Repeated Differential Trip While Charging a 50 MVA Station Transformer
At a 2×210 MW thermal station, a 50 MVA, 132/33 kV station transformer (YNd11, OLTC on the HV winding) is charged from a strong 132 kV bus to feed the unit auxiliary buses. On each of the last three chargings it has tripped on transformer differential protection within 50–70 ms of breaker closing, every time with the 33 kV side unloaded. The relay disturbance record shows HV secondary currents of 0.52, 0.52 and 0.51 A, LV currents zero, differential current 0.28 pu, second-harmonic content 11% and fifth harmonic 1.2%. CT ratios are 400/1 A on the 132 kV side and 2000/1 A on the 33 kV side; both ratio-correction factors in the relay are set to 1.00 and the vector compensation is set internally to YD11. Ratio, winding-resistance, insulation-resistance and oil BDV results from the last overhaul are within limits. The relay's second-harmonic blocking is set at 15% and its unrestrained element at 8 pu. The unit has a six-day outage window and the auxiliary buses are presently on standby supply.
1. Verify whether the CT ratio compensation is correct, and decide whether the ratio mismatch can explain this particular trip.
Read the worked answer
Check the ratio compensation first.
- HV rated current = 50,000 kVA / (√3 × 132 kV) = 218.7 A; through a 400/1 CT this is 0.547 A secondary.
- LV rated current = 50,000 / (√3 × 33) = 874.8 A; through a 2000/1 CT this is 0.437 A secondary.
- Required correction = 0.547 / 0.437 = 1.25. With both channels set to 1.00 the relay compares unequal quantities and will see about 25% spill current on load.
But it cannot explain this trip. At charging the LV side is open, no load current flows, and the measured HV current (0.52 A ≈ 0.95 pu of rated) is the transformer's own magnetising inrush, not a load current. A ratio mismatch produces mismatch proportional to through-current, and here the through-current is zero.
Conclusion: the trip is inrush-related, with the second-harmonic restraint failing to block because the measured 11% is below the 15% setting. The 1.25 ratio correction is nevertheless a genuine defect that must be corrected before the transformer is loaded, otherwise it will maloperate on load or fail to detect an internal fault with adequate sensitivity.
2. The shift engineer proposes lowering the second-harmonic blocking setting to 10% so that the transformer can be charged today. Evaluate this proposal and give the correct diagnostic sequence.
Read the worked answer
Reject the proposal as unsafe. Lowering the blocking threshold to 10% means the relay will treat genuine internal faults accompanied by CT saturation (which also produces second harmonics) as inrush and may refuse to trip. It treats the symptom, not the cause, and it blinds the main protection of a 50 MVA transformer.
Low second harmonic can arise from: (i) deep CT saturation driven by the DC offset of the inrush, especially with a strong source, point-on-wave closing and remnant core flux; (ii) a wrong CT secondary connection or vector setting so that the relay sees a spurious differential quantity; (iii) an actual inter-turn or core fault, which produces mainly fundamental current.
Sequence before any further charging:
- DGA on oil (key gases, acetylene, ethylene) and repeat ratio, winding resistance and magnetising current at reduced voltage; add SFRA to rule out winding/lead distortion.
- Verify CT polarity, single-point secondary earthing, CT class and connected burden, and confirm the star/delta compensation matches YNd11.
- Correct the ratio correction to 1.25 and record the change.
- Only if all tests are healthy, treat it as inrush with poor harmonic signature - demagnetise the core and charge with the protection in service.
3. Assuming all offline tests are healthy, state how you would re-energise the transformer and what records would prove the issue is closed.
Read the worked answer
Do not bypass the differential relay and do not leave it in alarm-only mode for a transformer of this rating without written approval from the protection wing.
Re-energisation plan:
- Demagnetise the core by reduced-voltage AC or controlled DC injection to bring remnant flux near zero - remnant flux is the usual reason a given unit trips repeatedly while an identical unit does not.
- Correct the ratio correction to 1.25 and confirm the vector setting, CT earthing and burden.
- Charge from the 132 kV side with the 33 kV side open (the source impedance and breaker should ideally have a closing resistor or controlled point-on-wave switching; if the station has a controlled-switching relay, select the optimum closing instant).
- Capture the relay disturbance file for every charging attempt and compare the harmonic signature with the earlier records; a repeatable signature identical in shape confirms inrush.
- On success, load in steps and check differential spill current at 25%, 50% and 100% load; repeat DGA at 24 hours and 7 days.
Evidence to close the case: corrected setting sheet signed by the protection engineer, the post-change disturbance records, the demagnetisation record, and DGA results confirming no acetylene or rising ethylene trend.