Partial discharge in MV switchgear — detection with UHF and TEV sensors

The guide on partial-discharge measurement and tan-delta on MV cables covers PD detection via the classic, charge-based measurement per IEC 60270 on a de-energised cable. This article covers a related but substantially different application: detecting partial discharge (PD) in an energised medium-voltage switchgear assembly, using UHF (ultra-high-frequency) and TEV (transient earth voltage) sensors, as described in IEC TS 62478 ("High voltage test techniques — Measurement of partial discharges by electromagnetic and acoustic methods").

Why a different measurement technique is needed for switchgear

The classic IEC 60270 charge measurement requires a controlled, de-energised test setup with a coupling capacitor directly on the conductor under test — suitable for a disconnected cable, but not practical for a complete, energised switchgear assembly with multiple panels, where de-energising for every periodic check would mean a significant service interruption. UHF and TEV detection instead measure the by-products of a PD pulse — an electromagnetic wave (UHF) or a brief voltage transient on the earthed enclosure (TEV) — and can do so while the assembly stays in service.

Operating principle: two different signals from the same source

UHF detection: a PD pulse lasts only a few nanoseconds and radiates electromagnetic energy up into the gigahertz range. A UHF sensor — often a built-in window antenna in gas-insulated switchgear (GIS), or an external sensor on air-insulated equipment — picks up this wave. Because the metal enclosure of a GIS largely shields the signal from external noise, UHF detection on GIS is particularly sensitive and low in interference.
TEV detection: in air-insulated metal-clad switchgear, an internal PD pulse causes a brief voltage transient on the inside of the earthed metal enclosure, which also manifests as a measurable TEV pulse on the outside through small openings (seams, ventilation grilles, cable entries). A TEV sensor is simply held against or mounted on the outside of the enclosure, without the assembly having to be opened.

Note: neither technique produces a picocoulomb-calibrated charge value the way the IEC 60270 method does. UHF and TEV measurements are primarily qualitative and comparative: they detect the presence and relative intensity of PD activity and are therefore mainly suited to trending and triage, not to an absolute conformity assessment against a pC limit.

Practical use

Periodic walk-round with a handheld UHF/TEV instrument across all panels of a switchgear assembly, comparing the measured signal levels with each other: a panel with a clearly higher level than the other, comparable panels points to local PD activity that warrants further investigation.
Permanently mounted sensors with data logging on critical switchgear, similar to the use of a power-quality datalogger, flagging a gradually rising trend before it leads to a breakdown.
Combination with acoustic or corona detection: on air-insulated equipment, TEV detection can be combined with UV-camera and ultrasonic corona detection of external corona phenomena, while UHF mainly detects internal PD in a closed enclosure that a UV camera cannot see.
Localisation within a switchgear assembly: by comparing signal strength at multiple measurement points around the same panel (for example with several TEV sensors or a time-difference analysis between UHF sensors), the likely PD source within the panel can be narrowed down further, as a supplement to — not a replacement for — a targeted offline measurement if there is doubt.

Practical relevance

During the periodic inspection of a medium-voltage switchgear assembly under NEN 3140, non-invasive UHF/TEV PD detection is a valuable addition to thermographic inspection: where thermography reveals heat generation from an elevated contact resistance, PD detection reveals incipient insulation degradation that at that point still causes no measurable temperature rise — often well before a breakdown occurs.

Common mistakes

Interpreting a measured UHF or TEV signal level as a calibrated pC value, while both techniques are primarily qualitative and comparative in nature.
Not recording a baseline measurement on a known PD-free installation, so a later measurement cannot be assessed against a known starting value.
Using UHF detection on air-insulated equipment without metal shielding, where sensitivity to external electromagnetic noise is considerably higher than on GIS.
Ignoring an elevated PD signal because the installation does not yet show a noticeable temperature rise on thermographic inspection, while PD activity signals an earlier stage of insulation degradation than a thermal effect.