A partial discharge is a small electrical discharge that bridges only part of the insulation, in a cavity, a sharp point or a contaminated surface. The test measures it by its apparent charge in picocoulombs with the electrical method of IEC 60270, or detects and locates it with UHF and acoustic sensors. It is the test that best reveals manufacturing or assembly defects in medium, high and extra-high voltage insulation before they end in a fault.
What a partial discharge is and what the test is
A partial discharge (PD) is a localized electrical discharge that does not bridge the two electrodes: it short-circuits only part of the insulation. It appears where the electric field exceeds the local strength: a gas bubble in a resin, a cavity in paper, a loose metal particle in a GIS, a sharp point on a conductor or a damp or dirty surface. Each discharge is very small, but it repeats every cycle, erodes the insulation and opens paths that over time end in a fault. To dig deeper into where a transformer’s hum comes from, read why a transformer hums.
The test applies AC voltage, normally above the service voltage at the factory, and detects the pulses those discharges produce: current pulses in the circuit, ultra-high-frequency electromagnetic waves and acoustic waves.
What it measures and what defect it detects
There are three families of methods:
- Conventional electrical (IEC 60270). A coupling capacitor and a measuring impedance pick up the current pulses. The result is the apparent charge in picocoulombs (pC), after calibrating the circuit with a charge calibrator. The inception and extinction voltages and the pattern of discharges against the voltage phase (PRPD) are also recorded, which helps recognize the type of defect.
- UHF. Ultra-high-frequency sensors pick up the electromagnetic waves. It is the usual method in a GIS substation and is also applied to power transformers with UHF sensors installed in the tank. What is specific to GIS is explained below.
- Acoustic. Sensors on the tank pick up the sound of the discharge and, by time differences, help locate it inside the transformer.
Defects it reveals: cavities and delaminations in resins, poorly impregnated paper, particles and protrusions, poor screen connections, sharp edges, surface contamination and, in cables, defects in joints and terminations.
The test does not measure moisture or the general aging of the insulation; that is done by tan δ and dielectric response.
In GIS substations
In a new GIS substation, insulation defects almost always come from assembly: a metal particle left inside a compartment, a protrusion on a conductor or on the enclosure, a poorly connected part left floating or a cavity in a spacer. That is why partial discharges are measured during the on-site withstand voltage test required by IEC 62271-203, with the UHF sensors that the manufacturer installs in the enclosure. The withstand voltage test says whether the insulation withstands; the PD measurement says whether a defect remains that could fail months later even though the test passed.
The IEC 60270 method in pC is impractical inside an on-site GIS, because of the noise and the large capacitance of the installation, and UHF cannot be calibrated in picocoulombs. So that a «clean» result means something, CIGRE brochure 654 (2016) describes a two-step sensitivity verification. In the laboratory, it determines which pulse injected into one sensor is equivalent, as a UHF signal, to a real reference defect. On site, that pulse is injected into one sensor and it is checked that the neighboring sensors detect it. Without that verification, it cannot be claimed that the installation is free of defects.
The pattern of discharges against the voltage phase helps recognize the defect, always as guidance:
| Typical GIS defect | Pattern feature (indicative) |
|---|---|
| Free particle | Scattered pulses with no fixed relation to phase; usually accompanied by a bouncing acoustic signal |
| Protrusion on conductor or enclosure | Pulses concentrated in one half-cycle, near the voltage peak |
| Floating electrode | High, regular pulses in both half-cycles |
| Cavity in a spacer | Pulses on the rising voltage, in both half-cycles |
Before declaring a defect, corona at the test source connections, radio and telephone signals and noise from the resonant system itself are ruled out. A confirmed signal is located by comparing amplitudes and arrival times between sensors, with acoustic support, and corrected before energizing. The patterns recorded at commissioning remain as a reference for monitoring over the life of the GIS.
Which equipment and voltage levels it applies to
| Level | Typical equipment | Factory | Commissioning | Maintenance |
|---|---|---|---|---|
| LV | Not applicable in practice | — | — | — |
| MV | Dry-type transformers, resin CTs and VTs, switchgear, cables and accessories | Yes, routine on several equipment types | Cables and switchgear, per contract | Online or offline depending on the equipment |
| HV | Power transformers, bushings, CTs and VTs, GIS, cables | Yes, routine | GIS (UHF) and transformers per contract | Online with UHF or acoustic |
| EHV | Transformers, reactors, bushings, GIS, cables | Yes, routine | GIS and transformers per contract | Monitoring and online campaigns |
Some examples of product standards that include it:
- Power transformers: IEC 60076-3 defines the induced voltage test with partial discharge measurement (IVPD) and an application guide in its Annex A.
- Dry-type transformers: IEC 60076-11 includes partial discharge measurement among the routine tests.
- Instrument transformers: IEC 61869-1 includes it as a routine test for equipment with Um of 7.2 kV or more, with permissible levels according to the insulation type and the system grounding. In its 2007 edition it applies to a CT or VT connected to ground, in a system with grounded neutral (earth fault factor up to 1.4). With liquid or gas insulation, the limit is 10 pC at Um and 5 pC at 1.2 Um/√3. With solid insulation, 50 pC and 20 pC. The current edition, from 2023, keeps the measurement as a routine test with its own level table.
- Bushings: IEC 60137 includes it among the routine tests.
Where it does not apply: in low voltage, because the electric field is not enough to produce discharges under normal conditions. Nor does it make sense to measure calibrated apparent charge on energized equipment: there the UHF and acoustic methods detect and locate, but do not replace the factory test.
The standard that defines it and how results are interpreted
IEC 60270 defines the terms, apparent charge, test circuits, instruments and their calibration. Its fourth edition (2025), retitled Charge-based measurement of partial discharges, replaced the 2000 edition, focuses on charge-based measurement and covers tests with AC voltage up to 500 Hz or with DC voltage. Electromagnetic and acoustic methods are covered in an informative annex, mainly for locating the source. For transformers, IEEE C57.127-2018 is the guide for the detection, location and interpretation of acoustic emissions, and CIGRE brochure 676 (2017) reviews factory and on-site measurements and proposes a three-step procedure for assessing a PD problem.
How it is interpreted:
- At the factory, against the maximum level in pC set by the product standard at the voltages defined in it. It is a pass or fail criterion.
- On site, by comparison with the commissioning fingerprint, by trend and by the PRPD pattern; the dBm or mV levels of UHF methods are not converted directly into pC.
- Inception and extinction voltage: a discharge that appears near the service voltage is more serious than one that appears only under overvoltage.
What can distort the result
- Noise and interference in the substation: corona effect on connections, radio, cellular telephony, converters. Background noise must be measured and documented before energizing the test object.
- External corona on test leads, connector edges or terminals without rings, which is mistaken for internal discharges.
- Calibration omitted or done with a different setup; in IEC 60270 every setup is calibrated.
- Surface moisture on bushings and terminals.
- Floating contacts or loose grounds that discharge on their own.
What to do if the result is bad
- Rule out external sources: connection corona, background noise, damp surfaces.
- Characterize the discharge: inception voltage, PRPD pattern and comparison between phases.
- Locate it with acoustic or UHF methods.
- Complement with dissolved gas analysis: hydrogen is a typical gas of discharges in oil.
- Agree with the manufacturer whether the equipment can stay in service with monitoring or must be repaired.
Equipment for this test
This test is done with equipment from recognized manufacturers; these guides explain what each one measures. In your quotation we recommend the test and the equipment your case needs.
- OMICRON MPD 600: partial discharge measurement
- OMICRON UHF 800: partial discharges in GIS
- OMICRON MONCABLO, MONGEMO and MONTESTO: PD in service
- HIKMICRO AI76 acoustic camera: leaks and discharges
Partial discharges with Tiza Group
Our team coordinates partial discharge measurements on GIS, transformers and cables in medium, high and extra-high voltage, in substations and industry, in Colombia or elsewhere on request, and interprets them together with the rest of the equipment’s tests. See the scope in power transformer testing and circuit breaker testing, or write to us to coordinate the measurement at the commissioning of your GIS.
Related guides
- Testing of primary substation equipment
- On-site withstand voltage test
- GIS substation
- Cable diagnostic tests
- Insulating oil analysis
- Power factor and tan delta test
- Why a Transformer Hums and When to Worry
References
- IEC 60270:2025, High-voltage test techniques – Charge-based measurement of partial discharges.
- IEC 60076-3, Power transformers – Part 3: Insulation levels, dielectric tests and external clearances in air.
- IEC 60076-11, Power transformers – Part 11: Dry-type transformers.
- IEC 61869-1:2023, Instrument transformers – Part 1: General requirements.
- IEC 60137, Insulated bushings for alternating voltages above 1 000 V.
- IEEE Std C57.127-2018, Guide for the Detection, Location and Interpretation of Sources of Acoustic Emissions from Electrical Discharges in Power Transformers and Power Reactors.
- CIGRE Technical Brochure 676, Partial discharges in transformers, 2017.
- CIGRE Technical Brochure 654, UHF partial discharge detection system for GIS: application guide for sensitivity verification, 2016.
- IEC 62271-203, High-voltage switchgear and controlgear – Part 203: AC gas-insulated metal-enclosed switchgear for rated voltages above 52 kV.
Frequently asked questions
What is a picocoulomb (pC) in a partial discharge test?
It is the unit of apparent charge defined by IEC 60270: the charge that, injected at the equipment terminals, produces the same reading as the real discharge. It is not the charge that moved inside the defect, which is why the system is calibrated in every setup with a charge calibrator.
Can partial discharges be measured with the equipment energized?
Yes. UHF and acoustic methods, and some electrical ones with sensors at the bushing tap or on cable screens, are used in service. They detect and locate activity, but do not give a calibrated pC measurement like the IEC 60270 test.
Why aren't partial discharges measured in low voltage?
Because at low voltage the electric field rarely reaches the value needed for a cavity or a sharp point to discharge. The test makes sense where dielectric stress is high: medium, high and extra-high voltage.