claude code, claude code 
Transformer differential protection compares the currents entering a protected zone with the currents leaving it, and trips when the difference is not explained by the transformer itself. Ratio mismatch between CTs, vector group compensation, tap position and inrush all shift that balance in a healthy transformer. Transformer differential protection testing proves that the relay restrains for those conditions and still trips for a fault inside the zone.
That makes the work a comparison exercise rather than a pickup check. The engineer reproduces the secondary currents the CTs deliver for through-load, external fault and internal fault cases, then checks the relay’s answer against the bias characteristic and the harmonic blocking logic.
What transformer differential protection testing proves
The relay does not subtract raw secondary currents. It applies ratio compensation, removes the vector group phase shift, then computes an operate current and a restraint current. The trip decision follows the bias characteristic: a slope, or a set of slope sections, that raises the minimum operate current as restraint grows.
Polarity comes first, because reversing one CT secondary makes a healthy load current look like an internal fault. Compensation settings come second, because a ratio or vector group setting that does not match the nameplate puts every test point on the wrong part of the characteristic.
Express every test point in per-unit
Amperes at the relay terminals mean little on their own. The relay’s settings are in per-unit of the transformer’s rated currents, and the same injected current is a different multiple of rated current on each winding. Every test point follows one chain: the per-unit value of winding current, multiplied by the transformer rated current on that winding, divided by the CT ratio, at the relay’s nominal secondary of 1 A or 5 A.
Writing that chain into the test sheet makes results reproducible. A point recorded as “2 A injected” cannot be repeated or compared with the relay’s event record; one recorded as a per-unit multiple can.
Differential test cases
Each case needs both windings energized in the proportion the compensation produces.
| Test case | What is injected | Expected relay behavior |
|---|---|---|
| Through-load | Balanced currents in both windings | No trip; operate current near zero as restraint rises |
| External fault | Balanced current in and out, raised in steps | Restrained; no trip inside the bias characteristic |
| Internal fault | Current into one winding only, repeated for each winding | Trips at the expected multiple of rated current |
| Bias characteristic | Operate and restraint pairs inside and outside the curve | Pickup follows published slopes within tolerance |
| Harmonic blocking | Fundamental with superimposed second harmonic | Blocks on harmonic, trips when it clears |
| CT circuit check | Single-ended injection, one phase at a time | Polarity, compensation and wiring confirmed |
Bias characteristic and harmonic blocking
The bias characteristic is tested as a curve, not a point. Two or three points per slope section are enough: one just inside the boundary where the relay should operate, one just outside where it should restrain. Points placed far from a boundary prove nothing about the slope between them.
Harmonic blocking needs a source that can superimpose harmonics. The blocking element holds the differential element back during energization, when inrush carries a large second harmonic component, and releases it as that component decays. Injecting the fundamental alone never exercises the logic. The VAE-660 superimposes harmonics from the 2nd to the 20th.
Frequently asked questions
Why inject both windings at once?
A three-phase set injects one winding at a time, so the other reads zero and every point looks like an internal fault. Six-phase equipment such as the VAE-660 drives both ends together, which is the only way to reproduce through-load and through-fault conditions. The six-phase relay testing guide covers when the extra channels are justified.
What is a through-fault stability test?
A balanced current is driven through the zone, in and out, at levels above the differential pickup, and the relay must not trip. The test is repeated at several multiples to walk up the bias characteristic. It catches reversed CT polarity, incorrect compensation settings and a bias slope too flat for the transformer’s through-fault duty.
How do you test harmonic blocking?
Inject an internal fault current with enough second harmonic to assert the block, confirm that no trip is issued, then reduce the harmonic and confirm that the relay trips on the same fundamental current. A block that never releases is as wrong as one that never asserts.
Running the tests in sequence
Verify the CT circuits single-ended, walk the bias characteristic with balanced injection, run the through-fault case, and finish with harmonic blocking and trip timing. Recording every set point in per-unit alongside the relay’s event record keeps the results reviewable months later.
The VAE-660 provides six current channels of 0-30 A each, and 1000 VA when paralleled into a single 0-180 A output for high-burden relays. A three-phase VAE-430 covers single-ended work on smaller transformers and the overcurrent elements backing up the differential scheme. Both belong to the secondary test instruments family; the distance protection relay testing guide covers the zone-based elements that see the transformer from the far end.
Send the transformer ratings, the CT ratios and the relay model ahead of the outage through the contact form or request a quote, so that transformer differential protection testing is prepared before the crew arrives.
