Maloperation of Transformer Differential protection

[RRaghunath]

Differential protection of a 31.5MVA, 33/6.9kV, Dyn11 transformer has maloperated in a case. The relay is numerical type and the recorded currents are balanced (about 230A) on the 33kV side of the transformer and match with the b-phase current (about 1150A-after applying transformation ratio)on the 6.6kV. The a- and c- phase currents are exactly half of the b-phase current (about 575A). The differential current in a-phase has exceeded the set value and caused the trip.
The investigation revealed an open circuit in a-phase ct secondary wiring on 6.6kV of the transformer. Every thing has been checked and found in order.
How to correlate the recorded currents with the trip or is there some other reason for the trip.

Any help.
Thanks in anticipation.

 

[jghrist]

Sounds like the 6.9 kV side CTs are connected delta and the open circuit was in the delta connection (although I would expect a 1/sqrt(3) difference not a 1/2 difference in the phase currents). Otherwise, an open circuit would result in zero current on one phase. Of course you will have a trip if the current is 1/2 on the secondary but not on the primary. This is the differential current that differential current relays are supposed to detect. Or am I misinterpreting something?

 

[jbartos]

Suggestion: Probably, the CTs are connected in the delta connection that became open due to open circuit in A-C delta winding portion. This would leave two CT secondary windings tied together at B. Then, the Ib=Ia+Ic.
Ia=Ic, Ib=2Ia=2Ic, and Ia=Ic=Ib/2
The CTs are assumed to be on individual CT cores, i.e. there is not one core for all three CTs.

 

[hsj]

For modern electronic and numerical relays the CT secondaries are always star connected. The relay has inbuilt logic to account for phase shift.

The relay operated correctly because there is unbalance in current between primary and secondary side.

It is not clear where the current was measured. Presumably inside the relay i.e.by the relay. It appears that two phases are acting as return path for the current in the third phase. The relay is receiving input from only one phase and two phase inputs are missing. Recommend to check secondary wiring.

 

[Bung]

I wouldn't be as bold as to say CTs are "always star connected" for numeric diff relays - we have a mixture of both star connection with ratio and vector group correction in the relay, and also exteranl correction at the CT. Generally when we replace an old externally corrected e-m relay with a numeric, we don't touch the CT wiring - it's too expensive, easily double the cost and time of the job.

But a diff operation is consistent with a broken CT connection - it will lead to a difference between HV and LV as far as the relay is concerned.

Bung
Life is non-linear...