Short answer: The routine AC withstand test applies 3.5 times the cable’s rated voltage (U₀) on medium-voltage cable, 2.5 times on high-voltage cable, and as little as 2 times on extra-high-voltage cable. The multiplier falls as the cable’s voltage class rises because the absolute test voltage would otherwise become impossible to generate and would damage the very insulation it’s meant to prove. Sensitive partial discharge measurement takes over the job of finding small defects.
That’s the whole story in one paragraph. The rest of this article explains it, with the exact figures from the three IEC cable standards.
Table of Contents
What is an AC withstand test?
An AC withstand test — also called a power-frequency voltage test or dielectric withstand test — applies an alternating voltage well above the cable’s normal working voltage for a set time. It’s a pass/fail proof test. If the insulation holds without breaking down, the cable passes. If it punctures, it fails.
It’s run at three points in a cable’s life:
- Routine test — on every manufactured length, in the factory.
- Type test — once, to qualify a cable design.
- After-installation test — on site, once the cable is laid and jointed.
This article focuses on the routine test, because that’s where the voltage-multiplier pattern is clearest.
The test voltage is always written as a multiple of U₀, the rated voltage between conductor and earth. Writing it that way lets one rule cover every cable size in a voltage class.
The pattern: routine AC withstand test voltages
Here’s what the three IEC standards call for on the routine factory test:
| Cable class | Standard | Test voltage | Hold time |
|---|---|---|---|
| MV, 6–30 kV | IEC 60502-2 | 3.5 U₀ | 5 min |
| HV, 45–150 kV | IEC 60840 | 2.5 U₀ | 30 min |
| EHV, 220–300 kV | IEC 62067 | 2.5 U₀ | 30 min |
| EHV, 330–500 kV | IEC 62067 | 2.0 U₀ | 60 min |
The multiplier drops as the voltage class climbs, and the hold time grows to make up for it. The trend even continues inside the EHV standard: a 220 kV cable gets 2.5 U₀, a 500 kV cable gets 2 U₀.
That looks backwards. The bigger, more critical cable gets the smaller multiple. Here’s why it’s the right call.
Why the AC withstand multiplier drops as voltage rises
1. The absolute voltage gets out of hand
Multipliers are easy to compare, but a test set produces volts, not multiples.
A 500 kV cable has a U₀ of 290 kV. At 2 U₀ you’re already generating 580 kV in the factory. Apply the medium-voltage multiplier of 3.5 U₀ to that same cable and you’d need over 1,000 kV — beyond most resonant test systems on a long, high-capacitance cable, and far beyond anything the cable will ever see in service. The multiplier has to come down just to keep the test physically possible.
2. Real overvoltages scale down in per-unit terms
A withstand test stands in for the worst voltage a cable meets in service — switching surges, temporary overvoltages, earth faults. On lower-voltage networks those events are large relative to normal voltage. On EHV networks they’re smaller in per-unit terms: the systems are tightly coordinated and surge arresters are sized aggressively. So a lower test multiple still covers the realistic worst case. The multiplier tracks the network, not the size of the cable.
3. The test must not wreck good insulation
This is the one that matters most for modern extruded (XLPE) cable.
EHV cable runs at high design stress — the insulation works hard even at normal voltage. Push a high multiple of U₀ through it and you risk starting the very faults you’re screening for: space-charge build-up, electrical treeing from a microscopic flaw. A medium-voltage cable has more headroom, so 3.5 U₀ for five minutes is a fair proof test. An EHV cable doesn’t, so hammering it at 3.5 U₀ would be a way to create weak cable, not find it.
4. The real screening moved to partial discharge
If the withstand test is dialed back, something has to catch the small defects. That’s the partial discharge (PD) test.
At high voltage the faults that kill a cable are tiny — a void, a contaminant, a rough interface at a joint. A brute-force withstand test is blind to those until they’re bad enough to puncture. PD measurement, run per IEC 60270 down to 5–10 pC, sees them while they’re still small. So the standards lean on sensitive PD detection as the main quality screen for HV and EHV cable, and let the withstand test play a supporting role. The lower multiplier is part of that shift — not a loss of rigor.
AC withstand vs DC and VLF testing
The withstand test can be run with different voltage types, and the choice matters for modern cable:
- AC (power frequency, 20–300 Hz) — the method the IEC cable standards specify for factory and most field testing. It stresses the insulation the way service voltage does.
- DC — once common for field testing, now discouraged for aged extruded cable. DC can inject space charge that damages XLPE and gives misleading results.
- VLF (very low frequency, ~0.1 Hz) — a practical field alternative for medium-voltage cable. It produces the AC-type stress without needing the huge test set that power-frequency AC demands on long cable runs.
For the routine factory tests in this article, power-frequency AC is the rule.
Pass and fail criteria
The criterion is simple and the same across all three standards: no breakdown of the insulation during the hold. The voltage is raised gradually to the specified value, held for the set time, and the cable must survive. There’s no partial credit — it either holds or it doesn’t.
For three-core cables tested with a three-phase transformer, the phase-to-phase voltage is 1.73 times the tabulated value.
Worked example: a 500 kV cable
- Rated voltage to earth (U₀): 290 kV
- Routine AC withstand test: 580 kV (2 U₀), held 60 minutes
- If the MV rule (3.5 U₀) were used instead: ~1,015 kV — impractical and damaging
Compare that to an 11 kV-class MV cable (U₀ ≈ 6 kV), which is tested at 21 kV (3.5 U₀) for just 5 minutes. Same family of test, opposite end of the multiplier.
FAQ
What is the AC withstand test voltage for cables? It depends on the voltage class. Medium-voltage cable (IEC 60502-2) is tested at 3.5 U₀, high-voltage cable (IEC 60840) at 2.5 U₀, and extra-high-voltage cable (IEC 62067) at 2.5 U₀ down to 2 U₀. U₀ is the rated conductor-to-earth voltage.
Why is the test voltage a multiple of U₀? Expressing the test as a multiple of U₀ lets a single rule cover every conductor size in a voltage class. The actual kV applied comes from each standard’s voltage table.
Is AC withstand testing the same as a hipot test? A hipot (high-potential) test is the general name for a high-voltage withstand test. An AC withstand test is a hipot test run with alternating voltage, as opposed to a DC hipot. The IEC cable standards specify the AC form.
How long is the AC withstand test held? On the routine factory test, 5 minutes at MV, 30 minutes at HV, and up to 60 minutes at EHV. The hold time gets longer as the voltage multiple gets smaller.
What happens if a cable fails the withstand test? The insulation breaks down — it punctures — and the length is rejected. A failure on a routine test means that drum doesn’t ship; a failure after installation means the fault is located and the joint or section is repaired before energizing.
Key takeaways
- The routine AC withstand multiplier falls with voltage class: 3.5 U₀ (MV) → 2.5 U₀ (HV) → 2 U₀ (EHV), with hold time rising from 5 to 60 minutes.
- It drops for three reasons: absolute test voltages become impossible to generate, real per-unit overvoltages are lower on big networks, and a high multiple would damage highly stressed insulation.
- Partial discharge testing, not a bigger withstand voltage, is what finds small defects in HV and EHV cable.
- The pass criterion is always the same: no breakdown.
So read it the right way round — the 500 kV cable tested at 2 U₀ is being screened more carefully than the 11 kV cable at 3.5 U₀, not less.