Transformer Primary Fuses Blown 2

[cslater]

A utility transformer at our site was brought online a few days ago and blew all three line side fuses after about 10 minutes. The transformer is a 67kV/12kV 10000 kVA unit. It was brought online unloaded and none of the secondary relays tripped.

The working theory is that either water got into the transformer, or there is some other short inside it.

I'm wondering if its more likely that the fuses weren't sized for inrush. I haven't heard yet how large the fuses are or if they're slow-blow.

Is it possible that the inrush could weaken them so that they would blow after 10 minutes unloaded? I've seen some mention of similar things on this forum but am not familiar enough with fuses to know if that applies here.

Are there other causes that we're not considering?

 

[davidbeach]

If the fuses lasted even seconds, let alone minutes, it wasn't inrush. Inrush is measured in cycles.

10MVA is rather large for fuses, a relay and a circuit switcher would have been a far better design, and the relay event report would provide all the information necessary to know what happened, but save a few penneys first cost and pay lots of dollars dealing with the "savings".

 

[Elmir]

Is this a new transformer installation or it is an existing transformer which was taken out of service for maintenance or a power shutdown?

If the unit was a new installation, was electrical testing, oil sample, inspection, and test of all auxiliary devices performed? Stardard commissioning procedures done by qualified persons followed to NETA Acceptance Testing Specification is recommended.

I have performed maintenance and commissioning on similar sized transformers protected by fuses at some of our customer sites. Test results are normally compared to factory test results or previous maintenance test results. Oil samples are sent to a lab for DGA and stanardard oil screen.

Testing after the blown fuse event will have to be completed to give some clue to the cause of the fault. How is the secondary side of the transformer connected; by cable or overhead line? Where is the secondary main feeder protetion located?

 

[cslater]

This is something that I'm not directly involved in, so my knowledge is spotty in places. Apologize for that.

I believe it's a new transformer. A reputable testing company came and tested it thoroughly before it was energized - everything passed. Oil is currently being sent for testing.

As far as I know, the secondary comes out underground and goes to a switchgear lineup with protective relays before going out to overhead 12kV lines.

 

[Zogzog]

If it was tested correctly my guess would be either that cable is bad or the primary fuses were incorrectly sized. Do you know what testing was done on the cable? More than a DC hipot I hope.

 

[oldfieldguy]

I'm thinking that if the time from energizing to fuse blowing was seconds, examine closely the cables and equipment connected to your secondary.

It sounds like a hard secondary fault. Current on these faults is limited by the transformer impedance, so it takes a bit of time for high side protection to operate. The most common failure is having crossed cables in multiple conductor per phase installations.

Ditto the DGA. If you have a good electrical testing company available have them do an immediate TCG (Total Combustible Gases)if the transformer has a nitrogen pad. If there was an internal fault, the TCG should show it immediately.

old field guy

 

[cslater]

It took approximately 10 minutes to blow the fuses. Also, there was no load on the secondary side - but I suppose the fault could have been between the transformer and the switchgear.

Can you point me to more information on what a TCG test is? I'm not familiar with that.

Thanks very much - everything so far has been very helpful!

 

[jghrist]

If all three phases blew, then a secondary cable fault is unlikely, unless it's something easily seen like grounding straps left on all three phases of the terminations.

 

[cslater]

Could water in the oil cause all three fuses to go at once? It seems like the failure would be more likely across one coil which should only blow 1 or 2 of the fuses, right?

 

[davidbeach]

There are two ways to get all three fuses "at once".

One is a hard three phase fault; fault to fuse clearing in less than a cycle or two. The other is that there was enough leakage to ground from all three phases that each fuse saw and cleared the fault independently around the same time. If the fuses really took 10 minutes to clear, you would had to have had fantastically well matched fuses, otherwise you would not have gotten more than two of them.

 

[cslater]

Just heard that the oil tests came back and there's no sign of anything in there.

I got a little more information that I hadn't heard before. This transformer was actually put in place a year ago and has run fine for that whole time. A few weeks ago a raccoon got into the substation and caused a fault, which blew out a bushing.

The bushing was replaced, along with a lightning arrestor on the secondary side.

The transformer was re-energized, with no load and - as stated before - all three primary fuses blew after 5-10 minutes.

This is not a transformer that I'm responsible for, but as an engineer it has me really curious.

 

[willhammett]

Probably a transformer problem. You can rule out the secondary side as previously stated unless a fault occurred across the switchgear(highly unlikely). If there is no load, the fuses would have no purpose if there is no thru put(unless the previously stated scenario had occurred). Although I would agree it could be a 3-phase fault as previously stated. Probably a winding fault in the transformer, or a fault somewhere along the line between the transformer and the switch gear.

 

[waross]

Or it may have been Charlie Raccoon looking for his friend Rocky.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[PHovnanian]

Like davidbeach said, three fuses clearing all at once after 10 minutes is a bit of a freak occurrence. A hard fault seems more likely. But it would have to be something that 'cooked' for 10 minutes before failing. I'd take a look at the secondary lightning arresters. Particularly the one that was replaced.

 

[racobb]

Do you know what fuses blew from the initial racoon?

Alan

 

[cslater]

racobb - That's a good question. I believe all three fuses were replaced, but I would have to confirm that.

PHovnanian - I wondered about the lightning arresters too. But if only one lightning arrester was replaced, how would that kill the other phases? Is there some way for a ground fault to put a high enough voltage on the ground buss and somehow transfer that upstream to the other phases?

willhammett - The current working theory is a winding fault in the transformer that didn't short until everything had warmed up a bit.

 

[oldfieldguy]

cslater-

"TCG" is Total Combustible Gases", typically performed in the field with a portable test set. It takes a sample from the transformer's gas space and checks it for combustible gases. Typical "good" transformers read less than 0.5% (IIRC, It's been a while) and transformers with internal problems will practically peg the meter.

In my field service days, the first test I did when I got called about a transformer coming offline was a TCG test. It's fast, almost instantaneous, and 24 hours faster than a 'rush' oil sample.

I missed the ten minute time you referenced. That colors my initial assessment, but since your oil tests (I'm assuming that includes dissolved gas analysis) show good, we're still looking at something on the secondary side of the transformer. The transformer impedance limits current and the fuses may have long melt times at low currents.

old field guy

 

[waross]

Another animal exploring the bushings could take out all three fuses. You have prior history!

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[PHovnanian]

But if only one lightning arrester was replaced, how would that kill the other phases?

The initial single phase to ground fault (one bad arrester) could cause a voltage imbalance that would cause a second marginal arrester to trigger. Or for a L-G fault on the outside of an arrester or bushing, the arc could propogate to involve another phase or two.

 

[PHovnanian]

Another animal exploring the bushings could take out all three fuses.

(Sniff, sniff.) Fried raccoon. Mmmmmmm!