Partial-Discharge Testing for Medium-Voltage Cable Systems

What kinds of PD evidence does each test stage provide?

Four kinds of evidence are often mixed together in procurement. Keep them separate in the specification, because each one answers a different question.

Type tests qualify a design under the scope and conditions of the relevant product standard. IEC 60502-4 specifies type-test requirements for accessories used with cables in its stated medium-voltage range. Successful evidence can support the accessory design represented by the tested construction, but it does not certify every field installation.

Routine and sample tests apply to manufactured cable lengths or production samples, as defined by the governing cable standard and purchase specification. Ask for the standard edition, cable identity, drum or length traceability, the test object, the result and the authorized sign-off. A generic report from a different construction is weak evidence.

After-installation or commissioning tests examine the installed cable system after transport, pulling, jointing and termination. This is the stage at which field workmanship and the complete electrical path can be assessed. The exact test set depends on voltage class, cable insulation, circuit length, owner practice and the governing standard.

Diagnostic PD testing detects, measures and may locate discharge activity. The IEEE 400-series guide for field diagnostic PD testing covers new or service-aged installed shielded cable systems, including cable, joints and terminations. It allows online testing or testing with an external voltage source within its scope. A diagnostic result must be read together with the test conditions and the noise environment.

Stage names follow the categories above. The governing standard and the purchase specification set the actual test set for each project.

Which standard applies to my system voltage?

IEC 60502-2 covers extruded-insulation power cables for fixed installations from 6 kV (Um 7.2 kV) to 30 kV (Um 36 kV). IEC 60502-4 covers type testing of accessories within the range stated in that publication. For systems above 30 kV, IEC 60840 addresses cable systems, cables alone and accessories in its defined voltage range, which extends up to 150 kV (Um 170 kV). These documents have different scopes, so naming “IEC cable” without the part, edition and rated construction is insufficient.

The IEC measurement standard for partial discharge defines terminology, charge-based PD measurement methods, calibration requirements and guidance on discriminating external interference. It is used by product specifications, but it does not by itself define every cable-system acceptance program. Projects following IEEE practice should identify the applicable IEEE 400-series document and the owner specification. The choice between online measurement and an external source, and the source waveform and frequency, affects what is stressed and how results are interpreted. Project-level guidance on choosing cable systems is covered in the medium-voltage cable buying guide.

What should the installed-system test plan include?

The test plan should identify:

• Circuit and voltage designation, insulation type, length and installation arrangement.
• Cable manufacturer and construction, with drum-to-route traceability.
• Joint and termination manufacturer, model, quantity and installer records.
• Bonding and grounding configuration during the test.
• Online or offline method and the external voltage source, if used.
• Applied voltage sequence and duration, taken from the governing procedure rather than improvised onsite.
• Sensor locations, coupling arrangement, calibration or sensitivity checks and usable frequency range.
• Background-noise survey and methods for separating interference from cable-system signals.
• PD magnitude, repetition or pattern data, estimated location and phase identification where applicable.
• Acceptance, investigation and retest rules, including who has authority to interpret ambiguous indications.

Handling damage that can affect the test result is discussed in cable installation, bending and pulling. This scope should be agreed before cable energisation and preferably before procurement. Otherwise the supplier, installer and testing contractor may each assume that another party owns the interfaces and test accessories.

Why is one PD number not enough?

PD readings depend on the source, coupling, propagation and attenuation along the cable, sensor position, bandwidth, calibration approach and external noise. A signal recorded at one end of a long circuit is not automatically comparable with a factory result or with a reading from a different instrument configuration.

A routine PD limit is often quoted for IEC 60502-2 cables: no discharge above 10 pC at 1.73 U0 (a value commonly cited from IEC 60502-2 and CIGRE material). U0 is the phase-to-earth voltage. Treat this as a factory routine reference that must be confirmed against the governing edition and purchase specification, not copied into a different test setup.

The report should therefore preserve raw or reviewable records, the test circuit, instrument settings, calibration or sensitivity evidence, background noise, applied voltage and the event location method. Where an indication is found, engineering review should consider whether it is repeatable, voltage-dependent, localized to an accessory or cable section, and distinguishable from interference.

What procurement mistakes should buyers avoid?

• Requesting “PD-free cable” without a test basis. Replace the phrase with the governing standard, test object, measurement method and documented acceptance criteria.
• Using cable-only evidence for a complete system. Field joints and terminations add interfaces and workmanship. Include them in the installation records and site-test boundary where the project requires system-level assurance. Check test reports against the edition named in the contract, as explained in cable standards and test reports.
• Confusing withstand and diagnosis. A withstand test asks whether the insulation survives a defined electrical stress. PD diagnostics seek discharge activity and location information. One result should not silently replace the other.
• Buying a test after installation. Cable length, grounding, access, energisation status and noise can constrain the feasible method. Define the test while the route, accessories and commissioning program can still be coordinated.

The correct test voltage, duration, method and acceptance criteria depend on system voltage, insulation, circuit design, owner requirements and jurisdiction. They must be set by the governing standard and a qualified cable-system engineer. This guide does not replace the purchased standards or the project test specification.

Frequently asked questions