SHUNT REACTORS

Page of

DESCRIPTION OF SHUNT REACTORS

General statement
Shunt reactors are inductive loads that are used to absorb reactive power to reduce the
over voltages generated by line capacitance.
An inductive load consumes reactive power versus a capacitive load generates reactive
power.
A transformer, a shunt reactor, a heavy loaded power line, and an under magnetized
synchronous machine are examples of inductive loads. Examples on a capacitive load are
a capacitor bank, an open power line and an over magnetized synchronous machine.
Although shunt reactors are inductive loads similar to transformers but they are different
than transformers in terms of construction and some electrical characteristics.
To describe the shunt reactors better, we need to look at the different designs of shunt
reactors and their electrical characteristics.

. Design of shunt reactors

Design and operation
Shunt reactors are mainly used in transmission networks. Their function is to consume
the excess reactive power generated by overhead lines under low-load conditions, and
thereby stabilize the system voltage. They are quite often switched in and out on daily
basis, following the load situation in the system. Shunt reactors are normally connected to
substation busbar, but also quite often directly to the overhead lines. Alternatively, they
may also be connected to tertiary windings of power transformers. The shunt reactors
may have grounded, or reactor grounded neutral.
Shunt reactors normally have iron cores with integrated air gaps. Due to the air gaps, the
iron cores cannot be significantly saturated, and the reactors therefore will have a
reasonably linear behavior during energizing events, for example.
Three-phase shunt reactors may consist of three separate single-phase cores, or they
could be of three-leg or five-leg design (alternatively shell type), see Figure and
Figure .
For transmission voltages the Five-leg core type or shell type are mainly used. They
make the three phases magnetically independent, while three-leg cores lead to magnetic
coupling between phases.
Any significant period of one or two phase excitation necessitates the provision of a
clearly defined return path for the zero-sequence flux created by the asymmetrical
excitation. Where single phase operation is likely to occur, e.g. in power systems
Page of

employing single pole auto-reclosing, there are two optional ways to achieve such a
return path for the zero-sequence flux. These are:
. To use a three phase -limb core (or shell type core).
. To use single phase units.
One major advantage with a five leg reactor (or shell type) compared with a three leg
reactor is that the construction to reduce vibrations and the long term use is much more
stable and stronger.

Figure Three-leg shunt reactor core

Figure Five-leg shunt reactor core with three wound limbs

Medium voltage reactors, connected to tertiary windings of transformers, in most cases

have air-insulated windings without iron cores.
Oil or Dry type shunt reactors
There are two types of shunt reactor groups one is the oil immersed type similar to
transformers the other type is the air core or core-less reactor.
The dry type of reactors are normally used up to . kV and often installed on the
tertiary of a transformer Error! Reference source not found.].
Page of

The design is divided in gapped core and core-less reactors. The gapped core has a
subdivided limb of core steel with air gaps inside the winding and no limb at all for the
core less concept. The gapped core gives compact design with low losses and low total
mass, low sounds and low vibrations.
Higher energy density can be achieved in a gapped core compared to an air core.
The slope of the permeability is greater in a gapped core versus an air core reactor.
The primary advantages of dry-type air-core reactor, compared to oil-immersed types, are
lower initial and operating costs, lower weight, lower losses, and the absence of
insulating oil and its maintenance. The main disadvantages of dry-type reactors are
limitations on voltage and kVA rating and the high-intensity magnetic field. There is no
magnetizing inrush current when the reactor is energized.
Unit ratings for existing single phase or three phase shunt reactors are:
Three phase up to: Mvar
Single phase up to: Mvar
System voltages up to: kV
Single Phase reactors
Single phase reactors are used when the power is above the limits for a three phase shunt
reactor.
Three phase (-leg or -leg shunt reactors)
Most three phase shunt reactors are designed with five limbs, because of a more robust
construction and to reduce vibrations over time, since shunt reactors should last -
years.
The unwound side limbs results in that the zero sequence impedance is equal to the
positive sequence impedance.
In a high voltage star connected shunt reactor the zero sequence reactance is dependent
on the core arrangement.
The physics are the same as the case with a star/star connected transformer. Under
symmetrical excitation the sum of momentary flux values in the three phases is zero. But
under earth fault conditions this is not the case and the resulting flux must find a way
back external to the three phase coils. In three limb reactors, this resulting flux would go
through the air from yoke to yoke; it means that the zero sequence reactance is lower than
the normal reactance and also non-linear, leads to higher zeros sequence current.
In some applications it is a distinct advantage if the reactor has high and constant zero
sequence reactance. This is the case when single pole reclosing is either required from the
beginning, causing zero sequence flux each time a single pole is opened, or considered
for future development of the system to limit the zero sequence flux. In certain cases (e.g.
Page of

line connected shunt reactors) it is also recommended that the zero sequence reactance be
tuned to a fixed, high value by the addition of an auxiliary neutral reactor.
As for transformers, a high zero sequence reactance requires a low reluctance unwound
return path in the magnetic circuit, leads to smaller zero sequence current which is
achieved with a five-limb core. In a reactor this will result in a next to absolute
decoupling of the phase limbs from each other because the wound limbs are gapped and
the outer limbs are not. This is easily verified by measuring non-induced voltage on the
other phases when one phase is energized Error! Reference source not found.].
Split winding
The split winding is used when the current have exceeded the maximum of mechanical
reasons in the construction, then two windings per phase will be parallel in the reactor.
Shunt reactors equipped with auxiliary power windings
Because shunt reactors are used to control voltage at the receiving ends of long Extra
High-Voltage and Very High-Voltage transmission lines, these reactors often are located
in remote regions that may not have an extensive or reliable distribution grid. Obtaining
reliable station power for the reactor switching station can be a problem. The highvoltage reactor application usually calls for oil-immersed reactors that look very similar
in appearance to power transformers. When designed with an air-gapped iron core, these
reactors can be equipped with a secondary core and winding such that a low-voltage can
be extracted from the high-voltage line, see Error! Reference source not found.] and
Error! Reference source not found.].
Shunt reactor in the neutral
For line connected shunt reactors, an additional single-phase reactor (neutral reactor) is
sometimes connected between neutral and ground. The purpose of this reactor is to
increase the overall zero sequence reactance of the overhead line. In this way, the fault
current is kept small in the event of single-phase line faults cleared by single-pole
opening of the line breakers. As a result, there will be a high probability that the arc at the
fault location is extinguished and that the reclosing operation is successful.
Variable shunt reactor (VSR), with tap-changer
Shunt Reactors are used in high voltage energy transmission systems to stabilize the
voltage during load variations. A traditional shunt reactor has a fixed rating and is either
connected to the power line all the time or switched in and out depending on the load.
Recently Variable Shunt Reactors (VSR) have been developed and introduced on the
market. The rating of a VSR can be changed in steps, The maximum regulation range
typically is a factor of two, e.g. from - Mvar. The regulation speed is normally in
the order seconds per step and around a minute from max to min rating. VSRs are today
available for voltages up to kV.
Page of

The variability brings several benefits compared to traditional fixed shunt reactors. The
VSR can continuously compensate reactive power as the load varies and thereby securing
voltage stability. Other important benefits are: - reduced voltage jumps resulting from
switching in and out of traditional fixed reactors, - flexibility for future (today unknown)
load and generation patterns, - improved interaction with other transmission equipment
and/or systems such as coarse tuning of SVC (Static Var Compensation) equipment, limiting the foot print of a substation if parallel fixed shunt reactors can be replaced with
one VSR, - a VSR can be used as flexible spare unit and be moved to other locations in
the power grid if needed.
Shell type core
The reactor design is said to follow either a core or shell type concept Error! Reference
source not found.].The difference in the two concepts was originally attributed to the
arrangement of the core. In the core type transformers the coils appear to surround the
core and in the shell type the core appear to surround the windings, see Figure .

Figure Reactor construction, Core type (A), Shell type (B).

This simple definition does not always hold. The way reactors are built today; a
description as follows would be more adequate:

In a core type reactor the core limb has a shape of a cylinder around which the
coils are arranged. For normal core type power reactors the coils too are
cylindrical and arranged concentrically. Each terminal is connected to one coil or
several coils in series. Further the coils are slid down around a pre-made core
limb to which yokes are connected after the windings are in position. Most often
the core limbs and yokes are in vertical position.
In the shell type reactor the separate coils have rectangular cross section and they
are wound in one plane. After the winding work the coils for one terminal are
stacked up on each other and connected together. The groups of coils are then in
turn, stacked together to form a winding packet for the complete circuit. The
packets for each phase are then raised to the upright and adjusted position it has in
the reactor. In and around these packets the core is now built up.

Page of

A five leg shunt reactor of core type has similar characteristics as a three leg shell type
reactor (to have a low vibration and noise level, low zero sequence current), so the reason
for a manufacturer to keep to a certain concept may today be historical.

. Electrical characteristics
The electrical characteristics of a shunt reactor that needs to be studied are:
Air-gap of the shunt reactor core.
Switching in a shunt reactor
Disconnection of the shunt reactor
Dominating harmonics
Hysteresis
Losses in shunt reactors

.. Air-gap of the shunt reactor core

To avoid saturation of the iron-core of the shunt reactor, small air gaps are distributed
along the core. The distribution of the magnetic flux intensity from the iron to the air-gap
can in that way handle larger H-field (magnetic flux intensity [A/m]). The air-gap is
seldom larger than a couple of millimeter. The magnetic flux intensity is much larger in
air than in iron approximately times.
In a transformer the aim is not to have any air-gap, therefore the slope in the B-H curve is
very steep with large hysteresis and remanence compared to a shunt reactor, however for
a reactor with air-gap the B-H curve becomes flatter and the hysteresis is very small with
practically no remanence. This leads to that there are small inrush currents and long DC
constants for the shunt reactor, compared to the transformer.
Air core reactors (no iron core), have small inductance L, high losses and high current in
the windings.
Oil immersed iron core reactors with air-gap, have higher inductance, smaller losses and
less current in the windings.
By introducing iron in the winding a higher inductance can easily be achieved without
increasing massively the number of turns as in air-core.

.. Inrush
When switching in air core or iron core reactors, long DC components up to second can
occur, some differences distinguishes between inrush in air respective iron core reactors.

... Switching in air core reactors

B
= k is
H
constant) a simple model with a breaker switched in can describe the electrical principles,
see Figure below Error! Reference source not found.]:

For an air core reactor where L is practically constant (i.e. permeability =

Page of

e = E sin( t + )

Figure Simple shunt reactor model

e uL
= iL
R
di
uL = L L
dt
di
eL L
dt = i
L
R
R
e
i L + i L =
L
L

eq.
eq.
eq.

eq.

The instantaneous voltage value when the breaker is closed.

e = E sin( t + )
eq.
The derived short current obtained through the limiting values
i L = i AC + i DC =
i = i AC
eq.
and L

t
=

t
=

E
=
Z

R
t

i L = i AC + i DC
sin( t + ) sin( ) e

Where the impedance is

Z = R + (L)

eq.

eq.

The first term in eq. states the time function for the steady state condition which is an
AC current and the second term indicates the transient condition which is a damped DC
current.
L
.
The time constant of the damping is =
R
The time constants for shunt reactors are longer than transformers and can be up to
second. Transformers have DC constants up to a couple of hundred milliseconds.
Page of

The inrush can therefore easily saturate a CT measurement of the current.

As can see from eq. the size of the short circuit current is depending on the voltage
phase angle at the instantaneous moment when the breaker is closed. If = then the
elapsed curving inward oscillation disappears (transient part) and the short circuit current
only consists of the steady state part.
With regard to the instantaneous phase angle of the voltage there is two cases of special
interests:
Case =
The short circuit is here done when the voltage run-through zero. In this case the current
reaches its absolute maximal instantaneous value, which is realized analytically when the
derivative of i k with regard to and t shall be zero in the maximum moment, i.e.
R
t

ik
= I k cos( t + ) cos( ) e L =

R
t

ik
R
= I k cos( t + ) sin( ) e L =
t
L

tan ( ) = tan( ) =

L
R

{ = , = , = , }

eq.

eq.

eq.

The maximum inrush current ik max appears when using realistic power system
impedances, approximately at the same time when the steady state current reaches its first

peak value, i.e. t

, see eq. below:

+
L
ik = I k + sin( ) e

eq.

We observe that the current time derivative is zero in the short circuit moment and that
the DC component initial value is not the maximal imaginable, since
in the
expression of .
Case
In this case has the steady state short circuit current the (negative) top value in the short
circuit moment and the DC component has in consequence hereby largest possible value.
Installed in eq. the short circuit current can be derived:

Page of

 RL t

i k = I k e
cos( t )

eq.

Maximum instantaneous value occurs likewise approximately in the AC components first

top value at time

and is:
eq.

Because the asymmetry in a certain (some) degree here is largest it is close to expect that
this maximum value not is much less then the theoretic correct value that can be derived
in the case = . This is also correct for a power network characteristic data and one can
with good approximation use this simple expression when deriving the inrush current.
Observe the ideal case with
damped case is:

when both cases lead to the same result. In this uneq.

Figure Aircore reactor

A short circuit current with DC component is often called unsymmetrical short circuit
current, and speaks about greater or less degree of asymmetry that is dependent on the
instantaneous time of the short circuit. With maximum asymmetry it is meant a current
which DC component has the same value as the AC components un-damped top value.
Observe that the word symmetry and asymmetry here is used in another meaning then
three phase symmetry and three phase asymmetry.
Page of

... Switching in oil immersed iron core reactors

For an iron core reactor where L is not constant, the permeability is depending on the
magnetic field intensity H [A/m] and magnetic flux density B [Vs/m] (i.e.
B
permeability =
) a simple model with a breaker switched in can describe the
H
electrical principles see Figure , Figure and Figure .
The magnetizing characteristic of a shunt reactor
The relation between voltage and current can be described with two lines one for the
unsaturated part and one for the saturated part. The intersection between is called knee
point and is usually located at % of the voltage amplitude. The saturated
region has a slope of % of the unsaturated region. Typical HV shunt reactor
magnetizing characteristic is shown in Figure Error! Reference source not found.].

Shunt Reactor Characteristic

Voltage [pu]

Current [pu]

Figure Typical magnetizing characteristic of a gapped core shunt reactor

Description of parameters
L = inductance [H]
N = number of windings
l FE = length of flux in iron yoke and limb [m]
= permeability in air [Vs/Am]

r = permeability constant iron

= total permeability [Vs/Am]
B = magnetic flux density [Vs/m]
H = magnetic flux intensity [A/m]

Page of

 = total flux [Vs]

AFE = iron area in limb [m]
Switching in a shunt reactor
The inductance L and permeability of the shunt reactor can be derived Error!
Reference source not found.]
A
AFE
eq.
L = N r FE = N
l FE
l FE

B
H

eq.

The differential equation when a shunt reactor is switched in follows

di ~
eq.
= (e Ri )
dt L
By using the flux [Vs] eq. and flux intensity H [A/m] in eq. into the eq. and
eq. the inductance L can be derived see eq.

= B AFE N

eq.

H l FE
eq.
N
d
eq.
L=
di
By introducing the inductance in eq. the magnetic flux over the inductance follows in
eq. :
di
di ~
eq.
=
(e Ri )
dt d
i=

d
= (e~ Ri ) = u L
dt

eq.

To derive the magnetizing voltage and the total magnetic flux the recursive equations
(Euler equations Error! Reference source not found.]) are used in eq. and eq. :
Starting conditions:
, B[] = BR , [] = BR AFE N

u L [k ] = e~[k ] R i L [k ]

eq.

[k + ] = [k ] + t u L [k ]

eq.

Page of

The magnetic field intensity (H) can be derived from the graph in Error! Reference
source not found.] after calculating of the magnetic flux density (B):
[k + ]
eq.
B[k + ] =
H [k + ]
AFE N
Calculate the magnetizing current iL :
H [k + ]
i L [k + ] =
l FE
N

eq.

Page of

Model of Shunt Reactor

E [V]

Im [A]

Um [V]

B - [T=Vs/m]

H [A/m]

Phi [Vs]

Figure Test result of inrush

E = Voltage over shunt reactor model [V]
Im = Magnetizing current trough shunt reactor [A]
Um = Magnetizing voltage over inductive part [V]
B = Magnetic flux density [T=Vs/m]
H = Magnetic field intensity [A/m]
Phi = Magnetic flux over the shunt reactor [Vs]

Page of

During inrush the permeability can move up over the threshold knee and cause a transient
current greater than I k ( to I k ) and after several seconds reaching steady state
current, see Figure and Figure .
As the shunt reactor is moving from saturation region to steady state region, the
permeability increases towards the constant value ( ) and the current decreases to steady
state value. The damping time also increases with less saturation as the permeability of
the shunt reactor moves towards steady state.
Both above statements can be seen from the following equations:
lj
iN =
eq.
r A
L
eq.
=
R

Figure Bus connected, kV, MVAr, oil immersed shunt reactor

Page of

Figure Bus connected, kV, MVAr, oil immersed shunt reactor

.. Shunt reactor disconnection

Disconnection of small reactive current was at one time regarded as a dangerous
operation because of the risk of current chopping and resulting switching overvoltage.
Modern surge arresters are fully capable of handling this condition, and besides, the
tendency of the circuit breaker to chop reactor current is not so pronounced for typical
HV shunt reactor rated current values Error! Reference source not found.].
However, the primary current chopping causes a transient, an exponentially decaying dc
current component in the CT secondary circuit (see Figure for a similar example).
This secondary dc current has no corresponding primary current in the power system; it is
caused by a discharge of the magnetic energy stored in the magnetic core of the current
transformer. However these discharge secondary currents are typically very small for
shunt reactors and pose no effect on the reactor protection schemes with numerical relays
Error! Reference source not found.]

Page of

.MVA, kV, Hz Reactor

[pu]

.
Cycles

IC
IcN

Figure Phase C winding currents during shunt reactor switching in and tripping out

.. Harmonics
The zero, nd and rd harmonics in a shunt reactor are described below.

... Zero Harmonic (DC)

The zero harmonic current (DC component, offset) appears at connection and
disconnection of the shunt reactor, with a time constant up to second for shunt reactors,
due to inherent low losses in a shunt reactor (small resistance compared to inductance).

... nd Harmonic
Inrush current in a shunt reactor doesnt appear as a differential current like that which
appears in a transformer, unless the CT saturates after some time due to long DC time
constant. The level of nd harmonic is small in shunt reactors compared to transformers.

... rd Harmonic
The rd harmonic is the dominant harmonic in shunt reactors during normal operating
condition, due to asymmetries in the reactor windings.
The rd harmonic can be seen in the neutral point of the shunt reactor or as residual using
all phases Error! Reference source not found.].

Page of

.. Hysteresis
There is practically no remanence in a shunt reactor compared to a transformer. The
small air gaps along the reactor winding create a thin hysteresis in the B-H curve for the
reactor and therefore very small remanence.

.. Losses
The fundamental losses in a shunt reactor are winding resistance and magnetization
losses, eddy current losses are also present but small in comparison.
The resistance loss is proportional to the weight of the winding material and to the square
of the current density. The magnetization loss in the core steel also rises by
approximately the square of the flux density.
The total loss is generally .% active power (W) of the total reactive power of the shunt
reactor distributed as follows Error! Reference source not found.]:
Resistance losses in winding, P = RI
-%
Core steel loss
-%
Eddy current losses, winding
and mechanical parts
-%

Page of

APPLICATION OF SHUNT REACTORS

Shunt reactors are used to compensate for large line charging capacitance of long high
voltage power transmission lines and cables.
Their major applications are:
Preventing over voltages that occur when the line is lightly loaded (Ferranti
Effect).
Providing voltage control.
Compensating for line charging reactive power demand of the open-circuit line.
Suppressing the secondary arc current for successful single pole reclosing.

. Connection to the Power System and Grounding Methods

The reactors are normally connected to power system in three locations. They can be
connected to Line, Bus or Tertiary winding of the power transformer or auto-transformer.

.. Line and Bus connected reactors

The line connected reactors are normally connected at both ends of the line as each end
can be energized or de-energized independently.
The shunt reactors can be connected directly to HV lines (see Figure ) or via circuit
switcher or circuit breaker to HV lines or buses depends on the application (see Figure
).
The permanently connected reactors are used to prevent overvoltages appear on long
lines due to lightly loading or open circuit. The switched reactors are used for voltage
control.
These reactors are normally grounded, solidly or via a neutral reactor (see Figure and
Figure ). The neutral reactor is used where single pole auto-reclose is applied, to
suppress the secondary arc current.

Page of

Line

Reactor

Figure Solidly grounded three phase reactor directly connected to line

Figure Three phase and neutral reactor connected to bus or line via circuit switcher or circuit breaker

Page of

.. Tertiary winding connected reactors

These reactors are normally ungrounded and can be switched via circuit switcher or
circuit breaker. These switching devices can be on supply side or neutral side of the
reactor (see Figure and Figure ).
Some utilities have used grounded reactors to reduce TRV (Transient Recovery Voltage)
duty of the switching breaker.

Figure Shunt reactor connected to transformer tertiary winding switching via circuit switcher or circuit
breaker on supply side

Figure Shunt reactor connected to transformer tertiary winding switching via circuit switcher or circuit
breaker on neutral side

Page of

. Effects of Shunt Reactors on Transmission Line Voltage

To better understand the effects of the shunt reactors, we can use the nominal- circuit of
a transmission line and compare the receiving-end voltage of a lightly loaded line with
and without shunt reactors.
Although the nominal- circuit do not represent a transmission line exactly and the
discrepancy between the nominal- and the actual line becomes larger as the length of the
line increases, it can be shown that the nominal- may represent long lines sufficiently
well if a high degree of accuracy is not required Error! Reference source not found.].

Figure Nominal- circuit of a transmission line

To derive Vs from the above circuit (Figure ), we note that the current in the
capacitance at the receiving end is VrY/ and the current in the series arm is Ir + VrY/,
then
Vs = (VrY / + Ir ) Z + Vr

eq.

Vs = ( ZY / + )Vr + ZIr

eq.

Is would be the summation of the current in the shunt capacitance at the sending end
which is VsY/, and the current in the series arm.
Is = VsY / + VrY / + Ir
eq.
Is = VrY ( + ZY / ) + ( ZY / + ) Ir

eq.

The equations eq. and eq. can be expressed in the following form:

Vs = AVr + BIr

eq.

Is = CVr + DIr

eq.

Where
Page of

A = D = ZY / +
B=Z
C = Y ( + ZY / )

eq.
eq.

eq.

Page of

A and D are dimensionless and B and C are in ohms and mhos, respectively.
Now let us look at an example of a line and using the above equations and compare the no load
receiving-end voltage before and after applying the shunt reactors.
Example: A single-circuit kV, mile transmission line has the following series
impedance and shunt admittance per mile:
z = .. /mi
y = . S/mi
Then
Z = z l = ..
Y = y l = . S
We can also derive the no load receiving-end voltage (Vr,nl) by substituting
Ir = in Vs equation eq. .
Vs = AVr , nl
Vr , nl = Vs / A

Now we need to calculate Vs and A.

To calculate Vs, we use the Vs equation eq. and assume the load on the line is MW at
kV with % power factor.
Ir = MW / kV = . A
Vr = . kV
A = ..
B = ..

Then
Vs = .. kV
and
Vr , nl = Vs / A = . kV

Now we calculate the no load receiving-end voltage for the same transmission line when
identical shunt reactors are connected at both ends of the line (see Figure ), compensating for
% of the total shunt admittance of the line.

Page of

Figure Nominal- circuit with shunt reactors added to both end of the transmission line

Vs would be the same, but A would change since adding shunt reactors changes the value of Y:
Z = ..
Y = ( .) . = . S
and
A = ZY / + = ..
Then
Vr , nl = Vs / A = . kV
This example shows that adding shunt reactors can limit the rise of the no load voltage at the
receiving end of the line from . kV to . kV.

Page of

SHUNT REACTOR FAULTS AND ABNORMAL CONDITIONS

The modes of failure differ from air-core to oil-immersed designs and this affects their protection
requirements and schemes.

. Fault types in Dry-type reactors

Three types of faults occur in dry-type reactor installations Error! Reference source not
found.]:
Phase-to-phase faults on the tertiary busbar, resulting in high magnitude phase current.
. Phase-to-ground faults on the tertiary busbar, resulting in a low-magnitude ground current,
dependent upon the size of the grounding transformer ground resistor.
. Turn-to-turn faults within the reactor bank, resulting in a very small change in phase current.
Phase-to-phase faults are a low probability fault for dry-type reactors because the reactors are
single phase units with relatively wide spacing between phases. The main cause of these phaseto-phase faults is when arcing from a failed reactor is not detected soon enough and the fault
ionization moves up into the tertiary busbar resulting in a phase to phase fault.
Since dry-type reactors are mounted on insulators which provide standard clearance and
insulation to ground, direct winding-to-ground faults are low probability as well and are
produced only when this neutral insulation is bridged by, for example, an animal. Damage done
by a winding to ground fault is determined by the grounding transformer/resistor impedance.
Turn-to-turn insulation failures in dry-type reactors begin s tracking from insulation
deterioration. Once the arc is initiated, these failures, if not detected quickly, cascade to the
entire winding because of the arc's interaction with the reactor's magnetic field. If the reactor
bank is ungrounded, the current in the healthy phase will increase to times normal phase
current and could thermally damage the un-faulted phases of the reactor bank.

. Fault types in oil immersed reactors

The oil-immersed reactor faults are broken into three categories:
.

High current phase-to-phase and phase-to-ground faults.

Turn-to-turn faults within the reactor winding.
Miscellaneous failures such as loss of cooling or low oil.

Because of the proximity of the winding with the core and tank winding-to-ground failures can
occur. The magnitude of this fault decreases as the fault is located closer to the neutral side of the
reactor. Turn-to-turn faults start out as a small change in phase currents but increase operating
temperature internal pressure, and accumulation of gas. If these are not quickly detected they will
evolve into a major fault.

Page of

. Failure rates of shunt reactors

The definition of failure rates yields;
Failure rate = no of failures / (total population * total unit years)
Actual failure rate data for reactors is not always kept by utilities. Failure rates of shunt reactor
may vary large from utility to utility in different countries and is affected by design, quality and
workmanship. For example a failure rate between .-.% of shunt reactors, may increase to
several percent during large expansion of the grid. Yearly maintenance of the shunt reactors and
bushings will keep the failure rate down.
Data from Canada and India indicates the distribution of failures can for example be
approximately -% bushing related, -% winding related, -% magnetic circuit, % terminals, and the failure origins may be distributed as % dielectric, % thermal, %
mechanical or others like unknown, chemical, geomagnetic induced currents.

. Turn to turn faults

Phase to Phase and Phase to Ground faults can be caused by turn-turn faults. The location of the
turn-turn fault is most likely in the windings closest to the high voltage part of the shunt reactor,
caused by for example an impulse voltage from electrostatic discharge like lightning storms.
Each winding on the shunt reactor can be seen as an inductance parallel with a leakage
capacitance and capacitance to ground. The inductive part acts stiff on inrush currents, and the
capacitive part causes an exponential distribution of voltage over the winding, with max at the
top due to high frequency.
The capacitive part consists of the insulation material e.g. paper. If the highest voltage difference
between the windings on top of the shunt reactor exceeds the capacitive insulation level, the
insulation material deteriorates and causes a turn-turn fault between two windings. A possible
way to protect for this is to design the shunt reactor with more insulation in the top and equip the
system with a surge arrester, to limit high currents.
Earlier problems with oil containing copper used in shunt reactors and transformers caused turnturn faults, today with improvements in oil quality, this special problem has disappeared.
Another cause of the turn-turn fault is vibrations. Vibrations create insulation material fatigue
which in turn reduces the level of insulation and can cause a turn-turn faults. Samples from oil
and material could tell the condition of the shunt reactor insulation.
Turn-turn faults can also be caused by excessive water in insulation paper, which can give raise
to water vapor bubbles when temperature increases, thus creating a low dielectric strength region
leading to electric arc.
The main risk for short-time failures is the reduction in dielectric strength due to the possible
presence of gas bubbles in a region of high electrical stress, which are the windings and leads.
These bubbles are likely to occur when the hot-spot temperature exceeds C for a reactor
Page of

with winding insulation moisture content above %. This critical temperature will decrease as
the moisture concentration increases.

Figure Equilibrium chart relating water vapor pressure over oil to water concentration in insulation (kraft) paper
vs. temperature.

The risk with excessive water in insulation paper can be mitigated by using an on-line
monitoring system with algorithms to determine water content in paper and bubbling
temperature, so as to issue warnings when the reactor is close to a dangerous condition, before a
turn-turn fault happens.

Page of

. Bushing failure
Overvoltages due to lightning impulses or even due to the reactor switching can bring about very
high dielectric stresses to the reactor bushings. Specifically in case of externally generated
overvoltages, the bushings will be the first ones to suffer the stress. This fact can lead to
bushings insulation deterioration, which ultimately would cause a phase-ground fault with severe
damages to the reactor itself or even to neighbor devices due to porcelain shards being thrown.
This is a severe risk also to people working close to the equipment.

Page of