Find the withstand test voltage, duration, governing standard and acceptance criterion for high voltage cable testing — VLF, AC site, AC factory or DC — together with the minimum insulation resistance for the cable length, corrected to the temperature you are testing at.
| Method | New | Maintenance | After repair | Factory |
|---|---|---|---|---|
| VLF 0.1 Hz | 2.0 U₀ | 1.5 U₀ | 1.73 U₀ | 1.73 U₀ |
| AC site withstand | 2.0 U₀ | 1.5 U₀ | 1.73 U₀ | 2.5 U₀ |
| AC factory (XLPE) | 2.5 U₀ | 2.5 U₀ | 2.5 U₀ | 2.5 U₀ |
| AC factory (EPR / PILC / PVC) | 2.0 U₀ | 2.0 U₀ | 2.0 U₀ | 2.0 U₀ |
| DC site (XLPE / EPR / PVC) | 3.0 U₀ | 2.5 U₀ | 3.0 U₀ | 3.0 U₀ |
| DC site (PILC) | 3.5 U₀ | 2.5 U₀ | 3.5 U₀ | 3.5 U₀ |
| Insulation resistance | No withstand voltage — a 1 to 10 minute megger test at 500 to 5000 V DC | |||
A newly installed 6/10 kV XLPE cable, 500 m long with a PE outer sheath, is to be commissioned by VLF testing. Ambient at the time of test is 30 °C.
| Method | Suits | Advantage | Limitation |
|---|---|---|---|
| VLF 0.1 Hz | XLPE, EPR | Portable set, service-like stress, allows tan δ | Not traditional for PILC |
| AC site withstand | All types | Closest to service conditions | Resonant set is large and costly |
| AC factory routine | All types | Full type and routine regime | Works only before installation |
| DC site | PILC only | Simple, compact equipment | Space charge damages extruded insulation |
| Insulation resistance | All types | Fast, non-destructive, trendable | Finds no weakness that only shows at stress |
| Test temperature | Correction factor | Limit for a 3.50 MΩ reference |
|---|---|---|
| 0 °C | 4.00 | 14.00 MΩ |
| 10 °C | 2.00 | 7.00 MΩ |
| 20 °C (reference) | 1.00 | 3.50 MΩ |
| 30 °C | 0.50 | 1.75 MΩ |
| 40 °C | 0.25 | 0.88 MΩ |
| 50 °C | 0.125 | 0.44 MΩ |
| Designation U₀/U | U₀ to earth | U between phases | VLF new test at 2 U₀ |
|---|---|---|---|
| 0.6/1 kV | 0.6 kV | 1 kV | 1.2 kV |
| 1.8/3 kV | 1.8 kV | 3 kV | 3.6 kV |
| 3.6/6 kV | 3.6 kV | 6 kV | 7.2 kV |
| 6/10 kV | 6 kV | 10 kV | 12 kV |
| 8.7/15 kV | 8.7 kV | 15 kV | 17.4 kV |
| 12.7/22 kV | 12.7 kV | 22 kV | 25.4 kV |
| 19/33 kV | 19 kV | 33 kV | 38 kV |
| 38/66 kV | 38 kV | 66 kV | 76 kV |
Test voltage is a multiple of U₀, the phase-to-earth rating, and the multiple depends on the method and why you are testing. A new cable given a VLF or AC site test is normally taken to 2 U₀ for 60 minutes, a maintenance test to 1.5 U₀ for 30 minutes, and a test after a repair to 1.73 U₀. Factory routine testing goes higher, at 2.5 U₀ for XLPE. A 6 kV U₀ cable given a VLF test after installation is therefore tested at 12 kV.
A direct voltage injects and traps space charge in the polymer, which does not dissipate when the test ends. When the cable is re-energised on alternating voltage, that trapped charge adds to the peak of the AC waveform and produces local field enhancement far above the design stress. The result is a cable that passes its test and then fails weeks or months later. This is why IEEE 400 treats DC as a legacy method for extruded insulation and why VLF at 0.1 Hz has replaced it. DC remains acceptable for paper-insulated lead-covered cable.
Very low frequency testing applies an alternating voltage at around 0.1 Hz instead of 50 or 60 Hz. Because the capacitive charging current of a cable is proportional to frequency, a 0.1 Hz source needs roughly one five-hundredth of the reactive power that a mains-frequency set would, so the equipment is small enough to carry to site. It stresses the insulation in a way that resembles service conditions without the space charge problem of DC, and it can be combined with tan δ and partial discharge measurement in the same connection.
Insulation resistance roughly halves for every 10 °C rise, so readings must be corrected to a common reference, normally 20 °C, before they can be compared with earlier tests. The correction multiplies the reference value by 0.5 raised to the power of the temperature difference divided by ten. A 3.5 MΩ limit at 20 °C becomes 1.75 MΩ at 30 °C and 14 MΩ at 0 °C. Comparing an uncorrected summer reading against an uncorrected winter one is the most common way a healthy cable appears to have deteriorated.
Cable ratings are written as U₀/U, for example 6/10 kV. U₀ is the rated power frequency voltage between one conductor and earth or the metallic screen, and U is the voltage between two phase conductors. In a three-phase system U = √3 × U₀, so a 6/10 kV cable has 6 kV to earth and 10 kV between phases. Withstand test voltages are always specified as a multiple of U₀, because the insulation wall between conductor and screen is what the test stresses.
Results are for guidance and preliminary planning. The governing document is always the cable manufacturer's recommendation and the project specification, which take precedence over the generic multipliers shown here. High voltage testing must only be carried out by competent authorised personnel under a written safe system of work.
multicalci.com — free engineering calculators