Need for Transformer Instantaneous Overcurrent
Need for Transformer Instantaneous Overcurrent
(OP)
I work for a utility that is reviewing its standard protection package for substation transformers at the moment.
At present, we provide each transformer with redundant differential relays from the same blue-box manufacturer. In each relay, we set timed overcurrent elements to back up the downstream bus, and high-set instantaneous overcurrent elements on the source side of the transformer. The instantaneous overcurrent is set above the maximum through-fault current for the transformer, so it will only operate for faults on the source side winding. The intent with the instantaneous overcurrent element is that, in conjunction with the Buchholz or fast gas relay, it provides an alternate algorithm to the differential elements that protects the majority of the transformer (we aren't as concerned about high-speed clearing of faults on the load-side winding as these have less impact on our transmission system).
My question is--are other companies (mainly utilities) still setting instantaneous overcurrents on transformers? On the one hand, the cost of setting them is negligible since the relay already includes the element and the calculation and testing is very straight-forward; on the other hand, transformer differential must be one of the protection elements that is least likely to fail to trip, and so adding an unnecessary protection element is just one more possible inadvertent trip. We've never had a phase overcurrent element misoperate (to my knowledge), but we have had issues with the occasional over-optimistically-set ground overcurrent elements.
Thoughts? Comments?
Thanks,
mgtrp
At present, we provide each transformer with redundant differential relays from the same blue-box manufacturer. In each relay, we set timed overcurrent elements to back up the downstream bus, and high-set instantaneous overcurrent elements on the source side of the transformer. The instantaneous overcurrent is set above the maximum through-fault current for the transformer, so it will only operate for faults on the source side winding. The intent with the instantaneous overcurrent element is that, in conjunction with the Buchholz or fast gas relay, it provides an alternate algorithm to the differential elements that protects the majority of the transformer (we aren't as concerned about high-speed clearing of faults on the load-side winding as these have less impact on our transmission system).
My question is--are other companies (mainly utilities) still setting instantaneous overcurrents on transformers? On the one hand, the cost of setting them is negligible since the relay already includes the element and the calculation and testing is very straight-forward; on the other hand, transformer differential must be one of the protection elements that is least likely to fail to trip, and so adding an unnecessary protection element is just one more possible inadvertent trip. We've never had a phase overcurrent element misoperate (to my knowledge), but we have had issues with the occasional over-optimistically-set ground overcurrent elements.
Thoughts? Comments?
Thanks,
mgtrp






RE: Need for Transformer Instantaneous Overcurrent
These days we mostly use 2 separate differential relays on all new installations, and wouldn't add an inst o/c trip to these relays.
RE: Need for Transformer Instantaneous Overcurrent
Ground elements maybe a little more of an issue, yes. But, if it's not a problem, and is of no additional cost, why are you interested in not using it?
And is the transformer not one of your biggest cost transmission elements?
RE: Need for Transformer Instantaneous Overcurrent
As mentioned, we do presently set these elements, but were curious as to whether industry thinking was going away from this.
Thanks,
mgtrp
RE: Need for Transformer Instantaneous Overcurrent
SEL-501 and 551 relays are pretty inexpensive.
RE: Need for Transformer Instantaneous Overcurrent
To counter this, 50 element is use with a much higher CTR. This 50 element would be faster than the differential in many cases, even though the differential was in theory faster.
So if you have low CTR's you want to keep the 50 element. But as a universal package, as I am expected to provide, I must include the 50 because some of the applications require it.
Does that help?