# Transformer routine tests before shipment: what IEC 60076 requires

Published 2026-09-09 by Toppower Asia Engineering Team · https://www.toppower-asia.com/blog/transformer-routine-tests-before-shipment/

Under IEC 60076-1, every transformer must pass routine tests before it leaves the factory: winding resistance, voltage ratio and phase displacement, short-circuit impedance and load loss, no-load loss and current, dielectric routine tests to IEC 60076-3 and, for liquid-immersed units, a leak test. Cast-resin units also need a partial discharge test to IEC 60076-11. Type and special tests apply only when the contract asks for them.

## Key takeaways

- Routine tests are done on every unit. Type tests prove a design on one representative unit. Special tests are carried out only when the purchaser asks for them in the contract.
- Measured values are checked against the guaranteed values using the tolerances in IEC 60076-1, Table 1. For example, total losses may be up to 10% above the guarantee.
- A complete factory test report records the method, the reference temperature, the instruments and their calibration, and the result of every test. A pass mark on its own is not enough.
- Witness inspection is worth paying for on large or critical units, provided the test plan, hold points and acceptance criteria are agreed in writing before production starts.

## What is the difference between routine, type and special tests?

IEC 60076-1:2011 (Edition 3.0, still the current edition of the [general part of the transformer series](https://webstore.iec.ch/en/publication/588)) sorts factory tests into three groups:

- **Routine tests** are carried out on every transformer. They show that the unit was built as designed and that nothing went wrong in assembly.
- **Type tests** are carried out on one transformer that represents others of the same design. They prove properties that do not change from unit to unit, such as temperature rise and sound level. If a type test certificate already exists for an identical design, it is often accepted in place of a new test.
- **Special tests** are any tests other than routine or type tests that are agreed between purchaser and manufacturer. Short-circuit withstand and frequency response analysis are typical examples.

The difference matters for cost and schedule. Routine tests are included in a normal transformer price. Type tests take test-bay time, and a short-circuit withstand test needs an external high-power laboratory. Specify only what you need in the enquiry; our [transformer RFQ checklist](https://www.toppower-asia.com/blog/transformer-rfq-checklist/) covers how to word this.

## Which routine tests does every transformer receive?

The table below lists the routine tests IEC 60076-1 requires for all transformers, with the additions that apply to larger units and to dry-type units.

| Test | What it proves | Applies to | Category |
|---|---|---|---|
| Winding resistance | Joints, brazing and tap-changer contacts are sound; gives the reference for loss correction | All | Routine |
| Voltage ratio and phase displacement | Correct turns on every tap; vector group (e.g. Dyn11) as ordered | All | Routine |
| Short-circuit impedance and load loss | Impedance and load loss meet the guarantee, corrected to reference temperature | All | Routine |
| No-load loss and current | Core quality and assembly; no-load loss meets the guarantee | All | Routine |
| Applied voltage (separate-source AC) | Insulation from each winding to earth and to the other windings | All | Routine (dielectric) |
| Induced voltage | Turn-to-turn and phase-to-phase insulation | All | Routine (dielectric) |
| Lightning impulse | Withstand against lightning surges | Um ≤ 72.5 kV: type; Um > 72.5 kV: routine | Type or routine |
| Leak test with pressure | Tank, gaskets and welds are tight | Liquid-immersed | Routine |
| Core and frame insulation check | Core earthed at one point only | Liquid-immersed with insulated core/frame | Routine |
| On-load tap-changer operation | Mechanism, motor drive and controls work | Units with OLTC | Routine |
| Ratio and polarity of built-in CTs | Protection and metering CTs wired correctly | Units with bushing CTs | Routine |
| Insulation resistance, capacitance and dissipation factor, DGA, no-load loss at 90% and 110% voltage | Fingerprint values for later diagnostics | Um > 72.5 kV | Routine |
| Partial discharge measurement | No harmful discharges in the cast-resin insulation | Dry-type (IEC 60076-11) | Routine |
| Temperature rise | Cooling meets the guaranteed limits | One unit per design | Type |
| Sound level (IEC 60076-10) | Noise meets the guarantee | One unit per design | Type |
| Short-circuit withstand (IEC 60076-5) | Withstands through-fault forces | When agreed | Special |
| Frequency response analysis (SFRA) | Mechanical fingerprint for comparison after transport | When agreed | Special |

### Winding resistance

Resistance is measured with DC on every winding and every tap position, with the winding temperature recorded so load loss can be corrected to reference temperature. A noticeable imbalance between phases points to a poor joint or tap-changer contact.

### Voltage ratio and phase displacement

The ratio is checked on all tap positions and the vector group is confirmed. This is what makes sure a Dyn11 unit really is Dyn11 and will run in parallel with your existing transformers. See our guide to [vector groups such as Dyn11 and Yyn0](https://www.toppower-asia.com/blog/transformer-vector-groups-dyn11-yyn0/) for why this matters on site.

### Short-circuit impedance and load loss

The low-voltage winding is short-circuited and rated current is passed through the high-voltage winding at reduced voltage. The voltage needed gives the impedance. The power drawn gives the load loss, which is then corrected to the reference temperature. For liquid-immersed units with Class A insulation the reference is normally 75 °C. For dry-type units it depends on the insulation thermal class.

### No-load loss and current

Rated voltage at rated frequency is applied to one winding with the other winding open. The result shows how good the core steel is and how carefully the core was stacked and clamped. The losses article explains [why no-load loss matters so much in total cost of ownership](https://www.toppower-asia.com/blog/transformer-losses-and-efficiency-regulations/).

### Dielectric routine tests

The applied voltage test stresses the main insulation, and the induced voltage test stresses the insulation between turns and between phases. IEC 60076-3 sets the test levels from the highest voltage for equipment (Um) of each winding. On distribution transformers up to 36 kV, lightning impulse is normally a type test unless the purchaser specifies it as routine.

### Leak testing for liquid-immersed transformers

The complete transformer, filled with oil, is pressurised above its normal operating pressure and held for a set time. Inspectors then look for seepage at welds, gaskets and fittings. For [hermetically sealed oil distribution transformers](https://www.toppower-asia.com/products/oil-immersed-transformer/hermetically-sealed-10kv-to-400v-oil-distribution/), the corrugated fins must also flex without permanent deformation.

### Partial discharge for dry-type transformers

IEC 60076-11 makes partial discharge measurement a routine test for dry-type transformers. The coil is briefly overstressed, the voltage is reduced to the measurement level, and the discharge is recorded. The standard limit is 10 pC unless another value is agreed. A void or crack in the resin shows up here long before it would fail in service, so it is the most useful single test on a [cast-resin dry-type transformer](https://www.toppower-asia.com/products/dry-type-transformer/630kva-20kv-0-4kv-copper-wound-cast-resin-dry-type/).

## What tolerances apply to the measured values?

IEC 60076-1, Table 1, sets how far a measured value may differ from the guaranteed value before the transformer is non-conforming.

| Item | Tolerance |
|---|---|
| Total losses | +10% of guaranteed total losses |
| No-load loss or load loss individually | +15% of each, provided total losses stay within +10% |
| Voltage ratio, principal tap, first winding pair | The lower of ±0.5% of the declared ratio or ±1/10 of the measured impedance (%) |
| Short-circuit impedance, principal tap, impedance ≥ 10% | ±7.5% of the declared value |
| Short-circuit impedance, principal tap, impedance < 10% | ±10% of the declared value |
| No-load current | +30% of the design value |

If your specification sets losses as absolute maximums with no positive tolerance, say so in the enquiry. And a unit that must meet Ecodesign or DOE limits must meet them in its own right: IEC tolerances do not loosen a regulatory limit.

## What should a factory test report contain?

A useful routine test report is a record that can be traced, not just a list of "pass" marks. Check for:

1. The transformer's identity: serial number, rating, vector group, cooling, standard and edition.
2. The ambient temperature and the winding or oil temperature at each measurement.
3. Measured values and corrected values (losses at reference temperature, resistance at reference temperature), shown side by side with the guaranteed values and the tolerance.
4. Ratio results for every tap, with the deviation as a percentage.
5. Dielectric test voltages, frequency and duration, and for dry-type units the partial discharge level in pC.
6. The instruments used, with serial numbers and the date their calibration expires.
7. Signatures from the test engineer, from quality control and, where applicable, from the witness.

For units above 72.5 kV, also ask for baseline dissipation factor, capacitance, insulation resistance, DGA and any SFRA trace, for comparison after transport and in later maintenance.

## How do you run a witness inspection or FAT?

A factory acceptance test (FAT) with the purchaser or a third-party inspector present takes some planning. A sequence that works:

1. **Agree the inspection and test plan (ITP)** when the order is placed. List every test, its standard clause, the acceptance criterion and whether it is a hold point (test cannot proceed without the witness) or a witness point (test can proceed after notice).
2. **Confirm the test bay can do the tests**, including load-loss supply capacity and an impulse generator rated for the BIL if impulse is ordered.
3. **Give realistic notice.** Ask for the test date two to three weeks ahead so travel can be arranged; a remote video witness is a sensible fallback.
4. **Inspect before energising.** Check the nameplate against the approved drawing, bushings, accessories, paint thickness and oil level. On dry-type units, look for cracks in the resin.
5. **Watch the instrument readings, not just the summary sheet.** Note the temperatures yourself and see the calibration certificates.
6. **Sign the punch list before you leave**, recording deviations, agreed corrective actions and owners.

On [substation assemblies](https://www.toppower-asia.com/products/transformer-substation/), extend the FAT to cover switchgear functional tests, interlocks, secondary wiring and the enclosure's degree of protection. A transformer test alone does not prove that the substation works.

Toppower Asia follows this process for every order: routine testing before release, witness inspection on request, and test reports issued with the shipping documents.

## Frequently asked questions

### Are type test certificates from another unit acceptable?

Usually yes, if the earlier unit is of the same design: same rating, voltages, core and winding construction, and cooling. The certificate must come from a recognised laboratory or a witnessed factory test. Ask for the design reference so you can check that the two designs match.

### Is lightning impulse a routine test on a 33 kV transformer?

Not under IEC 60076-3 by default. For windings with Um ≤ 72.5 kV, lightning impulse is a type test. Many utilities still specify it as a routine test on each unit, which is allowed if it is stated in the order.

### What partial discharge level is acceptable for a cast-resin transformer?

IEC 60076-11 sets 10 pC as the limit unless the purchaser and supplier agree a different value. Some purchasers specify a lower limit, often 5 pC. A lower limit needs a low-noise test bay and should be agreed before the order is placed.

### What tests should be repeated on site after delivery?

Typically: insulation resistance, ratio, winding resistance, and an oil breakdown voltage test on liquid-immersed units. Large transformers should also be checked against the factory SFRA trace and the readings from the transport impact recorder. Comparing site results with the factory report is how you detect damage in transit.

### Can a transformer with losses above the guarantee be accepted?

Within the IEC tolerance it conforms to the standard. Many contracts still apply a financial penalty per kW of excess loss, calculated with the same A and B factors used in the bid evaluation. Above the tolerance, the purchaser can reject the unit.

If you are preparing a tender, send us your single-line diagram, transformer data sheet or ITP, and we will return a technical proposal with the routine test scope and witness options. [Contact the Toppower Asia engineering team](https://www.toppower-asia.com/contact/).
