Transformer Differential Trips 1

[racobb]

Looking forward to input!

In summer 2008 we commissioned a 20/27/33 MVA, 115-25 kV, Delta-Wye substation transformer. Subsequently we have had 3 differential trip events on the unit and have yet to determine the cause. Except for thes ethree events the transformer has otherwise been in service, we have had faults on circuits with no differential issues.

First event - Fall of 2009
Don't remember the type of trip. Transformer mfg came in and found several of the stud connections on the feed through bushings very loose. Loose such that some of the ring terminals were not even hand tightened! Some burning on one feed through and it was replaced.

Tests
DGA for the oil
TTR for the transformer
Ratio for the CT's
Review of the diff settings
All tests were acceptable and the unit was returned to service with the thought that CT connections were the problem.

Second event - Fall of 2010
Restrained Diff trip on B phase and lockout of high side circuit switcher. Again mfg was called in. Found all CT connections tight on feed throughs but one exhibited severe burning and was replaced.

Tests Again
DGA for oil
TTR for the transformer
Ratio for the CT's
Review of diff settings
Review wiring for grounds/proper connections, etc.
Diff relay sent to mfg for complete testing....found ok
Diff relay mfg reviewed settings and found no issue

Third event - Fall of 2012
Another restrained trip on phase B and lockout. Diff trip and lockout of the high side circuit switcher for fault on distribution circuit. Owner opened all distribution breakers, reset lockout relay, and closed circuit switcher thus re-energizing transformer! YES a definite NO-NO! Then closed one un-faulted circuit and diff tripped and locked out station. This trip is assumed to be from inrush on a circuit that had approx. 50 to 75 amps of steady state load. Mfg called in again. No CT or feed through issues found.

Test again
DGA for oil
TTR for transformer
Ratio for CT's
Review wiring and settings for diff relay and again checked for inadvertent grounds
Diff relay mfg had 2 engineers review settings....no issues found
Excitation test of CT's this time....ok so far
Awaiting results of oil tests.

We have approx 5,000 amps of available fault duty at the transformer low side.

Differential relay is a Basler BE1-87T with no memory so event history is not available.
High side Ct's are 600/5 MRBCT set on 200 amp tap
Low side Ct's are 1200/5 MRBCT det on 1000 amp tap

Any Ideas....additional tests? I know there are more tests we could do for the health of the unit, but given the issues we have found with the Ct's over time, we have been concentrating on that. Short of shutting down the station in the months of Sept and Oct each year, don't know what else to do. If it is a false trip, it is the first one I have ever experienced.

Would love to get some input!

Thanks

Alan

 

[vandal06]

My first thought would be to replace the relay with a unit that has event capturing.

Regarding your latest trip, if it was inrush, there is a settings problem. I'm not familiar with the Basler unit but generally most are provided with harmonic blocking to restrain the differential elements on inrush.

 

[davidbeach]

I'd agree, need a relay with event capture ability. Pretty much no excuse left for buying relays that don't. If you want to stay with Basler, they have transformer relays with event capability as do many other vendors.

Differential on picking up a feeder circuit can mean one of two things.

The preferred possibility is that the relay is wired wrong and it tripped on a through "fault" where the fault was the cold load inrush.

The non-preferred possibility is that something is loose in the transformer and moves on inrush and you did actually have a differential event.

But in either case, you need a new relay.

 

[DTR2011]

I would suggest Primary Injection Testing of the protection scheme. You will need a 3 phase service (480V preferred), either via station service (separate feed) or a generator. You will also need a phase angle meter. When properly set up, the primary injection test will cause 3 phase current to flow through the transformer and thus the CTs feeding the differential relay. Once set up, the test only takes a few minutes. It will verify Ratio, Polarity and proper secondary wiring (phasing). By pulling the trips on the 86T, you can see where the problems are.

Please see the SEL paper "Commissioning of Protective Relay Systems" on the instructions and calculations required for the test.

Paper

 

[vandal06]

To expand on what David said, the newer relays with event targeting allow you to monitor in real time the operate and restraint quantities. Sometimes if the transformer is loaded lightly enough you won't know of a potential CT mis-wiring until it trips on an external event, like a feeder through fault David mentioned. If everything is wired correctly during load you will see predominantly restraint quantities with just a tiny bit of operate (due to CT errors, transformer magnetization current, and others). I've caught wiring mistakes or incorrectly set winding compensation settings that way.

 

[cranky108]

Years ago the company I was with had a simular problem with a BE1-87T. Never had a problem with any others of the same model. The relay was replaced by Basler engineer with a newer model relay, and we never had the problem again.

Not that I would blame Basler for an errant relay. We also had a problem with an SEL relay, getting stuck with a large value of current on one input of one phase. The event report showed nothing other than the large value of current on one input. The relay was sent back and SEL never could make the problem happen again.

Conclusion: Electronic devices can have hidden damage, and failures that can't be located. Replace them and move on.

Also reconigise some transformers may not produce enough 2nd harmonic noise to pickup the relay harmonic blocking.
Or more recently I found CT's with a low C curve set at half winding to feed EM differential relays.

Another conclusion: Sometimes the problem may not be the relay.

 

[racobb]

Thanks to all so far. I made myself a promise in about 2009 never to install a critical relay unless it had memory for an event/fault history with vector data. This unit has been loaded to approx. 5000 kW during its life so far. We have had numerous system disturbances with breaker trips to lockout and no diff trip. I know it seems strange, but the trips have occurred in late Sept or early Oct. in 3 out of 4 years....2009, 2010, and 2012. Sure seems odd to me!

vandal06.....giving your post some serious thought....Thanks Just don't understand why we can go 2 years with numerous system faults and not see a diff trip.

Alan

 

[racobb]

On an additional note, the single phase excitation tests went well today with no problems.

Alan

 

[dpc]

System faults may not create the right conditions for misoperation. A transformer through fault should create a misoperation if the CT wiring is incorrect.

If the max load has been 5 MVA on a 33 MVA unit, you might still be below the pickup for the restrained element at this load.

I'd still be suspicious of wiring errors until it's proven correct. There are so many ways to mis-wire a differential relay. I've tried them all, and verified that only one connection is correct.

But arcing on the CT leads seems like a bad thing.....

 

[pwrtran]

One incident we had that caused 87 misoperation was water problem inside transformer mech box during the wet season. The leakage current from the terminal blocks due to the water caused the Idiff then 87 trip. Your three incidents all happened in fall, how the weather like?

 

[norcalnewb]

Have you performed a load check on the secondary CT wiring to confirm expected inputs into the relay against a known quantity? We had two Basler 87T relays that operated fine with no trips for 3 or 4 years on two 35 MVA 69/13.8 kV transformers on a M-T-M substation. Then 2010, we had the substation single ended as we were doing some work on the 69 kV system and we had a transformer differential trip. We had operated the substation on a single main several times in the past with no issues, so we did all the transformer tests assumming a failure and found nothing. A full test of the relay showed no problems either. During all of this, we had of course put the sub on the other transformer. A day or two later, that relay tripped as well. Without data recording on the relays themselves, we were a bit stuck. We then started looking at all the wiring, including load checks on the transformer diff circuits. We found significant imbalances in the currents which ended up being caused by multiple grounds in the CT wiring (I will say the station was commissioned a few years before I started with the company). As I mentioned, this station operated for 3 -4 years without incident but had an issue the whole time. We did end up replacing the Baslers with another relay with data capture to help with future issues.

 

[racobb]

dpc....agree that there are many wrongs and only one right. To clarify, the system faults I referred to are on the distribution system fed by the transformer, so they are through faults.

pwrtran....cabinet heaters are working and all is clean and dry. Most recent trip was on a beautiful dry Sunday morning here in the southeast US.

norcalnewb....We have checked the wiring and grounds 2 or 3 times.

Aside from open or shorted turns in a ct, which should show up on ratio and/or excitation test, does anyone know another test to check a ct? I am concerned about damage from the original loose connections and subsequent arcing. I know we have some instrument transformer experts here. Ct's are all C400.

Thanks to all of you!

Alan

 

[stevenal]

Had a similar problem with a set of ABB 87Ts. It would be secure during energization of the protected unit, but trip during the inrush when downstream load was connected. Secure again when load was steady. Checked CTs, wiring, grounding, swapped draw-out relays; all with no joy. Replaced it with a microprocessor relay so we could at least look at the events, but no further events have occurred.

 

[norcalnewb]

Basler has a spreadsheet on their website that can be used to predict operate and restraint currents using measured CT currents. You might consider downloading and running it using measured CT secondary values.

 

[racobb]

norcalnewb,

We did use it prior to commissioning and have checked since. Have not input measured values though. Plan for now is to replace the relay and do another complete wiring check and testing.

Thanks

Alan

 

[oldfieldguy]

Are you absolutely sure you don't have a transformer issue?

One of my clients had an on-going problem with a transformer tripping at seemingly random times. The transformer, its CT's, its relays, all parts were checked by TWO separate testing companies. Nothing was found. Finally, the manufacturer was called in. They replicated the same tests, then after finding nothing, drained the oil, opened up the top, looked in, found nothing. Pulled a radiator and sent a flexible borescope into the lower regions of the transformer and found the culprit: A core ground wire had come loose and in certain conditions of oil flow (forced oil cooling) it would whip around and touch one of the windings. The transformer would trip off, the pumps would shut down, and the wire would fall back out of contact, so nothing showed up on the test.

Sometimes when you hear hoofbeats in the street and look out the window, it really IS zebras.

old field guy

 

[racobb]

No I am not completely sure. But for now the transformer mfg. stills owns it since it is still under extended warranty. The mfg. has blessed its return to service, but I want to make sure I have data the next time.....if and when! BTW it is ONAF.

Thanks!!!

Alan

 

[vandal06]

I'm glad you are getting a new relay. The event reporting capabilities will be of great value if/when the next trip occurs.

 

[zer0sequence]

One quick question: does this transformer feed any single phase loads? Is there proper zero sequence removal from the wye side winding for any SLG faults or heavy loads that may arise on a single phase? This zero sequence removal can be done through either settings or ct wiring.

 

[racobb]

This is in a substation that feeds outgoing utility distribution feeders with all combinations of loads. Zero sequence removal is done with the settings. I was reminded today that for 2 to 3 months the transformer had between 15 and 20+ MVA on it while another station was down for maintenance. Since that is not the standard configuration, I had forgotten and thought it had only 3 to 5 MVA load.

Thanks

Alan