What partial discharge is
In a medium voltage cable the XLPE insulation sits between two semiconductive screens: the conductor screen on the inside and the insulation screen on the outside, with a copper tape or wire screen over it. Between them the electric field is high and, in a good cable, smooth. A void in the insulation, a contaminant particle, or a protrusion where a screen meets the XLPE concentrates that field. When it is strong enough, small electrical discharges occur across the defect without bridging the full insulation. That is partial discharge (PD).
PD does not cause immediate failure, which is exactly why it matters. Each discharge erodes the insulation around the defect, and over time the damage can grow until the cable breaks down in service. A voltage withstand test may pass a cable with a small defect; a PD test is designed to find it. For that reason PD measurement is a routine test for screened MV cable, made on every manufactured length. Low voltage cables to IEC 60502-1 do not have the same requirement, because they have no screened field to control.
The routine test in IEC 60502-2
IEC 60502-2 covers extruded-insulation power cables in voltage classes from 3.6/6 kV to 18/30 kV. Its routine PD test is measured with the method in IEC 60885-3. The test voltage is raised to 2 U0 and held for a short time, then reduced to 1.73 U0, where the cable must show no discharge above the declared sensitivity. For routine testing that sensitivity is 10 pC or better. U0 is the rated voltage between conductor and screen: 8.7 kV for an 8.7/15 kV cable.
Alongside PD, the routine voltage test applies 3.5 U0 for 5 minutes without breakdown. The table below shows those multiples worked out for each class, rounded, for orientation only; the test plan always takes its values from the standard and the order.
| Voltage class | U0 | PD measurement at 1.73 U0 | Voltage test at 3.5 U0, 5 min |
|---|---|---|---|
| 3.6/6 kV | 3.6 kV | about 6.2 kV | about 12.6 kV |
| 6/10 kV | 6 kV | about 10.4 kV | 21 kV |
| 8.7/15 kV | 8.7 kV | about 15.1 kV | about 30.5 kV |
| 12/20 kV | 12 kV | about 20.8 kV | 42 kV |
| 18/30 kV | 18 kV | about 31.1 kV | 63 kV |
The partner catalogue behind our MV power cable page covers classes up to 8.7/15 kV, with a reference 8.7/15 kV Cu/XLPE/CTS/PVC 3x95 mm2 construction; higher classes are matched to a different plant with the test capacity to suit.
Why sensitivity is the number to watch
A PD result of "no discharge detected" means nothing on its own. It means no discharge above whatever the system could detect at that moment. If the test room is electrically noisy, or the system is poorly calibrated, the detection threshold may sit far above 10 pC and real defects pass unseen. That is why the standard ties the requirement to a declared sensitivity, and why a good report states it.
A usable PD record per drum shows:
- Drum number, cable designation and length tested.
- Calibration with a known charge before the test, and the resulting sensitivity in pC.
- Background noise level during the test.
- Voltage profile: the 2 U0 step and the measurement at 1.73 U0.
- The measured PD level, and the pass or fail statement against the declared sensitivity.
- Test date, equipment identification and the signature of the person responsible.
A failed PD result also deserves a clear procedure. A plant may retest after cleaning or re-making the terminations used for the test, because discharge at a poorly prepared test termination can look like discharge in the cable. That is legitimate, but the record should show both results and the reason for the retest. If discharge persists, the length should be rejected or the defective section located and cut out, with the remaining lengths retested and recorded under new drum numbers. Ask the plant how often lengths fail PD and how they are handled; a credible answer, with records, says more about the insulation line than any brochure.
Where PD defects come from
Most PD sources are created in production, which is why the audit of an MV plant looks closely at the insulation line. Contamination in the XLPE compound, poor handling of compound in an open area, an irregular interface between the screens and the insulation, and moisture or voids from incomplete processing are all typical causes. Damage after production matters too: a cable bent below its minimum bending radius, crushed on a badly stored drum or with its ends left unsealed so moisture enters the screen can develop discharge later even after passing the factory test. The factory audit checklist lists what to look for on site, and drum damage prevention covers the storage side.
Witnessing and site tests
For utility and engineering projects it is common to witness PD and voltage tests on selected drums before shipment. Witnessing works best when the drums are chosen by the buyer on the day, the calibration step is observed, and the inspector takes a copy of the raw results. Allow a test day in the schedule; the time a PD test takes depends on the number of drums and the room, and is planned with the plant.
After installation, the project may run its own commissioning tests with joints and terminations in place. Those are separate from the factory routine test and are specified by the project engineer and the network owner. A clean factory PD record gives a baseline: if site tests later show discharge, the problem is more likely to be in an accessory or in installation than in the cable itself. Our utility and grid page covers how MV cable fits a distribution project.
Putting PD into the order
Write the PD requirement into the spec sheet explicitly: routine PD test on every drum to IEC 60502-2, measured to IEC 60885-3, with the declared sensitivity and full records delivered with the shipping documents. Add witnessing if the project needs it. Lead time for MV cable depends on copper booking, the insulation line schedule and routine testing, and is confirmed in writing for each order; see MOQ and lead time. To scope a medium voltage order, send the BOQ through the quote form.