Differential Trip on Transformer

[anggap]

Recently, one of my oil filled transformer 13.8/4.16 kV, 5 MVA tripped by 87T (GE SR-745 Relay). This transformer worked in parallel with another identical 5 MVA transformer.
Relay's event record and phasors strongly indicated that there was single phase to ground fault in phase B LV side (attached below).
But insulation resistance test on the transformers and cables, WRM, TTR and DGA analysis shown no abnormality,

Is it possible that fault on transformer but the electrical and DGA test still shown good result? Thanks in advance

 

[davidbeach]

335A doesn’t look like fault current. Looks like something didn’t switch properly; either only one leg made when switching in or one leg failed to open when switching out. Check all of the compensation factors in the relay settings.

 

[anggap]

Thanks for your response davidbeach

Neutral connection of LV winding connected to a 6 ohm Grounding Resistor. Based on calculation, the maximum fault current (single phase to ground fault) on LV winding should be around 400 A. Prior trip there wass no load under this transformer, it only supplied 4.16 kV busbar.

 

[jghrist]

Actually, to me it looks like an external ØB ground fault, assuming that this is a delta (winding 1) -wye (winding 2) transformer with a source on the delta side. The ØB fault on the wye side would result in line currents on ØA and ØB on the delta side. If it were an internal fault with a source on the delta side only, there would be no current in the winding 2 CTs. I agree with David that you should check the compensation factors.

There appears to be a lot of winding 2 CT saturation that could have caused an operation, but the relay output didn't occur until after the saturation stopped.

 

[davidbeach]

My comment that it didn't look like fault fault current was based on seeing higher load current on that phase at an earlier time. The fact that the transformer is grounded through a resistor should have been mentioned in the original post.

 

[anggap]

I attached the drawing below for the details, apologize I didn't attach it earlier.
The CT's are located inside 13.8 kV and 4.16 kV switchgear. Previously, I suspected that the fault located on phase-B LV cable. But the insulation resistance test shown good result for both HV and LV cable.

How to check the compensation factor in GE SR-745 relay?

 

[stevenal]

I'm seeing a polarity problem. For the external b to ground fault jghrist described, I believe the two B phase currents should be 180 degrees out of phase rather than in phase as shown. Perhaps the compensation is correct, but the wiring is wrong.

I note the primary 13.8 kV CT secondaries are grounded, but the 4.16 kV CT secondaries are not. Suggest grounding them.

 

[RRaghunath]

Guess it is a new transformer just commissioned and load feeders from 4.16kV switchboard are yet to be energised.
Hope you would have done differential protection stability test on the transformer as part of commissioning tests.
If yes, then as suggested above, please review the CT secondary star point earthing. The star point of 4.16kV CTs as well shall be earthed and that too in the panel where the 87T relay is located.
That is to say, if the 87T is located in 13.8kV panel, the 4.16kV CT star point shall also be earthed in 13.8kV panel and not in 4.16kV panel.
Hope you have ruled out the possibility of 4.16kV bus fault (You said all outgoing feeders from busbar are OFF).

Rompicherla Raghunath

 

[anggap]

Well noted to ground CT star point at 4.16 kV side.
This transformer and the protection relay has been in service for many years, and I've observed the differential current at full load condition. So I think there is nothing wrong with the wiring except the neutral poit of 4.16 kV side CT.

Analyzing the phasor, I think the fault should be located between transformer with breaker as illustrated in attached sketch. I would appreciate any comment on this

 

[jghrist]

Analyzing the phasor, I think the fault should be located between transformer with breaker as illustrated in attached sketch. I would appreciate any comment on this

What is the source of the IB2 coming from the 4.16 kV Swg? If the Winding 2 CTs are in the switchgear, then there will be no current in them for a fault between the transformer and the switchgear.

 

[anggap]

What is the source of the IB2 coming from the 4.16 kV Swg? If the Winding 2 CTs are in the switchgear, then there will be no current in them for a fault between the transformer and the switchgear.

This Transformer is paralleled with another 5 MVA transformer, IB2 from 4.16 kV should be supplied from another transformer. Attached below the illustration.

*Note: I tried upload the picture in this post (using upload image feature) but always failed.

 

[stevenal]

anggap,

I agree with your diagram for a through fault, assuming there are sources on both sides. Draw in the differential CTs next. If CT polarities are all pointing away from the transformer as indicated in your 3 line, then the two B phases should be 180 out. Magnitudes looks okay for a through fault, but the angle does not.

 

[anggap]

anggap,

I agree with your diagram for a through fault, assuming there are sources on both sides. Draw in the differential CTs next. If CT polarities are all pointing away from the transformer as indicated in your 3 line, then the two B phases should be 180 out. Magnitudes looks okay for a through fault, but the angle does not.

Stevenal, looking at the phasor, is it safe to say that the fault should be located inside the protection zone?

 

[stevenal]

Yes, the fault being fed by two identical sources to account for the magnitudes matching the ratios.

Are the CTs external to the transformer? The three line indicates they are. What's the nature of the circuit between the L2 transformer terminal and the CT on that phase? Could the fault have occurred here?

 

[anggap]

Yes, the CTs are located in 13.8 kV and 4.16 kV switchgear. Transformer HV and LV terminals are connected to switchgear through CLX cable - XLPE insulation (cable length approximately 100m each). The cables installed on cable tray.

Yes, I think the fault could have occurred on the cable. I have checked the insulation using 5kV insulation tester but it shown good result.

 

[stevenal]

Upon review, I see the 3 line and fault illustration diagrams are inconsistent in their CT polarities. I was using the 3 line in my earlier analysis.

Any exposed live parts at the terminals? Fully insulated?

 

[mgtrp]

Assuming that there's no parallel feed transformers or zero-sequence sources on the LV distribution network, the LV CTs should only see fault current for a faults that are downstream of them (not on the transformer). But, if the fault is downstream of the CTs, then the differential should not have picked up as this is outside of the protected zone. Therefore, the CTs are probably miswired or the compensation in the relay is wrong resulting in a false operation.

The actual fault might be on your cables or switchgear, but if you've tested those and they passed then likely not. The actual fault is probably somewhere on the LV distribution network.

Cheers,
mgtrp

 

[anggap]

Stevenal,
thanks for noticing the mistake. The correct one is 3 line diagram.
The terminals at 4.16 kV switchgear are not fully insulated. I have checked the terminal but no sign of fault/arcing. The only area that's not checked yet is CT area, wich located between 4.16 kV swg terminal with VCB, but the area can't be accessed without de-energizing the busbar (the 4.16 kV busbar still energized, fed from another transformer).
I'm gathering the evidence to convince the plant manager that it was actual fault occured inside the protection zone, because he insist that it was CT miswiring/ compensation error.

Mgtrp, there is identical transformer parallel with this transformer.