We have a transformer differential relay protection for a start-up transformer at a Nuclear Plant. The differential scheme for this transformer is achieved using GEH-2057 with BDD16 relays. The transformer is a 345 kV/13.8 kV wye-wye with a tertiary delta. We have a concern that we may in future have an open phase fault on the primary of the transformer with and without ground (Conductor falling to ground from transformer primary side). With regards to that I have a couple of questions.

i) Under what fault conditions (line to ground or line fault without ground) on the transformer primary side not within the zone of the differential scheme, would you expect the differential protection relay to trip. Is it possible for the differential scheme to cause the trip during line to ground fault on the primary 345 kV side of the transformer (not within the differential zone) ?

ii) Due to tertiary delta, on a ground fault on the primary side, the primary of the transformer see extremely high zero sequence currents but the secondary of the transformer almost sees none. Therefore, phase currents on the primary and secondary side of the transformer are much different (after normalizing with turns ratio). Would this trip the differential scheme.

iii) Further during the line to ground fault on the transformer primary the current direction will reverse so as to flow from transformer primary into ground. Would the differential relay consider the change in direction and result in tripping?

The relay is set with 25% slope tap setting.
TH=4.2 and TL=8.7

 

Amrit, please confirm the following conditions:

1) Grid-side winding (345kV) is wye-connected, and mid-pt is solidly grounded

2) Gen-side winding (13.8kV) is resistance or otherwise impedance-grounded

3) Teritiary winding (?kV) is delta-connected, but:
3a) its terminals are outside the tank
3b) it is unloaded
3c) if unloaded, 1 terminal is solidly grounded

4) The Grid and Gen-sides are equipped with a CB, not fuses

Regards,
Phil Corso

 

Phil,

Thanks for your help!

1. Grid side (Wye 345 kV) neutral is solidly grounded.

2. Gen side or Load side actually is 13.8 kV grounded through a resistor 2.0 ohms resulting in negligible zero sequence currents during fault in primary (Also due to tertiary delta)

3.Tertiary delta winding is internal and unloaded. No connection outside the tank.

4. Grid side and load side is equipped with CBs not fuses

The transformer is a start-up transformer for a Nuclear Plant. The transformer is the preferred supply for the safety buses in the plant. It is not the main transformer for the generator.

Thanks a lot,
Amrit

 

Amrit...

one last question; what is size of the xfmr?

Regards,
Phil

 

Phil,

The size of the transformer is 75 MVA, Core Type Construction, FOA

Thanks,
-Amrit

 

Amrit... First some general comments related to your open-phase concern:

A) During normal operation of the transformer, Phase-fault Differential-relays will only trip if the open-phase condition results in a substantial current-unbalance within the differential zone, i.e., primary to secondary CTs.

B) Similarly, during normal operation of the transformer, Ground-fault Differential-relays (if existing) will only trip if ‘enabled’ by detection of current in the neutral-to-ground connection… regardless of grounding method.

C) It is true that during energization there could be considerable inrush-current magnitude and unbalance (regardless of source, i.e., primary or secondary)! However, tripping is intentionally delayed for several cycles by the harmonic-detection element in the relay.

D) Is your concern about the recent NRC directive covering the Open-Phase Problem?

Regarding your specific questions related to unsymmetrical faults… before I can make meaningful comments I need a simple SLD and the protective relay wiring as currently installed! And, since both primary and secondary neutrals are involved, I need to the know their impact on Xo flow! Do you have a fault-study illustrating current distribution?

Phil ([email protected])

 

Phil,

Thanks a lot for your help and being this prompt. Yes this is one of the NRC concerns, which in turn is a consequence of Byron Open Phase event and subsequent open-phase analysis required for other plants.

Although, differential relay is not a big part of it. Its more for my curiosity and understanding as to why the transformer differential relay did not trip when Byron Nuclear plant had line to ground fault and the fault itself went undetected for long until the degraded voltage relay on the safety buses picked it up.

I should be able to provide you required material. I will work to simulate it and get back to you soon.

 

Amit... the proposed NRC "fix" for the 'Open-Phase' phenomenon appears to consider such faults while systems are "loaded."

Would you be interested in Open-Phase detection during initial energization of assets before they carry load?

Phil

 

Phil,

Not completely certain however, I would believe you would want to detect an open-phase during any of three conditions:

i) During safety injection signal.

ii) At times when the transformer is lightly loaded. For the main transformer feeding the plant loads through backfeed, it would be close to no load (1000 MVA capacity as compared to 10 or 15 MVA load)

iii) During loaded conditions.

-Amrit

 

Amrit... OK!

Then to summarize, neither phase or ground-fault differential protection can reliably detect open-phase conditions outside of the diff'l zone.

The rather simple method I would have proposed is Phase comparison.

Phil Corso

 

Phil,

Thanks a lot for your help. I agree with you. I was initially not able to evaluate the normalized phase currents on the primary and secondary of the transformer to be equal for open-phase to ground condition.

The cause for this was the high percentage of zero sequence currents on the transformer primary and much lower on the transformer secondary. However, as the the differential transformers CTs are Delta connected (for wye-wye transformer), the zero sequence currents can be ignored.

Finally, when I calculated the phase currents using only positive and negative sequence currents, they were evaluated to be equal on the primary and secondary (normalized).
This follows the same result as yours.

Do you agree with my reasoning?

Thanks again for all the help!!

-Amrit

 

Amrit...

Would you mind sending me a copy of your calcs?

Phil

 

Phil,

I am emailing you my excel sheet and wiring diagrams.

-Amrit

 

do you find a solution for this case yet?

 

Pandc,

you have to ask Amrit!

Phil Corso

 

Pandc,

Search Internet for an EPRI 2013 Report:

"Nuclear Maintenance Application Center:
Development and Analysis of an Open Phase Detection Scheme for Various Configurations of Auxiliary Transformers!"

Phil Corso

 

Pandc,

I just learned of a solution by ALSTOM. It uses an optically-coupled current-sensor to monitor the Xmfr's no-load current.

It was proven by a collaborative effort between ASTHOM and FP&L, in Florida.

Phil Corso

 

Yes many solutions are close to implementation stage for this problem. I know of three that are close to implementation. However, I do feel there is a big room for improvement in most of the solutions given I have worked closely with many of them.

Thanks,
Amrit