Capacitive Switching

Capacitive Current Switching for HV Circuit Breakers

How high-voltage circuit breakers are applied to capacitive current switching — the restrike and inrush stresses that arise when capacitor banks, filter banks, unloaded cables and unloaded lines are switched, the ratings and voltage factors that follow, and the measures used to limit the transients. Follows IEEE Std C37.012-2022, with APS engineering interpretation.

Reading time ≈ 24 min

Section 1

Overview and Scope

Capacitive current switching is one of the recognised switching duties for alternating-current high-voltage circuit breakers rated above 1000 V. The steady-state current to be interrupted is small — normally no more than a few hundred amperes — so if only current magnitude is considered the interruption looks like a light duty. The real difficulty lies elsewhere. After opening, the trapped charge on the capacitive load creates a recovery voltage across the breaker contacts; if the contact gap cannot withstand it, a restrike may occur. During closing, the same capacitance draws a high-magnitude, high-frequency inrush current. These effects can stress the breaker, the capacitive equipment and surrounding plant, and a restrike can inject damaging overvoltages or high-frequency transients into the network.

This guide follows the application methodology of IEEE Std C37.012-2022 and sets out the factors that govern breaker selection, the ratings and voltage factors involved, and the measures available to limit the resulting transients. This page provides the common application framework; the following pages then treat capacitor banks, unloaded cables, unloaded lines, voltage factors, fault conditions, filter banks and mitigation devices in more detail. It is written as a standalone reference; for the detailed treatment of one specific application see the APS note on switching of shunt capacitor banks, and filter banks are covered in Section 12 below.

Per-unit convention

1 per-unit (pu) voltage is taken as the crest of the phase-to-earth operating voltage, consistent with IEEE Std C37.012-2022. A single restrike can produce a load overvoltage of up to about 3.0 pu. A reignition — meaning re-establishment within approximately a quarter-cycle — normally produces little or no overvoltage, because the voltage has not had time to build up significantly.

Where capacitive currents arise

Four applications generate the capacitive currents a breaker must switch. They share the same interruption physics but differ in current magnitude, in the influence of nearby plant, and in the shape of the recovery voltage.

Table 1 — Applications that produce capacitive switching duty.

Application	Source of the Capacitive Current	Application Notes
Shunt capacitor banks	Discrete reactive-compensation banks connected bus-to-earth or bus-to-neutral.	Current set by bank size; short connections make the Ferranti effect negligible. Frequently switched, so class C₂ is normally required.
Unloaded cables	Distributed shunt capacitance of an open-ended cable.	High capacitance per unit length; long cables can approach the breaker’s rated capacitive current.
Unloaded transmission lines	Distributed shunt capacitance of an open-ended overhead line.	Charging current rises with length and voltage; the Ferranti effect increases it further (Section 4).
Filter banks	Tuned or damped harmonic-filter branches.	Recovery voltage need not follow a 1 − cosine shape and may contain harmonics, which changes the breaker requirement (Section 6).

Transients on energisation

Energising a capacitive load is almost always accompanied by transient voltages and currents. The principal ones are the inrush current as the capacitance charges, the overvoltage caused by the system responding to the momentary voltage collapse when a bank is connected, and the travelling-wave overvoltages that arise on lines and cables. Because the actual magnitude of these transients is strongly system-dependent, type tests can reproduce the duty only up to the point of clearing, reigniting or restriking; they cannot replicate the full system response. This is why classification is expressed as a probability of restrike rather than as an absolute guarantee, and why coordinated protection — typically surge arresters — is common practice.

Section 2

Selecting the Circuit Breaker

The capacitive-current switching capability of a breaker depends on its rated voltage and frequency, on the particular application, and on the power-system earthing arrangement. Selecting a rating for a given duty means working through a defined set of inputs, because each one influences either the recovery voltage across the breaker or the probability and consequence of a restrike. In simple terms, the application identifies the type of duty, the earthing and fault conditions define the recovery-voltage severity, and the calculated capacitive current confirms whether the breaker rating is adequate.

Table 2 — Inputs that determine the capacitive-switching rating.

Input	Why It Matters
Application	Capacitor bank, cable, transmission line or filter bank — fixes the current magnitude and the recovery-voltage shape, and sets the recommended restrike class.
Power frequency	50 Hz or 60 Hz; 60 Hz is the more severe stress after interruption (Section 3).
System earthing arrangement	Effectively earthed or non-effectively earthed — a primary driver of the recovery voltage and of the voltage factor \(k_{c}\).
Bank neutral earthing	Solidly earthed, unearthed or impedance-earthed — determines whether phases interact and thus the peak recovery voltage.
Presence of earth faults	Single- or two-phase-to-earth faults present during switching raise the recovery voltage and the required voltage factor.

From the application the required restrike class (C₀, C₁ or C₂) follows. From the earthing arrangement, bank neutral treatment and fault presence, the recovery voltage across the breaker is fixed, and from that the appropriate single-phase test voltage factor \(k_{c}\) is chosen. These two selections — class and \(k_{c}\) — are the core of a capacitive-switching specification.

Section 3

Rated Voltage and Frequency

The operating voltage must not exceed the rated maximum voltage, which is the upper limit for operation. Caution is also warranted when applying older breakers that were never tested to the modern capacitive-switching procedures — their in-service restrike behaviour is simply unknown.

Note — 60 Hz is more onerous than 50 Hz

After a capacitive interruption the recovery voltage rises at power frequency and reaches its first peak about half a cycle later. At 60 Hz that peak arrives at roughly 8.3 ms, against 10 ms at 50 Hz, so the contact gap has less time to build up dielectric strength. Consequently a capability demonstrated at 60 Hz covers the 50 Hz requirement for the same voltage factor \(k_{c}\); the reverse is not automatically true, so tests performed at one frequency to cover the other must be examined carefully.

Section 4

Rated Capacitive Current

Preferred values of rated capacitive switching current are tabulated in IEEE Std C37.04. The preferred line and cable values cover most switching cases; for capacitor banks a preferred value plus three alternate values are given, intended to span the majority of in-service ratings without covering every case. The charging current of a line or cable follows directly from its shunt capacitance:

\[ I_{c} = 2\pi f\, C_{0}\, \frac{U}{\sqrt{3}} \approx i_{\ell}\,\ell \]

\(I_{c}\): per-phase charging (capacitive) current
\(f\): power-system frequency
\(C_{0}\): positive-sequence shunt capacitance per phase
\(U\): line-to-line operating voltage
\(i_{\ell}\): charging current per unit length (A/km)
\(\ell\): line or cable length

Transmission lines and cables

For very long lines and cables the unloaded current can exceed the tabulated preferred values, and the manufacturer should be consulted; often currents up to the continuous current rating are acceptable. The Ferranti effect — the voltage rise along an open-ended line — raises the charging current above the simple product above, and that increase is not covered by the preferred values.

Worked example — a 550 kV line

A 550 kV overhead line draws roughly 1.1 A/km at 50 Hz and 1.3 A/km at 60 Hz. Ignoring the Ferranti effect, a 500 km line therefore presents about 550 A at 50 Hz and 650 A at 60 Hz. The Ferranti effect would push both figures higher. A larger current is not itself a problem for a modern breaker, but the higher peak recovery voltage that accompanies interruption may be.

Capacitor and filter banks

Bank current is set by the size of the bank, and the connections to the capacitors are usually short, so the Ferranti effect is negligible. Where a bank is rated above the preferred C37.04 value, experience shows that capacitive current is not the limiting parameter up to the breaker’s continuous current rating. For altitudes above 1000 m the capacitive current need not be corrected, provided it does not exceed the corrected rated continuous current.

General-purpose versus definite-purpose breakers

The current used to rate a device for capacitor duty is not the nominal bank current. Allowances must be added for the operating overvoltage (typically about 10%), the capacitor manufacturing tolerance (about 105–115% of nominal) and any harmonic loading, so the effective switching current sits well above the plain bank rating.

Against that derated duty, ANSI/IEEE rating tables draw a sharp line between two categories:

A general-purpose breaker carries only a modest capacitive current — a 1200 A, 15 kV general-purpose breaker is rated about 250 A capacitive, and a 362 kV general-purpose breaker about 250 A for line or isolated-bank switching.
A definite-purpose capacitor-switching breaker carries much more — of the order of 630 A at 15 kV, for isolated or back-to-back duty — but is subject to explicit inrush limits: traditionally the inrush magnitude and frequency had to stay within roughly 15 kA and 2000 Hz at medium voltage, and 25 kA and 4250 Hz at 362 kV, for back-to-back switching.

So even a definite-purpose breaker needs its back-to-back inrush reduced — by a current-limiting reactor or controlled closing — where those limits would be exceeded.

Modern gas interrupters are far less sensitive to inrush frequency, and judge the duty by the inrush-current integral instead; the duty still has to be checked (see the APS note on energising capacitor banks). SF₆ breakers have a much better capacitive-switching capability than the older oil and air-blast designs and are now used almost exclusively at high voltage.

Section 5

The Voltage Step on Energising

Connecting a capacitive load to a predominantly inductive supply raises the power-frequency voltage — the voltage step. The rms voltage step is the quantity planners use when sizing banks, and it is bounded by network planning limits that tighten as the system becomes stiffer. The voltage step is a steady-state or power-frequency rms change; it should not be confused with the high-frequency transient overvoltage that may occur during switching.

Table 3 — Typical steady-state voltage-step planning limits.

System Level	Typical Maximum Voltage Step
Distribution	about 5%
Sub-transmission	about 2.5%
Bulk power	about 1% to 1.5%

For a single bank the voltage step is inversely proportional to the square of the harmonic order of the natural inrush frequency:

\[ \frac{\Delta U}{U} \approx \frac{1}{h^{2}} \]

\(\Delta U/U\): per-unit rms voltage step
\(h\): harmonic order of the natural inrush frequency, \(h = f_{i}/f\)

A 1% step therefore corresponds to a natural inrush frequency near the 10th harmonic (600 Hz on a 60 Hz system), and a 4% step to the 5th harmonic. These are steady-state planning rules; the actual magnitude depends on the system and the capacitive load, but the relation is a useful sanity check when the inrush frequency is known.

Section 6

Recovery Voltage, Earthing and the Voltage Factor

After interruption the load-side charge is trapped while the source-side voltage keeps following the system, so the recovery voltage across the open contacts swings to at least twice the crest of the phase-to-earth voltage. How much more than twice depends on the system earthing and the bank neutral treatment, which is captured by the single-phase test voltage factor \(k_{c}\). The multiplication factors for single-phase tests are given in IEEE Std C37.09; they range from unity for effectively earthed systems to 1.7 for unearthed systems in the presence of single- or two-phase-to-earth faults. The voltage factor \(k_{c}\) is therefore not a generic safety margin; it is selected from the system earthing, the bank neutral arrangement and the fault condition.

Table 4 — Single-phase test voltage factor \(k_{c}\) by condition.

Condition	Voltage Factor \(k_{c}\)
Effectively earthed system	1.0
Non-effectively earthed system (no fault)	Between 1.2 and 1.4, depending on neutral and earthing conditions
Unearthed system with a single- or two-phase-to-earth fault	up to 1.7

Note — filter banks

The recovery voltage across a harmonic-filter-bank breaker need not follow the smooth 1 − cosine waveshape of a plain capacitor bank; it can contain harmonic components. This distorted recovery voltage may cause occasional reignitions. Those reignitions can be acceptable as an economical solution, but if they are not, a higher-performance breaker or controlled switching should be applied.

Section 7

Restrike Probability and Breaker Classes

Because every breaker has some restrike probability in service, a genuinely restrike-free breaker cannot be defined. The workable concept is a restrike classification tied to a specific test procedure. Three classes are defined in IEEE Std C37.04, with the design-test criteria set out in IEEE Std C37.09 and C37.100.2.

Table 5 — Restrike-probability classes and where each is applied.

Class	Restrike Probability	Typical Application
C₀	Undetermined	Acceptable where restrikes are not a concern, often distribution-line capacitors below 38 kV.
C₁	Low	Often acceptable at 100 kV and below, and for infrequently switched lines and cables.
C₂	Very low (an order of magnitude below C₁)	Recommended for capacitor-bank breakers and for lines and cables switched more than about 30 times a year.

The classes differ mainly in their test programmes. Class C₂ tests are performed on preconditioned contacts — preconditioning being three interruptions at 60% of rated short-circuit current — and the C₂ programme uses a greater total number of tests, with many more operations near the minimum arcing time, than C₁. In other words, C₂ is demonstrated by more severe and more numerous tests than C₁.

The service level of restrike probability still depends on operating conditions such as the number of operations per year, network condition and maintenance policy, so no single common probability level can be attached to a class.

The intended service probabilities are roughly 1 restrike in 50 operations for class C₁ and 1 in 500 for class C₂, but the demonstration is statistical: only 168 single-phase test operations are required for C₂, while a C₂ breaker may see a switching life of 2000 operations or more. It cannot be guaranteed that fewer than about four restrikes occur across that life, so a capacitive-switching installation must still be designed to withstand a few restrikes during its life.

Type tests also hold the recovery voltage for 0.3 s to expose any delayed restrikes, which in practice do not occur beyond about 0.2 s.

Those raw test counts overstate the field restrike rate, and by a known margin. A type test interrupts at the minimum arcing time — the worst case for a restrike — whereas in service the arcing time is randomly distributed. The ratio between the two is an acceleration factor set by the scatter of the breakdown voltage: about 6 for the narrow scatter of SF₆ and about 4 for the wider scatter of vacuum. Because a class C₂ programme runs on the order of 76–104 operations at that worst-case arcing time, the demonstrated “fewer than one restrike” performance extends to roughly 330–480 random field operations (CIGRE TB 817). A separate, smaller acceleration (about 1.5) applies to the making duty, because tests close within a narrow window around the voltage peak while field closing is random. The class is therefore a deliberately conservative, worst-case statement of field behaviour.

Key idea

A class is a laboratory statement of probability. The energies in a test circuit differ from those in a real network, so even the class C₀ requirement that multiple restrikes not be observed cannot be assured in service, and the bank’s overvoltage protection during a type test may differ from that in the field. Select the class from the service conditions, the number of switching operations expected per year, and the consequences of a restrike for both the network and the breaker — not from the test result alone.

Section 8

Interrupting Time

For capacitive switching, the interrupting time is the interval between energising the trip circuit at rated control voltage and interruption of the main circuit in all poles on an opening operation. For some designs the time to interrupt capacitive current can be longer than the rated interrupting time — oil circuit breakers are a common example — and for breakers fitted with opening resistors the interrupting time of the resistor current may be longer still. These differences matter when the switching is coordinated with protection or with controlled-switching timing.

Section 9

Transient Overvoltages and Their Effects

The main application concern is the transient overvoltage that a restrike can generate during opening. The transient overvoltage factor is defined as the ratio of the transient voltage that appears between a disconnected breaker terminal and the neutral of the disconnected capacitance, to the operating line-to-neutral crest voltage just before opening.

This factor matters because the transient it describes has to be withstood by the surrounding insulation, clamped by any surge arresters, kept out of the control wiring, and managed through the earthing system. The class of breaker selected should therefore be coordinated with the insulation capability of the other components on the system.

Transients are produced both by restrikes when de-energising and by the energisation of capacitive loads. Their effects include the following:

insulation degradation and possible failure of substation equipment;
operation of surge arresters;
interference in substation control wiring;
increased step potentials in substations (transient ground rise);
undesired tripping of, or damage to, sensitive electronic equipment.

During back-to-back switching the magnetic fields of the high inrush currents — in the conductors or in the earthing grid — can induce voltages in control cables by capacitive and electromagnetic coupling. These can be minimised by shielding the cables and using a radial cable configuration so that inductive loops are not formed, and metal-oxide varistors can be fitted on control circuits to limit the induced voltages.

Switching of capacitor banks

Because most class C₁ and C₂ breakers have a low probability of restrike, the majority of switching transients on a capacitor bank are generated on energisation. Their effects appear both locally and at remote points on the system. The high-frequency inrush current of back-to-back switching also stresses other plant: wound-type current transformers, for example, see turn-to-turn insulation stressed by the high rate of rise of current and the voltage developed across circuit inductance.

Table 6 — Local and remote effects of capacitor-bank switching transients.

Local Effects	Remote Effects
Voltage transients producing dielectric stress on nearby equipment.	Transfer of capacitively coupled fast transients through transformer windings.
Electrical, mechanical and electromechanical forces from the inrush current.	Reflection of travelling-wave transients on open-ended or transformer-terminated lines.
Transient earth-potential rise of the earthing system from the inrush current.	Excitation of near-resonant portions of the network by the oscillatory transient.

Phase-to-phase overvoltages at a remote transformer

A less obvious effect is a phase-to-phase overvoltage appearing at a remote location. Where a bank is switched on a bus that reaches a transformer through a transmission line, the energising surge travels along the line and roughly doubles at the high-impedance transformer terminal. Because the surges launched at energisation are of opposite polarity on two of the three phases, the remote transformer sees the two doubled surges together as a phase-to-phase stress. Surge arresters — normally connected phase-to-earth — clamp the phase-to-earth component to their protective level, but the phase-to-phase overvoltage can be about twice that level, so phase-to-earth arresters alone do not bound it. Restrikes in the device controlling an ungrounded-wye bank can likewise generate high phase-to-phase voltages.

Switching of lines and cables

When energising a line or cable, higher overvoltages arise if the circuit is pre-charged from a preceding opening — the classic case being auto-reclosing onto trapped charge. The reclose can catch the line at an unfavourable polarity, and the resulting overvoltage may damage insulation. This is one reason line breakers used for frequent switching are held to class C₂ and are often combined with controlled closing or pre-insertion resistors. For the detailed treatment of cables — the charging current, the screened/belted recovery voltage, and single versus back-to-back inrush — see the APS note on unloaded cable switching; for lines — the Ferranti effect, the C₁/C₀ recovery voltage, shunt compensation and delayed current zeros — see switching unloaded transmission lines.

Section 10

Limiting the Overvoltages

Several established methods are used to reduce either the overvoltages, the inrush and outrush current transients, or both. They differ in what they limit — inrush current, remote overvoltage, or both — and in the complexity and availability penalty they add, so the choice follows from which stress dominates the application. For the device-level detail — controlled switching, pre-insertion resistors, current-limiting reactors, solid-state and hybrid diode switches, with their timing, sizing and cost — see the APS note on devices for limiting capacitor-switching transients.

Table 7 — Overvoltage- and inrush-limitation measures.

Measure	What It Limits	Notes
Current-limiting reactors / transient-limiting inductors (TLIs)	Inrush-current transients	Reduce the back-to-back inrush current and its \(di/dt\); they do not limit remote overvoltages.
Pre-insertion resistors	Inrush current and remote overvoltages	A basic, widely used solution on transmission breakers; fitted to the breaker, adding complexity that may or may not reduce availability.
Pre-insertion reactors	Inrush current and remote overvoltages	Usually fitted to circuit switchers; complexity and availability effect similar to pre-insertion resistors, depending on design.
Point-on-wave (controlled) closing	Both, local and remote	Closes each contact near a voltage zero across it, reducing inrush and overvoltage together; the controller adds complexity and can affect availability.
Surge arresters	Overvoltage magnitude	Clamp transient overvoltages to the arrester’s switching-surge protective level at its terminals.

Section 11

Inrush and Outrush Currents

Inrush is the transient charging current that flows during energisation; it is a making duty that stresses the switching device as it closes. When a bank is energised onto a bus that already carries other banks or long cables, the charging current from those elements flows through the making device as well — the back-to-back case — and is limited only by the small inductance between banks. For a single bank the peak and natural frequency of the inrush follow from the source strength and the bank rating:

\[ i_{i,\text{peak}} = \sqrt{2}\,\sqrt{I_{sc}\,i_{1}} \qquad\qquad f_{i} = f\,\sqrt{\dfrac{I_{sc}}{i_{1}}} \]

\(i_{i,\text{peak}}\): peak inrush current
\(f_{i}\): inrush (natural) frequency
\(I_{sc}\): source symmetrical short-circuit current (A rms)
\(i_{1}\): bank power-frequency current (A rms)
\(f\): power-system frequency (Hz)

Outrush is the transient discharge current from an already charged capacitive element into a fault or another low-impedance path. When the breaker between the bank and the fault is fully closed, it is usually not in jeopardy; the concern arises when a breaker closes into a nearby fault with significant charged capacitance behind it (a reclose, or closing with safety earths left on by mistake). This can happen on breakers never intended for capacitive making, such as general-purpose oil breakers, which are sensitive not only to the peak current but also to its rate of rise, \(di/dt\).

Damaging outrush events are rare, and to be serious — particularly below 100 kV — the fault must be quite close to the station; the higher the system voltage, the further out a fault can be and still matter. A risk analysis is therefore the appropriate way to weigh the rare likelihood of a damaging outrush against the cost of mitigation, and it should also account for any requirement imposed by the utility regulator regardless of the calculated probability.

Note — the cost of transient-limiting inductors

TLIs limit outrush \(di/dt\) and magnitude and are often a good solution, but they are not free of side effects. They raise the transient recovery voltage seen by the breaker, which can be offset with a capacitor-to-earth or a capacitor-and-arrester across the TLI, at further cost.

Table 8 — Unintended consequences of adding transient-limiting inductors.

Consequence	Effect
Raised capacitor voltage	Increased voltage on the capacitors and possibly shortened capacitor life.
Shunt harmonic-filter branch	The TLI-plus-capacitance forms a shunt filter branch, which may cause harmonic overload of the capacitive load, especially banks and long cables.
Fast transient recovery voltage	With a fault where the TLI is most of the fault impedance, a fast TRV can result that gas breakers may not withstand.

Section 12

Filter Bank Switching

This section gives only the application-level distinction between a filter bank and a plain shunt capacitor bank; the detailed switching behaviour is treated in the dedicated filter-bank page. A harmonic-filter bank is the fourth capacitive application, and it switches differently from a plain shunt bank. A filter bank is not purely capacitive: it is an LC circuit — capacitors in series with reactors, and sometimes resistors — tuned to control harmonics and improve power quality. The two common forms are single-tuned filters, which present a low impedance at one harmonic (say the 5th or 7th) to absorb that current, and de-tuned filters, deliberately offset from any major harmonic to give power-factor correction while avoiding resonance between the bank and the network. For the detailed switching duty — why the large tuning reactor makes energising milder but de-energising more onerous, with the inrush, recovery-voltage and restrike equations — see the APS note on switching of harmonic filter banks.

How the switching duty differs

Because the filter contains inductance as well as capacitance, its resonant behaviour changes every part of the switching duty:

Inrush is usually lower than a plain bank, because the series reactor limits it — though the waveform can still carry oscillatory components. The reactor is sized for the harmonic tuning, so this inrush reduction is a beneficial side effect, not its design purpose.
Transient recovery voltage is more complex: the reactor and capacitor interact, so the recovery voltage can combine a power-frequency component with high-frequency and resonant oscillations, making de-energisation more onerous than for a simple bank.
Pre-strike on closing at an unfavourable point can excite the LC resonance and drive an overvoltage across the filter components.
Restrike on de-energising can be more severe than in a plain bank, because the LC circuit amplifies the oscillation and stresses the capacitor units, reactor insulation, breaker contacts, arresters and connected plant.
Outrush into a nearby fault depends on the filter size, the distance to the fault, the circuit inductance and the reactor values, and must be checked for breaker rating, protection and thermal duty.

Why the recovery voltage is not a clean 1 − cosine

On a plain bank the line-side recovery voltage is a trapped dc charge, giving the smooth 1 − cosine of Section 6. On a filter bank the reactor and capacitance ring together, so the recovery voltage carries harmonic and high-frequency content. That distorted waveshape can provoke occasional reignitions — acceptable as an economical solution in some cases, but otherwise a reason to specify a higher-performance breaker or controlled switching.

Table 9 — Shunt capacitor bank versus filter bank — switching and design characteristics.

Item	Shunt Capacitor Bank	Filter Bank
Main purpose	Reactive power and voltage support	Harmonic filtering, sometimes reactive support
Main components	Capacitors	Capacitors and reactors
Electrical behaviour	Mainly capacitive	Resonant LC behaviour
Inrush current	Often high	Usually lower to moderate
TRV shape	Relatively simpler	More complex, often high-frequency oscillations
Pre-strike concern	Important	Important; may excite resonance
Restrike consequence	Overvoltage, possible voltage doubling / escalation	Potentially more severe due to LC resonance
Value of controlled switching	High	Very high
Protection complexity	Moderate	Higher
Need for a harmonic study	Usually lower	Essential

Section 13

Common Mistakes

The recurring errors in capacitive-switching specification are less about the interruption itself than about the transients around it.

Common mistakes to avoid

Treating capacitive switching as trivial because the current is small — and overlooking the restrike overvoltage and the making inrush that actually govern the duty.
Specifying a restrike class without the earthing and fault conditions, so the wrong voltage factor \(k_{c}\) is applied to the recovery voltage.
Assuming a class C₂ breaker is restrike-free; it is only very low probability, and an installation should still tolerate a few restrikes over its life.
Ignoring the Ferranti effect on long lines and cables, so the true charging current is underestimated.
Applying a filter-bank breaker as if the recovery voltage were a clean 1 − cosine shape, when harmonics in the recovery voltage may cause reignitions.
Fitting transient-limiting inductors to solve outrush without checking the raised capacitor voltage, the harmonic-overload path they create, and the faster transient recovery voltage they impose.
Re-using a general-purpose or oil breaker for capacitive making and forgetting its sensitivity to outrush peak current and \(di/dt\).

Section 14

Key Points

Application checklist

Identify the application — capacitor bank, cable, line or filter bank — because it fixes the current magnitude, the recovery-voltage shape and the recommended restrike class.
Establish the system earthing, the bank neutral treatment and any earth faults; from these fix the recovery voltage and the single-phase voltage factor \((k_{c})\), which ranges from 1.0 to 1.7.
Select the restrike class based on application and switching frequency: C₂ is normally used for capacitor banks and frequently switched lines or cables; C₁ may be acceptable for infrequent switching at 100 kV and below; C₀ should only be used where restrikes are not a concern.
Check the rated capacitive current, remembering the Ferranti effect on long lines and cables and the altitude allowance above 1000 m.
Treat 60 Hz as the more onerous frequency; a 60 Hz capability covers 50 Hz for the same \(k_{c}\), but not automatically the reverse.
Match the limitation measure to the dominant stress: TLIs for inrush and outrush current, pre-insertion resistors/reactors or point-on-wave closing for both current and overvoltage, and surge arresters to clamp the overvoltage magnitude.
Assess outrush exposure for breakers that may close into a nearby fault with charged capacitance behind them, and run a risk analysis before committing to TLIs.
Consult the manufacturer wherever ratings are approached or exceeded, and coordinate the breaker class with the insulation capability of the surrounding plant.

Section 15

References and Further Reading

The standards and technical brochures behind this guide.

IEEE Std C37.012-2022, IEEE Guide for the Application of Capacitive Current Switching for AC High-Voltage Circuit Breakers Above 1000 V. New York, NY, USA: IEEE, 2022.
IEEE Std C37.04-2018, IEEE Standard for Ratings and Requirements for AC High-Voltage Circuit Breakers with Rated Maximum Voltage Above 1000 V. New York, NY, USA: IEEE, 2018.
IEEE Std C37.09-2018, IEEE Standard Test Procedures for AC High-Voltage Circuit Breakers with Rated Maximum Voltage Above 1000 V. New York, NY, USA: IEEE, 2018.
IEEE Std C37.100.2-2018, IEEE Standard for Common Requirements for Testing of AC High-Voltage Circuit Breakers Above 1000 V. New York, NY, USA: IEEE, 2018.
IEEE Std C37.010-2016, IEEE Application Guide for AC High-Voltage Circuit Breakers Above 1000 V. New York, NY, USA: IEEE, 2016.
IEEE Std 1036-2010, IEEE Guide for the Application of Shunt Power Capacitors. New York, NY, USA: IEEE, 2011.
CIGRE, Shunt Capacitor Switching in Distribution and Transmission Systems, Technical Brochure 817. Paris, France: CIGRE, 2020.
J. C. Das, Transients in Electrical Systems: Analysis, Recognition, and Mitigation. New York, NY, USA: McGraw-Hill, 2010.

Part 1 Reading now

Capacitive Current Switching for HV Circuit Breakers

The application hub: restrike classes C0/C1/C2, the voltage factor k _c, line and cable charging currents, inrush and outrush, and the overvoltage-limitation measures.

Series progress 1 of 8

Part 2

Energising Capacitor Banks: Inrush and Back-to-Back Switching

Single versus back-to-back inrush, the continuous-current rating, a full 138 kV worked example, capacitor-unit withstand and the mitigation options.