Short-Circuit Test Reports: Proving Similarity to the Offered Transformer

1. The report must belong to an engineering argument

Do not accept a short-circuit certificate as proof that every transformer in a manufacturer's catalogue has the same withstand capability. Ask for the complete report, the tested design identity and a signed comparison with the offered unit. IEC 60076-5 covers external short-circuit withstand, separating thermal calculation from evidence for dynamic effects. Its public scope identifies test and theoretical evaluation routes; it does not give a purchaser permission to extrapolate from matching kVA alone.

This review is for a procurement engineer deciding whether the evidence is sufficient to release a design. It is not a substitute for the specified standard, the manufacturer's detailed calculations or the contractually agreed acceptance route. A factory routine-test schedule and a short-circuit evidence schedule answer different questions.

2. Establish the actual duty and identity

Freeze the offered rating, winding arrangement, voltage ratio, vector group, frequency, impedance and tapping range. Add source impedance, maximum fault level, relevant fault locations, system X/R and protection clearing times. Contributions from parallel transformers or generators can change the duty. State whether these inputs represent the present network or its planned reinforcement.

The tested unit needs serial number, drawings and revision identifiers, test laboratory identity, test circuit and report date. Obtain oscillograms and recorded currents, durations and tap positions, together with pre-test and post-test measurements and inspection findings. A one-page pass statement cannot show which design was exercised or whether a later winding change was incorporated.

3. Illustrative current screening

Consider an illustrative 1000 kVA, 400 V three-phase transformer with 6% impedance. Neglect upstream impedance and assume rated voltage at the principal tap:

I rated = 1000000 / (√3 × 400) = 1443 A

I symmetrical ≈ 1443 / 0.06 = 24050 A

If another design has 5% impedance at the same rating, the corresponding screening current is 28860 A. The current ratio is 1.20, so an I²-based force screening gives 1.44. This is an engineering warning, not a calculated winding stress or a standard acceptance threshold. Geometry, leakage field, axial imbalance and peak asymmetry still require design analysis. The example excludes source impedance, tap variation and resistance; it cannot set a test current or relay clearing time.

4. Compare construction, not catalogue labels

Ask the OEM to identify every changed feature and explain its consequence. Equal impedance does not establish equal mechanical strength. A shorter winding, changed conductor temper or revised end support can matter even when electrical routine tests look similar. Conversely, different ratings do not automatically invalidate evidence if a detailed, accepted design assessment establishes the applicable relationship.

5. Separate thermal and dynamic acceptance

A satisfactory thermal calculation does not prove that a winding resists displacement during the first asymmetric peak. Dynamic assessment must address the relevant force directions and failure mechanisms; its inputs must agree with the drawings and material specifications. Do not invent a universal acceptable percentage difference in reactance from an abstract of a paywalled standard.

Record the acceptance route in the purchase order: test of the offered unit, justified similarity assessment, or another explicitly agreed evaluation route. Define who reviews confidential drawings and calculations. If the tender only provides a certificate, list the missing evidence as an open technical deviation rather than silently treating it as a pass.

6. Release and delivery evidence

Use a traceability register that connects approved design revisions, purchased materials, factory inspections and the delivered serial number. Require a change-control notification when winding or support details change after review. Include responsibility for closing deviations and the effect on delivery release; a commercial warranty cannot replace missing engineering evidence.

Coordinate the factory acceptance plan with an appropriate SFRA baseline. SFRA can support subsequent condition comparison, but a transport baseline is not a short-circuit withstand demonstration. Retain the accepted comparison and report with the asset record so a future uprating or network fault-level change can be reassessed.

7. Procurement FAQs

Is a laboratory certificate enough? It can establish a reported result for a tested unit; the complete report and design mapping are still needed for an offered unit.

Does matching kVA and impedance prove similarity? No. Winding geometry, support strength, peak duty and manufacturing traceability also determine the engineering case.

Can routine tests replace short-circuit evidence? They establish other properties. Agree the withstand evidence route separately before placing the order.

8. References

• IEC 60076-5:2006