MV Protection Trip

[alingstone]

A 33kV feeder of ours connects to a remote substation B with a single 33/11kV Dyn11 transformer connected at B, and an 11kV distribution board at B. The relay on the 33kV feede CB at Substation A has Overcurrent & Earth Fault protection. There was a phase to phase fault and the 11kV CB at substation B opened on OC prot. About 2 ms AFTER the correct opening of the 11kV CB at sub B, the 33kV feeder at sub A tripped on 33kV earth fault prot, which we find strange. The 11kV feeder which faulted runs for about 2 miles on conjoint poles carying the incoming 33kV supply with the 11kV underbuilt bu about 2.5 metres. Any ideas as to why this may have happenned?

 

[racobb]

Did they trip on inst or time? If the inst trip on the 33 kV breaker could "see" the fault, it had to trip. Also check your device TCC curves. Phase to phase on the 11 kV side will decrease the time margin between the two device curves.

Also, did it happen on the first trip of the 11 kV relay or was it possibly after some reclosing operations?

Alan

 

[rbulsara]

The 2ms difference indicates that both breakers started opening at the same time, just that other took 2ms longer to trip. The trip action has already taken place. ( Also how did you synchronize the clocks in the two relays?, but that is a different question)

It probably has to do with the sensitive E/F setting on 33 kV side and/or also the type of E/F scheme employed. Not enough info. Harmonics during a fault can fool a realy to think that there is unbalance between phase currents, showing up as E/F. Difficult to say without any waveform captures.



Rafiq Bulsara

http://www.srengineersct.com

 

[davidbeach]

The phase-phase fault on the 11kV system released enough hot gasses/plasma that rose into the 33kV and provided a path for one of the 33kV phases to flash over. The 33kV ground relay did not see the 11kV fault, but responded to a 33kV fault caused by the 11kV fault.

 

[slavag]

Hi.
Please add information.
What are a current amd time setting on the 11kV side OC function?
What is a current and time setting of the 33kV side EF function?

What is a tipe connection to EF input ? dedicated toroid, Holmogen connection.

It's known result of phase to phase fault on the Y ground side.

Best Regards.
Slava

 

[alingstone]

33 kV OC set to 330A (33 kV) with a 0.4 TMS Standard Inverse, 33 kV EF set to 26 A at 1 sec DT.
11 kV OC Set to 210A (11 kV) with a 0.15 TMS Standard Inverse. The 33 kV HV CB on the 33kV side of the Dyn11 transformer is set to 20 A EF and it didn't see a thing, so i'm leaning towards some sort of mechanical interaction.

All CTs measure the neutral spill of the CT circuits for EF, no fancy coils etc.

 

[ScottyUK]

Outside chance: winding movement caused by the through fault resulted in a primary earth fault? It would show up in the transformer DGA if this had occured inside the tank.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[davidbeach]

If I'm correct in my 1 Jul 09 9:59 post then the 33kV relay at the transformer would not have seen the ground fault. The fault current would have had to been such that the 11kV overcurrent relay took about 1 sec to clear the fault. Had it been much faster the fault would have been gone and the 33kV would not have tripped.

 

[slavag]

What David saied.
Are you have DR file of event.
11kV OC is IDMT, EF of 33kV is DT with low current setting.
You must check time grading.

 

[davidbeach]

I wouldn't worry about time grading or coordination as I would call it. Any fault that involves only the 11kV will never be seen by the 33kV ground relay. The only way to get there is two independent but simultaneous faults (how likely is that) or a single fault that involves both systems - a "cross country" fault. If I'm right, both systems worked as they should have for that fault, a fault the involved both systems.

On the other hand, I'd never want to set primary protection at 1 second. Both systems could probably be sped up.

 

[alingstone]

Thanks for all your help, it remains a mystery. ScottyUK, if there had been a movement of the winding to earth then our transformer dif prot would have operated and it didn't. 11kV relay operated in 0.45 secs as the fault current was around 2kA (do the maths guys). So the 33kV earth fault trip must have been completely independant of the phase trip on the 11kV electrically. I'm happy with my grading. DaVIDBEACH I'M AGREEIN WITH YOU, but can't speed up times as there are so many downstream relays to consider etc. When we upgrade to 66kV on that line i'm putting in line dif prot, cheers.

 

[odlanor]

A phase-to-phase on a wye side of a Dy transformer has somewhat the appearance of a phase-to-ground on the delta. At 33 kV there is one large current relative to two smaller currents,which is characteristics of phase-to ground faults , but THERE IS NO zero sequence current.
I dont know the conclusion what may have happened...

 

[racobb]

OOPS...Missed the fact that it was a 33 kV EF trip initially!

Alan

 

[slavag]

odlanor!!!!
Exectly. It's not zero sequency, it's some kind of unbalance
and reasidual or Holmogren connection see it as ground-fault.

 

[davidbeach]

No, no, no. There is nothing that can happen on the wye side of a delta-wye transformer that will appear to be a phase-ground fault on the delta side. Nothing. I don't know what a Holmogren connection is, but a residual connection only measures zero-sequence, all positive- and all negative-sequence currents are canceled. The 33kV ground relay, if it is truly a ground relay, did not see the 11kV fault. End of story.

On the other hand, if the 33kV relay is actually a negative-sequence relay looking at imbalance, that would be a different matter.

The 11kV tripping in 0.45 sec (still way too long for something that close to the 11kV source) then the hot gas from that fault established the ground fault and the ground fault was able to continue without the support of the 11kV fault.

 

[slavag]

David. I know it's not seems, but:
Ia+Ib+Ic=0, if Ia+Ib+Ic not =0 you have ground-fault trip, and not important what is a reason:
really ground fault, unbalance, inrush or disconnected CT wiring. If you have dedicated toroid, of course no trip, but residual connection it is something other.
And two phase fault on the wye side is unbalnce voltage on the delta side and as result of unbalance voltage-unbalance current.
BTW, what is 0.45sek trip from/to, alingstone , are you have event list, DR fail.

Best Regards.
Slava

 

[davidbeach]

Sorry slavag, there is nothing you can do on the wye side of a delta wye transformer such that a residual connection on the delta side will see any current. Obviously a disconnected CT will cause current in the residual because in that case the residual isn't a sum of all three phases.

But any time you have the sum of all three phases flowing into a delta winding, and there is no fault between the CTs and the transformer, the only residual current you can ever see will be due to CT mismatch. Never from the other side of the transformer.

 

[odlanor]

alingstone.
What is the power of 33/11kV transformer and 33kV CT ratio?

 

[rbulsara]

davidbeach:
could harmonics or CT saturation play a role?

Rafiq Bulsara

http://www.srengineersct.com

 

[davidbeach]

Maybe, but probably not. More than once we've had faults on distribution circuits underbuilt with transmission lines and had the transmission lines trip for the distribution fault due to flash over caused by the products of the distribution fault. It just surprises me that people are looking for more exotic explanations for something that is just a "feature" of hanging both distribution and transmission on the same poles.