Skip to content

Case study

Transformer failure investigation — industrial distribution system

A recently commissioned transformer failed on re-energisation and tripped the site's main substation. Root cause established by event data and teardown.

Western Canada · Forensic engineering · Electrical failure

We were retained after a recently commissioned distribution transformer failed on re-energisation and took an industrial site's main substation off line.

Challenge

The site owns its substation and medium-voltage distribution. After a brief utility outage, its unit substations were re-energised one at a time. One faulted, and the fault was not cleared locally: upstream feeder, bus and differential protection all operated.

Three explanations were open: a defect in the transformer, a protection scheme that failed to coordinate, or an error in commissioning or switching. Each pointed to a different party.

Approach

  • Interviewed those involved and built a millisecond timeline from relay event files, waveform captures and supervisory control records
  • Traced the path of the fault current to locate its origin
  • Attended the detanking and teardown of the failed unit alongside the other parties, and measured the coil clamping assembly
  • Reviewed the protection coordination studies against IEEE 242
  • Reviewed dissolved gas analysis for the failed unit and for others of the same type
  • Reviewed the factory acceptance and site commissioning test records against the applicable CSA and IEEE standards

Findings

The fault originated in the transformer. The event data placed it inside the unit at the moment of re-energisation.

The coil clamping system did not restrain the windings. Teardown showed winding deformation, damaged insulation and loose blocking, and measurement showed the clamp tightened unevenly. Axial forces from inrush current are of the same magnitude as those from a short circuit, and the governing standard requires a transformer to withstand them.

The protection design was excluded. The coordination studies were consistent with accepted practice.

The test records had gaps, though none caused the failure. These were reported separately from the cause.

Why this matters

Gas trends on other transformers of the same type led to a recommendation for continued monitoring. A causation finding on one unit became a risk position on the rest of the fleet.

Services applied

Forensic engineering and loss consulting · Forensic electrical engineering · Protection coordination studies

Insurance and legal · Industrial, mining and manufacturing · All work

Talk to us about a claim

More work

Talk to us about your project

Send us the scope and the constraints. We will tell you plainly whether we are the right firm for it — and if we are not, we will say so.