When a circuit-breaker interrupts current, the duty does not end at current zero. The voltage across the open contacts recovers, and the gap must withstand it — and if the gap fails again, several different events may follow. Re-ignition, restrike, late breakdown and non-sustained disruptive discharge (NSDD) are often confused, but their consequences differ. A voltage collapse across the contacts is not automatically a restrike: the current that follows decides what the event really is.
Reading time ≈ 20 min · Practical explanation for capacitive-current and fault-current switching
When a circuit-breaker interrupts current, the current does not simply disappear and the duty does not immediately end. After current zero the voltage across the open contacts recovers, and the contact gap must withstand that recovery voltage. If the gap fails again after interruption, several different events may occur — and although they are often confused, their consequences are not the same. The most important terms are re-ignition, restrike, late breakdown and non-sustained disruptive discharge (NSDD).
The distinction matters in circuit-breaker type testing, capacitive-current switching studies, vacuum circuit-breaker application, and the interpretation of test oscillograms. A voltage collapse across the contacts is not automatically a restrike: the current behaviour after the breakdown must also be considered. In practical terms, the question is not only whether the gap breaks down, but whether the breakdown develops into sustained current in the main circuit.
The four events — at a glance
Re-ignition is an early breakdown of the interrupter gap close to current zero.
Restrike is a breakdown that leads to current resumption in the main circuit.
Late breakdown is a delayed disruptive discharge after the circuit-breaker has already interrupted the current.
NSDD is a disruptive discharge that self-recovers and does not result in power-frequency current or main capacitive load current.
The voltage waveform alone is not enough to classify the event. The current path, the current duration and whether main-circuit current resumes must be checked.
Abbreviations and symbols used on this page
NSDDNon-sustained disruptive discharge
TRVTransient recovery voltage
RVRecovery voltage
RRRVRate of rise of recovery voltage
CBCircuit-breaker
VCBVacuum circuit-breaker
SF₆Sulphur hexafluoride (interrupting gas)
GISGas-insulated switchgear
HFHigh-frequency
ACAlternating current
\(X_0/X_1\)Zero- to positive-sequence reactance ratio
puPer-unit
EMTElectromagnetic transient (study)
CIGREInternational Council on Large Electric Systems
IECInternational Electrotechnical Commission
IEEEInstitute of Electrical and Electronics Engineers
\(C_n\)Neutral-to-earth stray capacitance
\(C_L\)Main load capacitance (capacitor bank, cable or unloaded line)
\(L_s\)Source-side inductance in the restrike / inrush path
Key idea
Voltage collapse is not the test. The classification depends on the current after breakdown — whether sustained current resumes in the main circuit.
Four distinct events. Re-ignition (early, within a quarter cycle), restrike (current resumes in the main circuit), late breakdown (a delayed dielectric breakdown), and NSDD (self-clearing, no main-circuit current).
Capacitive switching is a dielectric duty. Trapped charge raises the recovery voltage; a restrike discharges the main capacitance and can cause high overvoltages.
NSDD is reported, not a failure by itself. Restrike can fail a type test. Vacuum late breakdown often self-clears as NSDD; SF₆ late breakdown outside the interruption window can be more serious.
Key terms used on this page
01Re-ignition
A breakdown of the gap very soon after current zero — within about a quarter of a power-frequency cycle.
02Restrike
A breakdown a quarter cycle or more after current zero in which current resumes in the main circuit.
03Late breakdown
A disruptive discharge after a relatively long delay during the recovery-voltage period.
04NSDD
A disruptive discharge that self-recovers and does not resume main-circuit current.
05Recovery voltage
The voltage across the contacts after interruption; its transient part is the TRV.
06Trapped charge
Charge left on a capacitive load after interruption, holding the load side at a fixed voltage.
07Disruptive discharge
A breakdown of the insulation between the open contacts.
08Neutral shift
A change in the load-neutral potential after one pole breaks down in a non-effectively earthed circuit.
09Capacitive-current switching
Switching of lines, cables, capacitor banks or filter banks — a dielectric-recovery duty.
10Field emission
Electron emission from a contact surface that can initiate breakdown in a vacuum gap.
11Self-clearing
A breakdown whose current is interrupted again very quickly, so the gap recovers.
12Back-to-back switching
Energising a capacitor bank near another already-energised bank, giving high inrush.
13Main-circuit current
Power-frequency current flowing in the main power circuit; its resumption after a breakdown means restrike, not NSDD.
14Main load-circuit current
For capacitive switching, the current that flows by discharging or charging the main load (capacitor bank, cable or line).
15Capacitive restrike current
The current when the main capacitive load discharges through the source inductance and the gap after a restrike.
16High-frequency breakdown current
A short, fast local current just after a breakdown; if it self-clears without main-circuit current, the event is an NSDD.
Section 1
Why this distinction matters
Late breakdown and NSDD are especially important for capacitive-current switching, but they can also follow other interruption duties, including short-circuit interruption. Getting the classification right is what separates a reportable, harmless event from a duty that can fail a type test or damage equipment.
Capacitive-current switching is relevant for:
unloaded overhead lines and unloaded cables;
shunt capacitor banks and harmonic filter banks;
cable-connected capacitor banks;
GIS or cable systems with significant capacitance;
medium-voltage vacuum switchgear;
high-voltage circuit-breakers applied to capacitive networks.
A dielectric-recovery duty, not a high-current duty
Capacitive switching differs from fault-current interruption. The current is usually much smaller, but the recovery voltage can be high because the load capacitance may retain trapped charge after interruption. That trapped charge creates a high voltage across the open contacts; if the gap then breaks down, the stored energy can produce high-frequency currents, voltage shifts or full restrike currents. Capacitive-current switching is therefore a dielectric-recovery duty more than a high-current interruption duty.
Section 2
What happens after current interruption
A circuit-breaker interrupts AC current at or near current zero. Immediately afterwards the voltage across the contacts begins to recover, and may include the power-frequency recovery voltage, the transient recovery voltage (TRV), trapped-charge voltage, high-frequency oscillations from stray capacitances and inductances, and neutral displacement in unearthed or non-effectively earthed circuits. The gap must withstand this voltage; if it fails again, a disruptive discharge occurs, and the consequence depends on whether the discharge self-extinguishes or develops into sustained main-circuit current.
Recovery voltage, TRV and RRRV — see the dedicated guide
This page focuses on what happens when the gap fails again. The recovery voltage itself — the power-frequency recovery voltage, the transient recovery voltage (TRV) and its rate of rise (RRRV), and how the first-pole-to-clear factor and earthing shape it — is covered in Transient Recovery Voltage — Guide ↗. The recovery-voltage waveform alone does not classify the event; the current that follows a breakdown does.
Section 3
Relevant standards and guidance
For IEC-based applications the main reference is IEC 62271-100 — High-voltage switchgear and controlgear — Part 100: Alternating-current circuit-breakers (the current consolidated reference is IEC 62271-100:2021+AMD1:2024 CSV). It applies to three-phase AC circuit-breakers for indoor or outdoor installation, operating at 50 Hz and/or 60 Hz, on systems above 1 kV, and refers to IEC 62271-101 for synthetic test methods.
For IEEE-based applications the relevant documents are IEEE C37.04 (ratings and requirements), IEEE C37.09 (test procedures), and IEEE C37.012 (application guidance for capacitive-current switching). IEEE C37.012-2022 is particularly relevant because it addresses capacitive-current switching theory and the concepts of restrike, re-ignition and NSDD, and covers different capacitive loads — lines, cables, capacitor banks and filter banks. CIGRE switching-equipment publications and papers on NSDD provide useful background, but IEC and IEEE documents should be used for project specifications and type-test acceptance criteria.
In short, each document has a distinct role:
IEC 62271-100 — AC circuit-breaker requirements and capacitive-current switching test interpretation.
IEC 62271-101 — relevant where synthetic testing is used.
IEEE C37.04 — the rating basis.
IEEE C37.09 — the test-procedure framework.
IEEE C37.012 — application guidance for capacitive-current switching, including re-ignition, restrike and NSDD.
Confirm the edition
Standard editions and clause numbering change over time. Confirm the exact edition of IEC 62271-100, IEC 62271-101, IEEE C37.012, C37.04 and C37.09 specified for the project, and check the circuit-breaker type-test report against it. Do not mix acceptance criteria from different standards without checking the project-specified edition and the manufacturer type-test report.
Section 4
Re-ignition
Re-ignition is a breakdown of the contact gap shortly after current interruption. It is normally associated with breakdown close to current zero, commonly within one quarter of a power-frequency cycle after interruption. A restrike, by contrast, occurs later, and is only classified as a restrike when the breakdown leads to resumption of current in the main circuit. Timing is a useful indicator, but the current behaviour decides the consequence.
Re-ignition is usually related to insufficient dielectric or thermal recovery during the interruption process. It may occur when the arcing time is too short, the contact gap has not opened enough, or the arc channel has not cooled sufficiently. In some test sequences re-ignition can be part of the interruption process and may not have the same consequence as a restrike, but whether it is acceptable depends on the test duty, the standard, the circuit-breaker class and the current that follows.
In short
Re-ignition occurs early, in the current-zero region, and is linked to the immediate recovery of the interrupter gap.
Section 5
Restrike
Restrike is a more serious event: the gap breaks down after interruption — a quarter cycle or more after current zero — and current resumes in the main circuit. In capacitive-current switching this is particularly important, because the load capacitance may discharge through the source inductance and circuit-breaker gap, producing high inrush-frequency current and severe overvoltages.
A restrike is not identified by a sudden voltage collapse alone — it must be associated with current developing in the main load circuit. For capacitive interruption, a restrike generally involves discharge of the main capacitive load, with current much larger and longer than the high-frequency current of a short NSDD. The practical test is simple: restrike means the main-circuit current returns; NSDD does not. In type testing, restrike is normally a serious event and can fail the test duty, depending on the applicable standard and class.
Section 6
Late breakdown
Late breakdown is a disruptive discharge that occurs after a relatively long delay following current interruption — much longer than the immediate current-zero recovery period, and (according to the literature) up to about 1 second after interruption. It is mainly a dielectric phenomenon: the open gap withstands the recovery voltage for some time, then suddenly breaks down.
Late breakdown is the initiating event
Late breakdown can lead to two different outcomes. If it develops into sustained conduction and current resumes in the main circuit, it becomes a restrike. If it is self-clearing and the gap recovers almost immediately, it is classified as an NSDD. Late breakdown is therefore the initiating event; the result may be either restrike or NSDD.
Section 7
Non-sustained disruptive discharge (NSDD)
In IEC terminology, an NSDD is a disruptive discharge associated with current interruption that does not result in the resumption of power-frequency current — or, for capacitive-current interruption, does not result in current in the main load circuit. This distinction is the heart of the matter. An NSDD can show a sudden collapse of the voltage across the gap, and can produce a short high-frequency current; but because the current does not develop into the main power-frequency or main capacitive load current, the event is not a restrike.
The literature states that NSDDs may occur during the recovery-voltage period following a breaking operation, that their occurrence is not a sign of distress of the switching device under test, that their number is not significant for interpreting the performance of the device under test, and that they should be reported in the test report when seen on an oscillogram, in order to distinguish them from restrikes.
How NSDD is treated in a type test
NSDD should be recorded in the test report when it is visible on the oscillogram, mainly to distinguish it from a restrike. Its occurrence is not normally treated as a sign of distress of the switching device by itself. A restrike is more serious, because it indicates current resumption in the main circuit.
In short
NSDD is a breakdown that self-recovers very quickly and does not lead to main-circuit current.
Section 8
Why NSDD is not the same as restrike
Warning — voltage collapse alone does not prove restrike
A sudden voltage collapse across the interrupter gap is not sufficient evidence of restrike — an NSDD can also collapse the gap voltage. The classification depends on the current that follows. If only a short local high-frequency current flows and self-extinguishes, the event is an NSDD. If power-frequency current or main capacitive load current resumes, the event is a restrike.
NSDD and restrike may both begin with a voltage collapse across the gap, so the voltage waveform alone can be misleading. The difference is in the current path and duration. In an NSDD, the current is usually a short high-frequency discharge associated with local stray capacitances and inductances around the gap; it may last only a few microseconds to a few tens of microseconds, and then the gap recovers. In a restrike, the breakdown develops into current in the main circuit — in capacitive switching, the capacitor bank, cable or line capacitance discharges through the system, and the current lasts long enough to be a main-circuit restrike current.
To identify the event correctly, look at:
voltage collapse across the gap;
current magnitude and duration;
whether power-frequency current resumes;
whether the main capacitive load discharges;
whether one phase or multiple phases are involved;
whether the event is self-clearing;
whether the oscillogram shows only a local high-frequency discharge or a main-circuit current.
A sudden voltage shift does not automatically prove restrike.
Section 9
Why vacuum and SF₆ behave differently
Late breakdown is usually associated with vacuum switching devices, though it is not limited to vacuum in theory. Vacuum gaps have very high dielectric strength, especially for small gaps, and recover insulation quickly — but vacuum breakdown strength is statistical, with a relatively wide spread for apparently similar conditions. Possible causes of late breakdown in vacuum interrupters include small metal particles in the gap, particles or solidified droplets detached during arcing, vibration from the opening mechanism, a sudden increase in field-emission current, local surface condition, microscopic protrusions, and contact erosion after repeated operations. A particle or emission site can initiate breakdown even after the circuit-breaker has already interrupted successfully.
The important point is that a vacuum gap can often interrupt the resulting high-frequency breakdown current very quickly, which is why many late-breakdown events in vacuum switchgear appear as NSDD rather than restrike. In SF₆ circuit-breakers, by contrast, successful interruption is strongly linked to gas pressure, gas flow, nozzle design and cooling at current zero. If a late breakdown occurs outside the normal interruption window, the circuit-breaker may no longer have the gas-flow conditions needed to interrupt the new current, so late breakdown in SF₆ equipment can be more dangerous and may develop into significant arcing.
Vacuum versus SF₆ — the key difference
Vacuum interrupters may experience late breakdown because the dielectric strength of a vacuum gap has a statistical character and may be affected by particles or field emission. However, the resulting high-frequency breakdown current can often be interrupted quickly, so the event remains an NSDD. In SF₆ circuit-breakers, late breakdown outside the normal interruption window can be more serious, because the interrupter may no longer have the required gas-flow and pressure conditions to interrupt the new current. (For vacuum switchgear at 36 kV and above, NSDD assessment has historically received particular attention, because the recovery voltage is higher and the dielectric behaviour of the vacuum gap becomes more critical.)
Section 10
Three-phase circuits and neutral shift
Symbols used in this section
\(\Delta U_n\) is the voltage collapse or neutral-shift effect caused by the first late breakdown; \(\Delta U_f\) is the resulting voltage shift in the neighbouring phases; and \(\Delta U_r\) is the increased recovery voltage that can cause a second gap to break down.
NSDD behaviour is more complex in three-phase circuits, especially when the load neutral is unearthed or non-effectively earthed. If one circuit-breaker pole breaks down, the voltage across that gap collapses; in a floating-neutral circuit this causes a neutral voltage shift, and the voltage across the other open gaps may increase. That increased voltage can stress the neighbouring phases — and if it becomes high enough, another gap may break down, giving a single-phase NSDD, a two-phase NSDD, or a restrike.
The following relationship shows how a breakdown in one interrupter gap can shift the voltage in the neighbouring healthy phases:
\[ \Delta U_f = \frac{k-1}{k+2}\,\Delta U_n \]
\(\Delta U_f\)
voltage shift in the neighbouring healthy phases
\(\Delta U_n\)
voltage collapse or voltage shift caused by the first interrupter-gap breakdown
\(k\)
ratio of zero-sequence to positive-sequence reactance of the capacitive load circuit, \(k = X_0/X_1\)
\(X_0\)
zero-sequence reactance
\(X_1\)
positive-sequence reactance
\(X_0/X_1\)
earthing-related sequence-reactance ratio that determines how strongly the neutral point shifts
A higher value of \(k\) means a weaker neutral reference and a larger voltage shift in the other phases.
For non-effectively earthed loads (\(k\) large) the neighbouring-phase shift can be large; for effectively earthed loads (\(k\to 1\)) it tends to zero. This is why NSDD or restrike development can be more severe in unearthed capacitive circuits than in effectively earthed ones — the earthing condition strongly affects the consequence of a late breakdown.
How an NSDD can develop into a restrike
The sequence is: first, one open pole experiences late breakdown, its voltage collapses and a short high-frequency current begins. Second, in an unearthed circuit the neutral shifts, increasing the recovery voltage \(\Delta U_r\) across one or more healthy gaps. Third, that increased voltage \(\Delta U_r\) may cause a second gap to break down. Fourth, if the high-frequency current is interrupted quickly by one of the gaps, the event remains a two-phase NSDD. Fifth, if the current continues and develops into main-circuit current, it becomes a restrike. So restrike in three-phase unearthed circuits often requires breakdown in at least two interrupter gaps, because there is no earth return path and a current loop must form; a single breakdown may produce only an NSDD if it self-recovers first. (In an effectively earthed circuit an earth return path exists, so a single-pole restrike can drive main-circuit current through earth.)
Section 11
Restrike and NSDD current paths
In a capacitive restrike, the main capacitive load discharges through the circuit after the gap breaks down. This is not a small local high-frequency discharge around the interrupter — it involves the main capacitor bank, cable or line capacitance, and in capacitor-bank switching it can produce high inrush-frequency currents and significant overvoltages. A practical indicator of restrike is at least a substantial part of a half-cycle of current associated with the discharge of the main capacitive load through the source inductance. For capacitive switching, the current waveform is essential for classification — voltage shift alone is not enough.
In an NSDD, the current path is local and short. A single-phase NSDD may involve current through the stray capacitance across the gap, the capacitance from the load neutral to earth, the capacitance of connecting cables, and local stray inductances. A two-phase NSDD may involve high-frequency current between two phases if a second gap breaks down due to neutral shift. These currents can be high-frequency, but they do not represent sustained main-circuit current — which is why an NSDD is reported but not treated as a restrike in IEC-based interpretation.
In summary, the current path is the deciding factor:
Single-phase NSDD — a local high-frequency current flows through stray capacitance and earth-related capacitance: the neutral-to-earth capacitance \(C_n\) plus the local gap and cable capacitance.
Two-phase NSDD — a high-frequency current flows between two phases after a second gap breaks down.
Restrike — the main load capacitance \(C_L\) discharges through the source inductance \(L_s\) and the circuit-breaker gap.
Key point: voltage shift is not enough; the current path and the current duration decide the classification.
NSDD-related overvoltages
Although an NSDD does not lead to main-circuit current, it can redistribute charge between the stray and load capacitances. In most cases the short duration of the NSDD current limits the overvoltage development. However, in special circuits — such as short cable connections to capacitor banks, or sensitive unearthed capacitive loads — the voltage excursion may still be relevant, so the consequence should not be ignored in sensitive capacitive installations.
Section 12
Classification and oscillogram interpretation
Table 1 — Classifying a post-interruption breakdown. The decisive factor is the current.
Event
When
Current After Breakdown
Consequence
Re-ignition
< ¼ cycle after current zero
Within the immediate recovery
Linked to immediate gap recovery
Late breakdown
Longer delay (up to ~1 s)
Depends — see below
Initiating event → restrike or NSDD
NSDD
During the recovery-voltage period
Short, local, high-frequency; self-clears; no main-circuit current
Reported, not normally a failure by itself
Restrike
≥ ¼ cycle, current resumes
Main-circuit / main capacitive load current
Serious; can fail a type test; overvoltages
Oscillogram review checklist
Does the gap voltage collapse?
Does any current flow after the voltage collapse?
Is that current high-frequency and short-duration?
Does power-frequency current resume?
Does the main capacitive load discharge?
Is one phase or are two phases involved?
Does a neutral shift increase the voltage on the healthy phases?
Does the event self-clear, or develop into a restrike?
If it is a short local high-frequency discharge with no main-circuit current, the event is consistent with an NSDD. If main-circuit current resumes, it is a restrike.
Section 13
Application checks and EMT studies
Application checklist
Applicable standard basis (IEC or IEEE) and the capacitive switching class / rating.
Type-test evidence for unloaded line, unloaded cable, capacitor-bank and (if applicable) back-to-back capacitor-bank switching.
Breaker technology (vacuum, SF₆ or other) and whether the network/load neutral is effectively earthed or floating.
Trapped-charge assumptions, recovery-voltage peak and possible neutral shift.
Likelihood of single-phase or two-phase NSDD, and the consequence of a restrike.
Surge-arrester location and rating, cable length between circuit-breaker and bank, damping / inrush-limiting reactors, and the sensitivity of connected equipment to overvoltage.
When an EMT study is useful
An electromagnetic transient study is useful when the capacitive switching duty is sensitive or unusual — large shunt capacitor banks, harmonic filter banks, cable-connected capacitor banks, GIS substations with significant capacitance, long unloaded cables, frequent switching, vacuum circuit-breakers at higher voltage levels, unearthed or non-effectively earthed capacitive loads, tight insulation-coordination margins, cases where NSDD-related voltage shifts may matter, or systems where restrike could damage equipment.
The EMT model should include the source-side inductance and capacitance, the load and cable capacitance and inductance, neutral-to-earth and circuit-breaker stray capacitance, local stray inductance, the earthing condition, surge arresters, trapped charge, the pole-opening sequence, and a late-breakdown event representation where required. The study should distinguish a short local NSDD-type discharge from a full restrike involving main-load capacitance discharge.
Section 14
Common mistakes, report wording and key message
Common mistakes
Treating every voltage collapse as a restrike — NSDD also collapses the voltage but does not produce main-circuit current.
Ignoring current duration — the current waveform is essential to distinguish NSDD from restrike.
Assuming NSDD has no consequence — it is not normally a failure by itself, but it can create voltage shifts and should be reported.
Ignoring neutral shift in three-phase unearthed capacitive circuits — one breakdown can raise the stress on the other phases.
Assuming vacuum circuit-breakers are always restrike-free — vacuum gaps can break down late, but often self-clear as NSDD.
Assuming SF₆ and vacuum have the same late-breakdown behaviour — their recovery mechanisms differ.
Using only power-frequency models for capacitive switching — NSDD and restrike involve high-frequency local circuits and trapped charge.
Suggested report wording
“Late breakdown, re-ignition, restrike and non-sustained disruptive discharge were considered as part of the circuit-breaker recovery-voltage assessment. These events are not equivalent and should be distinguished based on the timing of the breakdown and the current that follows. Re-ignition refers to an early breakdown of the contact gap close to the current-zero region. Restrike refers to a breakdown that results in current resumption in the main circuit; in capacitive-current switching it is associated with discharge of the main capacitive load and can produce significant overvoltages. NSDD refers to a disruptive discharge associated with current interruption that does not result in resumption of power-frequency current or, for capacitive interruption, current in the main load circuit; it may produce a short high-frequency current and a voltage shift but self-recovers and should be distinguished from restrike. For IEC-based assessments, IEC 62271-100 should be used for AC circuit-breaker requirements and capacitive switching test interpretation; for IEEE-based assessments, IEEE C37.012 provides application guidance for capacitive-current switching, including re-ignition, restrike and NSDD, while IEEE C37.09 provides the test-procedure framework. The type-test report should be reviewed to confirm the applicable capacitive switching class, test duties and treatment of NSDD/restrike events. Where the circuit includes large capacitor banks, filter banks, long cables, unearthed capacitive loads or vacuum circuit-breakers at higher voltage levels, an electromagnetic transient study may be required to assess trapped charge, neutral shift, restrike current and possible NSDD-related overvoltages.”
Key message
Re-ignition, restrike and NSDD are not the same event, and the difference is not just the voltage collapse — the current after breakdown must be checked. NSDD is a self-clearing discharge that does not lead to main-circuit current: it is reported when observed, but not normally treated as a restrike. Restrike is more severe because current resumes in the main circuit, and in capacitive switching it can discharge the main capacitive load and produce high overvoltages. Vacuum interrupters may show late breakdown and NSDD because vacuum dielectric strength is statistical, but the high-frequency current is often interrupted very quickly; SF₆ interrupters recover differently, and late breakdown outside the normal window can be more serious. Always check the applicable IEC or IEEE standard, the capacitive switching class, the type-test record, the recovery voltage, trapped charge, the neutral earthing condition, and whether the event is NSDD or restrike based on the current waveform.
A practical five-part series on the fault-interruption and switching duties of high-voltage circuit-breakers — from the terminal fault and overhead-line faults, through transformer- and reactor-limited faults, to out-of-phase switching and post-interruption breakdown events.
05
05Part FiveReading now
Late Breakdown, Re-Ignition, Restrike & NSDD
Why a voltage collapse is not automatically a restrike — re-ignition, restrike, late breakdown and NSDD distinguished by the current that follows, for capacitive switching.
