Routine tests are performed on every transformer before it leaves the factory and confirm that the unit as built matches its specification. They do not prove the design. Understanding which is which is what separates a test certificate you can rely on from one that merely looks reassuring — and it is the single most common misunderstanding in transformer procurement.
Routine, type and special tests
| Category | Performed on | What it proves | Typical examples |
|---|---|---|---|
| Routine | Every individual unit | The unit as built matches its specification | Winding resistance · ratio and vector group · impedance and load loss · no-load loss and current · dielectric routine tests |
| Type | One representative unit of a design | The design meets the standard | Temperature rise (IEC 60076-2) · lightning impulse (IEC 60076-3) · short-circuit withstand (IEC 60076-5) |
| Special | Only when agreed between purchaser and manufacturer | A specific characteristic the contract calls for | Partial discharge · switching impulse · sound level measurement · agree scope at contract stage |
The routine test set, test by test
1. Measurement of winding resistance
Confirms conductor continuity, joint quality and that the windings match the design. Measured on each winding at each tap position, corrected to a reference temperature so results are comparable. It is also the baseline against which future maintenance measurements are compared, which is why a properly recorded value matters long after commissioning.
2. Voltage ratio and phase displacement (vector group)
Confirms the turns ratio on every tap and, critically, the vector group — Dyn11 and Yyn0 are not interchangeable, and getting this wrong on a paralleled or replacement unit is a serious fault. Measured at every tap position, not just the principal tap.
3. Short-circuit impedance and load loss
Measured with one winding short-circuited and current raised to rated value. This produces the impedance voltage that governs fault level and voltage regulation, and the load loss that dominates lifetime cost on a heavily loaded transformer. Corrected to reference temperature.
4. No-load loss and no-load current
Measured at rated voltage and frequency on the open-circuited winding. No-load loss is drawn continuously whenever the transformer is energised, so this is the figure that matters most on lightly loaded duty — rural feeders, renewable plants, and anything energised 24/7.
5. Dielectric routine tests (IEC 60076-3)
Two withstand tests prove the insulation system as built: the separate-source AC withstand test on each winding to earth and between windings, and the induced AC voltage test which stresses turn-to-turn and phase-to-phase insulation at enhanced voltage and frequency. Passing these is what allows the unit to be energised safely.
6. On-load tap-changer tests
Where an OLTC is fitted — as on our SZ11 series — the tap changer is tested for correct operation across its full range and the winding resistance is recorded at every position. OLTCs are the most mechanically complex item on a transformer and the most common source of later faults, so this record is worth keeping.
What routine testing does not prove
This is the part buyers most often get wrong. A complete routine test certificate does not tell you that:
- The transformer will survive a short circuit. That is IEC 60076-5, a type test on a representative unit or a design verification — not a routine test on your unit.
- Temperature rise is within limits. That is IEC 60076-2, also a type test.
- The unit will withstand a lightning impulse. That is IEC 60076-3, a type test.
- The declared sound level has been measured on this unit. Sound measurement per IEC 60076-10 is a special or type test unless agreed otherwise.
None of this is a criticism of routine testing — it does exactly what it is meant to do. But if your acceptance procedure assumes the routine certificate covers the design questions above, you have a gap. Ask at contract stage which type and special test documents your receiving authority requires, so they are produced with the order rather than reconstructed afterwards.
How to read a test certificate
| Element | Why it matters |
|---|---|
| Unit serial number | Ties the certificate to the physical transformer — without it the certificate proves nothing |
| Standard and edition tested to | IEC 60076 and ANSI/IEEE C57.12 give different limits; the edition changes tolerances |
| Measured values with units | A pass/fail-only certificate is not adequate evidence |
| Specified value alongside measured | Lets you check the tolerance yourself |
| Reference temperature for correction | Losses and resistance are temperature dependent; uncorrected figures are not comparable |
| Ambient conditions at time of test | Context for the corrected values |
| Instrument identification and calibration status | Traceability — a routine audit requirement on utility programmes |
| Date and signature | Accountability and warranty start |
Tolerances
IEC 60076-1 defines tolerances for losses and for impedance voltage, and the two are not the same figure. Because these values have been revised between editions, confirm the applicable tolerance against the edition referenced in your contract rather than relying on a remembered number — and state the tolerance explicitly in your purchase specification so there is no ambiguity at acceptance.
Witness testing and commissioning
- Witness testing is worth the trip on a first order from a new supplier, and on any unit above a modest rating. You confirm the set-up, the instrumentation and the actual readings before shipment — far cheaper than discovering a discrepancy after the unit is on the water. Agree the witness scope and attendee days at contract stage.
- Commissioning testing on site is not the same thing. Factory routine testing confirms the transformer before transport; commissioning confirms it after transport, installation and connection, typically covering insulation resistance, ratio, winding resistance and oil condition. Both are needed, because transport and installation can change what was measured in the factory.
Frequently asked questions
What is the difference between a routine test, a type test and a special test?
A routine test is performed on every individual transformer and confirms the unit as built. A type test is performed on one representative unit of a design and confirms the design — temperature rise, lightning impulse and short-circuit withstand. A special test is performed only when agreed, such as partial discharge measurement. Routine tests alone do not prove the design.
Does a routine test certificate prove the transformer will survive a short circuit?
No. Short-circuit withstand is a type test governed by IEC 60076-5, performed on a representative unit or demonstrated by design verification. Ask for the type test report or design verification for the specific design you are buying.
Which routine tests are performed on every transformer?
Under IEC 60076-1: winding resistance, voltage ratio and phase displacement, short-circuit impedance and load loss, no-load loss and no-load current, and the dielectric routine tests of IEC 60076-3. Where an OLTC is fitted it is tested too. Confirm the exact list against the edition in force.
What tolerances apply to losses and impedance?
IEC 60076-1 defines them and they are not the same figure. Confirm against the edition referenced in your contract, and state the tolerance explicitly in the purchase specification.
Should I attend a witness test?
For a first order from a new supplier, yes. It lets you confirm the set-up, instrumentation and measured values before shipment. Agree the scope at contract stage.
What should a test certificate show?
Serial number, rating and ratio, standard and edition, ambient conditions, measured values with units for every test, the specified values for comparison, instrument identification and calibration status, date and signature. Pass/fail alone is not adequate.