Shunt Capacitor Bank Design and Protection Basics
Shunt Capacitor Bank Design and Protection Basics
Protection Basics
P: (877) 322-5800
F: (877) 322-4774
info@cedengineering.com
SHUNT CAPACITOR BANK DESIGN AND PROTECTION BASICS
Introduction
Shunt capacitor units are typically used to deliver capacitive reactive compensation or
power factor correction. The use of shunt capacitor units has gained popularity
because they are quite affordable, simple to install and commission and can be placed
anywhere in the electrical distribution system. Its usage has additional advantages on
the electrical distribution system such as: enhancement of the voltage at the load side,
increased voltage regulation, decrease of power losses and decrease or
postponement of investments in electrical transmission network. The primary
weakness of the shunt capacitor units is that their reactive power generation is relative
to the square of the voltage, and accordingly when the voltage is low and the electrical
system needs them most, they are delivering the least amount of the reactive power.
The capacitor unit is the essential element of a shunt capacitor bank. The capacitor
unit consists of individual capacitor segments, connected in parallel/series
arrangements, within a steel case. The internal discharge element is a resistor that
decreases the unit residual voltage to 50V or less in 5 min. Capacitor units come in a
range of voltage ratings (240 V to 24,940V) and ratings (2.5 kvar to about 1,000 kvar).
Figure 1. Typical capacitor bank arrangement
Protection of shunt capacitor units calls for knowledge of the advantages and
restrictions of the capacitor unit and related electrical devices that include: individual
capacitor elements, bank switching equipment, fuses, voltage and current sensing
elements. Capacitors are meant to be run at or below their rated voltage and frequency
since they are highly sensitive to these parameters; the reactive power produced by a
capacitor element is relative to both of them (kVar ≈2𝜋𝜋𝜋𝜋𝑉𝑉 2 ). Standard sizes of the
capacitors elements made for shunt interconnection to AC electrical systems are given
in IEEE Std 18-1992 and Std 1036-1992. These standards also give application
guidelines. These standards specify that:
• Shunt capacitor units are not supposed to be suited for continuous service at
up to 135% of the rated reactive power made by the mixed impacts of:
The function of fuses for protection of the shunt capacitor elements and their location
(inside the capacitor unit on each element or outside the unit) is a significant topic in
the design of shunt capacitor banks. They also impact the failure modality of the
capacitor element and impact the setting of the capacitor bank protection. Depending
on the usage, any of the described arrangements are appropriate for shunt capacitor
elements:
A fault of a capacitor element welds the foils together and causes short circuit currents
to flow between capacitor elements arranged in parallel in the same group. The
remaining capacitor elements in the bank stay in operation with an increased voltage
across them than before the fault. If a second element breaks down the procedure
duplicates itself causing an even greater voltage for the remaining elements.
Sequential faults within the same bank will make the fuse to trip, unplugging the
capacitor element and suggesting the failed one.
Externally fused shunt capacitor units are assembled using one or more series groups
of parallel-connected capacitor elements per phase as shown in Figure 2. The
unbalance signalling level reduces as the number of series groups of capacitors is
raised or as the number of capacitor elements in parallel per series group is increased.
Nevertheless, the reactive power rating of the separate capacitor element may require
being smaller since a minimum number of parallel elements is needed to allow the
shunt capacitor bank to stay in operation with one fuse or unit out.
Phase A
Phase C
Phase B
Each capacitor element has a fuse inside the capacitor element. The fuse is a basic
part of the wire sufficient to limit the current and capsulized in a wrapper that can resist
the heat generated by the arc. Upon a capacitor element fault, the fuse takes out the
struck element only. The remaining elements, linked in parallel in the same
arrangement, stay in operation but with a somewhat increased voltage across them.
A common capacitor bank that uses capacitor elements with an internal fuse is shown
in Figure 3. Generally, shunt capacitors using capacitor elements with internal fuses
are assembled with less capacitor elements in parallel and more series groups of
elements than that utilized in banks employing elements with internal fuses. The
capacitor elements are typically large because the whole unit is not anticipated to
break down.
Phase A
Phase C
Phase B
Fuseless shunt capacitor units - The capacitor elements for capacitor banks without
fuses are the same as those with external fuses. To make a bank, capacitor elements
are arranged in series chains between phase and neutral, as displayed in Figure 4.
The protection is founded on the capacitor elements (inside the unit) breaking down in
a shorted mode, causing short circuit in the group. Once the capacitor element breaks
down, it welds, and the capacitor unit stays in operation. The voltage across the broken
capacitor element is then split among left over capacitor element groups that are
connected in the series. For instance, in the case where 6 capacitor units are
connected in series and each unit consists of 8 element groups in series, there is a
total of 48 element groups connected in series. If one capacitor element breaks down,
the element is bridged, and the voltage on the left-over elements is 48/47 or around a
2% increment in the voltage. The capacitor bank remains in service; nevertheless,
consecutive break downs of elements will cause removal of the bank. The design
without fuses is not typically used for system voltages lower than about 34.5 kV. The
cause is that there shall be more than 10 elements connected in series so that the
capacitor bank does not have to be taken away from operation for the breaking down
of the one element, since the voltage across the left-over elements would increase by
a factor of about E (E – 1), where E is the number of elements in the chain.
The discharge energy is insignificant since no capacitor units are linked directly in
parallel. The additional benefit of units without fuses is that the unbalance protection
does not have to be stayed to achieve coordination with the fuses.
Unfused Shunt Capacitor Units - Opposite to the fuseless arrangement, where the
units are linked in series, the unfused shunt capacitor bank applies a series/parallel
arrangement of the capacitor units. The unfused arrangement would typically be
utilized on units below 34.5 kV, where a series chain of capacitor units are not practical
or on higher voltage units with small parallel energy. This arrangement does not need
as many capacitor units connected in parallel as a bank with external fuses.
As a universal rule, the minimum number of units linked in parallel is such that the
isolation of one capacitor unit in a group should not induce a voltage unbalance
enough to place more than 110% of the rated voltage on the staying capacitors of the
group. Similarly, the minimum number of series linked groups is such that the total
bypass of the group does not subject the ones staying in operation to a lasting
overvoltage of more than 110%.
The maximum number of capacitor units that may be put in parallel per group is
determined by unlike conditions. When a capacitor bank unit breaks down, the
remaining capacitors in the same parallel group hold some amount of charge. The
charge will disappear in a form of a high frequency transient current that goes through
the broken down capacitor unit and its fuse. The fuse holder and the broken down
capacitor unit should resist this discharge transient.
The discharge transient from a big quantity of parallel capacitors can be serious
enough to tear the broken down capacitor unit or the expulsion fuse holder, which may
result in damage to the next units or induce a serious bus break down within the bank.
To derogate the probability of breaking down of the expulsion fuse holder, or damage
of the capacitor case, the standards enforce a limit to the overall maximum energy
stacked away in a parallel linked group to 4,659 kVar. In order not to breach this
boundary, more capacitor groups of a lower voltage rating linked in series with less
units connected in parallel per group may be an adequate answer. Nevertheless, this
may decrease the sensitivity of the unbalance detection system. Dividing the bank into
two parts as a double Y may be the favored answer and may provide a better
unbalance detection system. Another option is the utilization of current limiting fuses.
The optimum arrangement for a shunt capacitor bank depends on the best usage of
the available voltage ratings of capacitor units, fuses, and protective relaying. Nearly
all substation units are linked wye. Distribution capacitor units, nevertheless, may be
linked wye or delta. Some units utilize an H arrangement on every phase with a current
transformer in the associated branch to discover the unbalance.
Grounded wye capacitor units consist of series and parallel-linked capacitor units per
phase and allow for a low impedance path to ground. Common bank arrangements
are shown in Figure 5. Benefits of the grounded capacitor units are:
• Low-impedance path for high frequency currents and so they can be utilized as
filters in arrangements with high harmonic content. Nevertheless, care is
needed to prevent resonance between the shunt capacitor bank and the
electrical system; and
• Diluted transient recovery voltages for circuit breakers and additional switching
devices.
Once a capacitor bank gets too big, thereby building the parallel energy of a series
group too big (above 4,650 kvar) for the capacitor units or fuses, the bank may be
divided into two wye parts. The features of the grounded double wye are similar to a
grounded single wye bank. The two neutrals need to be linked with a common link to
ground. The double Wye arrangement provides a safe and quicker unbalance
protection with an uncomplicated, uncompensated relay, since any zero sequence
component system impacts both wyes evenly, but a broken down capacitor unit will
come out as an unbalanced in the neutral. Time coordination may be needed to grant
a fuse, in or on a broken down capacitor unit, to blow. If it is designed without a fuse,
the time delay may be adjusted short since no fuse coordination is needed. If the
current through the string outperforms the uninterrupted current capability of the shunt
capacitor unit, more chains need to be lent in parallel.
More units grounded single wye More units grounded double wye
Common bank systems of ungrounded wye shunt capacitor bank are presented in
Figure 6. Ungrounded wye units do not allow zero sequence currents, third harmonic
currents, or big capacitor discharge currents during system ground failures to flow.
Another benefit is that over-voltages coming out at the CT secondary sides are not as
high as in the situation of grounded arrangements. Nevertheless, the neutral needs to
be insulated for full line voltage, since it is instantly at phase potential when the shunt
capacitor bank is turned on, or when one capacitor element breaks down in a bank set
up with an individual group of units. For units above 15kV this may be costly.
More units ungrounded single wye More units ungrounded double wye
Shunt capacitor bank units that are connected in delta arrangement are typically
utilized only at distributions voltages and are set with a single series group of
capacitors rated at line-to-line voltage. With only one series group of units, no
overvoltage happens across the staying capacitor units from the isolation of a failed
capacitor unit. Hence, unbalance detecting is not needed for protection.
H arrangement
Few bigger shunt capacitor bank units use an H arrangement in each phase with a
current transformer linked between the two legs to cross compare the current down
each leg. As long as all capacitors are fine, no current will go through the current
transformer. If capacitors fuse trips, current will go through the current transformer.
This bridge arrangement can be very sensitive. This configuration is utilized on big
units with many shunt capacitor bank units connected in parallel.
• Protection of the shunt capacitor bank against failures happening within the
shunt capacitor bank involving those inside the shunt capacitor bank unit
• Protection of the shunt capacitor bank against system disturbances and
failures. The adopted protection scheme for a shunt capacitor bank depends
on bank arrangement, whether or not the shunt capacitor bank unit is grounded,
and the nature of system grounding.
The protection of shunt capacitor units against internal failures includes protective
equipment in a coordinated system. Commonly, the protective devices found in a
shunt capacitor bank for internal failures are: separate fuses, unbalance protection to
give alarm, and overcurrent devices for bank failure protection.
Removal of broken down capacitor element by its fuse ends in increasing the voltage
across the left over elements which causes an unbalance within the bank. A
permanent overvoltage, higher than 10%, on any element shall be stopped by means
of protective relays that disconnect the bank.
Unbalance protection typically delivers the primary protection for arcing failures within
a shunt capacitor bank and other irregularities that may harm a shunt capacitor bank
unit. Arcing failures may induce significant damage in a small fraction of a second. The
unbalance protection should have the lowest possible intentional detainment in order
to derogate the amount of harm to the shunt capacitor bank in the case of external
arcing.
In most shunt capacitor bank units, an external arc inside the capacitor bank does not
end in enough change in the phase current to function the primary failure protection
(typically an overcurrent relay). Sensitivity demands for proper shunt capacitor bank
protection for this condition may be very extensive, especially for shunt capacitor
banks with many series arrangements. The requirement for sensitive demands can
lead to the development of unbalance protection scheme in which particular voltage
or current logistics of the shunt capacitor bank are observed and cross compared to
the bank balance circumstances. Shunt capacitor bank unbalance protection is given
in many different forms, which depends on the capacitor bank configuration and
grounding method. A number of unbalance protection systems are utilized for
internally fused, externally fused, fuseless, or unfused shunt capacitor banks.
For an effective unbalance protection it is crucial to comprehend the break down mode
of the shunt capacitor element. In externally fused, fuseless or unfused capacitor
banks, the broken down element is short-circuited by the weld that normally happens
at the point of fault (the element breaks down short-circuited). This short circuit leaves
out of operation the complete group of elements, thereby raising the voltage on the
remaining shunt capacitor bank groups. Few capacitor element collapses may happen
before the external fuse takes out the complete unit. The external fuse will function
once a shunt capacitor bank unit becomes basically short circuited, thereby isolating
the broken unit.
Shunt capacitor banks that are internally fused have separate fused capacitor
elements that are disjointed when an element failure happens. The danger of
sequential failures is minimized since the fuse will set apart the broken element within
a few cycles. The level of unbalance brought in by an element fault is less than that
which happens with units that are externally fused (since the quantity of capacitance
taken away by a blown fuse is lower), and therefore a more sensitive unbalance
protection system is needed when units with internal fuses are used.
C C C-ΔCx C C
C-ΔCx
CT
For specific shunt capacitor bank arrangements some failures within the shunt
capacitor bank will not induce an unbalance signal and will remain unnoticed. For
instance:
• rack-to-rack failures for shunt capacitor units with two-series groups linked
phase-over-phase and using neutral voltage or current for unbalance
protection; and,
In reality, the unbalance detected by the unbalance relay is the outcome of the tripping
of separate capacitor elements and the inherent system and shunt capacitor bank
unbalances. The primary unbalance that appears on all shunt capacitor bank
arrangements is due to system voltage unbalance and capacitor production margin.
Secondary unbalance mistakes are brought in by detecting equipment margins and
change and by relative variations in capacitance due to deviations in shunt capacitor
unit temperatures in the shunt capacitor bank. The underlying unbalance mistake may
be in the direction to stop unbalance relay functioning, or to induce a delusive
functioning.
If the inherent unbalance mistake reaches 50% of the alarm value, compensation
should be allowed for in order to properly alarm for the fault of one element as defined.
In some situations, a different shunt capacitor bank arrangement can enhance the
sensitivity without providing compensation. For instance, a wye connected shunt
capacitor bank can be divided into a wye-wye shunt capacitor bank, thereby
duplicating the sensitivity of the protection and extinguishing the system voltage
unbalance result.
The time detainment of the unbalance trip relay needs to be minimized to cut down
harm from an arcing failure within the bank arrangement and forestall exposure of the
remaining shunt capacitor elements to overvoltage circumstances outside their
allowed boundaries.
The unbalance trip relay needs to have sufficient time detainment to avert false
functioning due to inrush, system ground failures, switching of nearby device, and non-
simultaneous pole function of the energizing switch. For the majority of usages, 0.1s
should be enough. For unbalance relaying schemes that would function on a system
voltage unbalance, a detainment somewhat longer than the upstream protection
failure clearing time is needed to avert tripping due to a system failure. Longer
detainments raise the possibility of catastrophic shunt capacitor bank faults.
With grounded capacitor elements, the fault of one pole of the shunt capacitor bank
switching equipment or a single phasing from a blown shunt capacitor bank fuse will
permit zero sequence currents to go in system ground relays. Shunt capacitor bank
relaying, including the operating time of the switching equipment, needs to be
interconnected with the functioning of the system ground relays to avert tripping
system load. The unbalance trip relay arrangement needs to have a lockout option to
advert accidental closedown of the shunt capacitor bank switching element if an
unbalance trip has happened.
To permit the impacts of inherent unbalance within the shunt capacitor bank, the
unbalance relay alarm needs to be set to function at about one-half the level of the
unbalance signal defined by the computed alarm terms that are based on an idealized
shunt capacitor bank. The alarm needs to have adequate time detainment to overrule
external disturbances.
59
59
VT
A system that evaluates the voltage between the capacitor neutral and ground using
a VT and an overvoltage relay with a third harmonic filter is displayed in Figure 8(a). It
is fairly straightforward but endures in the case of system voltage unbalances and
inherent unbalances. The voltage-sensing equipment is typically a voltage transformer
but it could also be a capacitive potential element or resistive potential element. The
voltage-sensing element needs to be chosen for the lowest voltage proportion
attainable, while still being in a position to resist transient and uninterrupted
overvoltage circumstances to get the maximum unbalance detection sensitivity.
Nevertheless, a voltage transformer applied for this purpose needs to be sized for full
system voltage since the neutral voltage can, under certain circumstances, increase
to as high as 2.5 per unit during switching.
An equivalent zero sequence component that eradicates the system unbalances can
be gained using three voltage-sensing elements with their high side voltage wye-
connected from line to ground, and the secondary side linked in a broken delta
arrangement. The voltage source VTs can be at a tap in the shunt capacitor bank or
utilize the VTs of the shunt capacitor bank bus.
Advanced digital protection relays can compute the zero sequence voltage from the
phase voltages as displayed in Figure 9(a), ridding of the requirement of extra auxiliary
VTs to get the zero sequence voltage. Fig 9(b) presents the same method but utilizing
the VTs on the shunt capacitor bank bus.
59 N 59 N
Calculated Calculated
Even though arrangements presented in Figure 8(b), 9(a) and 9(b) get rid of system
unbalances, they do not get rid of the inherent capacitor unbalance. Protection
arrangement that eliminates the system unbalance and even up for the inherent
capacitor unbalance is presented in Figure 10.
Phase C
Phase A
Phase B
Other
phases
X1 X3
Neutral
Unbalance
VT Relay
VT
Ungrounded shunt capacitor bank elements can be divided into two equal elements.
This bank arrangement inherently evens up for system voltage unbalances;
nevertheless, the impacts of the producers’ capacitor margin will impact relay function
unless steps are taken to even up for this error.
The neutral current is one-half of that of a single grounded shunt capacitor bank of the
same rating. Nevertheless, the current transformer ratio and relay size may be picked
out for the hoped sensitivity since they are not subjected to switching surge currents
or single-phase currents as they are in the grounded neutral arrangement.
Arrangement where the neutrals of the two capacitor parts are ungrounded but linked
together is shown in Figure 12. A voltage transformer, or potential device, is utilized to
evaluate the voltage between the capacitor bank neutral and ground. The relay needs
to have a harmonic filter.
VT 59
Figure 12
An unbalance in the shunt capacitor bank will induce current to flow in the neutral.
Protection arrangement founded on a current transformer used on the link between
the shunt capacitor bank neutral and ground is shown in Figure 13(a). This current
transformer has unusual high overvoltage and current demands. The ratio is chosen
to provide both decent overcurrent capability and appropriate signal for the protection.
The current transformer output has a burden resistor and a sensitive voltage protection
relay. Due to the existence of harmonic currents (especially the third, a zero sequence
harmonic that goes through the neutral-to-ground link), the protection relay needs to
be tuned to cut down its sensitivity to frequencies other than the power frequency.
The voltage across the burden resistor is in phase with the neutral-to-ground current.
This neutral-to-ground current is the vector amount of the three-phase currents, which
are 90° out of the phase with the system phase-to-ground voltages. This arrangement
may be counterbalanced for power system voltage unbalances, by taking into account
the 90° phase shift, and is not uncommonly suitable for very large capacitor elements
rquiring very sensitive adjustments.
Each time the shunt capacitor bank is switched on, momentary unbalanced capacitor
charging currents will diffuse in the phases and in the capacitor neutral. In the case
when the parallel shunt capacitor bank is already in function, these currents can be on
the order of thousands Amps, causing the relay to malfunction and the CT to break.
VT
59 N
Calculated
CT R 59
An unbalance voltage protection arrangement for single grounded wye linked shunt
capacitor banks utilizing capacitor tap point voltages is shown in Figure 13(b). An
unbalance in the shunt capacitor bank will induce an unbalance in the voltages at the
tap point of the three phases. The protection arrangement comprises of a voltage
sensing element linked between the capacitor’s intermediate point and ground on each
phase. A time detainment voltage relay with a third harmonic filter is linked to the
broken delta secondary sides. Modern digital protection relays utilize the computed
zero sequence voltage instead as presented in Figure 13(b).
51 CT
Figure 14.
The arrangement utilizes two voltage transformers per phase: one linked to a tap on
the shunt capacitor bank; the other, at the bank bus for single wye elements; or, for
double wye elements, at a similar tap on the second bank. By cross comparing the
voltages of both VTs, a signal responsive to the loss of single capacitor elements or
units is gained.
The shunt capacitor bank tap voltage is found by linking a voltage-sensing element
across the ground end parallel group (or groups) of capacitors. This may be a midpoint
tap, where the voltage is evaluated between the midpoint of the phase and ground.
Alternatively, the tap voltage may be evaluated across low-voltage capacitors (that is,
a capacitive shunt) at the neutral end of the phase.
Phase B
Phase A
Phase C
59
X3 X3
VT VT
Figure 15. Voltage differential scheme for grounded single wye shunt capacitor bank
For commissioning, after testing that all shunt capacitors are operational and no fuses
have functioned, the voltage levels are initially adapted to be the same. The initial
difference signal between the shunt capacitor bank tap voltage and the bus voltage
(for single wye elements) signals is zero. Also, the shunt capacitor margin and initial
system voltage unbalance is counterbalanced. If the system voltage unbalance should
change, the protection relay system is still counterbalanced because a given percent
change in bus voltage leads to the same percent variation on the capacitor bank tap.
Any later voltage deviation between the capacitor tap voltage and the bus voltage will
be due to unbalances induced by the loss of capacitor elements within that specific
phase. For double wye elements, the tap voltage is cross-compared to the other wye
tap voltage.
Phase C
Phase B
Phase C
59
X3 X3
VT VT
Figure 16. Voltage differential arrangement for a grounded double wye shunt
capacitor bank
There are specific failures within the shunt capacitor bank that the unbalance
protection will not sense, or other ways are needed for its clearance.
Generally, individual phases of a shunt capacitor bank are made on different structures
where phase to phase failures are improbable. Nevertheless, conceive an ungrounded
single wye shunt capacitor bank with two series groups per phase where all three
phases are placed on a single steel structure. A mid-rack failure between 2 phases as
presented in Figure 17 can happen and will go undiscovered. This failure does not
induce an unbalance of the neutral voltage (or neutral current if grounded) as the
healthy voltage is counterbalanced by the 2 other faulty phase voltages.
The most effective protection for mid-rack phase to phase failures is the negative
sequence current. Tripping shall be detained to coordinate with other protection relays
in the system.
Time overcurrent protection relays for phase and ground are needed to supply
protection for phase and ground failures on the connecting feeder (or bus) between
the bank bus and the first capacitor element.
Directional overcurrent protection relays looking into the shunt capacitor bank are
favored to avert malfunction of the TOC 51N for unbalance system failures.
The shunt capacitor bank may be subjected to over voltages leading from abnormal
system operating conditions. If the system voltage surpasses the shunt capacitor bank
capability the bank needs to be taken out from function. The removal of the shunt
capacitor bank brings down the voltage in the locality of the shunt capacitor bank
abbreviating the overvoltage on other system elements. Time delayed or inverse time
delayed phase overvoltage protection relays are utilized.
Once disconnected from the system a shunt capacitor bank cannot be put back into
service instantly due to the electrical charge present within the shunt capacitor
elements, otherwise catastrophic harm to the circuit breaker or switch can happen. To
speed up the discharge of the shunt capacitor bank, every shunt capacitor unit has a
resistor to discharge the trapped charges within 5 minutes. Under voltage or
undercurrent protection relays with timers are utilized to sense the shunt capacitor
bank going out of function and forestall closing the breaker until the set time has
passed.
Conclusions
The protection of shunt capacitor bank units uses straightforward, well known relay
protection methods such as overvoltage, over-currents. Nevertheless, this needs the
protection engineer with a solid apprehension of the shunt capacitor bank units, its
scheme, and shunt capacitor bank design problems before embarking in its protection.
Unbalance is the most crucial protection in a shunt capacitor bank, as it gives fast and
efficient protection to ensure a long and reliable life for the bank. To achieve its goal,
unbalance protection needs high degree of sensitivity that might be hard to achieve.
The main objectives for the design of a shunt capacitor bank and its associated
protection schemes have been presented in this course. The latest “IEEE Guide for
the Protection of Shunt Capacitors Units” needs to be the leading standard when
carrying out a protection arrangement for a shunt capacitor bank.