Combining a withstand test with partial discharge (PD) measurement turns a pass-or-fail proof into evidence that can also reveal incipient defects: the insulation is stressed with an alternating voltage while discharge activity is measured, so a defect that has not yet caused breakdown can be detected during the same test.

VLF and resonant sources are both used for monitored withstand testing on cables and other high-capacitance assets, and the value of the combination depends on the source, the noise environment and the interpretation discipline. A monitored withstand test says more than a plain proof, but its PD findings are indicators that require pattern interpretation, not automatic verdicts.

Why a Monitored Withstand Test Adds Diagnostic Value

A plain withstand test only reports breakdown: the insulation either held the voltage or it failed. A defect that is severe enough to cause breakdown will fail the test, but a developing defect that has not reached that point produces no signal in a plain proof. Monitoring the voltage application with PD measurement closes that gap in part: discharge activity at a defect site can be detected while the voltage is applied, which means the test can flag a joint, termination or section that needs attention even though it survived the withstand level.

The monitored approach is most valuable where installation defects are the target, such as acceptance testing of new cable circuits, because workmanship defects in joints and terminations are exactly the kind of localised problem that PD measurement can reveal under stress. It is also used in maintenance programmes where the asset history or diagnostics suggest that a withstand proof alone would leave the real question unanswered. The cost is complexity: PD measurement in the field is sensitive to noise, and the result must be interpreted by someone who can separate real discharge from interference.

Which Sources Suit PD Measurement: VLF, DAC and Resonant

PD measurement requires an alternating stress to drive discharge activity, so the source family matters. VLF sources energise cables at low frequency and are used for monitored withstand testing under cable testing guidance, with the discharge behaviour interpreted in the context of the low-frequency waveform. Damped AC (DAC) sources apply an oscillating voltage near the cable’s natural frequency and are specifically designed to combine withstand stress with PD measurement. Resonant sources provide power-frequency or near-power-frequency stress, which is the most service-representative environment for PD observation where the asset and specification call for it.

The source choice follows the asset and the governing guidance rather than a preference for one technology. Each source has its own PD measurement context, including the frequency of the stress, the test duration and the noise environment, and the PD results are not directly interchangeable between sources. The selection framework in the withstand method article places the source families; this article explains how the PD layer behaves within them.

Separating Real PD from Noise and Interference

The central difficulty of field PD measurement is separating genuine discharge activity from electrical noise. Corona from nearby conductors, switching transients, radio interference and poor connections in the measuring circuit can all produce signals that look like PD, and a false positive can send a maintenance team chasing a defect that does not exist. The separation is built on pattern recognition: genuine PD tends to appear at consistent phase positions of the applied voltage and to show characteristic patterns that repeat, while many noise sources are less phase-coherent or appear across the whole cycle.

Practical measures improve the signal quality before interpretation begins: use the correct sensor and connection for the asset and source, check and stabilise the grounding arrangement, perform a noise survey before the test, and record the background level so that it can be subtracted or compared. When a PD signal appears, repeat the measurement to confirm it is repeatable, change the test conditions where the procedure allows, and compare the pattern with known PD characteristics. The noise handling rules are developed in the partial discharge noise article in the PD series; the principle here is that no PD finding should be acted on from a single unverified observation.

What PD Findings Do and Do Not Prove

A PD finding during a monitored withstand test is evidence that discharge activity occurred at a detectable level under the applied stress. What it does not prove, by itself, is the exact nature or severity of the defect, the remaining life of the insulation, or the certainty that the asset will fail at a predictable time. PD measurement indicates that a defect site is active; its significance is judged from the pattern, the level relative to the background and the guidance for the asset type, and it is confirmed or localised by follow-up measurements.

The relationship between the withstand result and the PD result must be stated honestly. An asset can pass its withstand test and still show PD activity that warrants investigation; an asset can fail its withstand test at a defect that PD measurement had already located. Neither outcome makes the other redundant. The monitored test provides two layers of evidence, and the report should present both with their limitations rather than merging them into a single comfortable conclusion.

Combining Results with Other Diagnostic Data

PD findings are most valuable when they are combined with the rest of the condition picture. Tan delta measurement provides a dielectric-loss indicator that can support or question a PD finding on cable insulation; insulation resistance and its derived ratios describe the general insulation condition; and the asset history explains whether the discharge activity is new or long-standing. A monitored withstand test that finds PD at a level above the asset’s own baseline is a stronger signal than the same level on an asset with no prior record, because the comparison with history gives the finding its meaning.

The combination also guides the response. A clear, repeatable PD signal at a joint location may justify immediate investigation or repair on a new circuit, while a marginal signal on an aged circuit may lead to closer monitoring and a shorter retest interval rather than an immediate outage. The response should be defined in the test objective before the test, so that the interpretation leads to an action instead of an argument.

When to Use PD-Monitored Withstand Testing

Use the table to decide whether the complexity of monitoring is justified. When the objective is only proof, a plain test is appropriate; when the objective includes finding defects, monitoring adds value that the specification should describe.

Executing a monitored test well is a teamwork discipline. The PD measurement requires a stable noise environment, which means coordinating with the site: nearby work that produces switching or corona should be paused or moved if it contaminates the measurement window. The operator watches the withstand voltage while a second person or a recording system watches the PD signal, and the two are correlated in time so that a discharge event can be related to the voltage level at which it appeared. When a signal appears at a specific voltage step, that information is recorded, because it helps the interpretation and the localisation later. A typical response sequence for a new circuit is: confirm the signal is repeatable, rule out noise by the pattern and the background record, localise the source with the available sensors or by sectionalising, and then decide whether the defect requires repair before energisation or can be re-tested and monitored. The decision is written into the test objective beforehand, so the team is not improvising the acceptance criteria at the end of a long day.

Frequently Asked Questions

Can partial discharge be measured during a VLF test?

Yes. VLF sources can energise a cable while PD measurement is performed, and monitored VLF withstand testing is used under cable testing guidance. The discharge behaviour is interpreted in the context of the low-frequency waveform, and the noise environment must be controlled for the measurement to be meaningful.

What is the difference between DAC and VLF for PD measurement?

DAC applies a damped oscillating voltage near the cable’s natural resonant frequency and is designed to combine withstand stress with PD measurement. VLF applies a low-frequency alternating voltage and is used for monitored withstand testing under its own guidance. The source choice follows the asset and the governing specification.

What should I do if PD is detected but the cable passes the withstand test?

Treat the PD finding as evidence of active discharge at a defect site and investigate it rather than accepting the withstand pass as a clean bill of health. Confirm the signal is repeatable, rule out noise, and use the pattern and asset history to decide whether immediate repair, localisation or closer monitoring is the right response.

For VLF and withstand test equipment, see the high voltage insulation and withstand testing page, and for PD method selection read the partial discharge series on the HVTesters site.