VA-TEK Engineering VLF withstand testing is often described as if the procedure were the same for every cable. It is not. The voltage band, the value of a partial discharge reading, and the risks of the test itself all depend on what the insulation is made of. A procedure written for XLPE and applied unchanged to an old paper-lead cable is a different test with a different meaning.
This article compares the three dielectrics a field engineer meets most often in medium-voltage cable — XLPE, EPR and paper-insulated lead-covered (PILC) — and sets out where the same VLF procedure should be adjusted.
Key takeaways
- The dielectric sets how the cable responds to DC, what ageing looks like, and how it behaves under discharge — so it sets the test.
- XLPE is the modern default and the case VLF was popularised for; DC testing is actively harmful to it.
- EPR is more forgiving of moisture and DC, but VLF remains the field method of choice.
- PILC is legacy and moisture-sensitive; VLF can be used, but interpretation of results needs more care, and the failure modes are different.
- Never carry a voltage band across cable types without checking the standard — the multiplier is not the whole story.
1. Why the dielectric decides the test
Three properties of the insulation material drive every downstream decision. First, how it responds to a unidirectional (DC) field — some materials trap charge and are damaged by it, others tolerate it. Second, what ageing physically looks like — voids, water trees, dried paper, corroded sheath — because that determines whether a diagnostic can see it. Third, how readily it discharges, which sets the value of a partial discharge reading.
| XLPE | EPR | PILC | |
|---|---|---|---|
| Insulation | Cross-linked polyethylene | Ethylene propylene rubber | Impregnated paper, lead sheath |
| Typical era | Modern default (1980s onward) | Industrial and heavy-duty runs | Legacy, mostly pre-1980s |
| Response to DC | Harmful — traps space charge | More tolerant | Historically tested with DC |
| VLF suitability | Reference case | Well suited | Usable, with care |
| Ageing mode | Water trees, voids | Less prone to water trees | Paper drying, moisture ingress, sheath corrosion |
2. XLPE: the case VLF was built for
Cross-linked polyethylene is the material that made VLF a field standard. Its problem with DC is space charge: a DC test drives charge into the insulation and leaves it there, and when the cable is later energised at AC the trapped charge distorts the field and can itself trigger failure. VLF keeps the field alternating, so no net space charge builds.
Ageing in XLPE is dominated by water trees — diffuse, tree-shaped degradation that grows from voids and contamination under the influence of moisture and field. Water trees are the reason a withstand test can pass a cable that later fails: the degradation is distributed, not a single defect, and it may not trip a proof test at the chosen voltage. This is exactly the situation where a diagnostic — tan delta or PD — adds information the withstand cannot. See VLF tan delta testing.
Temperature matters too. The dielectric strength of XLPE falls as it warms, so the cable should be at a known, recorded temperature, ideally near ambient, when tested. A test run on a hot, recently loaded cable is not comparable to one run cold.
3. EPR: more forgiving, same method
Ethylene propylene rubber is selected where flexibility, heat resistance and moisture tolerance matter more than the lowest cost. EPR is less prone to the water-tree mechanism than XLPE and is more tolerant of DC, which is why some legacy DC test records on EPR cables are not as alarming as the same records would be on XLPE.
For field testing, the practical difference is small: VLF at 0.1 Hz remains the method of choice, and the voltage bands follow the same logic of new versus aged. Because EPR tolerates a wider range of conditions, the caution that applies to aged XLPE — test lower — is less critical, but the principle of not over-stressing an old cable still holds.
4. PILC: legacy, and a different animal
Paper-insulated lead-covered cable is the oldest of the three and is still in service in many networks. Its insulation is oil-impregnated paper inside a lead sheath, and its failure modes are not the same as extruded cable: the paper can dry out, moisture can enter through a damaged or corroded sheath, and the impregnating oil can migrate.
Historically these cables were tested with DC, and they tolerate it far better than XLPE does. VLF testing of PILC is possible and is used, but the results need more care:
- The voltage band is not the XLPE band. A level that is routine for XLPE may be inappropriate for aged paper insulation.
- Partial discharge interpretation is harder. The composite paper and oil structure discharges differently from extruded insulation, so a PD reading carries less transferable meaning.
- The cosine-rectangular waveform has historically been used on PILC alongside sinusoidal; whichever is chosen, the record must state it.
Where a PILC cable is the subject, follow the standard the network owner specifies rather than importing an extruded-cable procedure.
5. What changes, and what does not
| Decision | Varies with cable type? |
|---|---|
| Waveform (sine as reference) | No — sinusoidal 0.1 Hz throughout; cos-rectangular only where specified |
| Test frequency (0.1 Hz default) | No — set by cable length, not material |
| Voltage band (2.5 U0 acceptance, lower for aged) | Yes — adjust per material and the governing standard |
| Value of a PD reading | Yes — strongest on extruded cable, weaker on PILC |
| Whether a diagnostic is warranted | Yes — water-tree-prone XLPE benefits most |
6. The rule that survives every material
Identify the cable type before writing the procedure. Confirm the rated U0 from the cable rather than the system voltage, place the cable in the right test class, and pick the voltage from the standard that governs the material. The single transferable principle across all three dielectrics is the same: a proof test proves the cable held at that moment, not that it is sound, and the older and more moisture-exposed the insulation, the more the test should lean toward diagnosis rather than a high-voltage pass or fail.
For the selection of the test voltage itself, see VLF test voltage selection; for the trade-off between a plain withstand and a diagnostic, see VLF withstand vs diagnostic testing.
