A cable fault is not over when the repair is finished. It is over when the repaired cable has been tested, the test has passed, and the circuit is restored with a record that will stand up. VLF is the tool that closes that loop, and the way it is used after a repair differs from an acceptance test in one important respect: the cable around the repair is not new.
This article covers the sequence from fault location to re-energisation, what voltage to apply to a repaired cable, and why testing the whole run rather than only the splice is usually the right call.
Key takeaways
- The repair is verified by testing the run, not just the splice — fault current and handling stress the adjacent cable too.
- Retest a repaired cable at a controlled level, generally at or near the maintenance voltage, not the new-installation acceptance level.
- Partial discharge is the diagnostic that confirms the new joint is clean; a passed withstand alone does not.
- Ground and discharge properly before anyone touches the conductor — VLF leaves stored energy in the cable capacitance.
- Record voltage, waveform, duration, leakage current and any PD so the next test on that circuit has a baseline.
1. The sequence
A typical repair close-out runs in five steps:
- Locate the fault and confirm the section to be cut out.
- Repair — remove the faulted section, install the new joint or joints.
- Re-instate the cable, complete the assembly and inspect it.
- Test with VLF — withstand plus, where available, partial discharge.
- Discharge, ground and re-energise with a completed record.
Step four is where VLF earns its place, and its voltage is the decision that matters.
2. Locating the fault — and where VLF fits
Fault location is usually done with time-domain reflectometry, arc-reflection methods, or a surge (thump) generator, all of which operate before the repair. VLF is not primarily a locating tool, but it has a role at the boundary: it applies a controlled high voltage to the cable, and on some circuits a VLF source is used to condition or break down a high-resistance fault so a locating method can then see it.
Once the fault is found and the section repaired, the cable must be proven again — and that is a VLF job.
3. What voltage to apply to a repaired cable
The instinct is to test the repair at the same voltage as a new installation, on the reasoning that the repair is new. That reasoning is only half right. The joint is new, but the cable on either side is the same cable that was in service — and if it is aged, a full acceptance voltage over the whole run uses the repair as an excuse to over-stress the good cable.
The usual practice is to retest at a controlled level, at or near the maintenance voltage, held long enough to prove the repair and the adjacent cable:
- Enough to verify the new joint and the reinstated ends.
- Not so much that the aged cable on either side is driven to its limit.
- Held long enough at 0.1 Hz for leakage current and any PD to stabilise and be read.
The exact level follows the governing standard and the cable’s history; the logic behind the band is set out in VLF test voltage selection.
4. Test the run, not only the splice
A repaired cable should be tested end to end where practical, not just across the new joint. Two reasons:
- The fault event itself stressed the cable adjacent to the failure — the fault current and the mechanical work of excavation act on more than the fault point.
- The repair work — cutting, re-terminating, re-laying — touches the cable around the joint.
Testing the whole run at a controlled level catches a second weakness that a splice-only test would miss.
5. Confirm the repair with partial discharge
A new joint can pass a withstand and still contain an active defect. The most reliable way to confirm that the repair is clean is to add partial discharge measurement: a well-made joint should be quiet, and a reading that is elevated, or that grows with voltage, points to a defect in the new work rather than in the cable.
This is the same reasoning as for any accessory — see VLF testing of cable joints and terminations — and the readings are interpreted as in VLF partial discharge testing.
6. Discharge and ground before anyone touches the cable
A tested cable holds charge. VLF charges the cable capacitance to the test voltage, and the source does not remove that charge when the test ends. Before the cable is touched, before grounds are removed, before re-energisation:
- Discharge through the set’s discharge path or an earthed rod, and hold the ground on.
- Apply working grounds at the point of work and at the far end.
- Treat the conductor as live until the grounds are proven in place — the capacitance of a long cable can hold a lethal charge after the set is switched off.
7. Close the record
The repair is not complete until the test is documented: cable identifier, rated U0, test voltage as a multiple of U0, waveform, hold time, leakage current, any PD reading, the temperature, and the result. That record is what turns the next outage’s numbers into a comparison instead of a one-off. The full set of field checks that make the test valid in the first place is in VLF test set verification and field calibration.
For the choice between a bare withstand and a diagnostic on the retest, see VLF withstand vs diagnostic testing.
Frequently asked questions
What test follows a cable repair?
First confirm the fault is cleared, then apply a VLF withstand to the repaired run. Where the repair itself is the concern, confirm it with partial discharge, which can single out an active defect the withstand alone would not.
Should I test only the splice, or the whole run?
Test the run, not only the splice. A repair can mask a second defect elsewhere on the circuit, and the acceptance argument is about the circuit, not only the point of work.
Is the cable safe to handle after the test?
Only once the set is de-energised and the cable has been discharged and grounded. A VLF test leaves the cable charged, so discharge and grounding are mandatory, not optional.
