SEL 587 Event Report 1

[rockman7892]

We have a SEL 587 relay being used as differential protection on a 5kV feeder between two 5kV tie-breakers. Tie-Breakers were tied through and in operation (only one source / not parallee) and we were running for about 15hrs or so and then all of the sudden the tie-breakers tripped. The tie breakers tripped due to an 87 differential pickup on phase A and C on the 587 relay. Attached is the event file from the 587 for this event.

From looking at this is appears that the phase A and C CT's are swapped for their phase angles are not 180deg apart. Phase B looks o.k. and the Phase A and Phase C inputs should be swapped in order for phasor current to sum to zero. Should this just be a matter of swapping phase A and C CT input on one set of CT windings?

The O87P setting in the relay is 0.1 and the CT's are 3000:5 CT's, therefore I believe differential pickup is 300A. My guess is that for the first 15hrs the phasor sum on phases A and B were below this 0.1 pickup and then all of the sudden a load or bunch of loads came on to cause the current phasor sum on A and C to go above .1 and thus trigger the 87 pickup.

When looking at the event file does IR corrospond to some restraint current?

I am trying to do the vector addition myself to see exactly what is going on with current phasors. When I add the IBW1 and IBW2 winding phasors I get a sum of 0 and when I add the IA's and IC's I come up with a value of .278 @ -75.4deg using the phasor magnitude and angles at .5 cycles. Is there a value or phasor in the event file that shows the phsor sums for each set of windings in order to verify my calculation.

 

[davidbeach]

IRW1 and IRW2 are the residual (ground) currents. Something real happened, not just a trip on load, otherwise the residual currents wouldn't be showing up. But you are also correct about A and C not being wired correctly. If you have an ABC system rotation the problem is at the winding 2 end; if you have an ACB (CBA) system rotation the problem is at the winding 1 end.

 

[rockman7892]

Thanks Davidbeach

The distribution setup is as follows:

We have two main utility transformers with the secondaries being wye 4.16kV. Each of these secondaries feeds a 4.16kV main switchgear lineup. If one of the utility transformers is down, we can close the tie-breaker to power both main switchgears off one transformer. This was the case when this even occured after the tie-scenario had been in place for 15hrs. After this event we went back to serving each main switchgear seperately off its respective transformer.

If you are saying that something "real" occured, then why would we not see it when we came back up with both buses and same loads on seperate transformers? Is it possible that the something real could be a fault in the cables connecting the two tie-breakers?

I see what you are saying about the residual currents showing up. They go as high as .1 to various points thorught the cycle. Could this be a function of incorrect CT wiring or does this actually indicate a fault current of some sort?

why would the IRW2 current go negative for a cycle then go positive before returning to negative again while the IRW1 current stayed negative the whole time? Both have magnitudes of .1.

 

[davidbeach]

Some of what is being seen in the residual could be an artifact of event report being 4 sample per cycle filtered data. Would be interesting to see the raw data. Have you reenergized the tie, or tested the tie? May not be anything of real concern, but it might. I'd have the line meggered and the CT wiring corrected as a minimum.

 

[rockman7892]

Davidbeach

No we have not re-energized the tie after this event, both main switchgears are being powered by their respective transformers. We need to energize this tie ASAP so that one of the utility transformers can be taken out of service so work on. I am hesitant to re-energize this tie until I am comfortable I know what happend in this event weather it is a nusiance trip or not, becasue if and when the tie trips it takes down one whole plant.

Is there a way to get and view the raw data that you mention?

As far as testing the tie, it was High-Potted about a year and a half ago and has been sitting de-energized ever since. Hi-pot results at the time seemed to look o.k.

 

[stevenal]

Two suggestions:

Always use the MET DIF command after the initial (small) load is applied. This will show you the operating currents, which should be near zero.

To correct the polarity issue, go back to your drawings first. Make sure they are correct before ensuring the wiring matches the drawings. If you simply start swapping polarities, you risk getting further out of sync with the drawings.

 

[rockman7892]

I'm thinking that the problem with the polarities is that both of the main switchgears come in with a different phase rotation from their respective utility transformers (dont ask me why). I am going to verify this. If I recall some time ago when we first tested the tie, the phase rotation was differnt on one main when energized through the tie then it was when energized from its transformer. To correct this I belive the cables between the two tie-breakers were rolled. This can explain the polarity issue. I will choose one phase rotation scheme and set the CT's to match.

As far as the zero sequence or residual current, could this be a result of charging current? The cables between ties are (9) sets of 1000MCM and are about 100-150ft. Seems kind of a short distance for for charging current to play a major factor.

 

[davidbeach]

If the residual were charging current it should be there from the very beginning and not appear when the load current went up. If the cables pass a quick megger test I wouldn't worry too much as it wasn't much residual. It is imperative though that you get the CT wiring problem figured out.

 

[rockman7892]

David

I see where the error is on the CT wiring and will definitely fix this problem.

In regards to the charging current it looks like you are saying that the zero sequence current appears to only show up at about 1.3 cycles about which time the load increases. If it was indeed charging current you are saying that it would have been there from cycle 0.

As far as the magnitude of the residual current it looks from this it had a secondary value of 0.1. Wouldn't this equate to about 300A with these 3000:5 CT's. I feel like I am incorrect on this magnitude since you are saying its not much to worry about.

The other thing that I find interesting on the attached curves is that the A and C phasors (incorrect wiring) dont appear to be exactly in phase with each other. Could the zero sequence current cause this, or is this a result of the A and C CT's being swapped?

 

[rockman7892]

I looked at the A and C phasors again and figured out the phase shift I mentioned above.

I do notice however that when the A and C phases are corrected the IA angle difference will be about 11 degrees off 180 and the IC angle difference will be about 16 degrees off of 180. I'm not sure that these would cause that much of a problem as far as a differential? What would cause these slight shifts?