TRANSFORMER PROTECTION

Introduction
Differential protection of transformer was introduced in lecture 2. Traditionally it
involves establishing circulating current through matched CTs installed on the primary
and secondary winding of the transformer. If there is no internal fault in the transformer,
zero current flows through the differential overcurrent element. However, in case of
internal fault, the CT secondary currents are not matched and hence the differential
current d
i
is not zero. This causes the overcurrent element to pic up and operate the
circuit breaers to isolate the transformer.
!election of CT Turns "atio
#et the transformer turns ratio given by
2 $
% N N and the corresponding CT ratio be given
by
$
% $ n and
2
% $ n as given in fig$. Then,
Current in CT & $ primary '
$
I
Current in CT & $ secondary '
$
$
n
I
Current in CT & 2 primary '
2
$ $
N
I N
Current in CT & 2 secondary '
2 2
$ $
n N
I N
If there is no fault, then with proper connections account for the CT polarity, we should
obtain circulatory current through CT secondary.
Hence,
2 $
i i
i.e,
2 2
$ $
$
$
n N
I N
n
I

or
$ $ 2 2
n N n N ((((((((((((((((((((((((((((((((()$*
If the transformer )to be protected* is woring on tap T as shown in fig 2, then the above
e+uality has to be modified as,
T n N n N
2 2 $ $
((((((((((((((((((((((((((((()2*
,-ample
#et the primary of the transformer winding has $... turns while secondary has /.. turns.
If the primary CT ratio is $..%/, find the CT ratio re+uired in the secondary side to
establish circulatory current scheme.
0nswer
$
N '$..., /..
2
N and
$
n '2.
1rom e+uation )$*, we get
2.
/..
2. $...
2
$ $
2

N
n N
n
Thus, a suitable secondary CT ratio would be 2..%/.
"emars$% !ometimes due to 3odd turns ratio4 involved in primary, it may not be possible
to obtain matching CTs on the secondary. In such situations 3au-iliary CTs4 are used
either on primary or secondary )or both sides* to obtain circulatory currents in absence of
internal faults.
"emar 2% The circulating current scheme described above has been traditionally used
with electromechanical and solid state relays. However in case of numerical relays, such
physical connections are no more re+uired. 5iven turns ratio
2 $
% N N and CTs ratio
$
% $ n and
2
% $ n , one can wor out the e-pected current in secondary of transformer )in
absence of internal fault*. Hence, au-iliary CTs become redundant and the transformer
connections are simplified drastically. 6ith numerical relaying most of the hardware
connections can be easily accounted through in software.
"emar 7% 6hen dealing with three phase transformers, the transformer connections lie
8(8 or Y also play a role in determining CT secondary interconnections to establish
circulating current scheme. This is because of the phase shifts typically of the order of
t7. that result in the line currents when we move from primary to secondary side of
the power transformer. Fig(((( shows the typical connections for Y 9 transformer ban
for establishing the circulatory currents. The study of the circuit brings out the following
important rule for interconnection of CT secondary for Y 9 transformers :If the power
transformer winding are connected in 8 configuration, use configuration for
corresponding CT secondary interconnections.
"emar 2% 6ith numerical relays such interconnection comple-ity can be easily handled
in software. 0fter, specifying the turns ratio and the phase shift from primary to
secondary, it should be possible to wor out the e-pected secondary differential current
by simple calculation.
"ole of ;ercentage Differential ;rotection
!o far our discussion has focused on an ideal transformer. However, practical
transformers and CTs pose additional challenge to protection.
)$* The primary of transformer will carry no load current even when the secondary is
opening circuited. This will lead to differential current on which the protection
scheme should not operate.
)2* It is not possible to e-actly match the CT ratio as per e+uation )$*. This would
also lead to differential currents under healthy conditions.
)7* If the transformer is used with an off nominal tap, then differential currents will
arise as e+uation )2* is not satisfied even under healthy conditions. However, tap
position can be read in numerical protection scheme and accounted by e+uation
)2*. This would mae the numerical protection scheme adaptive.
To prevent the differential protection scheme from picing up under such conditions a <
differential protection scheme is used. )see Fig*
It improves security at the cost of sensitivity. =otice an offset of m
I
to account for the no
load current. The current on the -(a-is is the average current of primary and secondary
winding referred to primary. It indicates the restraining current while the corresponding
difference represents the differential current. The differential protection will pic up if
magnitude of differential current is more than fi-ed < of the restraining current.
Complications of >agnetizing Inrush
Fig 3
6e now plan to show that differential protection scheme will misclassify the inrush
current as fault current. This is because during inrush, the secondary current is negligible
)zero if secondary is open circuited*, while primary current can be as high as $. & 2.
times full load current. Thus some ind of restraint function is re+uired to inhibit the pic
up on inrush. Traditionally, this restraint is based upon second harmonic content in
primary which is high during inrush conditions. 0lternatives in numerical relaying also
include voltage restraint used in integrated substation protection scheme and flu- restraint
scheme. ?efore, discussing these schemes in detail, we will review the origin of
magnetizing inrush phenomenon.
Consider the circuit in fig 7. The switch is closed at t ' .. ?y 1araday4s law,
$
sin
m
d
V t
dt

, with
$ $ m
N
((((((((((((((()/*


t
m
t
t d t
V
d
. .
$
* ) sin

@ cos $ A * . ) * )
$ $
t
V
t
m

* . ) @ cos $ A * )
$

+ t
N
V
t
m
and
* . )
2
$
ma-

+
N
V
m
#et for simplicity, assume that
. * . )
. I.e. core is initially demagnetized. Then, the
post energisation flu- wave form in the core is shown in fig 2.
This wave form is +uite different from what we obtain by steady state analysis. During
steady state analysis, we can replace
dt
d
by
j
and corresponding steady state phasor
e+uation is,
$ $
V j N

i.e,
B
B
$
$
V
j N

$
N
V
m
m

and
*
2
sin) * )

t t
m
It is thus clear that pea flu- during energization is twice as high as the steady state pea
sinusoidal value.
=ow,
knee
m
B
A

, where knee
B
is the nee point flu- density of core. Thus, during
energization, the core is driven deep into saturation. The resulting H and hence
magnetizing current can be very high )up to 2. times full load current*. This current is
nown as inrush current of transformer. !o far, we have assumed
. * . )
. ?ut in
general,
* . )
depends upon the remnant flu- in the core and H can be anywhere
between
m

and
m
+
. Hence,
ma-

during energization will vary from

m

to
m
7
.
Fig 5 visualizes the nature of the inrush current seen in the core.
Cur analysis so has a discrepancy. 0s per our analysis flu- and current wave shape in fig
2 should persist till infinity. i.e. it should represent the steady state flu- and current
waveform in the core. However, we also now that steady state flu- and current
waveform as per Fig 4a. The resulting discrepancy is a conse+uence of our assumption
viz, core is ideal and winding non resistive. If we model the winding resistance we would
observe that the voltage available at the ideal transformer primary terminals would reduce
drastically due to large voltage drop in the winding impedance during inrush condition.
Conse+uently, the effect should propagate and reduce flu- in the core. If the flu- in the
core reduces, so would magnetizing current. Thus, each cycle would have smaller
magnetization current pea and the magnetizing current and flu- would finally approach
the one produced by steady state phasor e+uation. Typical inrush current waveform is
setched in Fig 6 below.
To summarize the discussion so far,
)$* Transformer when energized can be subDected to large inrush currents.
)2* Inrush current can trip differential protection of transformers.
)7* !ome methodology to detect inrush has to be devised and transformer differential
protection has to be inhibited from tripping during this condition.
Detection of Inrush Current and Cvere-citation Condition
0ppreciable differential current can result due to either inrush or overe-citation of
transformer. 6hen a transformer is overe-cited then from the relationship
m
f N V 22 . 2
$ $

, we can infer that pea sinusoidal flu- is large. This implies that
transformer core will be driven into saturation for an interval in each half cycle. Due to
this, there would be distortion from the sinusoidal magnetization current. During
saturation, the corresponding magnetizing current can be +uite large and on the resulting
differential the protection system may issue a trip decision by confusing overe-citation
for an internal fault. Hence, in practice restraint has to be provided for both
overe-citation and inrush current condition. There are three possible ways in which this
can be achieved.
(1)Second Haronic Re!"rain"
0nalysis of current waveform indicates that inrush current is rich in second harmonic and
current during overflu-ing has large fifth harmonic component. Thus, if we compute the
second and fifth harmonic current in 2 $
I I I
d

)see fig $*, then, we can provide
following logic for restraining operation of differential protection. "estraining operation
of differential protection if
$*
$ 2
E E I I <
)=o load magnetization current*
or
2*
2 2
2 $ 2 $
I I I I +
<

A1or standard differential protection@

)6here

is the slope of the < differential characteristics*

or
7* d dH
I I >
2 A1or restraining inrush current@
)6here 2 dH
I
is the magnitude of second harmonic*
Cr
2* d dH
I I >
/ A1or restraining overe-citation@
)6here / dH
I
is the magnitude of fifth harmonic*
Typically, and

are the percentage harmonic restraint and would depend upon type
of transformer and steel. 1or numerical relays, )to be discussed in later lectures*, the
design of anti aliasing filters also affects the choice of above parameters. During any
transient )including the fault condition*, harmonics develop much more rapidly than the
fundamental and hence, typically numerical relays are restrained for about a cycle
indirectly by these transients. Typical setting for

used in practice are $., 2. or 7.<.

(#) $o%"age &a!ed Re!"rain"
In an integrated substation protection system, it is reasonable to assume availability of
bus voltage measurement for transformer protection without e-tra cost of FT. Then,
voltage measurement can be used to restrain the operation of differential protection
scheme. Traditionally, this has been referred in literature as :tripping suppressorG and it
suppresses tripping function. If the voltage signal is high the relay is restrained if
E E
d
I <
BBBBBBBBBBBBBBBB )a*
E E E E
d r
I I <
BBBBBBBBBBBBBBBBBB )b*
> E E V
BBBBBBBBBBBBBBBBBBBBBBBB)c*
0nd
p t
v
> E E
BBBBBBBBBBBBBBBBBBBB )d*
The terms a, b and c are self e-planatory.
The last restraining function v
t
is nown as the transient monitor function. The transient
monitor function is used to +ualify the purity of data. Typically, in a numerical relaying
set up, with moving window algorithm, there are a certain number of windows in which
both pre fault and post fault data points are present. 0ny phasor computation done with
them is meaningless because the window does not fit with either prefault or post fault
signal. Hnder such condition, the residual error i
r
)
th
i sample* between estimated
)reconstructed* and measured signal is +uite high. 1or a half cycle window with I
samples in it,

I
$
E E
k
k v
r t
. 6hen v
t
is greater than threshold value

, then decision
maing should be inhibited.
(3) F%'( Re!"rain" Sc)ee
Fig * shows that the flu- current plane is associated with no fault and internal fault
regions. It is difficult to evaluate the actual flu- in the core, because the initial condition
is unnown. It depends upon the remnant flu-. 1ortunately, to distinguish no fault from
the faulted region, we are interested in the slope d
i
curve rather than the actual
values. This can be easily wored out as follows.
#et the voltage at the terminal of the transformer be v)t*, current i)t* and let # be the
leaage inductance of the winding. If we neglect the resistance of the winding, then,
dt
t d
dt
t di
L t v
* ) * )
* )


2
$
2 $ 2 $
) * ) * ) * A ) * ) *@
t
t
t t v t dt L i t i t

using trapezoidal rule of integration.

2 $
2 $ 2 $
) * ) *
) * ) * A ) * ) *@
2
v t v t
t t t L i t i t
+

L
i i
v v t
i i di
d
k k
k k
k k
k k
k

1
]
1

,
_

$
$
$
$
2
Cperation on the unsaturated region of magnetizing curve produces large value of slope
di
d
. 6hile the fault or no fault )saturated* regions have smaller
k
d
di
_

,
slopes, we can
now distinguish internal fault from the saturation condition by following counter scheme.
If current differential indicates trip and
<
,
_

k
di
d
, increment counter.
$
r r
k k +
If . >
r
k and
>
,
_

k
di
d
, then
$
r r
k k
If .
r
k and
>
,
_

k
di
d
, then
.
r
k
If there is an internal fault in the system, then
r
k will monotonically increase and once
r
k crosses a nown threshold, trip decision will be issued. Cn the other hand, during
inrush or over e-citation the
di
d
will alternate between low slope and high region
depending upon whether the core is in saturation or not. Thus, the counter
r
k will
indicate a small tooth ind of behavior with
ma-
r
k below the threshold. Hence, operation
of the different protection scheme would be restrained in this region.
Three ;hase and Three 6inding Transformers
6e have illustrated the basic principle so far using a single phase transformer for
simplicity. However, in practice, we use both three phase two winding and three phase
three winding )primary, secondary and tertiary* transformers. The basic principle of
differential protection is the same but we now have to scale up to multiple phases. 1or a
three phase )two winding* transformer, there would be 7 trip currents )one per phase* and
three restraining currents )one per phase*. 1or three phase three winding transformers,
two restraining per phase are re+uired.
m