Key takeaways
- IEC 60076-1 requires routine tests on every transformer, including winding resistance, ratio, impedance, losses, and voltage withstand.
- Type tests such as temperature rise and lightning impulse are performed on a representative unit of a design, not on every transformer.
- Special tests, including partial discharge, sound level and SFRA, are performed only when specified in the purchase order.
- Typical acceptance tolerances allow around plus/minus 10% on measured losses and plus/minus 0.5% on ratio, but the purchase order should state exact values.
- A complete documentation pack (test reports, nameplate, drawings, oil analysis) should be received and reviewed before the transformer ships or is energised.
Factory acceptance testing (FAT) verifies that a transformer meets its design and contractual specification before it leaves the works. IEC 60076-1 defines three tiers of testing — routine, type, and special — with routine tests mandatory on every unit, type tests performed once per design, and special tests added only when specified. Understanding which tests apply, and the tolerances that decide pass or fail, is essential before signing off a purchase order or attending a witness test.
Routine tests (performed on every unit)
Routine tests under IEC 60076-1 apply to every transformer manufactured and typically include, in a common test sequence:
- Measurement of winding resistance — DC resistance of each winding, used to calculate I²R loss component and detect loose connections or broken strands.
- Measurement of voltage ratio and check of vector group/polarity — confirms the turns ratio matches the nameplate across all tap positions and that phase relationships (e.g. Dyn11) are correct.
- Measurement of short-circuit impedance and load loss — establishes the percentage impedance used for protection coordination and parallel operation, and quantifies load (copper) loss at rated current.
- Measurement of no-load loss and current — establishes core loss and magnetising current at rated voltage.
- Measurement of insulation resistance — insulation resistance between windings and to earth, typically using a megohmmeter, often supplemented by polarisation index for larger units.
- Applied voltage (dielectric) test — a power-frequency withstand voltage applied between windings and earth to prove insulation integrity.
- Induced voltage withstand test — an over-frequency, over-voltage test induced through the winding itself, checking inter-turn and inter-layer insulation, often combined with partial discharge measurement on larger units.
- Oil tests (liquid-immersed units) — dielectric breakdown voltage, moisture content, and dissolved gas baseline (DGA) before energisation.
Each of these is intended to be performed on every transformer, since manufacturing variation between units, not just design, is what routine testing is meant to catch.
Type tests (performed once per design)
Type tests demonstrate that a transformer design meets performance requirements and are performed on a representative unit rather than repeated for every transformer of that type, unless the purchase order specifically calls for a type test to be repeated.
- Temperature rise test — confirms winding and top-oil temperature rise stays within the limits for the insulation class (e.g. 65°C winding rise for ONAN oil-immersed designs) at rated load, run over an extended period until thermal stability is reached.
- Lightning impulse test — applies a standardised impulse waveform (1.2/50 microseconds) to simulate a lightning surge, verifying the insulation system withstands transient overvoltage without failure, per IEC 60076-3.
Because type tests are destructive-adjacent in stress and time-consuming, buyers should confirm whether the offered design already holds a valid type test certificate for the rating and voltage class in question, rather than assuming a fresh type test will be run.
Special tests (performed only when specified)
Special tests are not included by default and must be explicitly stated in the purchase order or technical specification.
- Partial discharge (PD) measurement — detects localised insulation defects via discharge pulses, increasingly requested even on distribution-class units for critical installations.
- Sound level (noise) measurement — verifies the transformer meets a specified sound power or pressure level, relevant near residential areas or sensitive sites.
- Sweep Frequency Response Analysis (SFRA) — records a baseline frequency-response signature useful for detecting winding movement after transport, short-circuit events, or over the transformer’s life.
- Zero-sequence impedance measurement — relevant for earth fault protection studies on certain winding configurations (e.g. star windings with an accessible neutral).
- Additional special tests can include short-circuit withstand verification, harmonic loss measurement, and auxiliary equipment functional tests.
Typical FAT sequence and duration
While the exact order can vary by factory procedure, a common sequence runs from the least invasive tests to the most demanding: visual and mechanical inspection first, followed by winding resistance, ratio and vector group checks, insulation resistance, short-circuit impedance and load loss, no-load loss and current, and finally the dielectric tests (applied voltage, then induced voltage, since the induced test stresses inter-turn insulation most severely and is typically performed last among the electrical routine tests). For a standard distribution transformer, a full routine test sequence typically takes a matter of hours; for larger power transformers with induced voltage and partial discharge measurement, or where temperature rise is included as part of the scope, testing can extend over one to several days. Buyers scheduling a witness visit should confirm the expected duration with the factory in advance, since travelling to witness only a partial sequence is a common and avoidable scheduling error.
Common causes of test failure or delay
The most frequent reasons a routine test result falls outside tolerance, or a witness test is delayed, are not exotic: loose bushing or tap-changer connections affecting resistance and ratio readings, incomplete oil filtration leaving moisture or gas content out of specification, calibration drift on test instrumentation not caught before the witness date, and simple scheduling mismatches where the unit is not fully assembled when the witness arrives. Agreeing a pre-test readiness checklist with the factory — confirming the unit is assembled, oil-filled (or ready for the dry test sequence), and instrumentation calibration certificates are current — before travel is booked reduces the risk of a wasted visit.
Witnessing the FAT
Witnessing arrangements should be agreed early, since they affect factory scheduling. Common options are: the buyer’s own engineer attends in person; an independent third-party inspection agency is appointed to witness and certify on the buyer’s behalf; or, increasingly, a video-witnessed test is conducted with a live feed and real-time reporting, which reduces travel cost for distant projects. Whichever option is chosen, the purchase order should specify which tests must be witnessed (often not all routine tests need witnessing) and the notice period the factory must give before the test date.
Acceptance criteria and tolerances
Measured results are compared against declared (guaranteed) values, within tolerances set by IEC 60076-1 unless the contract specifies tighter limits. Typical tolerance bands include:
| Test parameter | Typical IEC 60076-1 tolerance |
|---|---|
| Total losses (no-load + load) | +10% of declared total |
| No-load loss (individual) | +15% of declared value |
| Load loss (individual) | +6% of declared value |
| Voltage ratio | +/-0.5% of declared ratio (or the smallest tap-step value, if less) |
| Short-circuit impedance (>2.5%) | +/-10% of declared value |
| Sound level | +3 dB above declared value |
These figures are typical and should always be confirmed against the specific IEC 60076-1 edition referenced in the contract, since tolerance definitions have been refined across editions. Where a measured result falls outside tolerance, the standard course is retesting after investigation, or rejection and rework, as agreed in the contract’s remedy clause.
Documentation pack
Before shipment (or, at minimum, before energisation on site), the buyer should receive and review a complete documentation pack:
- Factory test report covering every routine test performed, with measured values against guaranteed values.
- Nameplate data sheet matching the physical nameplate.
- General arrangement and connection drawings.
- Oil test certificate for liquid-immersed units, including DGA baseline.
- Type test certificates for the design, where applicable and not repeated on this unit.
- Any special test reports agreed in the contract (PD, SFRA, sound level, etc.).
Retaining this pack is also valuable later: baseline DGA and SFRA results are the reference point against which future maintenance or fault-investigation tests are compared, so a missing or incomplete FAT record can complicate diagnosis years into service.
Summary
A transformer FAT is not a single event but a structured sequence of mandatory routine tests, design-level type tests, and optional special tests, each with defined tolerances against guaranteed values. Agreeing the test scope, witnessing method, and tolerance basis in the purchase order — before manufacturing begins — avoids disputes at the point of testing and gives a documentation pack that remains useful throughout the transformer’s service life. See our related article on loss capitalisation for how routine test loss figures feed into total cost of ownership calculations.
How MARS can help
MARS performs routine tests — winding resistance, ratio, insulation resistance, no-load and load loss, applied and induced voltage withstand, and partial discharge — on every transformer manufactured, with type tests and third-party or customer-witnessed FAT available on request. Test reports and documentation packs are provided ahead of shipment. To discuss FAT scope and witnessing arrangements for a specific order, see our products page or contact us.
Frequently asked questions
What is the difference between routine, type and special transformer tests?
Routine tests are performed on every single transformer manufactured, as required by IEC 60076-1. Type tests are performed once on a representative unit to prove a design, not repeated per unit. Special tests are performed only when requested in the contract, such as partial discharge measurement or sound level testing.
What tests are included in a standard transformer FAT?
A standard IEC 60076-1 factory acceptance test typically covers winding resistance measurement, voltage ratio and vector group check, short-circuit impedance and load loss, no-load loss and current, insulation resistance, and applied and induced voltage withstand tests, plus oil tests for liquid-filled units.
What loss tolerance is normally accepted on a transformer FAT?
IEC 60076-1 typically permits positive tolerances of around 15% on individual no-load loss, 6% on individual load loss, and roughly 10% on total losses, with tighter limits on combined loss and ratio; the purchase order should restate the applicable tolerances explicitly.
Do I need to witness the factory acceptance test in person?
Witnessing is optional under IEC 60076-1 but common for larger power transformers or critical projects. Options include attending in person, appointing an independent third-party inspector, or requesting a video-witnessed test with real-time reporting, agreed in the purchase order.
What is SFRA testing and when is it needed?
Sweep Frequency Response Analysis (SFRA) records a transformer's frequency response signature to detect winding deformation or displacement. It is not a routine test but is increasingly requested as a special test to create a baseline for comparison after transport or a later fault investigation.
What documents should be included in a transformer test report pack?
A complete pack typically includes the factory test report covering all routine tests performed, the nameplate data sheet, general arrangement drawing, oil test certificate (for liquid-filled units), and, where applicable, type test certificates and any special test reports agreed in the contract.
Tags: FAT factory acceptance testing IEC 60076 routine tests type tests
