3.6 Excitation Current, Core Insulation, and Applied and Induced Potential Tests

Excitation Current, Core Insulation, and Applied and Induced Potential Tests

Quick Answer: Level III task 3.1a.2, "Perform complex electrical tests," names four measurements Level II does not require: core insulation resistance, insulation power-factor on all windings, power-factor on bushings, and excitation current in accordance with the test equipment manufacturer's published data. It also requires turns ratio "at all de-energized tap and LTC positions" and an overpotential test. This section covers excitation, core insulation, and the two families of potential test.

---

1. Exciting current — reading a pattern, not a number

What it is. With the low-voltage winding open-circuited, AC voltage is applied to one high-voltage phase at a time and the resulting current is measured. That current is the magnetizing current required to establish flux in the core, plus core loss current. It is typically well under one percent of rated current.

Why it works as a diagnostic. Exciting current is exquisitely sensitive to anything that changes the magnetic circuit or the winding turn count:

• Shorted turns create a short-circuited loop that the applied flux drives current through, sharply raising the measured exciting current.
• Core problems — shifted laminations, shorted laminations, an unintended core ground — change the reluctance of the magnetic path.
• Tap changer contact problems — misalignment, coking, high resistance, a bridging position — change the effective turns in circuit.

The pattern rule. Exciting current has no universal acceptance value, because it depends on core design, and this is exactly what the exam tests. On a three-phase core-form transformer the two outer phases (typically H1-H2 and H2-H3) share a similar magnetic path length and read similar to each other, while the center phase has a shorter path and reads noticeably lower. The expected signature is therefore high–low–high, or two similar and one lower.

Because the value is design-dependent, the strong comparisons are phase-to-phase within the unit, against the unit's own baseline, and against a sister unit. Compare at the same tap position and same test voltage; changing either invalidates the comparison.

Practical cautions: the low-voltage winding must be genuinely open, not merely unloaded; any residual magnetism from a prior DC test (winding resistance, for example) will distort the reading, so demagnetize the core after DC testing before measuring exciting current; and readings are taken at the same tap for every phase.

2. Core ground insulation resistance

Every transformer core is intentionally grounded at exactly one point. That single ground drains electrostatic charge that would otherwise build up and discharge, without creating a closed loop.

The failure mode: if a second, unintended ground appears — from a shifted lamination, a piece of debris, a failed clamp insulation, a sludge bridge — the core now has a closed conductive loop linking the main flux. Circulating current flows in that loop, producing localized overheating, which shows up in a DGA as a thermal fault signature.

The test:

1. Lift the core ground. The transformer usually brings the core ground out to an accessible external bushing or link for exactly this purpose. On units without external access, the test requires an internal inspection.
2. Apply 500 V to 1,000 V DC between core and tank with a megohmmeter.
3. A healthy core reads high — commonly hundreds or thousands of megohms; a value near zero means a second ground.
4. Restore the core ground. A transformer returned to service with the core ground still lifted is an immediate hazard: the floating core accumulates electrostatic charge and discharges to the tank, generating gas and eventually damaging insulation.

That final step is the exam point. Forgetting to restore the ground is a real field error with real consequences.

3. Applied potential versus induced potential

These are two different tests that stress two different insulation systems, and confusing them is a classic error.

Why induced testing needs elevated frequency. Core flux density is proportional to voltage divided by frequency (V/Hz). To induce, say, twice rated voltage at 60 Hz you would drive twice rated flux, saturating the core, drawing enormous exciting current, and overheating the unit. Raising the frequency in proportion holds V/Hz constant and keeps flux at its normal level. The elevated frequency also limits the permitted test duration, since core and winding losses rise with frequency — which is why induced test times are specified in cycles rather than a flat one minute.

In field practice, service-aged transformers are rarely subjected to full factory-level potential tests. NETA and IEEE C57.12.90 guidance is that field overpotential testing on in-service units is performed at reduced levels relative to factory values, because the test is a genuine proof test that can push marginal insulation to failure. That trade-off — a test that can cause the failure it is looking for — is exactly the judgement Level III and IV candidates are graded on.

4. How these fit the diagnostic sequence

The tests are ordered so that non-destructive, low-stress measurements come first and the results gate whether the higher-stress tests are performed at all:

1. Visual and mechanical inspection, nameplate, gauges.
2. Insulation resistance, DAR/PI.
3. Winding resistance (DC — remember to demagnetize afterward).
4. Turns ratio at all tap positions.
5. Excitation current.
6. Power factor / dissipation factor on windings and bushings.
7. Insulating liquid tests and DGA.
8. Only if all of the above are satisfactory: overpotential testing.

A unit that fails an early step is not advanced to an overpotential test. Applying a proof-level overvoltage to insulation already known to be degraded destroys the transformer and answers nothing.

Exam trap: A question describes a transformer with a suspected shorted turn and asks which test will confirm it. An applied potential test will not — it puts every turn at the same potential, so a turn-to-turn short experiences no stress and produces no indication. Excitation current or an induced potential test is what detects it.