SEL 587 and Transformer Differential 5

[Buckeye46]

A little too eager to wait until Monday when the protection engineers get back to work so I figured I would see what this forum knows about an issue we had on Friday with a 230/60kV transformer. It is brand new, has been in service for over two weeks through high and low load. Friday the generation plant that's off of this substation opens a remote breaker that forces all load to go into my substation and thru parallel banks, one of them being the one in question. As soon as that occurred, the SEL 587 tripped on an 87A. I am attaching images of the event reports. They don't make any sense. I could type an essay in here but I guess I will see if anything jumps out to you guys as to why this occurred. The primary CT waveform does look like its almost clipping.. but CT saturation shouldn't be an issue since it wasn't near its rated capacity by any means. Nor was it fault level current. The restraint and operate currents are pegged at 0.5 which makes no sense to me. Any ideas?

 

[davidbeach]

Could you post the actual event file?

 

[Buckeye46]

yes give me 5 minutes i will post everything I have.

 

[Buckeye46]

I am not too familiar with sending files and not images over a forum so if this isnt good for you let me know a better way

https://drive.google.com/open?id=1OIxARHBySAYfS_7W94pl9dzkqTvj41J9

 

[HamburgerHelper]

I would bet on sympathetic inrush. The opening of their breaker produced a voltage step change that created a dc offset that saturated your transformers CTs and caused the differential to operate. The waveform looks asymmetrical, indicating dc saturation (maybe 2nd harmonic) and not AC and it appears to be pulling out of saturation as the dc componet decays.

------------------------------------------------------------------------------------------
If you can't explain it to a six year old, you don't understand it yourself.

 

[Buckeye46]

The issue I have is this can't happen every time they open the remote breaker whether for a requested outage or to add SF6 gas. The 587 is set very reasonably with a 0.3 pu and 40% slope 1 with slope 2 set to off. 2nd harmonic blocking is 15%

 

[RRaghunath]

Is 2nd harmonic set for
# cross blocking and also is set to
# block as long as the I2h lasts (>15%).
I am not familiar with SEL relays. But, I have seen the above factors leading to relay maloperation in other make relays.

Rompicherla Raghunath

 

[Buckeye46]

I am the commissioning engineer for the transformer, hence why I am eager to discover the issue since everyone likes to point the blame on me missing something during commissioning, so I only know enough protection knowledge to be deadly. I know there is a way to view harmonics thru the raw event files, that is just where I seem to get a little behind. I am attaching two images, one of the differential settings in the 587 and the other a snip of an event report that shows the harmonic blocking asserting. It seems like it only picked up after the 87 tripped.

 

[davidbeach]

There's a pretty fundamental error in the settings - the relay is set for ABC phase rotation and the system is ACB rotation. The internal compensation is going the wrong direction; getting operate currents equal to restraint current points to something pretty significant.

 

[Buckeye46]

This is something I have been harping on as well for the past few days... You are correct the PHROT in the settings is set for ABC and not for ACB. Someone managed to tell me it wasn't an issue and I had a hard time believing that setting was compensating for the phase offset. Where I made a big mistake was during load checks I did not log into the relay and calculate the mismatch. I would of see this right away and knew something was wrong. But how in the world could this not have operated over the course of two weeks????? It just didn't see enough load?

 

[davidbeach]

Well, "Someone" was wrong. The compensations for the delta need to use the correct phases and they're not presently.

It's just like telling the relay that it has a YDAB connection when it really has a YDAC connection, would "Someone" have also said that wasn't a problem? That's basically what the relay is seeing, you could change TRCON to YDAC and I think it would work correctly, but then there would be two mistakes in the settings instead of none.

It's nearly impossible to figure out why the trip happened when it did. The 587 event reports are in secondary current with a 100mA resolution, that means that the event report has steps of 40A primary and it's impossible to find the nuances.

I find it very surprising that anybody would use that relay for a transformer of that size. Given the other options, I wouldn't use it at all; used to be that it was a decent relay for small transformers but now the 787 fills that bill very well and provides far more information about what's going on. For that transformer though, the only SEL relay I'd consider is the 487E. It's definitely more expensive than the 587 but both are just rounding errors compared to the total cost of installing that transformer.

 

[Buckeye46]

We are using a 487E as well as a 587. Currently the transformer is protected by 2 421's for high and low side lead protection, the 587 for primary differential, 2 551's for neutral protection on the transformer and ground bank, and the 487E for an overall differential. Thanks for your help David. I was on this path but didn't want to inform protection engineers of their mistakes until I was certain. They can get somewhat egotistical occasionally haha

 

[davidbeach]

Wow, sounds like a mess. Once upon a time we might have done something similar, but certainly wouldn't today. The 587 and both 551s could be replaced by a single 487E and have total redundancy for everything.

 

[Buckeye46]

The issue is I work for a larger company, so its a lot of people stuck in their ways and if you aren't near the headquarters you aren't going to change that anytime soon. I always get the Swiss cheese analogy with protection redundancy. More slices from different blocks of cheese the less likely anything falls thru haha. Some of the senior guys out east even want to throw in GE relays...

 

[HamburgerHelper]

Sounds like someone had the bright idea of replacing each mechanical relay with a microprocessor relay. I would like to think it was an intern. That or someone is trying to pump up their capital investment.

------------------------------------------------------------------------------------------
If you can't explain it to a six year old, you don't understand it yourself.

 

[Mbrooke]

Dumb question- brain fart- how do you open the file in post #4?

 

[JG2828]

Dumb question- brain fart- how do you open the file in post #4?

7-Zip is a really nice opensource program for working with zip files. I like it much more than WinZip. 7-Zip also has their own archive format that uses the .7Z file extension. It is similar to .zip but has added features. One useful feature is when you encrypt a .7z file with a password, you can select if you want to also encrypt and hide the file names, .zip files do not let you hide the file names within the zip file.

https://www.7-zip.org

 

[Buckeye46]

Yeah 7-zip is the way to go. Once your in, to actually view .cev files, you will need acSELerator Analytic Assistant software. I believe without a license you can still view it with limited functionality

 

[Mbrooke]

Thanks to both- I will download 7zip and give it a try.

 

[marks1080]

Had something similar happen at a station I was working on... Protection a 75MVA unit.... normal ABC rotation...

When construction were building the thing they had some space/clearance issues and ended up swapping a couple phases with local P&C finding no reason to change the settings... needless to say the unit eventually tripped.