Buyer's guide

Switchgear testing: the evidence a buyer should require

1 September 2026 12 minute read

Switchgear testing is not one event. A buyer meets it as three separate occasions — design verification, routine verification, and site acceptance testing — each producing a different document, proving a different thing, and each easy to conflate with the others. Knowing which document to ask for, and when, is the actual buying decision.

Three occasions, three different documents

All three occasions get called "switchgear testing" in casual conversation, which is exactly why they get confused. Design verification happens once, on a representative sample of a design. Routine verification happens on every single unit before it leaves the works. Site acceptance testing happens after installation, once the equipment is wired into the buyer's own system. A fourth occasion — ongoing maintenance testing through the equipment's working life — sits outside the purchase itself but belongs in the same picture, and is covered later in this article.

Design verification: proving the design once, not the unit

Design verification does not have to mean a physical test for every characteristic. The relevant low-voltage standard allows a manufacturer to verify a characteristic by testing, by comparison against an already-verified reference design, or by calculation and assessment. A buyer should not assume that every line on a design-verification certificate reflects a laboratory test performed on that exact configuration.

The other point worth holding onto is what the resulting document is actually tied to. It documents a representative sample of the design, not the specific order. A certificate proves the design family passed; it is the buyer's job to confirm that their own order — rating, configuration, options — falls inside what that certificate actually covers. This site's busbar-manufacturer guide touches the design-versus-routine distinction briefly, in the context of busbar sizing; this article works through it in full, across both the low- and medium-voltage standards.

Low-voltage assemblies under IEC 61439-1

For low-voltage switchgear and controlgear assemblies, IEC 61439-1:2020 clause 10 sets out what design verification has to establish. Two items carry the most weight for a buyer sizing equipment: temperature rise stays within limits at rated current (clause 10.10), and the busbars and internal connections withstand the short-circuit current the assembly is rated for (clause 10.11). For more on how that short-circuit evidence is actually produced and checked at assembly level, see how to choose a busbar manufacturer.

Clause 10 also treats dielectric properties, protection against electric shock, mechanical operation, and the enclosure's degree of protection as separately verified characteristics — each is its own line item, not a single bundled pass.

Medium- and high-voltage switchgear under IEC 62271-200

For metal-enclosed medium-voltage switchgear, the equivalent one-off proving exercise is the type-test programme defined in IEC 62271-200:2021. It includes short-circuit making and breaking performance of the switching device, capacitive-current switching where the application calls for it, and an internal-arc classification test of the enclosure.

The same standard also defines Loss of Service Continuity (LSC) categories — a classification of how much of the assembly has to be de-energised or opened to reach one functional unit for maintenance. LSC is a design characteristic a buyer can specify up front, and design verification is what confirms the stated category was actually achieved.

Routine verification: what leaves the works on every unit

This is the occasion the buyer's own purchase order is actually tested against. Routine verification is mandatory on every assembled unit before it leaves the works — a distinct and separate exercise from the one-off design verification above.

Low-voltage assemblies

Clause 11 of IEC 61439-1:2020 sets a specific checklist that applies to every assembly: a visual check of the degree of protection, verification of clearances and creepage distances, a protective-circuit continuity check, a check of wiring correctness together with a functional test, a dielectric power-frequency withstand test, and confirmation of mechanical operation.

Traceability follows directly from the wording of that clause. Because it applies to "every assembly," a report that cannot be matched to a serial or works number does not evidence that the requirement was met for the actual unit being shipped — it could belong to any unit of the same design.

Medium- and high-voltage switchgear

For metal-enclosed medium-voltage switchgear, IEC 62271-200:2021 applies routine tests to every transport or shipping unit: a power-frequency dielectric withstand test and a measurement of the resistance of the main circuit. Where the design uses a sealed gas or liquid insulating or interrupting medium, a tightness test is added.

Routine verification happens on the assembly alone, before it is wired into anything. That is precisely the gap that site acceptance testing exists to close.

On-site switchgear testing: acceptance after installation

Acceptance testing performed after installation — commonly termed field or site acceptance testing — is a recognised category in its own right, distinct in purpose from the manufacturer's routine works testing. It assesses whether the equipment, as actually installed, is suitable for initial energisation.

Categories described within this kind of programme include system function tests, thermographic survey, electromagnetic field testing, and, for medium-voltage equipment, an online partial discharge study. What ties them together is that they exercise the switchgear as actually wired and interconnected on site — including interlocking with other apparatus and the site's own cabling — which a factory routine test, performed on the assembly alone before dispatch, cannot address.

One figure worth flagging rather than asserting: it is common practice to run a site dielectric test at a reduced voltage compared with the factory routine test, but no citable source in this research gave a specific standard-backed ratio. Treat that number as a question for the supplier, not a fact to expect on a data sheet.

Switchgear maintenance and testing after commissioning

A third occasion sits outside the purchase transaction altogether: testing through the equipment's working life, addressing condition rather than proving a design or a shipped unit. NFPA 70B exists as the recognised standard for this ongoing maintenance-testing work, and it is cited repeatedly, section by section, as the applicable reference throughout a U.S. federal maintenance manual for power circuit breakers.

That the obligation is recurring rather than one-off is not just a standards point. A U.S. federal agency maintains a published schedule of required periodic maintenance tasks and test frequencies for circuit breakers and switchgear at its own facilities — treating re-testing after commissioning as scheduled work, not a discretionary extra. Separately, U.S. federal workplace-safety regulation requires that electrical equipment be kept free from recognised hazards likely to cause death or serious injury, and maintained so that internal parts are not damaged or contaminated — a general legal basis for periodic checking that stands independently of any voluntary standard.

Even the instruments used in this maintenance work are treated as a maintained item. A federal facility-maintenance manual lists calibration of test equipment as its own scheduled maintenance category, sitting alongside the switchgear it is used to test.

The tests a buyer should know by name

Six checks recur across nearly every electrical switchgear testing programme, at the works or on site, low voltage or medium voltage. Knowing what each one actually measures is what lets a buyer read a report critically rather than just check that a box has been ticked.

Dielectric and power-frequency withstand

This is an applied AC voltage at supply frequency, held for a short specified duration, between live parts and to earth. It is a pass/fail test at a single stress level — it either survives the applied overvoltage without breakdown or it does not — rather than a measurement of a continuous quantity. Because this test (together with the routine test performed on every unit) is how a rated voltage claim is actually demonstrated, it sits alongside the fault-current figures a buyer specifies for the same equipment; see how available fault current is worked out and specified for the ratings this testing verifies.

Insulation resistance

This is a different check from dielectric withstand, even though both concern insulation. A DC megohmmeter measurement produces a numeric resistance value, used to screen for gross insulation defects, contamination or moisture rather than to prove the insulation survives a defined overvoltage. The measurement is taken for one minute at a manufacturer-set test voltage, or a minimum of 1000 V DC where no manufacturer value is available, first with the device closed and then repeated with it open. Where the manufacturer publishes no minimum acceptable value, a floor of 1 MΩ is used as the pass/fail threshold.

Main-circuit resistance

Main-circuit resistance testing measures the electrical resistance of the current path through a closed switching device or breaker contact. That value is obtained by injecting a known DC current — a minimum of 100 A, so the reading is stable and repeatable — and measuring the resulting voltage drop across the contact. Because contact-resistance values are measured in micro-ohms, a general-purpose digital multimeter does not have the resolution to read them meaningfully; the test requires a dedicated low-resistance ohmmeter or four-point tester built specifically for this kind of measurement.

The point of the test is early detection: a resistance that has drifted upward relative to the original factory value, or that differs markedly from the same assembly's other poles, points to a developing high-resistance joint or worn contact before it produces localised heating in service — often well before the equipment shows any other symptom.

Protective-circuit continuity

This check confirms a low-resistance path exists between the enclosure and other exposed conductive parts of the assembly and its earthing terminal. The purpose is straightforward: under an internal fault, the assembly should not present a touch-voltage shock hazard to a person in contact with it.

Functional and mechanical operation

This exercises the switching devices, interlocks, and any doors or shutters — manually or electrically — to confirm they operate as intended. It is a deliberately different kind of check from the electrical measurements above: it catches assembly or wiring errors that a resistance reading or a withstand test would not reveal at all.

Partial discharge measurement (medium voltage)

Partial discharge is a localised electrical discharge that only partially bridges the insulation between conductors. Because it can occur at normal operating voltage without immediately causing breakdown, its presence is treated as an early indicator of an insulation defect — one a brief, higher-voltage withstand test would not necessarily catch, since that test only proves the insulation survives one short overvoltage, not that it is free of a slowly developing weakness. For that reason, partial discharge measurement is applied as an additional, non-destructive diagnostic for medium-voltage switchgear insulation, not as a substitute for the mandatory withstand test. No general numeric acceptance threshold is stated here; ask the supplier what theirs is.

What a test report has to state to be worth anything

The requirement is not a calculation, but it is worth reading concretely — because the difference between a report that actually evidences a claim and one that only looks like it does comes down to a short, specific list of what is present.

Consider two routine test reports for the same hypothetical low-voltage assembly, rated 630 A, side by side. Neither is a real document; both are constructed here purely to illustrate the checklist.

A report that states enough:

• Assembly identified by works order and serial number, e.g. "Works order 4471, unit 3 of 6"
• Standard and clause tested against: "IEC 61439-1:2020, clause 11"
• Applied test values recorded: dielectric withstand applied and held for the specified duration without breakdown; insulation resistance measured at 500 V DC, closed and open, both readings stated in MΩ
• Ambient conditions at time of test noted: temperature and, where relevant, altitude
• Instrument used identified by model and serial number, with its calibration due date stated

A report that typically omits what matters:

• Identifies only "the design" or a model number, with no serial or works number tying it to the specific unit shipped
• States "tested to IEC 61439-1" with no clause reference, so it is unclear which checks were actually performed
• States only "passed" against each item, with no applied test value or reading recorded
• No ambient conditions given, despite insulation-related readings being sensitive to them
• No instrument identification and no calibration status

The first report lets a buyer verify the unit in front of them was actually checked, against a known clause, under known conditions, with a known instrument. The second only asserts that something happened. Four things separate them: per-unit traceability, a specific clause reference, recorded values rather than a bare pass mark, and instrument identity with calibration status — and each of those four is independently supportable. Per-unit traceability follows from the "every assembly" wording of the routine-verification clause itself. Ambient conditions matter because the standard's rated-condition provisions — including ambient temperature — sit alongside, and are meant to be read together with, the verification clauses. A dedicated four-point instrument is what the maintenance literature specifies for a credible contact-resistance reading in the first place. And calibration of test equipment is treated, in its own right, as a scheduled, documented maintenance item — not an incidental detail.

Questions to put to a supplier

• Which of the three occasions does this specific certificate cover — design verification, routine verification, or site acceptance?
• Is the routine test report tied to my order's serial or works number, or only to the design?
• Which standard and which clause was each test performed against?
• What test values were actually recorded, not just a pass/fail mark?
• What instrument was used, and what is its current calibration status?
• For medium-voltage equipment, is partial discharge measurement included, and against what acceptance criterion?

Questions buyers ask

What is the difference between a type test and a routine test?

A type test (design verification) is carried out once on a representative sample of a design, by testing, comparison or assessment. A routine test is carried out on every individual unit before it leaves the works. The two produce different documents and prove different things.

Does every piece of switchgear get individually tested before shipment?

Yes. Routine verification is mandatory on every assembled low-voltage assembly and every medium-voltage transport unit before dispatch, covering checks such as dielectric withstand and main-circuit resistance. A buyer can ask for the report tied to their own unit's serial number.

Is partial discharge testing required for medium-voltage switchgear?

It is a recognised additional diagnostic for medium-voltage insulation, not a substitute for the mandatory dielectric withstand test, and it is included among the categories covered in acceptance-testing standards. No general numeric acceptance threshold is asserted here — ask the supplier for theirs.

What should a switchgear test report state to be useful?

The specific assembly by serial or works number, the standard and clause tested against, the applied test values, ambient conditions, and the instrument used with its calibration status. A report tied only to "the design" rather than the shipped unit does not evidence that routine verification was met for that unit.

How often should installed switchgear be re-tested?

This is treated as a distinct, recurring obligation rather than a one-off: a maintenance-testing standard and published periodic schedules exist for exactly this purpose, separate from factory or site acceptance testing. Ask the supplier or a maintenance provider for the applicable interval.