distance relay element set t o re?:
beyond the remote end of the protected line is used:
send an intertripping signal t o the remote end. Howev;
it is essential that the receive relay contact is monito!:
by a difectional relay contact t o ensure that trippl!
does not take place unless the fault is within
protected section; see Figure 12.8. The instantane6
contacts of the Zone 2 unit are arranged t o send
signal, and the received signal, supervised by zone;
operation, is used to energise the trip circuit.
scheme is then known as a 'permissive over-rc
transfer tripping scheme' (sometimes abbreviate1
'POP'). 'directional comparison scheme', or 'permi:
ov~rreachdistance p r o t e c f i o ~scheme'.






. . . . . .

w Signal


.: . .







s i g n a ~ i v c ~ F ~ f p ' p ~


(a] Signal logic

The above scheme using Zone 2 relay elements is often

referred to as a POP 22 scheme. An alternative exists
that uses Zone 1 elements instead of Zone 2 , and this is
referred to as the POP 2.1 scheme.

Since the signallinc channel is keyed by over-reaching Zone

2 elements, the scheme requires duplex communication
c'nannels - one frequency for each direction of signalling.
If distance relays with mho charac~eristicsare used, the
scheme may be more advantageous than the permissive
uhder-reaching scheme for protecting short lines,
because the resis:ivc coverage of the Zone 2 umt may be
greater than that o f Zone 1.
To prevent opera:ion under current reversal conditions in
a parallel feeder circuit, i t is necessary to use a current
reversal guard tirnrr to inhibit the tripping of the forward
Zone 2 elements. Otherwise maloptration of the scheme
may occur under current reversal conditions. see Section
11.9.9 for more details. It is necessary only when the
Zone 2 reach is set greater than 1 5 0 0 f o of the protected
line impedance.
The timer is used to block the permissive trip and signal
send circuits as shown in Figure 12.9. The timer is
energised if a signal is received and there is no operation
of Zone 2 elemer,:~. An adjustable time delay on pick-up
(1,) is usually set to allow instantaneous tripping to take
place for any ir,ternal faults, taking into account a
possible slower operation of Zone 2. The timer will have
operated and blocked the 'permissive trip' and 'signal
send' circuits by the time the current reversal takes place.
The timer is de-energised i f the Zone 2 elements operate
or the 'signal received' element resets. The reset time
delay (),I of the timer is set to cover any overlap in time
caused by Zone 2 elements operating and the signal
resetting at the remote end. when the current i n the
healthy feeder reverses. Using a timer in this manner
means that no extra time delay is added in the
permissive trip circuit for an internal fault.

In the standard permissive over-reach scheme, as with

the permissive under-reach scheme, instantaneous
clearance cannot be achieved for end-zone fau& under . !-: , .t <. .
weak infeed or breaker ppen cqnditions. ~ o o v e r c b m e.:..,.;-.. . . . .:
. ,.:
.. . .
this disadvantage, two possibilities exist.
. . . . . . .c-..:.



. -

The Weak lnfeed Echo feattire available i n s ' 6 m e 1.:::

protection relays allows theremote relay toecho the trip .
signal back to the sending relay even if the appropriate
remote relay element has not operated. This caters for
conditions of the remote end having a weak infeed or
circuit breaker open condition, so that the relevant
remote relay element does not operate. Fast clearance
for these faults is now obtained at both ends of the line.
The logic is shown in Figure 12.10. P, time delay (T,)is
required in the echo circuit to prevent tripping of the
remote end breaker when the local breaker is tripped by
the busbar protection or breaker fail protection
associated with other feeders connected to the busbar.
The time delay ensures that the remote end Zone 2
element will reset by the time the echoed signal is
received at that end.



'POP' signal


lap12 exe





Signal transmission can take place even after the remote

end breaker has tripped. This gives rise t o the possibility
of continuous signal transmission due to lock-up of both
signals. Timer T, is used to prevent this. After this time
delay, 'signal send' is blocked.

initiated a t any end of the protected section.

A variation on the Weak lnfeed Echo feature is to allow

tripping o f the remote relay under the circumstances
described above, providing that an undervoltage
condition exists, due to the fault. This is known as the
Weak lnfeed Trip feature and ensures that both ends are
tripped i f the conditions are satisfied.

The arrangements described so far have used the signalling

channel(s) to transmit a tripping instruction. I f the
signalling channel fails or there is no Weak lnfeed feature
provided, end-zone faults may take longer t o be cleared.

to two variants of the scheme.

Blocking over-reaching schemes use an over-reaching

distance scheme and inverse logic. Signalling is initiated
only for external faults and signalling transmission takes
place over healthy line sections. Fast fault clearance
occurs when no signal is received and the over-reaching
Zone 2 distance measuring elements looking into the line
operate. The signalling channel is keyed by reverselooking distance elements (23i n the diagram, though
relay used].
which zone is used depends o'n-the
An ideal blocking scheme is Figure 12.11.

as a 'directional comparison blocking scheme' or a

'blocking over-reach distance protection scheme:

(a1 D ~ s r a n c c l r ~ r ncharaclcristics

- 12 -

Chacncl In 5 c ~ i c c

a Signal scnd


A fault at F 1 is seen by the Zone 1 relay elements at,,

both ends A and B; as a result, the fault is cleared
instantaneously at both ends o f the protected l i n q

Zone 1 elements.

Signalling cquipmcnl
-End A

Signalling cquipmcnl

-End B

(cl Signalling arrangcmcn(




Page 1 9 9

.+:: .: .:,
L-,;,;.3.:;+;:,;. ..:. .

- - 1.
-+.:<.I.:.; - 4 . ~:-..;
52 . ;.


nal transmission takes place, since the

ernal and the fault is cleared i n Zone 1 time a t
after the short time lag (STL) a t end A.

relay elements a't end B associated w i t h

the 2 2 elements a t end A f r o m tripping, the

neously by the protection o n line section B-C,

lity o f Zone 2 elements initiating tripping and the

I c ~ k i n g Zone 3 elements failing t o see an
I fault.
This would result i n instantaneous
ing for a n external fault. When the signalling
nel is used for a stabilising signal, as i n the above
takes place over a healthy line section
is used. The sigrialling channel
uld then be more reliable when used i n the blocking
e-than, i n tripping mode.

I n a practical application, the reverse-looking relay

elements may be set w i t h a forward offset characteristic
t o provide back-up protection for busbar faults after the
zone time delay. It is then necessary t o stop the blocking
signal being sent for internal faults. This is achieved b y
making the 'signal send' circuit conditional upon nonoperation o f the forward-looking Zone 2 elements, as
shown i n Figure 12.1 3.


_-:.. _.*_ ..
.. .. .,.: -1.. '
:,>: ,:..,
$ ..:
,.- ,.~.%....

Blocking schemes, like t h e permissive over-reach

scheme, are also affected b y the current reversal i n the
healthy feeder due t o a fault i n a double circuit line. If
current reversal conditions occur, as described i n section
11.9.9, it may be possible for the maloperation o f a
breaker on the healthy line t o occur., To avoid this, the
resetting o f the 'signal received' element provided i n the
blocking scheme is time delayed.
The timer w i t h delayed resetting (t,) is set t o cover the
time difference between the maximum resetting time o f
reverse-looking Zone 3 elements and the signalling
channel. So, if there is a momentary loss o f the blocking
signal during the current reversal, the timer does n o t
have time t o reset in the blocking mode t r i p circuit and
n o false tripping takes place.



. .

:'&ientialt h a t , t h e operating times o f the various

.bPskilfully co-or'dinated f o r all system conditions,
@thatsufficient time is always allowed for the receipt
K=:blocking signal from t h e remote end o f the feeder.
#.this is n o t done accurately, the scheme may t r i p for an
$ernal f a u l t or alternatively, the end zone tripping
imes may be delayed longer than is necessary.

This is s ~ m i l a rt o the BOP 22 scheme described above.

except that an over-reaching Zone 1 element is used i n
the logic, instead o f the Zone 2 element. It may also be
known as the BOP Z1 scheme.

f the signalling channel fails, the scheme must be

$ranged t o revert t o conventioaal basic distance
uotection. Normally, the blocking mode trip circuit is
upervised b y a 'channel-in-service' contact so that the
locking mode trip circuit is isolated when the channel is
ut of service. as shown i n Figure 12.1 2.

The protection a t the strong infeed terminal will operate

for all internal faults, since a blocking signal is not
received from the weak infeed terminal end. In the case
of external faults behind the weak infeed terminal, the
reverse-looking elements a t that end will see the fault
current fed from the strong infeed terminal and operate,
initiating a block signal t o the remote end. The relay a t
the strong infeed end operates correctly without the
need for any additional circuits. The relay a t the weak
infeed end cannot operate for internal faults, and so
tripping of that breaker is possible only b y means o f
direct intertripping from the strong source end.









. 12.



The permissive over-reach scheme described i n Section ;:.,:. : -,-.. .

12.3.4 can be arranged t o operate o n a directional :.;:+-.):..:,:?
. .
comparison unblocking principle by providing additional .;: . . . .,.:;,
circuitry i n the signalling equipment. I n this scheme ,:::''' -;',.":
. . . . ....
(also called a 'deblocking overreach distance protection

(b] Solid % t a l c logic of rcnd circuit



;Chap12 exe



Page 200

scheme'), a continuous block (or guard) signal is

transmitted. When the over-reaching distance elements
operate, the frequency of the signal transmitted is
shifted t o an 'unblock' (trip) frequency. The receipt o f the
unblock frequency signai arid thc pera at inn crf overreaching distance elements allow fast tripping to occur
for faults within the protected zone. In principle, the
scheme is similar t o the permissive over-reach scheme.
The scheme is made more dependable than the standard
permissive over-reach scheme by providing additional
circuits i n the receiver equipment. These allow tripping
to take place for internal faults even i f the transmitted
unblock signal is short-circuited by the fault. This is
achieved by allowing aided tripping for a short time
interval, typically 100 to 150 milliseconds, after the loss
of both the block and the unblock frequency signals.
After this lime interval, aided tripping is permitted only
i f the unblock frequency signal is received.
This arrangement gives the scheme improved security over a
blocking scheme, since tripping for external faults is possible
only i f the fault occurs within the above time interval of
channel failure. Weak lnfeed terminal 'conditions can be
catered for by the techniques detailed i n Section 12.3.5.
In this way, the scheme has the dependability of a
blocking scheme and the security of a permissive overreach scheme. This scheme is generally preferred when
power line carrier is used, except when continuous
transmission of signal is not acceptable.

On normal two-terminal lines the main deciding factors in

the choice o f the type of scheme, apart from the reliability
o f the signalling channel previously discussed, are
operating speed and the method of operation of the
system. Table 12.1 compares the important characteristics
o f the various types of scheme.


Spccd of opcrarion


Spccd with in-wnicc t e t i n g


SuiUblc lor auto-rcclosc


Scarify against
malopcraticm duc l a :
Cumnr r r v c m l

Lou of communica~~ons
Wrak l n l r r d l O ~ nCB

Not as fast

A$ fast


Special fcalurcs rcqulrcd

Sprcial fcaturcs rcquircd

Sprcial fcaturcs rquircd


Special fcalurn rcquircd

Modern digital or numerical distance relayrare provided

with a choice o f several schemes i n the same relay. Thus
scheme selection is now largely independent of relay
selection, and the user is assured that a relay is available
with all the required features to cope'with changing
system conditions.








Busbar Protection

.?$ ,:*+;++,






~ . . . y * ~ * - ' ,;

.<'-. .
. . .c.. . ...



,. :.
.... ..
&,<: :;* s:..;


. . . .......


... i:,*.i




The protection scheme for a power system should cover

the whole system against all probable types of fault.
Unrestricted forms of line protection, such as overcurrent
and distance systems, meet this requirement, although
faults i n the busbar zone are cleared only after some
time delay. But if unit protection is applied to feeders
and plant, the busbars are not inherently protected.

....'. .-:.,.
,.: .-;..\ .--... '
.:.... . . . .







...; .

:;;../, .

Busbars have often been left without specific protection,

for one or more of the following reasons:
a. the busbars and switchgear have a high degree of
reliability, to the point of being regarded as
intrinsically safe
b. i t was feared that accidental operation o f busbar
protection might cause widespread dislocation of
the power system, which, i f not quickly cleared.
would cause more loss than would the very
infrequent actual bus faults
c. i t was hoped that system protection or back-up

protection would provide sufficient bus protection

i f needed
I t is true that the risk of a fault occurring on modern
metal-clad gear is very small, but i t cannot be entirely
ignored. However, the damage resulting from one
uncleared fault, because of the concentration of fault
MVA, may be very extensive indeed, up to the complete
loss of the station by fire. Serious damage to or
destruction of the installation would probably result i n
widespread and prolonged supply interruption.
Finally, system protection will frequently not provide the
cover required. Such protection may be good enough for
small distribution substations, but not for important
stations. Even i f distance protection is applied to all
feeders, the busbar will lie i n the second zone of all the
distance protections, so a bus fault will be cleared
relatively slowly, and the resultant duration of the
voltage dip imposed on the rest of the system may not be
With outdoor switchgear the case is less clear since.
although the likelihood of a fault is higher, the risk of
widespread damage resulting is much less. In general
then, busbar protection is required when the system
protection does not cover the busbars, or when, in order

.: .










- .





Page 234

t o maintain power system stability, high-speed fault

clearance is necessary. Unit busbar protection provides
this, with the further advantage that if the busbars are
sectionalised, one section only need be isolated to clear
a fault. The case for unit busbar protection is i n fact
strongest when there is sectionallsation.

The majority of bus faults involve one phase and earth,

but faults arise from many causes and a significant
number are interphase clear of earth. In fact, a large
proportion of busbar faults result from human error
rather than the failure of switchgear components.
With fully phase-segregated metalclad gear, only earth
favlts are possible. and a protection scheme need have
earth fault sensitivity only. In other cases, an ability to
respond to phase faults clear of earth is an advantage,
although the phase fault sensitivity need not be very high.

incidence, amounting to no
fault per busbar in twenty years, it is clear that u
the stability of the protection is absolute, the degr
disturbance to which the power system is li
subjected may be increased by the insta
protection. The possibility of incorrect operation h
the past, led to hesitation in applying bus protectio
has also resulted in application of some very co
systems. Increased understanding of the response of
differential systems t o transient currents enable
systems to be -applied w i t h confidence i n their
fundamental stability.
The theory of differential
protection is given later i n Section 15.7.
Notwithstanding the complete stability of a correctly
applied protection system,
number of reasons. These are:
a. interruption of the secondary circuit of a
transformer will pr
might cause trippin
relative values of circuit load and effective setting.
It would certainly do so during a through fault,
producing substantial fault current i n the circuit in

Although not basically different from i t h e r circuit

protection, the key position of the busbar intensifies the
emphasis put on the essential requirements of speed and
stability. The special features o f busbar protection are
discussed below.

b. removal of busbar faults in less time than could be

achieved by back-up line protection, with the
object of maintaining system stability
Some early busbar p r o t e c t i ~ n schemes used a low
impedance differential system having a relatively long
operation time, of up to 0.5 seconds. The basis of most
modern schemes is a d~fferentialsystem using either low
impedance biased or high impedance unbiased relays
capable of operating in a time of the order of one cycle
at a very moderate multiple of fault setting. To this must
be added the operating time of the tripping relays, but an
overall tripping time of less than two cycles can be
achieved. With high-speed circuit breakers. complete
fault clearance may be obtained in approximately 0.1
seconds. When a frame-earth system is used, the
operating speed is comparable.

..',... .



b. a mechanical shoc
cause operation, although the likelihood
occurring- with modern numerical sch
c. accidental interference'with the relay, aris
a mistake during maintenance testing, may lead to

Busbar protection is primarily concerned with:

a. limitation of consequential damage

In order to maintain the high order of integrit

for busbar protection, i t is an almost invariable practice I
to make tripping depend on two independent
measurements of fault quantities. Moreover, if the
tripping of all the breakers within a zone is derived from
common measuring relays, t w o separate elements must -:
be operated at each stage t o complete a tripping i
operation. Although not
the relays are separated
reasonable accidental m
relays simultaneously is possible.



The two measurements may be made by two similar

differential systems, or one differential system may be
checked by a frame-earth system. by earth fault relays
energised by current transformers in the t
neutral-earth conductors or by overcurr
Alternatively. a frame-earth system may be checked by
earth fault relays.
If two systems of the unit or other similar t y
they should be energised by separate
transformers in the case of high impedan
differential schemes. The duplicate ring CT
mounted on a common primary

. .,.

The 5tability of bus pr02ection is of paramount

importance. Bearing in mind the low rate of fault

l J 4









Page 2 3 5

be maintained t h r o u g h o u t the
he case of l o w impedance, biased differential
mes t h a t cater for unequal ratio CTs, the scheme
be energised from either one or t w o separate sets o f

'isthen n o more t h a n t h a t o f normal circuit protection, so

no duplication is required a t this stage. N o t least among
the advantages o f using individual tripping relays is the

Security o f both stability and operation is obtained by

providing three independent channels (say X, Y a n d Z)
whose outputs are arranged i n a 'two-out-of three'
voting arrangement, as shown i n Figure 15.1.

A number o f busbar protection systems have been

a. system protection used t o cover busbars

through a common multi-contact tripping relay.

b. frame-earth protection
c. differential protection

d. phase comparison protection

e. directional blocking protection


? t h e section switch trips b o t h the adjacent zones. This

has sometimes 'been avoided i n the past b y giving the

zone on the faulty side o f the section switch

. ... . . .. . . .. . .


. . . . . . :;..: . . . . . . .


Of these, (a) is-suitable for small substations only, while

(d) and (e) are obsolete. Detailed discussion o f types (b).
and (c] occupies most o f this chapter.






Early forms o f biased differential protection for busbars.

such as versions o f 'Translay' protection and also a
scheme using harmonic restraint, were superseded b y
unbiased high impedance differential protection.




is obtained a t the expense of seriously delaying the bus

This practice is therefore
not generally favoured. Some var~ationsare dealt w i t h
later under the more detailed scheme descriptions. There
are many combinations possible, b u t t h e essential
principle is that n o single accidental incident of a
secondary n a t u r e shall be capable o f causing a n
unnecessary trip o f a bus section.

k?.., protection for all other faults.


,; >\




Security aqainst maloperation is only achieved by

increasing the amount o f equipment that is required t o
function t o complete a n 0peration;'and this inevitably
increases the statistical risk t h a t a tripping operation due
t o a f a u l t may fail. s u c h a failure, leaving aside the
question o f consequential damage, m a y result i n
disruption o f t h e power system t o an extent as great, or
greater, than would be caused by a n unwanted trip. The
relative risk of failure o f this kind may b e slight, but i t
has been thought worthwhile i n some instances t o
provide a guard i n this respect as well.

The relative simplicity o f the latter, and more importantly

the relative ease w i t h which its performance can be
calculated. have ensured its success u p to the present


But more recently the advances i n semiconductor

technology, coupled w i t h a more pressing need t o be able
t o accommodate CT's o f unequal ratio, have led t o the
re-introduction o f biased schemes, generally using static
relay designs, particularly for the most extensive and
onerous applications.
Frame-earth protection systems have been i n use for
many years, mainly associated w i t h smaller busbar
protection schemes at distribution voltages and for
metalclad busbars (e.g. SF6 insulated busbars). However,
i t has often been quite common for a unit protection
scheme t o be used i n addition, t o provide t w o separate

-. .


means o f fault detection.

The different types o f protection are described i n the
following sections.


-..... . .
... .






.: .


.. . .. .

Past 2 3 6

. . . . . . . . .


System protection that includes overcurrent or distance

systems will inherently give protection cover t o the
busbars. Overcurrent protection will only be applied t o
relatively simple distribution systems, or as a back-up
protection, set t o give a considerable time delay.
Distance protection will provide cover for busbar faults
w i t h its second and possibly subsequent zones. In both
cases the busbar protection obtained is slow and suitable
only for limiting the consequential damage.


Switchgear framc



fault rclay

," 0 -


. - --:

. . . .:'. .


... .

The only exception is the case of a mesh-connected

substation, i n which the current transformers are located
a t the circuit breakers. Here, the busbars are included, i n
sections, in the. individual zones o f the main circuit
protection, whether this is o f unit type or not. I n the
special case when the current transformer_s are located
on the line side o f the mesh. he circuit protection will
n o t cover the busbars i n the instantaneous zone and
separate busbar protection, known as mesh-corner
protection, is generally used - see Section for



...L..: ..'

check rclav



. . . ,! "






3 . 2 : .j;,?!,!!,


b. earth current flowing to a

system cannot flow into or out of the



- 15.

Frame leakage protection has been extensively u s e d i n

the' past i n many different si~uations. There are several
variations of frame leakage schemes available, providing
busbar protection schemes w i t h different capabilities.
The following sections schemes have thus been retained
for historical and general reference purposes. A
considerable number of schemes are still i n service and
frame leakage may provide an acceptable solution i n
particular circumstances. However, the need t o insulate
the switchboard frame and provide cable gland
insulation and the availabiliry of alternative schemes
using numerical relays, has contributed t o a decline in
use o f frame leakage systems.

The iwitchgcai must b e insulated i s a

standing it on concrete. Care must
founddtion bolts do not touch the
sufficient concrete must be cut
permit grouting-in w i t h no risk o f touching metalwork :;.$
The insulation to earth finally achieved will not be high, $
a value of 10 ohms being satisfactory.


When planning the earthing arrangements of a frame- .',(:?:

leakage scheme, the use of one common electrode for $
both the switchgear frame and the power system neutral '.$.?
point is preferred, because the fault path would :$;
otherwise include the two earthing electrodes i n seriel
I f either or both of these are of high resistance or have
inadequate current carrying capacity, the fault current
may be limited to such an extent that the protection
equipment becomes inoperative. In addition, if the >;&
electrode earthing the switchgear frame is the offender,
the potential of the frame may be raised t o a dangerous
value. The use of a common earthing electrode of
adequate rating and low resistance ensures sufficient
current for scheme operation and limits the rise i n f r a m e i s
potential. When the system is resistance earthed, the ;g
earthing connection from the switchgear frame is made



This is purely an earth fault system and, i n principle.

involves simply measuring the fault current flowing from
the switchgear frame t o earth. A current transformer is
mounted on the earthing conductor and is used to energize
a simple instantaneous relay as shown in Figure 15.2.
No other earth conncctions of any type, including
incidental conncctions t o structural steelwork are
allowed. 'This rcquircmcnt is so that:
a. thc principal carth connection and current
transformer arc not shunted, thereby raising the



earthing electrode.


This risk is small i n practice





Page 237



. . . . .






:, framcZonc



Zonc 11

,- 1- - - -





- - - -1- -

rcsistancc to caYn









,. ; .

Under external fault conditions, the current I , flows

through the frame-leakage current transformer. If the
insulation resistance is too low, sufficient current may
, .
: flow t o operate the framc-leakage relay, and, as the check
....feature is unrestricted, this will also operate t o complete
trip circuit. The earthresistance between the G r t h i n g
:-'electrode and true earth is seldom greater than IR,so
with 10R insulation resistance the current I , is limited to
10% o f . t h e total earth fault current I , and 12. For this
for the
reason, the recommended minimum se;:ing
scheme is about 30% of the minimum earth fault current.



Trip L

Trip i;

Trip iI1

If it is inconvenient t o insulate the section switch frame

o n one side, this switch may be included i n t h a t zone. It
is then necessary t o intertrip the other zone after
approximately 0.5 seconds if a fault persists after the
zone including the section switch has been tripped. This
IS illustrated i n Figure 15.5.



All cable glands must be insulated, to prevent the

circulation o f spurious current through the frame and
earthing system by any voltages induced i n the cable
Preferably, the gland insulation should be
provided i n t w o layers or stages, with an interposing
layer of metal, t o facilitate the testing of the gland
insulation. A test level o f 5kV from each side is suitable.


Zonc I1



Zonc H

Section 15.6.1 covered the basic requirements for a

system t o protect switchgear as a whole. When the
busbar is divided into sections, these can be protected
separately, provided the frame is also sub-divided, the
sections mutually insulated, and each provided w i t h a
separate earth conductor, current transformer and relay.
Ideally, the section switch should be treated as a
Separate zonc, as shown i n Figure 15.4, and provided
w i t h either a separate relay or t w o secondaries o n the
frame-leakage current transformer, w i t h an arrangement
t o t r i p b o t h adjacent zones. The individual zone relays
trip their respective zone and the section switch.


Zonc G



L - - - - - - J

Trap J

Trip A'



.. ,i ?..

' .:







Page 238




For the above schemes to function it is necessary to have

a least one infeed or earthed source of supply, and i n the
latter case it is essential that this source o f supply be
connected t o the side o f the switchboard not containing
the section switch. Further, if possible, it is preferable
that an earthed source of supply be provided on both
sides of the switchboard, in order to ensure that any
faults that may develop between the insulating barrier
and the section switch will continue to be fed with fault
current after the isolation o f the first half o f the
switchboard, and thus allow the fault to be-removed. Of
the two arrangements, the first is the one normally
recommended, since i t provides instantaneous clearance
of busbar faults on all sections of the switchboard.


. ., . .;....,

..: i



as operation due to mechanical shock or mistakes made ,.,$

by personnel. Faults in the low voltage auxiliary wiring
must also be prevented from causing operation by !#
passing current to earth through the switchgear frame ':..$:
A useful check is provided by a relay energised by the,
system neutral current, or residual current. I f the neutral ,f
check cannot be provided, the frame-earth relays should ':
<! .
have a short time delay.

When a check system is used, instantaneous relays can $

be used, with a setting of 30% of the minimum earth ,$
fault current and an operating time at five times setting 3
. ..\,
of 15 milliseconds or less.
Figure 15.7 shows a frame-leakage scheme for a
metalclad switchgear installation similar to that shown
in Figure 15.4 and incorporating a neutral current check !
obtained from a suitable zero sequence current source, ;s'<
such as that shown in Figure 15.2.



It is not generally feasible to separately insulate the

metal enclosures of the main and auxiliary busbars.
Protection is therefore generally provided as for single
bus installations, but with the additional feature that
circuits connected to the auxiliary bus are tripped for all
faults, as shown in Figure 15.6.

Trip relays

A b a r r i c r s
. .

Zonc G



Zonc 1,


74 Alarm canccllat~onrelay

CS5 Control sclcctor swftch protcct~on~n/prolccl~on

L, Busbar prolcclcon In scrvtcc lamp

I, Busbar protcclton oul of scrv~ccl ~ m p

L , lrlpptng supply hcallhy lamp
I , , Alarm and ~ n d ~ c a l ! osupply
hcalthy lamp



! 5 ;.


lk::,cc( r?<p;2,",! ::>I:






rrar;c.-lrc.oyc. ,<!~c:nr



' J



On all but the smallest equipments, a check system

should be provided to guard against such contingencies




The protection relays used for the discriminating an

check functions are of the attracted armature type, wit;
two normally open self reset contacts. The tripfln
circuits cannot be complete unless both ,,.:r
discriminating and check relays operate; this is becay:
the discriminating and check relay contacts
connected in series. The tripping relays are
attracted armature type.



. . .-


Paae 2 3 9


fis usual t o supervise the satisfactory operation o f the

The scheme may consist of a single relay connected t o

the bus wires connecting all the current transformers in
parallel, one set per circuit, associated with a particular
zone, as shown i n Figure 15.8(a). This will give earth
fault protection f o r the busbar. This arrangement has
often been thought t o be adequate.

h t c c t i o n scheme with audible and visual alarms and

&$cations for the following:


busbar faults




ifi service

c. busbar protection out o f service

If the current transformers are connected as a balanced

group for each phase together with a three-element
relay, as shown i n Figure 15.8(b), additional protection
for phase faults can be obtained.

d. tripping supply healthy



e. alarm supply healthy

enable the protection equipment-of each zone t o be

out of service independently during maintenance
periods, isolating switches - one switch per zone - are
'.provided i n the trip supply circuits and an alarm
cancellation relay is used.

The phase and earth fault settings are identical, and this
scheme is recommended for its ease of application and
good performance.

. .. .. .. . . . . . .. .. .. .. .. ................ . . .

. . . .; + ? ! ,,,:. ,;: . L i.
. 3.

'The Merz-Price principle is applicable to a multi-terminal

Zone such as a busbar- The principle is a direct
i . application of Kirchhoffs first law. Usually, the
2f- circulating current arrangement is used, in which the
current transformers and interconnections form an
1;....-analogue of the busbar and circuit connections. A relay
$.F?cQ?nected -across t h c . 0 bus .wires represents a fault
the-prima-w.system i" the analogue and hence i s .
until i f a u l t occurs o n t h e busbar; it then
. .
an input that, in principle at Icast, represents
;::the fault current.



> . . . . .


Each section o f a divided bus is provided with a separate

current system. ~h~ zones so formed are
switches, so that a fault
over-lapped across the
the latter
tripthe two adjacent zones. hi^ is
illustrated in~i~~~~
Tripping two zones for a section switch fault can be
avoided by using thetime-dela~ed technique
However instantaneous
. .
operation is the .
- preferied
- . .



= - f




. . . . . .

Zonc B






a) Basic circulating currcnt schcmc (carth fault protcction only)


Zonc C



Typical fccdcr circuits

F:z!,r< ! 5.9: zo,:<.>

1,: ~ 1 r ~ : c ~ f ~ : l l ~

for d<!l,!,l<. b,,,

. . . . .
. . . . . . . .


For double bus installation, the two busbars will be

treated as separate zones. 'The auxiliary busbar zone will ;:::::.',;..-.' ....:',...
overlap the appropriate main busbar zone at the bus .:.;. . . . - .: .:,';::-. .
.. -,, ; coupler.
..,.;:s....,.:.I,.; ,.:.......
;.: -:;:,:.
,,:- .:. . :;?
Since any circuit may be transferred from one busbar to ;:+?i<~:l3:+;;!f:.
. . . .. . . . .
the other by isolator switches, these and the associated :;.!$.:;?&
?; l/..,;~L:::?:,
tripping circuit must also.,be switched to the appropriate .. :,..::<
..: :'..+~


Diffcrcntial r c h y


b) P ~ P S Cand carth fault circulating currcnt schcmc using

thrcc-clcmcnt rclay


-. . . - . . . . . .


---. i

15.9: C i r c o l o r i n g c u r r c n r schcn;r


,-;:;,:z.;z.'::<:::. :. .

I, . .

.,I . +:.*.
.:, , -e;,>.
..,.; *;;-a,..
:. &

~ , r - . ~ p
l r.rrrri..






'. :&;.;:.;,.,:, Y!?'..

, - -,.cp,;;:<,
. ..+8s;+,$,.$%, ......>






. .


. . . . . .




.. .

. -








:.. :

Figure 15.10(a) shows the ideal arrangement i n which,...:

both the circuit and busbar zones are overlapped leaving!:
no region o f the primary circuit unprotected.

zone by 'early make' and 'late break' auxiliary contacts.

This is t o ensure that when the isolators are closing, the
auxiliary switches make before the main contacts of the
isolator, and that when the isolators are opened, their
main contacts part before the auxiliary switches open.
The result is that the secondary circuits of the t w o zones
concerned are briefly paralleled while the circuit is being
ransferred: these t w o zones have i n anv case been
. :-:. . .
.. united through the circuit isolators during the transfer
. - .....
?:?>:$;: operation.

a small region o f the priniary circuit unprote

unprotected region is typically refcrred t o as the 's
zone: The fault shown will cause operation o f the busbar
protection, tripping the circuit breaker, but the fault will
continue t o be fed from the circuit, if a source of p
is present. It is necessary for the bus protecti
intertrip the far end of the circuit protection, if the
is o f the unit type.

:..\ ..

. . ......
. . . .

Ideally, the separate discriminating zones should overlap

each other and also the individual circuit protections.
The overlap should occur across a circuit bteaker, so that
the latter lies i n both zones. For this arrangement i t is
necessary to install current transformers on both sides o f
the circuit breakers, which is economically possible with
many but not all types o f switchgear. With both the
circuit and thc bus protection current transformers on
the same side o f the circuit breakers, the zones may be
overlapped a t the current transformers, but a fault
between the CT location and the circuit breaker will not
.be completely isolated. This m a t t e r is important i n all
. . . . . :.
. . . . .:., ..
. ...
. .1 .. .... . . switctigear,to '.wl%ch these . conditions apply, and is
:: .i:-;.::,
paaiculaily the case of outdopr switchgear
- .
. .
. .. -. . where
:se.parately 'mbirnted. muiti-secondary current
. .. .
,!s. - transformers are generally 'used. The conditions are.
shown i n Figure 15.10.

With reference -to Figure 15.10(b), special 'short

technique may be used, particularly when th

the fault is i n the switchgear connections
generator; the latter is therefore tripped electrical
shut down on the mechanical side so as to



L a


The protection of c o h e

gives rise t o additional considerations
location. A single mesh corner is s
. . . .


Norc I : Only 1 connection lo (hc rncsh corncr pcrmiltcd

(a) CT arrangcrncnlr for protection including mesh corner




;, :.g
, ,

a. Currcnt lransformcrs mounted on bolh rider of brcakcr

-no unprolcclcd region
b. Currcnt transformers mounted on c ~ r c u side
~ l only of brcakcr

Nocc 2: Multiplccircuils may bc conncclcd

lo rhc mesh corncr
(b1 CT arrangcmcnlr for protccrion additional mesh corncr prolcclion required

-law11 shown not cleared by circuil prolcction


c , ) a . 7 ~s t l i c c . : c c cr,i,r!! f > * c g k c r *):.:,.

. -..- .



. . ..

. . ... . .. . ..





Page 241



: I.

.. . .


onnection t o the mesh is

An equivalent circuit, as i n Figure 15.12, can represent a

irculating current system.


-- ---...-.. -. - -.-...... -. -

--..--- -. - .... -.......


hout any means of determining the faulted
'connection. Protection CT's must therefore be located on . :
each connection, as shown in Figure 15,11(b). This leaves

R a

hown i n Figure 15.11 (b).



. ..... . .



considerations that have to be taken into account are

detailed i n the following sections.

............: : . . ! 2 . E : : ~ ; G < ! ! ? : > i! : : < ; : ; i




!.:.~;:, -:c3 .c, > . .."


: .

G: c ~ ~ f < i i ~ Z <~U ; ~< < LO: : :Z


~ h , current
are replaced inthe diagram by
of flux

is not' detrimental as long as it

linear range of the


ideal current transformerj feeding an equivalent circuit . ............

. that represents the ,magn,etising losses and seco":da,iy=


.resistance, ..and

a i s o . t h e resistan,&



..,: ..

df .;I.:-.;.-



the convecting .leads. ~h~~~. circuits

can then be ,:GI~. . ..:..1
interconnectedas shown, with a relayconnected to the. ':'.'.
junction points t o form the complete equivalent circuit




region of the
characteristic; this
not in itself
a spill
output from a pair of balancing current transformers
are identical and equally
b e of
A group of
transformers, though
the same design, will not be completely identical, but a
important factor is inequaliOl of burden in the
case of a differential system for a busbar, an external
fault may be fed through a single circuit, the current
being supplied to the busbar through all other circuits.
The faulted circuit is many times more heavily loaded
than the others and the corresponding current
transformers are likely to be heavily saturated, while
those of the other circuits are not. Severe unbalance is
., therefore probable, which, with a relay of normal burden,
could exceed any acceptable current setting. For this
reason such systems were at one time always provided
dclay. This practice is, however, no longer


Saturation has the effect of lowering the exciting

and is assumed
take place severeiy in
current transformer H until, at the limil. the shunt
impedance becomes
and the " can produce
output, This condition is represented by a short circuit.
shown i n broken line, across the exciting impedance. it
should be noted that this is not the equivalent of a
physical short circuit, since it is behind the winding
resistance .



Applying the Thkvenin method of solution, the voltage

developed across the relay will be given by:



RR+R,.,i +RUN



. . . . . . . . . .. . . .
. .. ,.:......c
. . ..:........ ,~,!<


The current through the relay is given by:

... ;$:;..>..
?, ...........
-. . ..:.


.: ,,.+,

;;. .




!5 . 2





.. ?




.. . .-... .



.., ..;,.:;.: ,.-.L,.,..,..: :\i_&~y>Tp

>, ,'.,.";*,.' jy;j
. ,?*<.~.?.>,?,
::-'. .;.
; .&d9".+$i,?,.

I . r . - # l * . .




If RR is small. IR will approximate to IF which is ?$@<@$@

unacceptable. On the other hand, i f RR is large IR
is jr.7~:c:22~
reduced. Equation 15.2 can be written, with little error. - ' ' ~ ~ ~ < ? ~ f % ? ~ .
.,+' ?.,.&
:;.:; ...1
as follows:
.:.-....,...... ...::>
.,.. . .
.. ,.;<;
. .A,,
. . ..,,,,':.k.j,;
,.: :.:

... .- . . . . . . . . . . .



R ~ + R f f i + R ~ ~ i

asible to calculate the spill current that may

ortunately, this is not necessary; an alternative
approach provides both the necessary information and the
technique required to obtain a high performance.


>,:. ~+;$;>;~,5~;~<.
,T<-27-,A... . ,..

- ~ , ( R u i +&H)
.,.f ">I!





I ;


..---.. . ..-

- ;a!i.,,i




I # = - -V,

Page 2 4 2

RL + RCI.= lead + CT winding resistance



... E g u o t i o n !5.3

or alternatively:

It remains to be shown that the setting chosen..


It is clear that,by increasing RR, the spill current ZR can be

reduced below any specified relay setting. RR is frequently
increased by the addition of a series-connected resistor
which is known as the stabilising resistor.
It can also be seen from Equation 15.4 that it is only the
voltage drop in the relay circuit at setting current that is
important. The relay can tie designed as a voltage
measuring device consuming negligible culrent; and
provided its setting voltage exceeds the value Vf of
Equation 15.4, the system will be stable. In fact, the
setting voltage need not exceed V/. since the derivation
of Equation 15.4 involves an extreme condition of
unbalance between the G and H current transformers
that is not coppletely realised. So a safety margin is
built-in i f the voltage setting is made equal to Vf

(range 0.7 - 2.0)

The current transformers will have an excitation curve

which has not so far been related to the

winding resistance, with the maximum secondary fa

current flowing through them. Under in-zone fa
conditions it is necessary for the current transformers
produce sufficient output to operate the relay. This will
be achieved provided the CT knee-point voltage exceeds.
the relay setting. In order to- cater for er
to specify that the current transformers
knee-point e.m.f. of at least twice the necessaq'settin
voltage; a higher multiple is of advantage i n ensuring
high speed of operation.

It is necessary to realise that the value of I/to be inserted

in Equation 15.4 is the complete function of the fault
current and the spill current IR through the relay, in the
limiting condition, will be of the same form. I f the relay
requires more time to operate than the effective duration
of the d.ct transient component, or has been designed
with special features to block the d.c. component, then
this factor can be ignored and only the symmetrical
value of the fault current need be entered in Equation
15.4. I f the relay setting voltage. V,, is made equal to VJ
that is, I/ (RL + RCr], an inherent safety factor of the
order of two will exist.

carrying primary cuiwnt or n

strictly speaking be vecto
arithmetically. t t can be expressed as:

IR = IS + l l I C s

cycle and with no special f~aturesto block the d.c.

component, it is the r.m.s. value of the first offset wave


eflective setting

number ofparallel - conilecred CT's

Equation 15.4 as:

Equoliorr 1 5 . 5

stabiliry of schetne

relay circuit voltage setting



p__.__-d A . t . r , r i . m

N ~ f w a r kP r . r r r t i . m



, .


a. phase-phase faults give o n l y 8 6 % o f the threephase fault current

. .-




':h resistance

reauce raulr currents somewhat

c. a reasonable margin should be allowed to ensure
that relays operate quickly and decisively -

~tis desirable that the primary effective setting should not

graceed 30% of the prospective minimum fault current.
earth fault, the minimum earth fault current should be
considered, taking into account any earthing impedance
iy that might be present as well. Furthermore, in the event
1 of the inter.f. is available
in the earth
g;? n u l r currenc. Ine prlmary operating current must
e not greater than 30/0 of the minimum
se earth fau;t current. In order to achieve
ed operation, it is desirable that settings should
r, particularly in the rase of the solidly
of the
power system. The transient
i n conjunction with wnfavourable residual
can cause a high degree of saturation and
utput, possibly leading to a delay of several cycles
time of the element.
91 to the natural


This will not happen to any large degree if the fault

current is a larger multiple of setting; for example, if the
fault current is five times the scheme primary operating
current and the CT knee-point e.m.f. is three times the
relay setting voltage, the additional delay is unlikely to
exceed one cycle.
The primary operating current is sometimes designed to
exceed the maximum expected circuit load in order to
reduce the possibility of false operation under load
current as a result of a broken CT lead. Desirable as this
safeguard may be, it will be seen that i t is better not to
the effective
current settingtoo much, as this
will sacrifice some speed; the check feature
inany case,
maintains stabilitv.
An overall earth fault scheme for a large distribution
board may be difficult t o design because of the large
number of current transformers paralleled together,
which may lead to an excessive setting. It may be
advantageousinsuch a case to provide a three-element
phase and earth fault scheme, mainly
to reduce the
number of current transformers paralleled into
one group.

Ems-high-voltage substations usually present no such

problem. Using the voltage-calibrated relay, the current
consumption can be very small.
A simplification can be achieved b y providing one relay
per circuit, all connected to the CT paralleling buswires:

Zonc R

.......... *.

. .:




Zonc AIZ




Zonc A I I
Bus wircs
Zonc A12
Bur wircl
Bur wi
Chcck 208
Bur wir

~ u l l rn aclay
w m e ar chcct

A11 First main burbar

A12 Second main burbar
K Rcrcwc burbar

L V ~ Cnr

r rcla

u m c ar chcci

--...,..-... ",
samc as chcck
5tabiliring Rcrirtor
Iligh lmpcdancc
Ctrculattng Currcnt




Page 2 4 4






Zonc indicating rclay

Alarm cancclla:~on relay
D.C. volts supcw~sionrelay
High impcdancc circulating current relay
Bus wires supcwision rclay

area and reduces the risk o f accidental operation.

i ..... :;. ; i':>,
: . : : .:;

Schemes for earth faults only can be checked by a frame-

. . .

. .-.


Zonc bus wircs shorting relay

Control selector switch
Indicating lamp protection in scrvicc
Indicating lamp pcorection out of scrvicc

subdivision being necessary. For phase fault sche

the check will usually be a similar type of scheme ap
to the switchboard as a single overall zone.

This enables the trip circuits t o be confined to the least

earth system, applied t o the switchboard as a whole, no



A set of current transformers separate from those use

the discriminating zones should be provided. No"
switching is required and no current transformers





U A,r.n.ri..

: zone i n bus-coupler and busCT Scconciarf Circuits

CT secondary circuit up to the

:tions will cause an unbalance i n
equivalent t o the load being carried by the
c i r c u i t Even though this degree o f
. ~ 'r i o u s output is below the effective setting the
'&dition cannot be ignored, since it is likely to lead t o
a- . b i l i t y under any through fault condition. .


h r v i s i o n can be carried out t o detect s;ch conditions

$ connecting a sensitive alarm relay across the bus
&es of each zone. For a phase and earth fault scheme,
mjnternal three-phase rectifier can be used t o effect a
bthmation of the bus wire voltages o n to a single alarm
see Figures 15.13 and 15.14.

It is possible that special circumstances
involving onerous conditions may over-ride this
convenience and make connection t o some other part o f
the ring desirable.
Connecting leads will usually be not less than 710.67mm
(2.5mm1), but for large sites or in other difficult
circumstances it may be necessary to use cables of, for
example 711.04mm (6mm1] for the bus wire ring and the
CT connections t o it. The cable from the ring to the relay
need not be of the larger section.
When the reserve bar is split by bus section isolators and
the t w o portions are protected as separate zones, it is
necessary to common the bus wires by means o f auxiliary
contacts, thereby making these two zones intb one when
the section isolators are closed.

. .:

he alarm relay is set so that operation does not occur

~iththe protection system healthy under normal load.
ubject t o this proviso, the alarm relay is made as
wsitive as possible; the desired effective setting is 125
imary amperes or 10% of the lowest circuit rating,
hichever is the greater.

This section provides a summary of practical

considerations when implementing a high-impedance
busbar protection scheme.

nce a relay of this order o f sensitivity is likely t o

!:! ::: :: ; .,I. .;:;!:':! .i:r:!j,.;:::,.
. . . ..
%rate;during thrdugh faults; time delay, typically of
I- :-';: - . i&;:s@cd"ds; is
. . avoid unn&cessary alarm,
. .
designed t o correspond to the switchgear rating; even if . . - :
jnals. . - .
-. .. ... - .. .
the available short-circuit power i n the systemis much

.8.5 krrangc:-- .i - ;





...... '..

;s shown in Equation 15.4 how the setting voltage for

liven stability level is directly related to the resistance
the CT secondary leads. This should therefore be
)t to a practical minimum. Taking i n t o account the
lctical physical laying of auxiliary cables, the CT bus
'es are best arranged i n the form of a ring around the
tchgear site.
double bus installation, the CT leads should be taken
:ctly t o the isolator selection switches. The usual
ting of cables on a double bus site is as follows:

a. current transformers to marshalling kiosk

b. marshalling kiosk to bus selection isolator auxiliary

less than this figure, it can be expected that the system

will be developed up to the limit of rating.


Current transformers must have identical turns ratios.

but a turns error of one in 400 is recognised as a
reasonable manufacturing tolerance. Also, they should
preferably be of similar design; where this is not possible
the magnetising characteristics should be reasonably



- .


Current transformers for use with

high impedance
protection schemes should meet the requirements of
Class PX of IEC 60044-1.
. I

The setting voltage is given by the equation


interconnections between marshalling k i d s to

form a closed ring

rclay for each zone is connected to one point of the

bus wire. For convenience of cabling, thc main zonc
ys will be connected through a multicorc cable
vcen the relay panel and the bus section-switch
rhalling cubicle. The reserve bar zone and the check
: relays will be connected together by a cable
ling t o the bus coupler circuit breaker marshalling

v,> I f ( R , + Red
. ,.,
.., ,





. .,..,..
.. .,.<

relay circuit voltage

- steady-state through fault curreirt


CT secondary ruinditrg resistntrce





., . , .
;3 *>:.





. : ,..!I
.. ;I$
,.i .

.: ,.wk;


: I$I.*

."..L,?.: '


', ,~
.: , I.,,.



Page 246

- ,.
. .





f5.&.6.<</:(:'<-3:?j17;.,;:.-..- ..~fa:I:;~:ie:-?:
, . . . ,.,r..;Ii!:,...'..r\



secondary condition is:

This is given by the formula

. - $2:




.The effective setting of the relay is given by

I f =fault current
= exciting
k current a t knee - point voltage


VK= knee - point voltage


= relay-circuit current settirlg
= CT
exritatiol~current at voltage seffi~lg
simple combination of burden and exciting impedance

n = number of CTS i n parallel

by the CT turns
For the primary fault setting multiply IR

These formulae are therefore to be regarded only as

guide to the possible peak voltage. With large current
transformers, particularly those with a low sesonda
current rating, the voltage may be very high, above

It is clear from Equations 15.4 and 15.6 that i t is

advantageous to keep the secondary fault current low;

this is done by making the CT turns ratio high. I t is
common practice to use current transformers with a
secondary rating of 1A. .
I t can be shown that there is an optimum turns ratio for
.. . ;.I the current transformers; thiswalue depends on all the
. .
but is generally about 200011.
' ' .
though a lower ratio, f o r instance 40011, is often
. employed; the use of the optimum ratio can result. in a
considerable reduction in the physical size of the current




. . ..





,. . .

, .,/..



15 .

When the burden resistance is finite although high, an


V p= peak voltage del)elopcd


VF= prospective voltagc i n abserlce ojsaturation

This formula does not hold for the open circuit condition
and is inaccurate for very high burden resistances that
approximate to an opcn circuit, because simplifying
assumptions used in the derivation of the formula are
not valid for the extreme condition.

,: . ,

c; ...?--

where C is a constant depending on dimensions and

a constant in the range 0.2-0.25.

voltage setting depends on t h e value of ~ ; . i ' n

keep the shunting effect t o a minimu
recommended to use a non-linear resistor with a value of


..-c.-. approximate formula for the peak voltage is:

VK =

v = CP

,: ,: .

Under in-zone fault conditions, a high impedance relay

constitutes an excessive burden to the current
transformers, leading to the development of a high
voltage; the voltage waveform will be highly distorted
but the peak value may be many times the nominal
saturation voltage.

ceramic non-linear resistor in parallel

having a characteristic given by:

Another approach applicable to the opcn circuit


Instantaneous attracted armature relays are used. Simple

fast-operating relays would have a low safety factor
constant in the stability equation, Equation 15.5, as
discussed in Section 15.8.1. The performance is improved
by series-tuning the relay coil, thereby making the circuit
resistive in effect. Inductive reactance would tend to
reduce stability, whereas the action of capacitance is to
block the unidirectional transient component of fault
current and so raise the stability constant.
An alternative technique used in some relays is to apply
the limited spill voltage principle shown in Equation
15.4. A tuned element is connected via a plug bridge to
a chain of resistors; and the relay is calibrated in terms
of voltage.

The principles of low impedance differential protection

have been described in Section 10.4. including the
principle advantages to be galned by the use of a bias.





Page 2 4 7

Most modern busbar protection schemes use

principles of a check zone, zone selection, and .

gements can still be applied. Current
transformer secondary circuits are not switched directly
by isolator contacts b u t instead by isolator repeat relays
after a secondary stage of current transformation. These
switching relays form a replica of the busbar within the
protection and provide the complete selection logic.

En .




;!< With some biased relays, the stability is not assured by.


the through current bias feature alone, b u t is enhanced

by the addition of a stabilising resistor, having a value
:; which may be calculated as follows.



The through current will increase the effective relay

i minimum operating current for a biased relay as follows:


= eflecfive 1nii1i111uii1o p r n t i ~ i g
- .


G = relay s e n i i ~ gc u n e i l t

It must be recognised though that the use of any technique

for inhibiting operation, to improve stability performance
for through faults, must not be allowed t o diminish the
abilii\i of the relay to respond to ir~iernaifauiu.

For an internal fault, and with no through fault current

flowing, the effective setting (IR) is raised above the
basic relay setting (Is) by whatever biasing effect is
produced by the sum of the CT magnetising currents
flowing through the bias circuit. With low impedance
biased differential schemes particularly where the busbar
installation has relatively few circuits, these magnetising
currents may be negligible, depending on the value o f Is.

. . .

The basic relay setting current was formerly defined as

the minimum current required solely i n the differential
circuit to cause operation - Figure 15.45(a].
This $@@:1i&3&.
approach simplified snalysis of performance. b u t was
considered t o be unrealistic, as i n practice any current -y*+c::7..~.
. .flowing i n the differential circuit must flow i n a t least
. .
one half o f the relay bias circuit causing the practical
minimum operating current always to be higher than the .... ,nominal basic setting current.. As a.. result,.-a:
: .... ._:................
later: :...-...,. +.- . . ..,L....r2,;
.-.' . definition, as shown i n Figure
. . . . . . . . . . . . . . . ..:-:.




From Equation 15.4, the value of stabilising resistor is

given by:



CT ratio


It is interesting to note that the value of the stabilising
resistance is independent of current level, and that there
would appear to be no limit to the through-faults
stability level. This has been identified [15.1] as 'The
Principle of Infinite Stability:
The stabilising resistor still constitutes a significant
burden on - the current transformers during internal
An alternative technique, used by the MBCZ system
described i n Section 15.9.6, is to block the differential
measurement during the portion of the cycle that a
current transformer is saturated. If this is achieved by
momentarily short-circuiting the differential path, a very
low burden is placed on the current transformers. In this
way the differential circuit of the relay is preventtd from
responding to the spill current.


Conversely, ifneedsto be. appreeiif& tkit

later definition of relay setting cu.rrent, whicti. flows . : -..$:;i=
;:!.:-.:...-:-::-..-.:-'-I through at least half the bias circuit, the notional mini- i ~ r :".rl
mum operation current i n the differential circuit alone
is somewhat less, as shown in Figure 15.15(b).
Using the d e f i n i t i w presently.applicable, the effective
minimum primary operating current


is generally much greater than Is,
the relay
= B k approximately.
effective current, IR

. - , ........
. . . . . . . . . - . . .;,
applying' .thi~.1:6:.....1<.~....~~...r.

(al Suocrxdcd definition

(bl Currcnt





Page 2 4 8

Unless the minimum effective operating current of a

scheme has been raised deliberately to some preferred
value, it will usually be determined by thy check zone,
when present, as the latter may be expected to involve
the greatest number of current transformers in parallel.
A slightly more onerous condition may arise when two
discriminating zones are coupled, transiently or
otherwise, by the closing of primary isolators. .,
It is generally desirable-to 'attain an effective primary
operating current that is just greater than the maximum
load current, to prevent the busbar protection from
operating spuriously from load current should a
secondary circuit wiring fault develop. This consideration
is particularly important where the check feature is either
not used or is fed from common main CTs.

For some low impedance schemes, only one set of main

CT's is required. This seems to contradict the general
principle of all busbar protection systems with a check
feature that complete duplication of all equipment is
required. but it is claimed that the spirit of the checking
principle is met by making operation of the protection-,
as directional',
dependent on two different criteria such .
. . . . ...
and d.iffere?tialmeasllrements.

isolators may provide the latter.

auxiliary relays within the protection. Theserelays form


them to be connected into this busbar replica.

transformers available per circuit Where the facility of
a check zone is still required, this can still be achieved
with the low impedance biased protection by connecting


.In the MBCZ scheme, described i n Section

15.9.6, the
provision of auxiliary CT's as standard for ratio matching
also provides a ready means for introducing the check
feature duplication at the auxiliary CT's and onwards to
the relays. This may be an attractive compromise when
3nly one set of main CT's is available.

. ..



. L. . . .
. . ,. . ....
. .

In low impedance schemes the integrity of the CT

secondary circuits can also be monitored. A current
operated auxiliary relay, or element of the main
protection equipment, may be applied to detect any
unbalanced secondary currents and give an alarm after a
time delay. For optimum discrimination, the current

. . .. ,

setting of this supervision relay must be less than that of

the main differential protection.

.. .

In modern busbar protection schemes, the supervision of

the secondary circuits typically forms only a part of a
comprehensive supervision facility.


.. .


. ..



..~. .

. -.
I .





~ , , : ,

It is a common modern requirement of low impedance

schemes that none of the main CT secondary circuits
should switched. in the previously convcptional manner,
to match the switching of primary circuit isolators.



range of CT mismatch.


:., .,:,;;<.:;

- TYl"


A separate module is used for each circuit breaker and

also one for each zone of protection. In addition to these

. ,.\ior::,


particular busbar installation. Additional modules can be

added at any time as the busbar is extended.

. ..

.! :;;i,

- 8 . .

. .. .~












F;scre i s . 17: J y 2 c :,!3CZ b:iI;o! i c : c r : ! < . n >ir::b,;i:;.i

t.'.:.. . berwccfl c8.rco8.:::?;+<,:?
o,>d p,:t;,:c:+,:,;

plug-in buswire connections




;I: figure


15.17 shows the correlation between the circuit

and the protection modules for a typical double
~ ~ % u s.b 8 r i n s t a l l a t i o nIn
. practice the modulesare mounted
@?nia,,multi-tier rack or cubicle:


$<<vemodules a r c interconnected via a multicore cable


ETthat is plugged i n t o the back o f the modules. There are

five main groups o f buswires, allocated for:

i. protection for main b u s b a ~

'cI "


ii. protection for reserve busbar

.r... .. . .


" ..
r .J

protection for the transfer busbar. When the

reserve busbar is also used as a transfer bar then
this group o f buswires is used

5. .;

iv. auxiliary connections used by the protection to

combine modules for some of the more complex
busbar configurations




protection for the check zone


E;: One extra module. n o t shown i n this diagram, is plugged

h.?: into the multicore bus. This is the alarm module, which
; Wntains the common alarm circuits and the bias resistors.
[<%-Thepower supplies are also fed in through this module.
if#^. f 5 3, ;;,.!


. >

@;,All zones o f measurement arc biased by the t o t a l current

B. .flowing t o or f r o m the busbar system via the feeders.
h i s cnsurcs t h a t all zones o f measurement w i l l have
similar fault sensitivity under all load conditions. The
bias is derived from the check zone and fixed a t 20%
, w i t h a characteristic gencrally as shown i n Figure
. lS.lS(b). Thus some ratio mismatch is tolerable.





The traditional method for stabilisinga differential relay

is t o add a resistor t o the differential path. Whilst'this
improves stability it increases the burden on the current
transformer for internal faults. The technique used hi
the MBCZ scheme overcomes this problem.
The MBCZ design detects when a 'CT is saturated and
short-circuits the differential path for the portion o f the
cycle for wk,ich saturation occurs. The resultant spill
current does not then f l o w through the measuring circuit
and stability is assured.
This principle allows a very low impedance differential
circuit t o be developed t h a t will operate successfully
with relatively small CT's.

. .

. .

If the CT's carrying fault current are n o t saturated there

will be ample current i n the differential circuit t o operate
the differential relay quickly for fault currents exceeding
the m i n i m u m operating level. which is adjustable
between 20%-200% rated current.

When the only CT(s) carrying internal fault current

become saturated, i t might be supposed that the CT
saturation detectors may completely inhibit operation by
short-circuiting the differential circuit. However, the
resulting inhibit pulses remove only an insignificant
portion o f the differential current, so operation of the
relay is therefore virtually unaffected.


Out of scrvicc



. .:







. . . . .-..
. . .
. . .;::
. . .j <;: .. '. .:.,



- ---

. . . . . . .. . .

As shown in Figure 15.18, each measuring module

contains duplicated biased differential elements and also
a pair of supervision elements. which are a part of a
comprehensive supervision facility.

This arrangement provides supervision of CT secondary

circuits for both open circuit conditions and any
. impairment of the element to operate for an internal
fault, without waiting for an actual system fault
to show this up. for a zone to operate it is
necessary for both the differential supervision element
and the biased differential elementto operate. F~~ a
circuit breaker to be tripped it requires the associated
: . l j .
zone to be ooerate,-j
and alsothe
.. *
. . -...
.."' :. -.






-r(:or;;r;ng o n i f

. ..


i -:





9::?$:i :i;vu,::r::

5 .3



. . .

cheek zonc

Main zonc



rhis is avoided by using a 'master/follower' arrangerneri<

By making the impedance of one of the measuring
elements very much higher than the other it is possibletp
ensure that one of the relays retains its original minimum
operation current. Then to ensure that both the ~
connected zones are tripped the trip circuits of the
zones are connected in parallel Any measuring unit can


have the role of 'master' or 'follower' as it is selectable by

means of a switch on the front of the module.

. , . .



Serious damage may result, and even danger to life, if,;

These schemes are qenerally based on the assumptiol



circuit breaker fails to open when called upon to do S$

To reduce this risk breaker fail protection schemes wefl
developed some years ago.

i +vc


to operate the two busbar sections as a single bar:-:&

' !:
fault currgnt will then divide betweenthe two meas&Gn,@
elements in the iatio'of their impedances. If both of th$
two measuring elements are of low and equal impedang
the effective minimum operating current of the scheme
will be doubled.
. .*<

. . . . . . .. . .



; 7..




When two sections of a busbar are connected together

by isolators i t will result in two measuring elements
being connected in parallel when the isolators are closed


i t has failed to function. The circuit breakers in the

stage back in the system are then automatically trippet
For a bus coupler or section breaker this would invc
tripping all the infeeds to the adjacent zone, a
that is included in the busbar protection scheme.


,. r!




Page 2 5 1

fibre optic link

Ccntral Unit

Systcm Communication Nctwork

PU: Pcriphcral Unit

CU: Ccntral Unit

... '

*;> .



. . . ., ..
. .





d. dead zone protection

. .

The application of numerical relay technology t o busbar

!.'.. protection has lagged behind that of other protection.
\?<functions. Static technology is still usual for such
, . .
. ::&hemes,
!-. . -~. . .
b u t numerical technology. is j o w ~ i e a d ' i l y ".'.
The very latest d e v e l o p m ~ ~ t si n t h e .
are included, such as extensive use o f a data
bus t o link the various units involved, and fault tolerance
against loss o f a particular link by providing multiple
i communications paths. The development process has
been very rigorous, because the requirements for busbar
j protection i n respect-of immunity to maloperation are
very high.
The philosophy adopted is one of distributed processing of
the measured values, as shown in Figure 15.20. Feeders
each have their own processing unit, which collects
together information on the state of the feeder (currents,
voltages. CB and isolator status, etc.) and communicates
it over high-speed fibre-optic data links to a central unit.
For large substations, more than one central unit may be
used, while i n the case of small installations, all of the
units can be co-located, leading to fhe appearance of a
traditional centralised architecture.


feederr. interface units at a

may be used
w i t h the data transmitted t o a single centrally
located peripheral unit. The central unit performs the
calculations required for the protection functions.
Available protection functions are:

a. protection
b. backup overcurrent protection
c. breaker failure

In addition, monitoring functions such as CB and isolator

monitorinq, disturbance recordinq and transformer
supervision ace.provided:
. :...
- . . . . .... . . . .. ... ... .


. . . . . . .

. .

Because .- o f , . the : , distributed :-topology. used, r.:l".'

:.:;.: .
- .
. . . . .
~ ~ n c h r o n i s a t i oon f the 'me&urernents. .taken-'.by
.-:;. .. .-pe;ifiheral u"its .is:of vital iipo.rt+ce. '.'rX1high&ability;
numerically-c6ntr;11e& oscillator is fitted-in'each o f the
central and peripheral units, with time synchronisation . . . . .
between them.
I n the event o f loss o f the
. .
synchronisation signal, the high stabilityof the oscillator
i n the affected feeder unit(s) enables processing o f the
~ncomingdata to continue without significant errors
: :2
. . .
until svnchronisation can be restored.





The peripheral units have responsibility for collecting the

required data, such as voltages and currents, and
processing it into digital form for onwards transmission
to the central unit. Modelling o f the CT response is
included, t o eliminate errors caused by effects such as CT
saturation. Disturbance recording for the monitored
feeder is implemented, for later download as required.
Because each peripheral unit is concerned only with an
individual feeder, the protection algorithms must reside
in the central unit.
The differential

algoriihm can be much more

sophisticated than w i t h earlier technology, due t o

improvements in processing power
addition to
calculating the sum of the measured currents, the
algorithm can also evaluate differences between
successive current samples, since a large change above a
threshold may indicate a fault - the threshold being
choscn such that normal load changes, apart from inrush
conditions do not exceed the threshold. The same







Page 2 5 2

considerations can also be applied t o the phase angles o f

currents, and incremental changes i n them.
One advantage gained f r o m the use o f numerical
technology is the a h i ! i q )ls easi!\t re-configure the
protection t o cater for changes i n configuration o f the
substation. For example, addition o f a n extra feeder
involves the addition o f an extra peripheral unit, the
fibre-optic connection t o the central u n i t and entry via
the M M I o f the new configuration i n t o the central unit.
Figure 15.21 illustrates th,e latest numerical technology

In considering the introduction o f numerical busbar

protection schemes, users have been concerned w i t h
reliability issues such as security and availability.
Conventional high impedance schemes have been one o f
the main protection schemes used for busbar protection.
The basic measuring element is simple i n concept and
has few componen:s. Calculation o f stability limits and
other setting parameters is straightforward and scheme
performance can be predicted w i t h o u t the need for
costly testing. Prac;ically. h i g h impedance schemes have
proved t o be a very reliable f o r m o f protection.

I n contrast, modern numerical schemes are more

.complex w i t h a much greater range o f facilities and a :\
m u c h high component count. Based o n low impedanc
bias techniques, and w i t h a greater range o f facilities t
,. <.+-.
set, setting calculations can also be more complex.
. ..:
However, studies o f t h e comparative reliability of
conventional high impedance schemes and .modern
.~..numerical schemes have shown that assessing relative
reliability is not quite so simple as i t might appear. The ->&?
numerical scheme has t w o advantages over its older ,;$;;i



a. there is a reduction i n the number. o f external

components such as switching and other auxiliary
relays, rcany o f the f u n c t i o n s o f which ?re
performed i n t e r n a l l y w i t h i n the software



b. numerical
monitoring features which provide alarm facilities
i f the scheme is faulty. In certain cases, siniularion
of the scheme functions can be (performed on line
f r o m the CT inputs through i o the tripping o ~ i p d t c
and thus scheme functions can be checked on a
regular basis t o ensure a f u l l operational mode is
available a t all times
Reliabil~tyanalyses using fault tree analysis methods
have examined issues o f dependability (e.q. the ability t o
operate when required) and security (e.g.-the ability n o t
t o provide spurious/indiscriminate operation). These
analyses have shown that:

a. dependability o f numerical schemes is better than

conventional high impedance schemes
b. security o f numerical and conven:ional
impedance schemes are comparable


I n addition, an important feature o f numerical schemes

is the in-built monitoring system. This considerably
improves t h e potential availability o f numerical schemes
compared t o conventional schemes as faults w i t h i n the
equipment and i t s operational state can be detected and
alarmed. W i t h the conventional scheme, failure t o reinstate the scheme correctly after maintenance may not
be detected u n t i l the scheme is required t o operate. In
this situation, i t s effective availability is zero until it is
detected and repaired.

15.1 The Behaviour o f Current Transformers subjected

t o Transient Asymmetric Currents a n d the
Effects o n Associated Protective Relays. J.W.
Hodgkiss. ClGRE Paper Number 329. Session
15-25 J'une 1960.


Motor Protection


serious loss of production may result.


The following table indicates typicall protection depending on the size of the motor. However,
other factors should be considered when selecting motor protection, for example importance of


1. Fuses
2. Fuses + ' ~ h e r r n a l
Overload + U N


Options - Stalling &


overcurrent +

phase and earth faults.

The protection must be able to distinguish between abnormal conditions and normal motor
operation. Therefore, it is important to understand the behaviour of the motor under certain
conditions to be able to apply protection successfully. For example, the magnitude and
duration of the starting current affects the application of overload protection; the magnitude and
maximum allowable duration of stalling current in relation to those of staring current determir.~e
whether separate stalling protection is required.


The tolerance to overload of motors depends on the motor design and differs considerably iri
Practice. The risk of damage of the insulation depends on the temperature. It is very difficult il
not impossible to cover all types and ratings of motors with different applications, variety (1;

Page 1

If a motor is assumed to be a homogeneou

and dissipating heat at a rate directly proportional
temperature at any instant is given by :


where T,,


= final steady teniperature

f = heating time constant

This assumes a thermal equilibrium in the form :

Heat developed = Heat stored + Heat dissipated.
Temperature rise is proportional to the current square.

Thus T

K I ,: - ! 1 - e

I, is that current \vhich produces the rated

temperature rise T,; when flo\vs
continuously in the motor.

For an overload current 1 the temperature rise is given by :











For the motor not to exceed the rated temperatu

rhe motor can withstand the current I is obtained by equating equations (1) and (2) with t = OC in


equatim (1).



!. .

I-lcncc K I , , -





t =







rload protection should satisfy the above r

:notor current or a percentage of it, depending on the motor design.

is an over-simplification to regard a motor as a homogeneous .body. It actually comprises

ral parts each with a characteristic surface area, mass, heat capacity, thermal conductivity
rate of heat production. The temperature rise of different parts or even of various point in
e same part may be very uneven. However, it is reasonable to consider that the current-time
lationship follows an inverse fashion.
b h i l e infrequency overloads of short duration may not damage the motor, sustained overload
;of a few percent may result in premature ageing and failure of insulation so that the time lag
!:..characteristic of the device is of vital importance in permitting the normal starting duty and
/:pmviding close sustained overload protection for the motor at the same time.



:A Direct-On-Line machine (DOL) will typically draw a starting current of approximately 6 times
\full load current for a period defined by the machines starting time. This is because the
:impedance of the machine is related to the slip frequency, which varies during start up; the
:impedance beirlg smaller at low speeds where the slip is larger.
With normal 3-phase supply, should a motor stall when running, or be unable to start due to
excessive load, it will draw a current equivalent to the locked rotor current. On the basis of
starting current being equal to locked rotor current it is not possible to distinguish between
3-phase stalling and healthy starting by monitoring the current alone.
In the majority of cases, the starting time of a normal induction motor is less than the maximum
stalling. time allowable to avoid excessive deterioration of the motor insulation.. Under this.
condition it is possible to discriminate on a tim.e basis between the two and provide' protection
against. stalling. In applications where the stalling time is less than the startingtime 'such' as
motors driving high inertia loads, it is more difficult to discriminate between a healthy start and a
stall condition. A separate stalling relay may be required depending on the type of overload
xotection relay used and the ratio of normal starting time to the allowable stall time.
The following conditions may be examined
lssume startrng current = stall current

~ S L

= maximum starting time

= niaximum allowable stall time

Thermal relay operating time at

the same current level < t , ~but

In this case the thermal relay can

Protect the motor against
3-phase stall, no separate
stalling relay is required.





Thermal relay operating time at

the same current level > t , ~and

In this case no stalling protection
is provided by the thermal
overload relay even though the
stall time is greater than the
starting time. A separate stalling
relay is required. If the difference
between tsL and t s ~
is adequate
to cater for relay errors a simple
single phase definite time overcurrent relay may be used.


'1 ----------------

. .,

Is current setting < locked rotor current hut > load current
ts time setting < t s ~but > t s ~

. . ..





OIC = overcurrent
TD = time delay
86 = trip relay
tsL > TD > tsT

- a a



. ..








In this case a separate stalling relay in the form of a definite time over-current relay and a shaft
monitoring device are required. The latter is used to check the motor speed while the relay
measures the motor current. Instead of the overcurrent relay a simple definite time delay relay
may also be used as shown below :

Page 4






Use of a tachoswitch monitor with a definite time delay relay:-



! '

The tacho contact will open

when the set speed (say 10% of
rated speed) is reached. It m x t
operate well within TD.
MSD = Motor switching device
auxiliary contact, closed
when the motor is
switched on.



TD < tSL


Use of a tachoswitch monitor with a definite time overcurrent relay:-

This offers more reliabil~ty

TD < tsL
OIC < stall current,
> load current


-0-OcFTRl P

Use of a 2-stage definite time overcurrent relay:-




Page 5

>t s ~
<t s ~
< stall current
> load current

No protection during motor starting

is continuously
energised when the motor is in

If tsL > tsT the same arrangement can be used in which case stalling protection is provided
during the starting period. This method provides additional advantage for motors with different
hot and cold stall times in that TD2 cah be set to less than hot stall time irrespective of cold stall :,




(TD1 + TD2)

< tsL (cold)


The supply voltage to a 3-phase induction motor can become unbalanced due to such reasons
as single phase load, imperfect transportation of feeders etc. The degree of unbalance is small
in normal installation except when onephase become ope'n circuited. This would not affect at
first sight, the motor to any large extent, but a small voltage unbalance could produce a much
larger negative phase sequence current in the winding due t o the relative small negative phase
sequence impedance of the machine compared with the positive phase sequence impedance.
Consider the following equivalent circuits for positive and negative phase sequence currents,
the magnetising impedance being neglected:
. ... .. .....,:
& ,?


....-:,: ..,.

.. .
. ..-.





- - '5.






' ' .

With positive phase sequence voltages a rotating field will be set up and the rotor will rotate in
the direction of rotation of the filed giving a slip s and slip frequency sf. With negative phase
sequence voltages the field will rotate in the opposite direction cutting a rotating rotor conductor
at almost twice the frequency. The actual frequency of negative phase sequence voltage and
current in the rotor circuit is (2 - s)f. From the equivalent circuits:
Motor +ve sequence impedance at a given slip s

= [ ( R ~+ R ' ~ )

+ (XI + X',


when s = 1 at standstill.

Page 6

Motor -ve sequence impedance at a given slip s

(R, +

' 22

' 22

+ (x, +


when s << 1 at normal running speed


The value of resistance is generally much less than the leakage reactance.


j neglecting the resistance term the motor -ve phase sequence impedance at normal running

speed can be approximated to the +ve phase sequence impedance at standstill.


; . ~ tnormal running speed :


+ve sequence impedance

-ve sequence impedance

starting current
normal load current

If a motor has a starting current 6 x the full load current, the -ve sequence impedance would be
about 116'~of the +ve sequence impedance.
Therefore if 1 pu +ve sequence voltage applied to the motor would produce 1 pu of +ve
;.;..sequence current, the same 1 pu of -ve sequence voltage would produce 6 pu - ve sequence
Consequently, if there is 5% -ve sequence voltage present in the supply it would result
. .
approximate 30% of -ve sequence component of current.


~. ..

'The ac resistance of the rotor conductor to the induced -ve sequence current is greater than the
dc resistance due to the higher frequency [(Z-s)fl causing skin effect. The heating effect of -ve
sequence current is therefore greater and increases the motor losses. The machine output
must be reduced to avoid overheating.
Because of the reversed rotation of the magnetic field due to -ve sequence current, a small -ve
torque is also produced.
As mentioned previously one unit of -ve phase sequence current has a greater heating effect
than one unit of +ve phase sequence current, this unequal heating effect should be taken into
account in the design of a thermal characteristic based on:

I equivalent =


where 11 = +ve sequence component

12 = -ve sequence component
N = a fixed constant
A typical value of n in motor protection relays is 6. This value has been carefully chosen to
provide adequate protection to both the stator and rotor windings for all designs of motor
without causing nuisance tripping.

Page 7




.-. .. ..

. ... -_


.. .



Assume a balanced 3-phase supply:

Normal starting current

IA 2 (VnN.Z)/2= Standstill impedance per phase of the motor

With one phase open-circuited ssy

C phase :

i.e, Starting current with one phase open circuited = 0.866 x normal starting current.

Page 8

i.e. + ve sequence current =

Similarly, I 2 =


- (ItA +

- normal starting current.

a IaB

normal starting current.

.'--. For

delta-connected winding motors the actual line starting current with one phase open circuit

e:. is the same as a wye-connected machine :


:Ir. :

For delta-connected winding motors the

line starting current with one
7 ;.
. . .:. . .phase
open circuit is the same as a
5;-bye-connected machine:-

ii;. actual

<f: -


Normal starting current =


,fi x z

Actual starting current = 21122



x normal starting current

= 0.866 x normal starting current

Note that one winding will carry twice the current in the other two windings


Page 9


On loss of one phase supply while starting the motor will remain stationary.. It has been sho
the motor will draw a current equal to 0.866 x the normal starting current. Therefore, if
measuring the total stator current is used it must have a time delay longer than the starting ti
of the motor. If the allowable stall time at that level of current is less than the starting ti
simi!ar arr=lr?gementsas in the case of 3-phase stalling protection have to be used.
However, it has also been shown that the negative phase sequence component present in
current is equal to half the normal starting current. A negative phase sequence curren
can therefore detect this condition. In the CTM relay an instantaneous negative
sequence current detector is fitted. It has a setting of 2-8 x rated current. If a setting of
normal starting current the relay will detect single phase stalling condition.


It is difficult to shown in simple mathematical terms the behaviour o f the motor when one ph
supply is lost with the motor running due to the complex nature of the s l i ~
calculation and
possibility of additional negative phase sequence current being fed into the motor from par
equipment. However, the following would happen:

Heating increases considerably due to high rotor losses caused by the -ve s


Output of motor is reduced and depending on the load it could stall altogether.


. ..


Motor current increases..


.. .


Ir: many installations such as lift motors and conveyors, protection is occasionally requ
ensure correct direction of rotation. Although not damaging to the motor this can be detri
to the process.
Under reversed phase sequence conditions the relay is designed to respond to the ex
negative phase sequence component of current. A number of methods can be
disconnect the motor from the supply during this condition
Instantaneous Negative Sequence Overcurrent Relay - This will respond very quic
load current is sufficient on the system.
Time Delayed Thermal Trip - As mentioned previous the thermal overload prot
influenced by the negative phase sequence component of the current, this elemen
more benificial for smaller loads.

. . .,




The disadvantage of the above methods is that in order for them to operate the motor must be:;
switched on, dpending on the inertia of the motor it may start to turn in the wrong direction.
this is unacceptable then a negative phase sequence voltage monitoring device can be used- *.
This device will monitor the phase rotation of the incoming supply to the motor and if interlocked.:
with the motor switching device will prevent closure onto a revese phase sequence supply. his':
approach is also used when the motor can only draw very low load currents.

Page 70



!for induction motor the torq;e
developed is approximately proportional to the square of the
Tapplied voltage. Low voltage level prevents motors from reaching rated speed on starting or
'may draw heavy current on losing speed. Some form of undervoltage protection is therefore
desirable with suitable time delay to disconnect the motors when severe low voltage conditions
tpersist for more than a few seconds. The time delay is required to prevent tripping on
' momentary voltage dips

.The majority of stator winding faults are the result of prolonged or cyclical overheating which
causes the insulation to deteriorate. Most faults are cleared by instantaneous earth fault
protection as the windings are generally surrounded by earthed metal. Sensitivity of the earth
fault relay is limited by the spill current from residually connected CTs during starting, usually
Most other faults are cleared by thermal or unbalance protection. Instantaneous overcurknt
units if fitted protect only against terminal flashovers and other heavy short circuits. This is
because of the high settings necessary to prevent maloperation on starting current surges. For
motors above say 1MW differential protection may be used to give high speed clearance of
phase and earth faults. This usually takes the form of high impedance differential or biased
differential. 6 current transformers are required with 2 per phase at the two ends of winding.


.:;,..-. . .

:'. .,.SELF
.. .




An alternative is to use self balance type different~alprotection arrangement

Using instantaneous current relays.

' ,

shown above

If conductors are placed reasonably bncentric w~thinthe window of the core balance current
W ~ t hthis low spill current and a

: transformers, spill current can be kept to a minimum.

1 ,

Page 11

reasonably indepenaence of CT ratio to full load a lower fault setting could be achieved than:.
conventional high impedance circulating current differential schemes.
Disadvantages :



the necessity of passing both ends of each phase winding through the CT and hence the
need for extra cabling on the neutral end.


to avoid long cabling position of CTs are restricted to the proximity of the machine output
terminals in which case the cable between the machine output terminals and controlling
switchgear might not be included within the differential zone.



;. $



. .:.






Where the motor is switched via a fused contactor, the interrupting capacity of the contactor ':,
must be taken into consideration. In general they will not be rated to break the maximum fault
current. In this case it is important to prevent the protection attempting to operate the contactor '%
above its maximum rating. This is usually achieved by disabling all instantaneous tripping .'.
elements and time co-ordinating with the associated fuije. This is illustrated in the following .?





Ts .





Ts > Tfuse at Icont.


Page 12


ings can suffer from both electrical and mechanical failure:
rical Interference - can result in an induced voliage and corresponding circulating
aring, it is important to take precautions against this, for example adequate
ng of equipment.

anical Failure - results in increased friction, generating heatirig and eventually failure of

f bearings are detailed below:

this type of bearing will cause the motor to come to a standstill'

immediately. The motor will draw a heavy current equivalent to the locked rotor
current. There is very little change that a relay monitoring the motor current can detect
bearing failure of this type before the bearing is destroyed. However, it is essential to
disconnect the motor before excessive winding damage. This may be covered in the form
of stalling protection.


Sleeve Bearings


Failure of this type of bearing is a very occurrence. If it occurs it will be indicated by

rature rise, vibration and increase in motor current in the order ~f 10% to20%.prior.



It is generally accepted that the bearing will need replacing following failure, however stall
protection will help niinimise damage to the motor itself. Unfortunately, in extreme cases
this .is not the case and distortion of the shaft may occur. One method used to prevent
this is direct temperature monitoring of the bearings using RTD's for example.

Out-Of-Step Protection

A synchronous motor decelerates and falls out of step when it is subjected to a mechanical
overload exceeding its maximum available output. It may also lose synchronism from a fall'in
field current or supply voltage. An out-of-step condition will subject the motor to undesirable
Overcurrent and pulsating torque leading to eventual stalling.
Two methods are available to detect out-of-step condition in a synchronous motor:

Field Current Method

The alternating component of current induced in the field circuit when the motor falls out
of step provides the basis for this method. One arrangement is to connect a reactor in
series with the field circuit to divert alternating current to a polarised field-frequency relay,
a coil of which is connected i n parallel with the reactor.

Disadvantages :

difficult to discriminate between alternating current induced by pole slip and

induced by'faults on the supply system or sudden swing of load.


certain faults introduce harmonics, in particular second harmonic alternating cu

in the filed circuit.

Power Factor Method

This method makes use of the change of power facto that occurs when, the motor
poles. When the motor loses synchronism a heavy current at a very low power fac
drawn from the supply.


r 1'

., .

Stator current on
loss of synchronism







Protection Against Sudden Restoration Of Supply


. -.;+

On loss of supply a synchronous motor-should be disconnected if the'supply couldbe restoied:;+

automatically or restored without knowledge of the machine operator. This is to .avoid the..$
. ....-...- +, 3
possibility of the supply beingarestoredout of phase with the motor generated emf.



Two ways of detecting loss of supply :





Overvoltaqe and Underfrequency

If the supply busbars have no other load connected and the motor is not loaded the motor,;;
terminal voltage could rise instantaneously to 20-30% on loss of supply due to the open ii:
circuit regulation of the machine. If the motor is loaded it will decelerate fairly quickly on :
loss of supply and the frequency of terminal voltage will fall.


Underpower and Reverse Power

Applicable when power reversals do not occur under normal operating conditions.


Underpower - arranged to look into the machine; applicable when there is a possibility of
no load connected on loss of supply.
Reverse power - arranged'to look away from the machine; applicable where there is
always load connected.
Time delay is required to overcome momentary power reversal due to faults etc.


'::!I A

. ~

Page 14

A C Motor Protection

A. C. Motor Protection

.. , .. ,


. '>;: .

There are a wide range of a.c. motors and motor

characteristics i n existence, because o f the numerous
duties for which they are used. All motors need
protection, b u t fortunately, the more fundamental
-problems affecting t h e choice of protection are
independent o f the type o f motor and the type o f load t o
which it is connected. There are some important
differences between the protection o f induction motors
and synchronous motors, and these are'fully dealt w i t h
i n the appropriate section.


. ...
. + .


. .


. .... .


:...~ ~.:~ ~ ~ ? , Y ~ ; ~ ~ ' ~ : I ;

:*.;;:.:': . . . . ..





M o t o r characteristics must be carefully considered when

applying protection; while this may be regarded as
stating the obvious, it is emphasised because it applies
more t o motors than t o other items o f power system
. .
For example, the starting and , s t a l l i n g .
- . ,
currents/times must be known when applying.overload. ?':.-.I.-;.--'.,-.:
protection, and furthermore the thermal'withstand o f
. .
the machine under, balanced and unbalanced loading -1.' - '
. ++
must be clearly defined.

- '



The conditions for which motor protection is required

can be divided into t w o broad categories: imposed
external conditions and internal faults. Table 19.1
provides details of all likely faults that require protection.


Extcmal Faults



-. .

- -,

Unbalanced rupplics

Bcating failurcs


Winding faults


Singlc phasing

Rcvcnc phasc wgucncc


In:crnal faults



The design of a modern motor protection relay must be

adequate t o cater for the protection needs of any one of
the vast range of motor designs in service, many of the
designs having n o permissible allowance for overloads. A
relay offering comprehensive protection will have the
following set o f features:
a. thermal protection

b. extended start protection

c. stalling protection

d. number o f starts limitation

heat a t a rate proportional to temperature rise. This is

the principle behind the 'thermal replica'
motor used for overload protection.

e. short circuit protection

f. earth fault protection

The temperature T a t any instant i s given by:

g. winding RTD measurementltrip

h. negative sequence current detection

T,,,,,, ( 1 - e-fk)


i. undervoltage protection


= final steady state temperature

s = heating time constant

j. loss-of-load protection

Temperature rise is proportional t o t h e curre

k. out-of-step protection
I. loss of supply protection


m. auxiliary supply supervision

(items k and I apply t o synchronous motors only)
I n addition, relays may offcr options such i s circuit
breaker condition monitoring i s a n aid t o maintenance.
Manufacturers may 3lso offer relays that implement a
reduced functionality t o that given above where less
comprehensive protection is warranted [e.g. induction
motors o f low rating).
The following sections examine each o f the possible
failure modes of a motor and discuss h o w protection may
be applied t o detect t h a t mode.



. u:

. ..?.. i.




. . ,.

... ..

. .

..... ..

The majority of winding failures are either indirectly or

directly caused by overloading [either prolonged or
cyclic), operation o n unbalanctd supply voltage, or single
phasing, which a l l lead through excessive heating t o the
deterioration of t h e winding insulation until an electrical
f a u l t occurs. The generally accepted rule is t h a t
insulation life is halved for each 10" C rise i n
tercperature above the rated value, modified by t h e
length of time spent a t the higher temperature. As an
electrical machine has a relatively large heat storage
capacity, it follows t h a t infrequent overloads of short
duration m a y n o t adversely a f f e c t t h e machine.
However, sustained overloads o f only 'a' few percent may
result i n premature ageing and insulation failure.




. 19 -

IR = current which, i f flowing continuously, pr

temperature T,,,,, i n the motor
~ t , ~ it can
~ ~be fshown
~ ~that,
~ for
, any overload
1 , the
tirne t for this current to flowis:

t x l o I [ (

should take into account both of these

typical equation for the equivalent current being:

lcq =

po5irive sequence current


negative sequence ,current


negative sequence rotor resistance

positive sequence rotor resistance

at rated spced. A typical value o f K is 3.

Finally, the thermal replica model needs t o take i n t o
account the iact that the motor will tend t o cool down
during periods of light load, and the initial state of the
motor. The rllotor will have a cooling time constant, T,.
that defines the rate of cooling. Hence, t h e final thermal
model can bc expressed as:



[k2 - A

j 2 -7

_ ...;



. ...:<c.&
;f i>,...
. *,






:... -,+,.:+.-,





I n general. t h e supply to which a motor is connec

The variety o f motor designs, diverse applications, variety

o f possible abnormal operating conditions and resulting
modes o f failure result i n a complex thermal relationship.
A generic mathematical model t h a t is accurate is
therefore impossible t o create. However. i t is possible to
develop an approximate model if it is assumed that the
m o t o r is a homogeneous body. creating and dissipating
.. -,?,.,.; .-,-..

7 - (I,?11)

Furthermore, the thermal withstand capability of the

motor is affected by heating i n the winding prior to a
It is therefore i m p o r t a n t t h a t t h e relay
characteristic takes account o f t h e extremes o f zero and
full-load pre-fault current known respectively as the
'Cold' a n d 'Hot' conditions.

contain b o t h positive a n d negative'- s

components, and both compbnents o f clrrrentgi
heating in thr motor. Therefore,. the, the!m


( 1 - e-'I7



\ . , ~ . o r h

P r . ( r . t i o .


,.,,(." ; . j ,


V A W I . - . , ; . ~

C i l r



Page 339

-: 9,.{ START!ST,?[.L

,_. . ..

= heating time c o ~ i s t a ~ t t

i a l stare of rr~oror[cold o r /lor)

?la/ setriug c l t r r e n t
into account the 'cold' and 'hot'
d i n IEC 60255, part 8.

ays may use a dual slope characteristic for the

and hence
values the
&time constant are required. Switching between
values takes place a t a pre-defined motor
may. be used to obtain better tripping
during starting
on motors that use a star~~~i~~
starting,the motor windings cam/
current, while inthe
condition, they carry
of the current seen by the relay. similarly,
~e motor is disconnected from the supply, the
i;g time
constant is set equal
to the coolingtime
iant TI-



f the

relay should ideally be matched t o the

be capable of close
a wide range of relay adjustment
. . protection,

When a motor is started, i t draws a current well i n excess

of full load rating throughobt the period that the motor
takes to run-up to speed. While the motor starting
Ciiiient iedijces somewhat as motor speed increases, i t is
normal in protection practice to assume that the motor
current remains constant throughout the starting period!.
The starting current will vary depending on the design of
the motor and method of starting. For motors started
DOL (direct-on-line], the nominal starting current can be
4-8 times full-load current. However, when a star-delta


. . . >. .


. . . ..

Should a motor stall whilst running. or fail to start, due

to excessive loading, the motor will draw a current equal
to its' locked rotor current. It is not therefore possible t o
distinguish between a stall condition and a healthy start
solely on the basis of the current drawn. Discrimination
between the t w o conditions must be made based on the
duration of the current drawn. For motors where the
starting time is less than the safe stall time o f the motor.
protection is easy to arrange.

. -


\ .

,. . . . . . .

. . . .. ..







.:. iZ--:.


;., ..

.... .

U .- .L1






.: .




A motor may fail to accelerate from rest for a number o f


:. . .



. .. .


.. . . .. . . .. . . .. . . .. .

. . . .


. . .. . . .

DOL starting


. . .

iijelay setting curves are i h o w n i n Figure 19.1.


.. .

. ..: . . ..
. ...

starter is used, the line current will only be 7 / f i o f the

However, where motors are used to drive high inertia

loads, the stall withstand time can be less than the
starting time. I n these cases, an additional means must
be provided to enable discrimination between the t w o
conditions to be achieved.

dd motor

=. -



loss of a supply phase

mechanical problems


low supply voltage


excessive load torque

A large current will be drawn from the supply, and cause
extremely high temperatures to be generated within the
motor. This is made worse by the fact that the motor is
not rotating, and hence no cooling due to rotation is
available. Winding damage will occur very quickly either to the stator or rotor windings depending on the
thermal limitations of the particular design (motors are



...- ..



said to be stator or rotor limited i n this respect). The

method of protection varies depending on whether the
starting time is less than or greater than the safe stall
time. In both cases, initiation of the start may be sensed
by detection of the closure of the switch i n the motor
feeder (contactor or CB) and optionally current rising
above a starting current threshold value - typically

1: -... . . .. . . . . . . . .

. - ......




.., ......,_ . . . . . .

..I, . . .



Thcrnnal cquivalcnl current I in term5 of thc currcnl

- -..

thermal lhrcrhold

9.:' TI:,.?!,-,,!




:.,,<;,,,:!: *



r , > / , j ~ , , ~ . . , :.~~, , , - . 3 ; s~T:c!!:::




.. - ...
-. . . . . . . ~ ~ . r . ~ r , .u. ~ . r . ~ a r i . r n G - i l *




..-.I ' .

. .

. ' .,



. . - . . . .. ..

. .. . .
. .
. . .
. . . ...
,..-. . .





: ?I




,. ,

. .. . -



.; . / .





, : . :.

..':P ' <





Page 340



200% o f motor rated current.

For the case of both

conditions being sensed, they may have t o occur within
a narrow aperture of time for a start to be recognised.
~ ~ c L ; arequirements

may exist for certain types of

motors installed i n hazardous areas (e.g. motors with
type o f protection EEx 'e') and the setting of the relay
must take these into account. Sometimes a permissive
interlock for machine pressurisation (on EEx 'p'
machines) may be required, and this can be conveniently
achieved by use o f a relay digital input and the in-built
logic capabilities. .


Protection is achieved by use o f a definite time

overcurrent characteristic, the current setting being
greater than full load current but less than the starting
current of the machine. The time setting should be a
little longer than the start time, but less than the
permitted safe .starting time of the motor. Figure 19.2
illustrates the principle of operation for a successful

successful start is used t o select relay timer used for the

safe run up time. This time can be longer than the safe
stall tjme, as there is both a (small) decrease in current
drawn by the motor during the start and the rotor fans



begin to improve cooling of the machine as i t

accelerates. I f a start is sensed by the relay through
monitoring current and/or start device closure, but the
speed switch does not operate, the relay element uses
the safe stall time setting to trip the motor before
damage can occur. Figure 19.3(a) illustrates the principle
of operation for a successful start. and Figure 19.3(b) for
an unsuccessful start.

CB Closcd

. :S$






















(a] Successful start

Ce Closcd

.. ..:.*:>:..-



.. . -

o ---J












,>,a>, ,>,>*C

rcnms .

ibl Unrucccssful start





.r .
. . ....-L.,4
. .-.-,,
. ..:?Pi
. .*.,

Should a motor stall when running or be unable to start?

because of excessive load, it will draw a current from t h e i 3 :
supply equivalent to the locked rotor current. I t is:3;g;
obviously desirable to avoid damage by disconnecting
the machine as quickly as possible i f this condition4@$



. I

Currcnt (p.u.


. . i : . .


For this condition, a definite time overcurrent

characteristic by itself is not sufficient, since the time
delay required is longer than the maximum time that the
motor can be allowed to carry starting current safely. An
additional means of detection of rotor movement,
indicating a safe start, is required. A speed-sensing
switch usually provides this function. Detection of a

Motor stalling can be recognised by the motor current:;@$

exceeding the start current threshold after a successful^:$$;
start - i.e. a motor start has been detected and the motor:+$?'
current has dropped below the start current threshold,::-%
within the motor safe start time. A subsequrnt r i a in$
motor current above the motor starting current:i;i?;
threshold is then indicative of a stall condition, and,,'@
tripping will occur i f this condition persists for
than the setting of the stall timer. An instantaneou~;i$:'
overcurrent relay element provides protection.


In many systems, transient supply voltage loss (typical!Y;$

up to 2 secmdr) does not rerult i n tripping of designate<&
motors. They are allowed t o re-accelerate upon@
restoration of the supply. During re-acceleration, the%?



--- .....





. . .

S t a a lockout

i ;
> .



Inhib. start time



draw a current similar t o the starting current for a period

that may be several seconds. It is thus above the motor
stall relay element c u r r e n t threshold.
The stall
protection would be expected t o operate and defeat the
?bject of the re-acceleration scheme.
motor protection relay w i l l therefore recognise the
'resence o f a voltage dip and recovery. and inhibit stall
lrotection for a defined period.
The undervoltage,
'rotection element (Section 19.11) can be used t o detect
he presence o f the voltage dip and i n h i b i t stall
'rotection for a set period after voltage recovery.
'rotection against stalled m o t o r s i n case of a n
nsuccessful re-acceleration is therefore maintained.
he time delay s e t t i n g is dependent o n the rec~elerationscheme adopted and the characteristics o f
ldividual motors.
I t should be established after
Vforming a transient stability study for the reSeleration scheme proposed.

. ...;:,,?fit,,..



:>:;:,:I. i ,!".:

Any motor has a restriction on the number of starts that

are allowed i n a defined period without the permitted
winding, etc. temperatures being exceeded. Starting
should be blocked if the permitted number o f starts is
exceeded. The situation is complicated b y the fact the
number o f permitted 'hot' starts i n a given period is less
than the number o f 'cold' starts, due t o the differing
initial temperatures o f the motor. The relay must
maintain a separate count o f 'cold' and 'hot' starts. By
making use o f the data held i n the motor thermal replica,
'hot' and 'cold' starts can be distinguished.
To allow the motor t o cool down between starts, a time
delay may be specified between consecutive starts (again
distinguishing between 'hot' and 'cold' starts). The start
inhibit is released after a time determined by the motor
specification. The overall protection function is ilhstrated
in Figure 19.4.

Cha~ ~ 1 9 - 3 3 6 - 3 5 12 0 / 0 6 / 0 2


Page 3 4 2

Motor short-circuit protection is often provided t o cater

f o r major stator winding faults and terminal flashovers.
Because of the relatively greater amount o f insulation
between phase windings, faults between phases seldom
occur. As the stator windings are completely enclosed in
grounded metal, the fault would very quickly involve
earth, which would then operate the instantaneous earth
fault protection. A single definite time overcurrent relay
element is all that is required for this purpose, set to
about 125% of motor starting current. The time delay is
required t o prevent spurious operation due to CT spill
currents, and is typically set at looms. I f the motor is fed
from a fused contactor, co-ordination is required with
the fuse, and this will probably involve use o f a long time
delay for the relay element. Since the object of the
protection is to provide rapid fault clearance t o minimise
damage caused by the fault, the protection is effectively
worthless in these circumstances. It is therefore only
provided on motors fed via circuit breakers.
Differential (unit) protection may be provided on larger HV
motors fed via circuit breakers t o protect against phasephase and phase-earth faults, particularly where the
power system is resistance-earthed. Damage to the motor
i n case of a fault occurring is minimised, as the differential
protection can be made quite sensitive and hence detects
faults i n their early stages. The normal definite time
overcurrent protection would not be sufficiently sensitive,
and sensitive earth fault protection may not be provided.
The user may wish t o avoid the detailed calculations
required of capacitance current in order to set sensitive
non-directional earth fault overcurrent protection
correctly on HV systems (Chapter 9) or there may be no
provision for a VT to allow application of directional
sensitive earth fault protection. here is still a lower limit
to the setting that can be applied, due to spill currents
from CT saturation during starting. while on some motors.
neutral current has been found t o flow during starting.
even with balanced supply voltages. that would cause the
differential protection to operate. For details on the
application of differential protection, refer to Chapter 10.
However, non-directional earth fault overcurrent
protection will normally be cheaper i n cases where
adequate sensitivity can be provided.

One o f the most common faults t o occur on a motor is a

stator winding fault. Whatever the initial form of the
fault (phase-phase, etc.) or the cause (cyclic overheating,
etc.), the presence of the surrounding metallic frame and
casing will ensure that i t rapidly develops into a fault
involving earth. Therefore. provision of earth fault
protection is very important. The type andscnsitivity of
protection provided depeAds largely' on the system
earthing, so the various types will be dealt with i n turn.

I t is common, however, t o provide both instantaneous

and time-delayed relay elements t o cater for major and ,:,slowly developing faults.

,... :. . . .


. . . .
;, , , . . :..
... E < - . . - . . z

....... ;"i

Most LV systems fall into this category, for reasons of

personnel safety. Two types o f earth fault protection are
commonly found - depending on the sensitivity required.
For applications where a sensitivity of > 20% of motor
continuous rated current is acceptable, conventional
earth fault protection using the residual CT connection
of Figure 19.5 can be used. A lower limit is imposed on
the setting by possible load unbalance and/or (for HV
systems) system capacitive currents.







Flow of



Care must be taken to ensure that the relay does not

operate from the spill current resulting from unequal CT
saturation during motor starting, where the high
currents involved will almost certainly saturate the
motor CT's. I t is common to use a stabilising resistor i n
series with the relay, with the value being calculated
using the formula:



I,, = starting current referred t o CT secondary

I,, = relay earth fault setting (A)
Rslah = stabilising resistor value (ohms]
= d.c. resistance of CT secondary (ohms)
= CT single lead rcstistance (ohms)




Page 3 4 3

CT connection factor

1 for star p t at CT

2 for star p t a t relay)

relay input restistance (ohms)
ffect of the stabilising resistor is t o increase the
ive setting of the relay under these conditions, and
delay tripping. When a stabilising resistor is used,
tripping characteristic should normally be
ntaneous. An alternative technique, avoiding the use
stabilising resistor is to use a definite time delay
cteristic. The time delay used will normally have to
und by trial and error, as it must be long enough to
ent maloperation during a motor start, but short
gh t o provide effective protection i n case o f afault.

If a more sensitive relay setting is required, it is necessav

t o use a core-balance CT. This is a ring type CT. through
which all phases o f the supply t o the motor are passed,
plus the neutral on a four-wire system. The turns ratio
o f the CT is no longer related to the normal line current
expected t o flow, so can be chosen t o optimise the pickup current required. Magnetising current requirements
are also reduced, with only a single CT core to be
magnetised instead of three, thus enabling low settings
to be used. Figure 19.7 illustrates the application o f a
core-balance CT, including the routing of the cable
sheath t o ensure correct operation in case o f core-sheath
cable faults.


rdination with other devices must also be considered.

Cablc b o x .

Cablc gland /sheath

g r o u n d connection

low the maximum system fault current - reliance is

b c e d on the fuse ir, these circumstances. As a trip
command from the reiay instructs the contactor t o open.
&re must be taken t o ensure that this does not occur until
the fuse has had time to operate. Figure 19.6(a] illustrates
'$correct grading of i h e relay with the fuse, the relay
k r a t i n g first for a range of fault currents in excess of the
nntactor breaking capacity. Figure 19.6(b) illustrates
&ect grading. To achieve th~s,i t may require the use o f
;liintentional definite :ime delay in The relay.


. Furc

Cablc gland

: Contactor
: capacnty

, .,

C / f rclay

: Contaclor




. .; .....'.,



r :';

' capacity



(b) Correct



::1.(. tnr;lrrq s!rc$:tv::,:!cr.:ro:

These are commonly found on HV systems, where the

intention is to limit damage caused by earth faults
through limiting the earth fault current that can flow.
Two methods of resistance earthing are commonly used:



. . .
. . .. .. ..

. ,..::
.I :. ,






. ,




Page 344


.. .

.. .


In this method, the value of resistance is chosen to limit

the fault current to a few hundred amps - values of
~ O O A - ~ O Obeing
typical. With a residual connection of
line crs, the
senjjliviti; pos;i';le is about !no/,
of CT rated primary current, due to the possibility of CT
saturation during starting. For a core-balance CT, the

I \



sensitivity that is possible using a simple non-directional

earth fault relay element is limited to three times the .:$
steady-state charging current of the feeder. The setting i j
shoild not be greater than about 30% of the minimum 'i
earth fault current expected. Other than this, t h e <,-.
considerations in respect of settings and time delays arc
as for solidly earthed systems.


'I, I


applying earth faults a t various parts of the system and

measuring the resulting residual currents.

O g m e HV systems, high resistance earthing is used t o
&rt the earth fault current t o a f e w amps. I n this case,
S;c system capacitive charging current will normally
&vent conventional sensitive earth fault protection
&ng applied, as the magnitude of the charging current
ill be comparable w i t h the earth fault current i n the
of a fault. The solution is t o use a sensitive
ircctional earth fault relay. A core balance CT isused in
rnjunction w i t h a VT measuring the residual voltage of
he system, with a relay characteristic angle setting o f
45' [see Chapter 9 for details). The VT must be suitable
the relay and therefore the relay manufacturer should
:consulted over suitable types - some relays require
)at the VT must be able t o carry residual flux and this
lies out use-of a 3-limb. 3-phase VT. F,. setting o f 125%
f the single phase capacitive charging current for the
hole system is possible usins this method. The time
:lay used is not critical but must be fast enough to
sconnect equipment rapidly in the event o f a second
~ r t hfault occurring immediately after the first.
'inimal damage is caused by the first fault, b u t the
,cond effectively removes the current l i m i t i n g
sistance from the fault path leading to very large fault

If it is possible t o set the relay to a value between the

charging current on the feeder being protected and the
charging current for the rest o f the system, the
directional facility is not required and the VT can be
dispensed with.
The comments made in earlier sections on grading with
fused contactors also apply.

1 alternative

technique using residual voltage detection

also posible, and is described in the next section.

rth fault detection presents problems on these systems

Ice no earth fault current flows for a single earth fault.
lwever, detection is still essential as overvoltages occur
sound phases and i t is necessary to locate and clear
t fault before a second occurs. Two methods are
ssible, detection of the resulting unbalance in system
arging currents and residual overvoltage.

A single earth fault results in a rise in the voltage

between system neutral and earth, which may be
detected by a relay measuring the residual voltage of the
system (normally zero for a perfectly balanced, healthy
system]. Thus, no CT's are required, and the technique
may be useful where provision of an extensive number of
core-balance CTs is impossible or difficult, due to
physical constraints or on cost grounds. The VTs used
must be suitable for the duty, thus 3-limb, 3-phase VTs
are not suitable, and the relay usually has alarm and trip
settings, each with adjustable time delays. The setting
voltage must be calculated from knowledge of system
earthing and impedances, an example for a resistanceearthed system is shown in Figure 19.10.

nsitive earth fault protection using a core-balance CT

-equired for this scheme. The principle is that detailed
Section 9.16.2, except that the voltage is phase shifted
+go' instead of -90'. To illustrate this, Figure 19.8
)WS the current distribution i n an Insulated system
ljected to a C-phase t o earth fault and Figure 19.9 the
ay vector diagram for this condition. The residual
-rent detected by the relay is the sum of the charging
'rents flowing in the healthy part of the system plus
healthy phase charging currents on the faulted
der - i.e. three times the per phase charging current
the healthy part of the system. A relay setting of 30%
this value can be used to provide protection without
: risk of a trip due to healthy system capacitive
~ r g i n gcurrents. As there is no earth fault current, it is
o possible to set the relay at site after deliberately




ff A . f . - . I i * n


Grading of the relays must be carried out with care, as

the residual voltage will be detected by all relays in the
affected section of the system. Grading has to be carried
out with this in mind, and will generally be on a time
basis for providing alarms (1" stage), with a high set
definite time trip second stage to provide backup.



.... -.



... ;





. .
. .


. ,
. . . . . .- ..

. . :
. . .: . .
.. . .. . . . . . .

. . .

' 8



., .,'4
. .., ,
, ' :,


tor positive sequence impedance at slip s

rice, a t


Modern motor protection relays have a negative

sequence current measurement capability, i n order to.
provide such prote~tion. The level of negative sequence
unbalance depends largely upon the type of fault. For
loss of a single phase at start, the negative sequence
current will be 50010 of the normal starting current. It is
more diff~cult to provide an estimate of the negative
sequence current i f loss of a phase occurs while running.
This is because the impact on the motor may vary widely,
from increased heating t o stalling due t o the reduced
torque available.





A typical setting for negative sequence current
protection must take into account the fact that the
motor circuit protected by the relay may not be the f!:.:'
?ource of the negative 5equence current. Time s h ~ u \ d b e =% . :
allowed for the appropriate protection to clear -the
.: 1:
source of the negative sequence current without
introducing risk of overheating to the motor being
This indicates a two stage tripping
characteristic, similar i n principle t o overcurrent
protection. A low-set definite time-delay element can
be used to provide an alarm, with an IDMT element used
to trip the motor in the case of higher levels of negative
seq"ence current, such as loss-of-phase conditions at
start, occurring. Typical settings might be 20% o f CT
' 19 '
rated primary current for the definite time element and
50010 for the IDMT element. The IDMT time delay has to
be chosen to protect.the motor while, if possible, grading
negative sequence relays on the system.
s may not incorporate two ekments, i n which
ingle element should be set to protect the
motor, with grading being a secondary consideration.



rf f -SJSI I





j ( X , + X;/

K,+ R ;

Negative sequence current is a t twice supply frequency.

Skin effect in the rotor means that the heating effect i n
the rotor of a given negative sequence current is larger
than the same positive sequence current. Thus, negative
sequence current may result in rapid heating o f the
motor. Larger motors are more susceptible i n this
respect, as the rotor resistance of such machines tends t o
be higher. Protection against negative sequence currents
is therefore essential.

indicates negative sequence quantities

R , + R;

. .

F o r the same motor,

negative sequence voltages i n excess of 17% will result
i n a negative sequence current larger than rated f u l l load

standstill (s=1.0), impedance



' leading to' excessive heating.

suffix p indicates positive sequence quantities


. . .
.. ....



speed is approximately equal t o the positive sequence

reactance at standstill. An alternative more meaningful
way of expressing this is:

;. ...,.

positive seq. impedance

negative seq. impedance

starting current
rated current

,,.,.:. .,

:., , 1 ', 1 ' .

: . . ; ;;\;

,,:j';..);:~?ib' :


. .,,,...
!. .
.:,;.>.:..; ..>*.,,, ::;,.
. ... .


. ,. ,+..;~:-..;..{.;


On wound rotor machines, some degree of protection ,..c,:.;~

p;.!:$>.j2;?'...;:,>:: ,.+?:
against faults i n the rotor winding can be given by an
instantaneous stator current overcurrent relay element.
As the starting current is normally limited by resistance
'- 1 p. ~+P",P".
to a maximum of twice full load, the instantaneous unit
, . ~-:!~'
can safely be set to about three times full load if a slight :,:!'6~!T!.
. ..,..
. .. I .


and i t is noted that a typical LV motor starting current is

GxFLC. Therefore, a 50j0 negative sequence voltage (due
to, say, unbalanced loads on the system) would produce
a 30010 negative sequence current in the machine,




.. . . .. . . . .. . ..
h f ~ f - . r lP , . r r r r ; . .



3 4 7


. . ~~..


. .


. ,.:c. ;...

. :.:.v.,,



Page 348





:br,kgat i m e delay o f approximately 3 0 milliseconds is
:;;.. +$:$.:&$.

It should be noted that faults occurring i n
:. .c ~ ~
., y,~.~ : : .t hi e rotor windinq would n o t be detected b y any
d~fferentialprotection applied t o the stator.


>*>.< x...% $,.;?.



.:. ..,....


A\.. ',!



RTD's are used t o measure temperatures o f motor

windings or shaft bearings. A rise i n temperature may
denote overloading o f the machine, or the beginning of
a fault i n the affected part. A motor protection relay will
therefore usually have the capability of accepting a
number of RTD inputs and internal logic t o initiate an
alarm and/or trip when the temperature exceeds the
appropriate setpoint(s). Occasionally, HV motors are fed
via a unit transformer. and i n the;e circumstances, some
o f the motor protection relay RTD in.puts may be
assigned t o the transformer winding temperature RTD's.
thus providing overtemperature protection for the
transformer without the use o f a separate relay.

factors in mind.

failure i n a mechanical transmission (e.g. conveyor belt),

or i t can be used with synchronous motors t o protect,

There are t w o types o f bearings to be considered: the

anti-friction bearing (ball or roller), used mainly on small
.:.. ,.' . motors (up to,.around 3 5 0 k ~ ) ,and the sleeve bearing.
. .a.
L . ,..?.. .. -used mainly on large motors:




The failure of ball or roller bearings usually occurs very

quickly, causing the motor t o come to a standstill as
pieces of the damaged roller get entangled with the
others. There is therefore very little chance that any
relay operating from the input current can detect
Searing failures of this type before the bearing is
completely destroyed. Therefore, protection is limited to
disconnecting the stalled m o t o r rapidly to avoid
consequential damage. Refer t o Section 19.2 on stall
protection for details of suitable protection.

+:;: .

- .




This is especially important for

system transients.
synchronous motor loss-of supply protection.

Failure of a sleeve bearing can be detected by means of

,,i:.:i&-i.j;, a rise i n bearing temperature. The normal thermal

.~ .. overload relays cannot give protection to the bearing
. . . .. . . ...+


.. ,, \. l- ! ; ,


. . . . . ... . . .

. .......
. . . .


..... . .

-., .




r:.:,: :



- 1 %





itself but will operate to protect the motor from

excessive damage. Use of RTD temperature detection, as
noted i n Section 19.9, can provide suitable protection.
allowing investigation into the cause of the bearing
running hot prior t o complete-failure.



Motors may stall when subjected t o prolonged

undervo!tage conditions. Transient undervoltages will
generally allow a motor to recover when the voltage is








Page 3 4 9

ied voltage to stator or field windings. Such a fall

n o t need to be prolonged, a voltage dip o f a few
nds may be all that is required. An out-of-step
ition causes the motor to draw excessive current
generate a pulsating torque. Even if the cause is
oved promptly, the motor will probably not recover
chronism, but eventually stall. Hence, it must be
m&iinecird from the supply.

i q e current drawn during an out-of-step condition is a t

5 v e r y low power factor. Hence a' relay element that
b o n d s t o low power factor can be used to provide
- The element must be inhibited during
rting, when a similar low power factor condition
rs. This can conveniently be achieved by use of a
!&finite time delay, set t o a value slightly i n excess of the
otor start time.

A low forward power relay can detect this condition. SeeSection 19.12 for details. A time delay will be required
to prevent operation during system transients leading t o
momentary reverse power flow i n the motor.


This section gives examples o f the protection o f HV and
LV induction motors.

Table 19.2 gives relevant parameters of a HV induction

motor to be protected. Using a MiCOM P241 motor
protection relay, the important protection settings are
calculated i n the following sections.

n e power factor setting will vary depending on the rated

tpwer factor of the motor. It would typically be 0.1 less
fh& the motor rated power factor i.e. for a motor rated
at 0.85 power factor, the setting would be 0.75.


If the supply to a synchronous motor is interrupted, it

a essential that the motor breaker be tripped as quickly
3s possible if there IS any possibil~tyo f the supply
3eing restored automatically or without the machine

Ratcd Voltagc

Ratcd frequency

Ratcd powcr

Stall withstand timc cuIdIhot

Slaning cumnt

Rrmincd starts culd/hot

CI ratio


start tirncal-








-< ..



*:: ...+'.'......;:. ;::+

2 :~.}:;~-'
. . . f;;:;25,75~i
. . . . . n.. . . . . . :.. ..1... . ji. .,.;.:..:;:.. . $1:


.55s .

.. .. .. ..

. . . . . . . . ... ..... ..:..z.

. .. . . .. .:. .. ... .. :.

5505 DOL

. . . . . .



Start tirnc@ f)(PYo+tagc







. .,

..: .>::. . .
. . . . . .Solid.
;! . .:<y:: .- 0
j.;-.:.: ,.~ii~ltBrcak&...'
." -1 . :..
. .% . .


his is necessary in order t o prevent the supply being

estored out of phase with the motor generated voltage.

methods are generally used to d'etect this condition,

to cover different operating modes of the motor.


he underfrequencv relay element will operate i n the

ase of the supply failing when the motor is on load,
!hich causes the motor to decelerate quickly. Typically,
NO elements are provided, for alarm and trip
i e underfrequency setting value needs to consider the
Iwer system characteristics. In some power systems.
ngthy periods of operation at frequencies substantially
:low normal occur, and should not result in a motor
ip. The minimum safe operating frequency of the
otor under load conditions must therefore be
:termined, along with minimum system frequency.

The current setting IT" is set equal to the motor full load
current, as it is a CMR rated motor. Motor full load
current can be calculated as 211A, therefore (in
secondary quantities):


Use a value of 0.85, nearest available setting

The relay has a parameter, K, to allow for the increased
heating effect of negative sequence currents. In the
absence of any specific information, use K=3.
Two thermal heating time constants are provided, r , and - . : :
r,. r, is used for starting methods other than DOL, . . ,.. .
otherwise it is set equal to r,. r , is set to the heating ,-.,j.::,..: .:.:::
time constant, hence r,=r2=25mins. Cooling time :::'
, \,p:,'. :. :-j-.' ""
constant r, is set as a multiple'of r,. With a cooling time
constant of 75mins,
.,, . \-. - .,.

is can be applied i n conjunction w i t h a time delay to

feet a loss-of-supply condition when the motor may

are a busbar with other loads. The motor may attcmpt

supply the othcr loads with power from the stored
letic energy of rotation.


... . . . . ..,.'..

r,= 3 x r ,

. . . .- . :..:.:
. . . . .:..




condition at starting.
19.1.I..? Ptotcctioti of

In accordance with Section 19.7, use a ~ e t t i n go f



S! r ~ i ~ , t < j :





b e carefully co-ordinated w i t h the fuse

re t h a t t h e contactor does n o t attempt t o break a
i n excess of i t s rating. Table 19.3(a) gives details
LV m o t o r and associated fused contactor. A
M P 2 l l m o t o r protection relay i s used t o provide

I,, = motor rated primary current

Ip = CT primary current
Hence, I b = 5 x 1 3 2 / 1 5 0 = 4.4A
W i t h a motor starting current o f 670% o f nominal, a
setting o f the relay thermal t i m e constant w i t h motor
initial thermal state o f 5.9% o f 15s is found satisfactory,
as shown i n Figure 19.14.



(a1 LV motor cxamplc data

M r l o a d tirnc dclay


timc dclay








.....5!> ..........
b. .

%b;cr3.3: ;L:!.<c:arpn!r<f
. . .




(a1 LV Motor Protcction








Re relay is set i n secondary quantities, and therefore a

hitable CT r a t i o has t o be calculated. From the relay
i a n u a l , a CT w i t h 5A secondary rating and a motor rated
:urrent i n the range of 4-6.4 when referred t o the
secondary of CT is required. Use o f a 15015A CT gives a
notor rated cuirent o f 4.4A when referred t o -the CT
iecondary, so use this CT ratio.

h e fuse provides the motor overcurrent protection, as the

Irotection relay cannot be allowed t o trip the contactor on
Wrcurrent in case the current to be broken exceeds the
mntactor breaking capacity. The facility for overcurrent
jrotection within the relay is therefore disabled.
1:; r - . . - .
. . . . . . . . . .. . . . .
The m o t o r is an existing one, and no data exists for it
x c e p t the standard data provided i n the manufacturers
This data does n o t include the thermal
heating) time constant o f the motor.
n these circumstances, it is usual t o set the thermal
jrotection so t h a t it lies just above the motor starting
;he current setting o f the relay, Ib , is found using the
orm mu la




3 A.r.r.1i.m

-- .


fcd cxamplc

l b l Rclay settings

- contactor

. . . . . . . . . . .


. . . . .
. . . . . . . . . . . . . . . . . . . .
....- . - . . . . . . . . .
. . . ..
....................... .
. . . . . . .
. . .. .. . .. ..........
.. .... ...... .. ..;:-.. . . . ........







[b] Rclay trip c h a r a c l c ~ ~ ~ t i c

The motor is built t o IEC standards, which permit a

negative sequence (unbalance) voltage of 1% on a
continuous basis. This would lead t o approximately 7010
negative sequence current-in the motor (Section 19.7).
As the relay is fitted only w i t h a definite time relay
element,a setting of 200,~ (from Section 19.7) is
appropriate, w i t h a tirnedelay of 2 s s to allowfor short
high-level negativesequence transients arising from
other causes.

The relay has a separate element f o r this protection. Loss

of a phase gives rise t o large negative sequence currents,
and therefore a much shorter t i m e delay is required. A
definite time delay o f 5s is considered appropriate.
The relay settings are summarised i n Table 19.3(b).



.::' .



Motor Protection Setting Criteria

and Tutorials
Page 1 of 38



Motor Details :




M ~ t o Rating
iri KW


Rated Voltage in KV

Motor Application


Motor Control by ( Contactor / Circuit Breaker )

Motor Full Load current
-: -


- ' '

230 Amps

CT Ratio

300/1 A

Type of Motor Starting


Starting current in Primary Arnps ( 100 % Voltage )

1 380

Starting current in Primary Amps ( 80 % Voltage )


Starting time in seconds ( 100 % Voltage )

Motor Starting time in seconds ( 80 % Voltage )

Thermal overload characteristics ( Available/

Stalling current in primary Amps


Hot stall withstand time in seconds


Cold stall withstand time in seconds


OALSTOM Limited, fnergy Automation & Information

Advanced Industrial Power System


Motor Protection Settins Crie

and Tutofi,
Page 2

Thermal Overload characteristic Curves

Times full load current

X 1.4
X 2.0

Cold Characteristics

Hot Characteristics

X 4.0
X 5.0
X Stalling current



OALSTOM Limited, Energy Automation & information

Advanced Industrial Power System


Motor Protection Setting Criteria

Thermal Overload Protection:

Set the pick-LIPat 103% of the rated current

= 230 x 1.03 = 236.9
to secondary
= 0..789 = 0.78
Set Is = 0.78
Calculation for the time constant at overload levels ( 1 <leq<2) times the
pick-up current
Cold Curve :

The operating time for the thermal overload characteristic available in the
relay is as follows :
t = T x In { ( PSM2)/( PSM2- 1 )

. .

. .


Therefore the required time constant can be calculated as :

T = t / { In { ( PSM2)/( PSM2 - 1)))

At 1.4 time rated current


For Cold Curve

PSM = 1.411.03 = 1.36, we require a operating time of 4000 x 0.7= 2800 ( 30
% margin )

= t / { In { ( PSM2)/( PSM2 - 1 )))

= 2800 / { In { ( 1.362)/( 1.362 - I ) ) ) = 3599 Seconds = 59.98 min

Calculation for the time constant at overload levels ( leq > 2) times the
pic k-up current
At 2 time rated current the relay curve is adiabatic.

we require an ope;ating time of 750 x 0.7= 525 ( 30 % margin )

OALSTOM Limited, Enemy Automation & Information


. - ;. 3
.. . . .,>:.I:
..., ,..


Advanced Industrial Power System


I. :

Motor Protection sewrn
. ..

(2)2x 525 = (1.03)2 XT

r = 1979.45 sec = 32.99 min


At 3 time rated current

dperating time of 190 x 0.7= 133 ( 30 % margin )

.. -


. -. ,

(3)2x 133 = (1.03)2 xr

r = 1128 sec = 18 min



_ I

; i.2

The minimum value of the time constant is considered. So the time

constant considered in this case is 18min


Based on the above setting, the hot curve based on the Hot to Cold R ~ I
.. .$
setting can be calculated as follows.
For Hot curve

Select HCR ( hot to cold Ratio ) = 0.66

. - -.

. . ..-,-54
. ... ..?>3
. .,-;<




For 1.4 times Full load Current time of operalion :

t = { r x In { ( PSM2)/( PSM2- 1 ) ) x 0.66

= (18 x In ( 1.362)/(1.362- 1 ) ) x 0.66
= 9.24 min = 554.4 seconds
For 2 times Full load Current time of operation :

For 3 times Full load Current time of operation :

OALSTOM Limited, Energy Automation & Information



ked Industrial Power System


Motor Protection Setting Criteria

and Tutorials
Page 5 of 38

pmparing the above values with the hot withstand characteristic of the
&tor, we find that the safety margin between relay curve and the motor
h e is clearly more than 30 %.

;"ce the following settings are recommended

Pick-up current
= 0.78
ne constant
= 18 minutes
HCR = 0.66
ling time constant
= 5 x Is ( normally adopted, if data


available )

bermal Alarm required = YES


3ALSTOM Limited, Energy Automation & Information







Mvcnced fndustrial Power System


Motor Protection Setting c&&$

and Tutorials':
Page 6 of a

Start Protection Maximum Motor starting current

at rated voltage


6 Times Full load Current

1 380 Amps

Maximum ~ o t o starting
at lower voltage of say 80%

80% of max starting curent

1 104 Amps

Prolonged Start Proteclion :

This protection will operate if the motor takes longer time to start. The
setting will b e based on the worst case type of voltage for starting, which
occurs when starting with low voltage
Therefore Current setting will be 80% of Max current during starting at low

IS, = 0.8 x 0.8 x 1380 / 300=

Set IS{- = 3.76


Time Setting will determine after how much time with this current will the
relay detect it as a prolonged start. This requires the value of the
maximum starting time, which is applicable, when the motor starts with
low voltage.
Is, = 8 + 4 seconds ( margin) = 12 seconds

Limited, Energy Automation & Information

lvanced Industrial Power System



Motor Protection Settit?=Criteria

and Tutorials
Page 7 of 38

of start limita.lions

!&' Number of Cold starts per hour Set = 3 as motor is rated for 4 starts per
Number of Hot starts per hour Set = 1 as motor is rated for 2 starts per hour
This setting can be changed
depending on how frequently we use the
Phase Sequence Startinq

As the motor most probably will be unidirectional, it is generally advised to

enable Reverse Phase sequence protection.

Set RPS as enabled

OALSTOM Limited, Energy Automation & Information

Mvancsc! !ndustrial Power System


Motor Protection Setting Cfie*

and Tutoriak
Page 6 of

Siait Protection Maximum Motor starting current

at rated voltage

6 Times Full load Current

1380 Ar;ips

Maximum Motor starting currenl

at lower voltage of say 80%

80% of max starting curent

1 104 Amps

Prolonged Start Protection :

This protection will operate if the motor takes longer time to start. The
setting will be based on the worst case type of voltage for starting, which
occurs when starting with low voltage.
Therefore Current setting will be 80% of Max current during starting at low

Isi = 0.8 x 0.8 x 1380 / 300=


Set 1st- = 3.76
Time Setting will determine after how much time with this current will the
relay detect it as a prolonged start. This requires the value of the
maximum starting time, which is applicable, when the motor starts with
low voltage.
1st = 8 + 4 seconds ( margin) = 12 seconds

OALSTOM Limited, Energy Automation & Information


.. . .

f .:

'{dvanced Industrial Power System



Motor Protection Settinz Criteria

Page 7 of 38

Number of start limitations


i Number of Cold starts per hour Set = 3 as motor i s rated for 4 starts per


,I: Number of Hot starts per hour Set = 1 as motor is rated for 2 starts per hour

This setting can be changed depending on how frequently we use the

li-Zeverse Phase Sequence Startinq



As the motor most probably will be unidirectional, it is generally advised to

enable Reverse Phase sequence protection.
Set RPS as enabled

OALSTOM Limited, Energy Automation & !nformation


Advanced Industrial Power System


Motor Protection Setting Criie

and Tutorial
Page 1 0 of S

Ne~ativePhase Sequence:
The negative phase sequence protection has to be graded with the NPS
withstand levels of the motor. In absence of this, it is recommended to
provide current setting equal to rated current.
Set current setting

1 .oo

Time characteristic-setting may be set to-definitetime as we de do not

have the inverse withstand level of motor.
The definite time setting can be set to 0.1 sec.
This means that i f the negative sequence settings reaches rated current,
the relay will operate instantaneously.

OALSTOM Limited, Energy Automation & Information

OALSTOM Limited, Energy Automation & Information




.. ...

.. . .
. .

. .

Advanced Industrial Power. System



Motor Protection Setting Critei

and Tutori~
Page 10 of I


Negative Phase Sequence:

The negative phase sequence protection has to be graded with the NPS
withstand levels of the motor. In absence of this, it is recommended to
provide current-settingequal to rated current.
Set current setting

1 .oo

Time characteristic .setting may be set to defiriite time as we de do not

have the inverse withstand level of motor.
The definite time setting can be set to 0.1 sec.
This means that if the negative sequence settings reaches rated current,
the relay will operate instantaneously.

OALSTOM Limited, Energy Awtomation & Information

strial Power System


.Motor Proiection Setting Criteria

and Tutorials
Page 1 1 of 38

ault setting shall be based on the standing leakage currents

ethod of system earthing. If the system is'resistance earthed,
be required and the required primary operating current may
based on system study.
ent case is assumed to be on a system which is solidly earthed,
balance leakage currents can be measured using standard
de of connection and earth fault setting shall be made
. Generally the leakage currents shall not exceed 5 % of the
t. Therefore the current setting may be set to 10%of the rated

Current setting

0.1 x full load Current x CTsec x1000


0.1-x230 x 1 x 1000


Set lo = 80mA
Time delay setting may be set to instantaneous which will be 0.1 sec (
100mA ) . This is an intentional delay and is used to prevent inrush currents,
which last for couple of cycles from operating the e/f element during


OALSTOM Limited, Energy Automation & Information


Advanced Industrial Power System

Negative Phase Sequence:

provide current setting equal to rated current.

Set current setting

1 -00

Time characteristic-setting may be set to definite time as we de do not

have the inverse withstand level of motor.




The definite time setting can be set to 0.1 sec.

* .I.$




This means that if the negative sequence settings reaches rated current,
the relay will operate instantaneously.

OALSTOM Limited, Energy Automation & Informatian




Advanced Industrial Power System


- -

-Motor Prokction Setting Criteria

and Tutorials
Page 1 1 of 38

Earth Fault Protection:



'The earth fault setling shall be based on the standing leakage currents
and the method of system earthing. If the system is resistance earthed,
CBCT may be required and the required primary operating current may
be studied based on system study.


As this present case is assumed to be on a system which is solidly earthed,

standing unbalance leakage currents can be measured using standard
residual mode of connection and earth fault setting shall be made
accordingly. Generally the leakage currents shall not exceed 5 % of the
rated current. Therefore the current setting may be set to 10%of the rated

Current setting

0.1 x full load Current x CTsec x1000


Set lo = 80mA
Time delay setting may be set to instantaneous which will be 0.1 sec (
100mA ) . This is an intentional delay and is used t o prevent inrush currents,
which last for couple of cycles from operating the e/f element during

OALSTOM Limited, Energy Automation & Information

Notes 1 C T S



'-';Current h n s f o r m e r s are among the most commonly used items of electrical apparatus and yet,
"~surprisingly,there seems to be a general lack of even the most elementary knowledge
$'concerning their characteristics, performance and limitations among those engineers who are
$,continually using them. The importance of current transformers in the transmission and
:&distribution of electrical energy cannot be over emphasised because it is upon the efficiency of
Ecurrent transformers, and the associated voltage transformers, th& the accurst=: metering and
" effective protection of those distribution circuits and plant depend.

Current and voltage transformers insulate the secondary (relay, instrument and meter) circuits
from primary (power) circuit and provide quantities in the secondary which are proportional to
those in the primary. The role of a current transformer in protective relaying is not as readily
defined as that for metering and instrumentation. Whereas the essential role of a measuring
transformer is to deliver from its secondary winding a quantity accurately representative of that
which is applied to the primary side, a protective transformer varies in its role according to the
type of protective gear it serves.
. Failure of a protective system to perform its function correctly is often due to incorrect selection

of the associated current transformer. Hence, current and voltage transformers must be
regarded as constituting part of the protective system and carefully matched with the relays to
fulfil the esseqtial requirements of the protection system.
There are two basic groups of current transformer, the requirements of which are often radically
different. It is true in some cases the same transformer may serve both purposes but in modern
practice this is the exception rather than the rule:
1. Measurement CT's - The measuring current transformer is required to retain a specified
accuracy over the normal range of load currents.
2. Protection CT's - The protective current transformer must be capable of providing an
adequate output over a wide range of fault conditions, from a fraction of full load to many
times full load.
Therefore they generally have different characteristics.


Various international standards are available. Such standards give information on the
classification, selection, error and operation of current transformers. They are a valuable source
of reference and can be used in conjunction with the relay manufacturer guide when selecting
the appropriate CT. The list below gives some examples:





IEC 185.1987
IEC 44-6:1992
IEC 186.1987
BS 7625
BS 7626
BS 7628
BS 3938:1973
BS 3941 :1975 ANSI C51.13.1978
CSA CAN3-C13-M83
AS 1675-1986

CTs and VTs
CTs and VTs

Page 1



Please note that the above are the applicable standards at the ti,me of print of this document an'
therefore they may vary.

. ....:...



. .

A current transformer consists essentially of an iron core with two windings. One winding is
connected in the circuit whose current is to be measured and is called the primary and the other
winding is connected to burden, and called the secondary. Two of the most basic construction of
current transformers are the bar type and wound type:
1. Bar Type - Sometimes referred to as 'Bushing Type'. Such current transformers normally
have a single concentrically placed primary conductor, sometimes permanently built into the.
CT and provided with the necessary primary insulation, but very often the bushing of a circuit
breaker or power transformer. At low primary current ratings it may be difficult to obtain
~ufficientoutput at the desired accuracy because a large core section is needed to provide
enough flux to induce the secondary emf in the small number of turns.




2. Wound Type -With this device it is possible to change the number of primary turns, thus
increasing the CT output voltage with altering the turns ratio. Therefore, for the same output
the wound CT is smaller in CSA than the bar type.

Page 2


: is

no official standard when it comes to defining the polarity of current transformers.

ver, most Engineers will use P I and P2 to define the primary winding and S1 and S2 to
the secondary winding. Generally speaking when P1 goes high S1 goes high. Therefore
current flows from P1 to P2 it is transferred and flows through the external circuit from S1
Typically P2lS2 is towards the Item of plant being protected.


3w current in the primary winding produces an alternating flux in the core and this flux
2s an e.m.f. in the secondary winding which results in the flow of secondary current when
~ndingis connected to an external closed circuit. -The magnetic effect of the secondary
~ t in
, accordance with fundamental principles, is in opposition to that of the primary and the
of the secondary current automatically adjusts itself to such a value, that the resultant
2tic effect of the primary and secondary currents, produces a flux required to induce the
necessary to drive the secondary current through the impedance of the secondary. In an
ransformer, the primary ampere-turns are always exactly equal to the secondary ampereand the secondary current is, therefore, always proportional to the primary current. In an
current transformer, however, this is never the case. All core materials, so far discovered,
e a certain number of ampere-turns to induce the magnetic flux required to induce the
sary voltage.
lost accurate current transformer is one in which the exciting ampere-turns are least in
rtion to the secondary ampere-turns. Exciting ampere-turns may be reduced in three
~ l ways:
By improving the quality of the magnetic material
Cold rolled grain oriented silicon steel (C.R.O.S.S.) has a magnetisation characteristic
with a knee point at 1.6 tesla.

. ,, ,

, .



Nickel steel (Proprietary name Mumetal) has a knee point of 0.7 tesla.

:! . . .




By decreasing the mean magnetic path of the core.






. 2.
."; Q:.




By reducing the flux density in the core.


:I. $'.&

,:.L .g+\;
:,; ,b#

': >;:g ;.
. . ,.



!, 2 .y
p,,: :,

Page 3







Protective relays are designed to operate from secondary quantities supplied from current
transformers and from voltage (or potential) transformers. The secondary output of these
devices is the information used by the relays to determine the conditions existing in the plan
being protected. It is necessary, therefore, that the secondary output of current and voltage
present a true picture to the relays of the conditions in the primary circuit during faults as well as -,.
during normal loads. Or, alternatively, that their performance be known under extreme
conditions so that any error in reproduction in the secondary circuit can be partially or completely
compensated for in the setting and characteristics of the relay.

In many applications, core saturation will .almost inevitably occur during the transient phase of a
heavy short circuit. The performance of the associated instrument transformers during faults is,
therefore, an important consideration in providing an effective relaying scheme. The relays and
their associated current transformers must be considered as a unit in determining the overall
performance of the protective scheme. Consequently, the characteristic of the current and
potential transformers at high currents and low voltage respectively, must be known. In any
current transformer the first consideration is the highest secondary winding voltage possible prior
to core saturation. This may be calculated from :
Ek = 4.44 x B A f N volts
Where :
Ek = secondary induced volts (rms value, known as the knee-point
N = number of secondary turns
= system frequency in hertz
A = net core cross-sectional area in square meters.
This induced voltage causes the maximum current to flow through the external burden whilst still
maintaining a virtually sinusoidal secondary current. Any higher value of primary current
demanding further increase in secondary current would, due to core saturation, tend to produce
a distorted secondary current.
The relevant circuit voltage required is typically :
Equation 1
Where :

= secondary current of ct in amps (assume nominal value, usually 1A or 5A)

= the connected external burden in ohms
= the ct secondary winding impedance in ohms
= the resistance of any associated connecting leads

In any given case, several of these quantities are known or can usually be estimated in order to
predict the performance of the transformers. From the ac magnetisation characteristic,
commonly plotted in secondary volts versus exciting current, Es can be determined for a
minimum exciting current. The equation for the relevant circuit voltage given above then
indicates whether the voltage required is adequate.

Page 4


at a bar primary type 200015A (CROSS core) current transformer having a core csa
square cm's is available with a secondary resistance of 0.31 ohm. The maximum
to which the transformer must maintain its current ratio is 40,000 amperes. It is
determine the maximum secondary burden permissible if core saturation is to be
Assume that the current transformer core will start to saturate at 1.6 tesla.

r From the data given :

N =

2000/5 = 400 turns

= 50 H z .

Secondary current (Is) with a primary current of 40,000A is given by

Knee point voltage Ek is-given as follows :

= 284 volts
Maximum burden permissible (including ct secondary resistance and lead burden) is equal to
= 2 84 ohms
284 / 100
Consequently, the connected burden including that of the p~lotscan be as high as 2.84 - 0.31 = 2
-53 ohms for negligible saturation in the core. Thus it mav be seen that the secondary burden
and the maximum available fault current are two important criteria in determining the
performance of a given current transformer.
A current transformer may operate satisfactorily :

At a high primary current where the connected secondary burden is low


At a lower primary current where the secondary burden IS high



'The primary current contains two components. These are respectively the secondary current
which is transformed in the inverse ratio of the turns ratio and an exciting current, which supplies
the eddy and hysteresis losses and magnetises the core. This latter current flows in the primary
winding only and therefore, is the cause of the transformer errors. It is, therefore, not sufficient to
assume a value of secondary current and to work backwards to determine the value of primary
current by invoking the constant ampere-turns rule, since this approach does not take into
account the exciting current. From this observation it may be concluded that certain values of
secondary current could never be produced whatever the value of primary current and this is of
course, the case when the core saturates and a disproportionate amount of primary current is
required to magnetise the core.

, ,;.

- ..C<,

The amount of exciting current drawn by a current transformer depends upon the core m
and the amount of flux which must be developed in the core to satisfy the burden require
the current transformer. The appropriate current may be obtained directly from the exciting
characteristic of the transformer since the secondary e.m.f. and therefore the flux develope
proportional to the product of secondary current and burden impedance.

The general shape of the exciting characteristic for a typicai yrade of CRZSS (cold rol!e
orientated silicon steel) is shown. The characteristic is divided into three regions, define
'ankle-point' and the 'knee-point'. The w o r k i ~ grange of a protective current transformer e
over the full range between the 'ankle-point' and the 'knee-point' and beyond, while a mea
current transformer usually only operates in the region of the 'ankle-point'. The difference in
working ranges between metering and protective current transformers stems from the radical
difference in their functions. Metering current transformers work over the range 10% to 1
load and it is even an advantage if the current transformer saturates for currents above this
range in order to provide thermal protection for the instruments. Protection current trans
on the other hand are required t o operate correctly at many times-rated current-.

,, . ; + i






. ,.

,' .




MMF ampere-turns per metre

The knee-point of the excitation characteristic is defined as the point at which a 10% increase in
secondary voltage produces a 50% increase in exciting current. It may, therefore, be regarded
as practical limit beyond which a specified current ratio may be maintained.




The current transformer magnetisationcurve, is usually expressed in terms of Kv and Ki which

when multiplied by the flux density in teslas and ampere-turns per cm respectively gives
corresponding volts and amperes :

Page 6

equation, the flux density 6 is in teslas and the core cross-sectional area is in squar

sity B is in teslas and the cross-sectional area is in square centimetres :

e exciting current le in amps can be obtained from the M'MF using the relationship:
le = Ki x


I depend on the units of MMF.

k l f the MMF is in ampere-turns per meter.





Ki = - where L is in metres


' Consider the case of a current transformer

ratio 10015A connected to an earth fault relay. Relay

: burden at minimum tap setting of 1O0/0 of rated current is given as 2 VA. Calculate the required
' values of Kv and Ki to provide the necessary output up to 10 times the plug setting, with :

A barprimaryjype current transformer and with


A wound primary (5 turns current transformer).

Assume the use of a CROSS core; B = 1.6 tesla.


Ring Type Current Transformer (Bar Frimaryj

,Relay current setting

Volts required to operate relay

-- -

Voits required at iO times the

the plug setting


0.5 ampere : ie.lO% of 5A.

= 4 volts


4 x 10 = 40 volts ignoring lead

burden and CT secondary
winding resistance

Therefore, 40 volts must correspond to the knee-point of the saturation curve which
represents a flux density of 1.6 tesla.
With a bar primary, secondary number of turns = 20

Assume stacking factor = 0.92

.-. Gross CSA

= 56.310.92 = 61.2 cm2

Assuming :
O.D. =
Depth =

18 cms
30 cms
10.2 cms

Wound Primary CT


--. . -

Assume current transformer is wound with 5 primsry turns :

l o o = 100
econdary turns = 5 x 5

49 = 4-44 x 53

x 1.6 100 x A. 7 9-4 (A in cm'j

7 .

m - le. csa

1 1.26
-= 12.24 cm'

18 cm
30 cm
2.04 cm

2 6 x 100
= 25
= 0.754 c m 1 turn


The secondary circuit of a current transformer should never be left open-circuited whilst primary
continues to flow. In these circumstances only the primary winding is effective and thus the
current transformer a highly saturated choke (induction) to the flow of primary
winding current. Thus a peaky and relatively high value of voltage appears at the secondary
output of terminals, endangering life, not to mention the possible resulting breakdown of
secondary circuit insulation.
In those cases where current transformers are associated with the "high impedance type" earth
fault relay the secondary circuit burden may have ohmic values up to several thousands of ohms.

Page 9

The errors of a current transformer may be considered as due to the whole of the primary current :&
not being transformed, a component thereof being required to excite the core. Alternatively, we .. '2
may consider that the whole of the primary current is transformed without loss, but that the
secondary current is shunted by a parallel circuit the impedance of which is such that the
equivalent of the exciting current flows there in. The circuit shown is the equivalent circuit of the ':1
current transformer.. The primary current is assumed to be transformed perfectly, with no ratio or
phase single error, to a current Ip/N which is often called 'the primary current referred to the
secondary'. A part of the curre'nt may be considered consumed in exciting the core and .this
current leis called the secondary excitation current. The remainder Is is atrue secondary
current. It will be evident that the excitation current is a function of the secondary excitatbn
voltage Es and the secondary excitation impedance Ze. It will also be evident that the secondary
current is a function of Es and the total impedance in the secondary circuit. This total impedance
consists of the effective resistance (and any leakage reactance) of the secondary winding and
the impedance of the burden.


primary current in amperes

current transformer ratio (primary to secondary amperes)
burden impedance of relays in ohms (r + jx)
current transformer secondary winding impedance in ohms (r + jx)
secondary excitation impedance in ohms Cjx)
secondary excitation current in amperes
secondary current in amperes
secondary excitation voltage in volts
secondary terminal voltage in volts across the current transformer terminal:
(input to the relay or burden)

Beyond the knee-point the current transformer is said to enter saturation. In this region the major
part of the primary current is utilised to maintain the core flux and since the shunt admittance is
not linear, both the exciting and secondary currents depart from a sine wave. For example, in
the case of a wholly resistive burden, correct transformation takes place until saturation flux
density is reached. The secondary volts and current then collapse instantly to zero, where they

Page 10

Notes Additional Analysis


understanding and working knowledge of System Analysis is very important to the

rotection Engineer as he must know how the system operates under load and fault conditions
efore choosing suitable relays to match the system parameters.
Analysis of load and fault conditions also provides useful information for :Choice of Power System Arrangement
Required Breaking Capacity of Switchgear and Fusegear
Application of Control Equipment
Required Load and Short Circuit Ratings of Plant
System Operation. Security of Supply, Economics
Investigation of Unsatisfactory Plant.Perforrnance

110 = R

jX = Zi (cos 0 + jsin 0) =

i B 1 and Z2 = iZ2!
JZ1l1Z21 LO, +(I2 and



= I-


l ejO

L O 1 -(I2


rotates a vector anti-clockwise through 90"

a = 1 L120

rotates a vector anti-clockwise through 120"

used extensively in symmetrical component analysis

a2 = 1 L240

a2 + a

1= 0


Current I flowing in direction shown produces a voltage drop in Z such that A is positive with
respect to B.


. - ..

Page 1


This is particularly useful when analysing large systems with several voltage levels. Before any j?s
system calculations can take place the system parameters must all be referred to common Q4+
base quantities. The base quantities are fixed on one part of the system and base quantities .-$
on other parts at different voltages will depend
on the ratio of intervening power transformers.
The base quantities used are :,.<.::yr<*

Base voltage




Base MVA

5 .-

phase to phase voltage in kV


three phase MVA






Other base auantities can then be established :Base impedance

Base current

Ib F



(kVb)' in ohms.

in M.



Per unit values are obtained by dividing actual values b y base values as follows :-

Per unlt ~mpedanceZp.u

Actual im~edance
Base impedance




Per rlnit v n l t a n ~kV-

Per nit MVA M V A -



Per unit current I P u



Percentage values are commonly used for transformer impedances and where per unit values
are very small. Percentage values are 100 times the equivalent per unit values.

Page 2


- ,





Find t h e fault current in e a c h section f o r a three p h a s e fault a t F.

Base kVb = 3 1
Base MVA


= kv;


= 50
= 2.42~


Z2.u 3 n
0.3 x - = 0.75,,
common base

0.1 p u











Page 3

The base voltage on each side of a transformer must be in the same ratio as the voltage ratio
of the transformer.



mrrect Selection of kVb

11.8 kV

irrect Selection of kVb 132 x 11.8 = 11-05 kV

132 kV

I 1 kV

132 kV

I 1 kV


Selection of kVh 11.8 kV
,,,-...- Correction

141 kV

141x11 =11.75kV




The per [,]nitimpedance of a transformer is the same on each side of the transforni e r
Consider a transformer with voltage ratio kVllkV2.



Actual impedance of the transformer viewed from side 1 = Zal

Actual impedance of the transformer viewed from side 2 = Za2.

zp.u 1

z a ~

Zp.u.2 -



= Zal


- Za2 X MVA



but Za2 = Zal x




Za1 x


= Z



Page 5


There are three basic laws :


Ohms Law


Kirchoffs Junction Law


At any junction (or node) CI = 0:


i.e. 1, + I2 + l3 = 0


' -


Kirchoffs Mesh Law

Round any mesh CE = CIZ

eg, in mesh (1):

El = il Z1 + il Z3

Page 6

- i2


These are derived from the circuit laws. The three most commonly used for system analysis
are Thevenins, Star/Delta Transform and Superposition Theorems.


Thevenins Theorem
This is useful for replacing part of a network which is noi of pan~cularinterest.

Any active network viewed from any 2 terminals can be replaced by a s~ngledrivirlg
voltage in series with a s~ngleimpedance where :Driving voltage

Open circuit voltage between terminals

Impedance of the network as viewed from the two
terminals with all driving voltages short circuited.

Example :

Where E' =

E, and Z' = ---






DeltalStar and StarlOelta Transform Theorems

Page 7

212 =







+ Z31

Superposition Theorem
In any linear network the current in any b r a ~ c h
different driving voltages is equal to the vector sum
voltage acting alone with the others short circuited.
Example :

l3 =



+ 13?
..,: ...-:

Page 8


In a balanced three-phase system, each of the three phases of any part of the system will have
currents and voltages which are equal and 120 displaced with respect to each other. To
maintain balanced operation, each Item of system plant must be symmetrical: i.e. have identical
impedances In each line, equal mutual impedances between phases and ground, and equal

etween two lines and ground.

circuit faults. They can arise

the operation of fuses.


Consider n-dimensional system of phasors.

Va = Val

+ Va2 + Va3 + ... + Van

Vb = Vbl

+ Vb2 + Vb3

+ ... +


. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . .

Vn = Vnl


+ Vn2 + Vn3 + ... + Vnn

Val , Vbl , etc. are phasors of the first set of balanced n-phase
system. Phasors are single spaced. .
Va2 , Vb2 , etc. are phasors of the second set of balanced n-phase
system. Phasors are double spaced.

Page 9

And so on.

Van , Vbn etc. are phasors of the uni-directional phasor system.

Take for example an unbalanced 5-phase system. V, , Vb , V, , Vd , V.,

First set of
aalanced Phasors

Second set of
Balanced Phasors

Fourth set of
Balanced Phasors

Fifth set of
Zero Sequence Phasors

Now consider an unbalanced three phase system. Va Vb , V.,



v,. .

v t>



. ..:$7
. .,

. , ,..
. T,.:


Three unbalanced phasors have been resolved into nine phasors.

Choose 'a' phase as the reference phase and replace Vas by Vao.

where a = 1.0 L120


It is convenient to delete subscript 'a' for the symmetrical components.

Add equations 1. 2,and 3

Multiply equation 2 by u amd equation 3 by u and add the resulting equations to

equation 1.

Multiply equation 2 by u2 and equation 3 by a and add the resulting equations to equation

Equations 1 to 6 can be re-written in matrix form

Re-write matrix equations 7 and 8 respectively as

[VP = [A]



Where Vp = phase components

Vs = sequence components

Resolve the following 3-phase unbalanced voltages into their symmetrical components.

Fig. 1




Fig. 2

/ .

Take a set of symmetrical three phase impedances (equally spaced, fully transposed, etc.)
carrying unbalanced phase currents L a , Ib and I,.

We may write the following equations.

Va = Zsla + Zmlb + Zmlc

where Zs = self impedance per phase


= mutual impedance betrween any ~ h a s e


Or, in matrix form

Resolving V and 1 phasors into their symmetrical components.


Page 15




I; [i i; [i ;*I


Z1 = Z, - Z,

Therefore, if the system is. symmetrical in its normal state the symmetrical co~~?poiaent
impedance becomes diagonal (equation 11) and, therefore, isolated sequence networks are .
obtained with impedances Z1, Z2 and Zo. These three networks will become interconnected
when an unbalance such as a fault or unbalanced loading is introduced. The manner of
interconnection will depend on the new constraints: i.e. the additional system connect~or~s.

Page 16



For static networks i.e. non-rotating plants, the positive and negative sequence Impedances are
the same. These are the leakage impedance of the transformers and the normal phase
impedance of the transmission circuits.
Zero sequence impedance of overhead line and cable circuits is determined by the return path
of the zero sequence currents through earth, earth wires or cable sheaths. The zero sequence
impedance is generally' greater than the positive and negative sequence impedance, being
usually of the order of two to three times the positive sequence value in the case of overhead
For transformers, if zero sequence currents have an available path and can flow, they will again
see the leakage reactance in each phase. If no path exists, an open circuit must be shown for
the particular windings in the zero sequence network. The flow of zero sequence current in any
winding is possible only if other windings provide a path for the flow of balancing zero sequence


Consider the transformer equivalent circuit in F~gure3 overleaf. The magnet~singimpedance Z ,

is of the order of 2000/~, compared to the leakage impedance ZIP + Z1, of about 10%.
Therefore, magnetising impedance can be ignored and the transformer can be represented in
the positive and negat~vesequence networks by a series impedance (=ZIP+ Z,,).

. .
J .

.. ....-


> .




primary winding leakage

Z1, = secondary winding leakage
Zm = rnagnetising impedance

Fig. 3 Transformer Equivalent Circuit

In the zero sequence network, although the leakage impedance is identical to the positive
sequence value (when zero sequence path is available) the zero sequence rnagnetising
impedance is dependent upon the transformer core construction and can be much lower. In
three-phase banks of single phase transformers and in three-phase shell cored transformers,
the zero sequence magnetising impedance is ,large and can be ,ignored as in the positive and
negative sequence networks. In three-limb core type transformers, however, the zero sequence
flux must be completed through the oil or tank. Owing to the high reluctance of the flux path,
zero sequence magnetising impedance is of the order of only 100% to 400%. However, this is
still high enough to be neglected in most fault studies, particularly when a delta winding Is

Page 17

Therefore, consider zero sequence circuit of transformer as a series impedance Zt. The mode
of connection of Z, to the external circuit is determined by taking account of each winding
arrangement and its connection or otherwise to ground.

Imaginary links 'a' and 'b' (see Figure 4) are used to derive the connections. If zero sequence
currents can flow into and out of a winding, for example a solidly earthed star winding, the
winding terminal is connected to the external circuit, that is link 'a' is closed.


Fig. 4



If zero sequence currents can circulate in the winding without flowing in the external circuit, for
example a delta winding, the winding terminal is directly connected to the zero bus, that is link
'b' is closed.



!i .


Example 1




Zero Sequence Equivalent

Circuit Connections

Page 18




i h e zero sequence impedance of a neutral earthing impedance Zn is 32..

can be readily understood from Figure 5 below.

At the neutral point the zero sequence currents I0 in the three phases combine to give 310 in the
neutral earthing impedance. The zero sequence voltage at the neutral point is given by


The reason for this

vo =


zo = vo = 32"

.. .

.. .

. . ..


Example 2

Transformer Connections



Zero Sequence Equivalent

Circuit Connections

Page 19




The positive sequence impedance of synchronous machines is the normal machine reactan
There are three defined values of positive sequence impedances, namely the synchron
transient and subtransient impedances and they are used according to whether steady st
transient or initial short-circuit values of current are required.


Unlike the non-rotating networks, the negative sequence impedance of the rotating plants is
equal to the positive sequence impedance. It relates to mmf at synchronous speed travelli
the opposite direction to the rotor. Its value is usually less than that of the positive sequ
, <.,.,~<

. .....


In the zero sequence network, the winding connection and earthing arrangement must be
considered as for transformers. Any earthing impedance will be seen by each phase and
therefore the correct voltages will be obtained if three times the impedance value is included in
the zero sequence network.



. .,?.
. %.

. s.


Typical turbo-generator sequence reactances are :

synchronous reactance
transient reactance
subtransient reactance
negative sequence impedance
zero sequence impedance



1.0 p.u.
0.15 p.u.
0.10 p.u.
0.13 p.u.
0.04 p.u.




>.+<;-, .%:.
. ..

. . .:. .,&xS!



. .

For anygiven fault there a% six quantities'to be mnsidered at the.fault point; Vat Vb, V,,
I,, It,, lc. If any three are known(provided they are not all voltagesor all currents) or.if
any two are known and two others known to have a specific relationship, then a
relationship between V1,Vp and Vo and 11, 12 and 10 can be established.

. :;T.><.>.
.. <. 6.
. ~ . >. -

. .


.., .,






. . .




. . ,.


.. .'...:.


These relationships are-called the circuit constraints.

- ...?

From the circuit constraints we can determine the manner in which the isolated sequence
networks can be interconnected.

,: . ..



,. . .


The relationships are derived with phase 'a' as the reference phase and the faults are
selected to be balanced relative to the reference phase. This yields the simplest
interconnection of the sequence networks. If this is not done the interconnections of the
sequence networks require additional transformations which are achieved by the
introduction of phase shifting transformers. This will be apparent in the case of
simultaneous faults where it is not possible for both the faults to be symmetrical about the
reference phase.








Page 20

:: :)

i t Faults


ine-to-ground faults, line-to-line faults, line-to-line to ground faults and three phase faul'ts all
,fall into the category of shunt faults.



, (b)

Figure 6 shows a system with a fault at F. The positive, negative and zero sequence
networks of the system are shown in Figure 7. The fault terminals for the positive
sequence network are F1 and N1, and the corresponding fault terminels for the negative
and zero sequence networks are F2, N2 and Fo, No respectively. It is at these terminals
that the interconnection of the networks will occur. In the denvation of sequence network
interconnections, it is convenient to show the sequence networks as blocks with fault
terminals F and N for external connections (F~gure8)
To derive the system constraints at the fault terminals, it IS convenient to imagine three
short conductors of zero impedance connected to the three line conductors at the point of
fault (F~gure9). The terminal conditions imposed by the different types of faults will be
applied to these imaginary leads, t h e potential to ground of which will be V, Vb and V,
and the currents ,I I b and I,. .,

Fig. 9


-Fis. 6 Single Line Diagram of Two Machine System


6 ~ ,

Pos~tiveSequence Network of System

@ Sequence & k k

of System

Fiq. 7 Sequence Networks of Faulted S y s t e ~


etwork Blocks

Page 22

Line to Ground on Phase 'A'

At fault point :

We know from section (2.2)that





But Va = 0

v, + v 2 + v o = o---------------------

We know from section (2.2) that



= 1/3(la+


= Ic= 0

l b +



10 = 113 ,1

Also, l1 = 113 (Ia + ulb + a I,)


= 113 (I,

lb +

= 113 1,

ul,) '= 113 ,I

Equations 3 & 4 are the CIRCUIT CONSTRAINTS. They suggest that the sequence netwo
are connected in series. -




Page 23

Line to Ground Fault through Fault Impedance ZF

At fault point :

We know from section (2.2) that


I0 = 113 ,I

since lb = 1,

= 0


:. 11 = 12 = 10 = 113 1,

We know

But V,

= I,Zf from constraint 2

But 1, = 310 from equation 3

Equations 3 & 4 suggests the following interconnections.














Page 24

Line to Line Fault on Phases 'B' and 'C'

At fault point :


We know l o = '113 (I, + I b + I,)

Substituting equations 2 & 3 into equation 4,


11= 1/3(Ia + a l b + u I,)


12 = 113 (I, + a I b + uI,)

= 1/3(a-a)

= -113 ( a - a )



:. I1 + 12 = 0

Substituting equation 1 into equation 7,

V1 = 113 (V,

- Vb)

Similarly V2 = 113 (V,

+ a Vb + a ~ i =
) 113 (V,

- Vb)

From equations 5, 6 & 8, the positive and negative sequence networks are in parallel but the
zero sequence network is unconnected.




Line to Line Fault on Phases 'B' and 'C' through Fault Impedance ZF

Page 25

At point of fault,

I, = 0
Ib +

lc = 0

Vb - Vc = IbZf

:. 10

= 113 (la + Ib + I,)

11 =

113 (I, + a l b + a 1,) = 113 (a - a ) lb

1 2 = 113(1, + a

= 0

Ib +


= -1/3(a-a)lb


I0 = 0
I1 +

I2 = 0

We know Ib = 10 + a I l + u12 )

--------------------- 5

Substituting equation 4 in 5

Vb = VO +

aLvl +


Vo + rxVl + cr V2

... V b - V C =

(a2 - a ) V 1

- (a2 - a ) V 2


Substitute equation 3 8 6 into 7,

Equations 4 8 8 suggest the following interconnections.



Line to Line to Ground Fault on Phases 'B' and 'C'

Page 26


At fault point :,

vc = 0

I, = 0

V2 = 113 (V,

+ a Vb + UV,)

= 113 V,

From equation 3 & 4, it can be concluded that the sequence networks are connected in parallel.





O N 1 -





Line to Line to Ground Fault on Phases 'B' and 'C' through Fault Impedance Z,
At fault point :,

Page 27


V1 = 113 (V,

v2 =

+ aVb +

113(~, +

2. .
a vc)


+ (a + a)Vbl = II~(v,-

aVc) = II~[V,


(Ib + )1, Zf .....................

.'. v1 = v2

= 113 [V, + (a + a)Vb] = 113 (Va- Vb)


Vl = 113 (2Vb + Vb) = Vb


Substitute equation 4 in 6

vo - v1 =


Equations 3, 5 and 7 suggest the following interconnections.







-- l o





Page 28





SERIES FAULTS (or Open Circuit Faults)


Figure 1 shows a system with an open circuit PQ. The positive, negative and zero
sequence networks of the open-circuited system are shown in Figure 2. Unlike the
case of shunt faults, the fault terminals for interconnection are P and Q, therefore
not I nvolving the neutral. The sequence equivalent network blocks (Figure 3) will
have terminals P and Q for interconnection. Terminal N is also indicated in the
blocks although it is not used for interconnections.


The terminal conditions imposed by different open circuit faults will be applied
across points P and Q on the three line conductors (see Figure 4). Therefore the
fault terminal currents will be IA, IB and Ic flowing from P to Q on the three
conductors, and the terminal potentials will be the potential across P and Q, i.e. V,

- V,',Vb - Vbl, Vc - V;

They will be represented by v,

Figure 4

Page 29

vb and vc respectively.