Large power transformers experience acceleration, vibration, shock, tilt, and handling forces during transport. Internal winding, leads, clamping structures, core, and insulation can move without producing an immediate open circuit or obvious ratio error. Power transformer frequency response analysis (FRA or SFRA) is a sensitive diagnostic method that compares the transformer’s impedance or transfer response over a broad frequency range.
FRA is most powerful as a comparison tool. A post-transport result can be compared with a factory baseline taken before shipment, a previous field trace, another phase, or a sister unit of the same design. The best evidence comes from a repeatable baseline on the exact transformer using documented connections.
What the response represents
Transformer windings, core, leads, bushings, capacitances, and structural paths form a distributed network of inductance, capacitance, and resistance. The measured response contains resonances and anti-resonances linked to that network. Mechanical displacement, deformation, loose clamping, lead movement, shorted turns, core grounding changes, or bushing differences can shift features.
Different frequency regions tend to have different sensitivities, but rigid universal boundaries are risky. Transformer size, design, test connection, and terminal condition influence the curve. Interpretation should use applicable guides, design knowledge, and multiple evidence sources.
Baseline quality matters
A factory shipping baseline should record instrument, calibration, test mode, source and measurement terminals, unused-terminal conditions, neutral and tank grounding, bushing connections, tap position, test leads, temperature, and active-part condition. If bushings or external links are removed for transport, note whether the baseline includes them.
Without a compatible baseline, phase-to-phase comparison can still identify asymmetry, but legitimate construction differences may exist. A three-limb core, regulating winding arrangement, lead routing, or tap changer can make one phase different from another.
Post-transport test setup
Lead length and placement strongly affect high-frequency results. Keep measurement leads close to the reference path and use the manufacturer’s recommended grounding braid. Clean terminal contact points. Route leads consistently between phases and document with photographs.
Confirm tap changer position and complete the same terminal connections as the baseline. Nearby grounded objects, connected cables, surge arresters, CTs, and bushings can alter capacitance. Test only under an approved isolated and grounded safety procedure.
Interpreting a difference
A shifted resonance is not automatically proof of damage. First repeat the trace without changing the setup and verify calibration, contacts, grounding, and tap position. Compare all phases and test modes. Review transport shock records, visual inspection, oil or pressure condition, and assembly changes.
Pattern changes that remain repeatable and cannot be explained by configuration deserve engineering assessment. The location and type of suspected movement may be inferred from affected frequency ranges and modes, but teardown decisions should not rely on a single curve alone.
Complementary tests
- Winding resistance and ratio on all tap positions.
- Excitation current and core-ground checks.
- Insulation resistance, power factor or tan delta, and capacitance.
- Leakage reactance or short-circuit impedance where practical.
- Bushing tests, oil tests, dissolved gas analysis, and visual inspection.
- Transport impact recorder review and mechanical inspection.
When to test
Useful checkpoints include before shipment, after site assembly but before energization, after a major through-fault, and when internal movement is suspected. A verified post-installation trace becomes a valuable field baseline for the transformer’s life.
BaoHui Tech recommends defining FRA baseline deliverables in the purchase specification: test modes, tap positions, digital file format, plots, instrument information, photographs, and setup notes. That turns a factory test into a practical condition-assessment asset.
Frequently asked questions
Can FRA prove a transformer is undamaged?
No single diagnostic proves absence of every defect. A stable, repeatable comparison supports the assessment and should be considered with other tests and transport records.
Can curves from different transformer designs be compared?
Only with caution. Design and geometry strongly shape the fingerprint. Same-unit baselines are preferred.