Substation Transformer Fire

[WorkSafe]

One of the large transformers in our substation failed early this morning. Many of of our buildings (3 million+ sqft industrial) are still without power. I don't know any specifics yet as to cause, but it made for a interesting day.

This is on Tinker AFB.

 

[chris kennedy]

Cool video, nothing taken closer?

 

[WorkSafe]

Cool video, nothing taken closer?

Hopefully I can get some better pics/video from our engineering tomorrow. The substation is probably a good 1/4-1/2 mile from where this video was taken. It's interesting because our facilities were in a scheduled outage around this time, so I wonder if this happened when they went to restore power.

 

[fmtjfw]

Took awhile for the power to go out.

 

[Sahib]

I don't know any specifics yet as to cause, but it made for a interesting day.

One sure cause (for such big flame) is the subject transfomer might be oil filled.

 

[GoldDigger]

But the "flames" died down immediately when the power finally cut off.

Tapatalk!

 

[don_resqcapt19]

But the "flames" died down immediately when the power finally cut off.

Tapatalk!

And makes a very important point...there is no way, other than stopping the flow of current, to extinguish an electrical fire. Fire extinguishing agents only put out the combustible materials that the electrical fire ignited, they do nothing to stop the actual electrical fire.

 

[sparkyrick]

Took awhile for the power to go out.

No kidding! I was surprised how long it took.

 

[Sahib]

But the "flames" died down immediately when the power finally cut off.Tapatalk!

Hmm. How can an arc flash, an explosion could exist that long?

 

[GoldDigger]

If there is no interrupting device on the source side that gets triggered, then the arc can continue as long as the "electrodes" at the ends of the arc survive.
There is a good chance that the eventual shutdown of the power source was a manual operation farther up the line.

Tapatalk!

 

[fmtjfw]

No kidding! I was surprised how long it took.

I'm guessing that the feed to the near building was from another source other than the failed equipment and that feed eventually burned down. Or that the relaying that finally cut the power off to the fire also cut the power to the building. Interesting.

 

[gadfly56]

Hmm. How can an arc flash, an explosion could exist that long?

Ever hear of carbon arc lamps?

 

[WorkSafe]

Still haven't gotten any new info on this incident. Hopefully in the coming weeks I should hear something. The substation (one of many on base) is owned by OG&E. I know at least 6 builds that lost power from it failing. It took them about 15 hours or so to restore power.

I know OG&E wanted it back online quickly because we pay them around $200,000/day in electrical costs.

 

[Sahib]

If there is no interrupting device on the source side that gets triggered, then the arc can continue as long as the "electrodes" at the ends of the arc survive.
There is a good chance that the eventual shutdown of the power source was a manual operation farther up the line.

Tapatalk!

No arc flash?
Only more or less steady arc fault?
It looks like a miracle for a steady arc to persist after an arc flash.

 

[GoldDigger]

Remember that the thing which usually terminates an arc flash event and stops it from being a continuous arc is the operation of an upstream protective device. In this case there was not an effective protective device feeding the substation, giving us an arc rather than a flash.
Clearly from the video the energy in the arc is changing in steps over time, probably as structure burned away.
Either the current finally hit the cutout level or manual action was taken.

Tapatalk!

 

[Sahib]

Remember that the thing which usually terminates an arc flash event and stops it from being a continuous arc is the operation of an upstream protective device. In this case there was not an effective protective device feeding the substation, giving us an arc rather than a flash.
Clearly from the video the energy in the arc is changing in steps over time, probably as structure burned away.
Either the current finally hit the cutout level or manual action was taken.

Tapatalk!

I think circuit impedance plays a role in making distinction between 'arcflash' arising, for example, out of a switchgear shortcircuit and steady arc as in a carbon arc lamp.

 

[iwire]

They need an AFCI breaker on that feeder ....

 

[Sahib]

Will AFCI breaker operate in a circuit with a carbon arc lamp?

 

[zbang]

Will AFCI breaker operate in a circuit with a carbon arc lamp?

By 'operate' do you mean that it will trip or that it won't trip?

Either way, arc lamps use either a ballast, transformer, or rectifier between the supply and the arc. If nothing else, that should filter most of the higher frequency components from getting back to the line, which would make the load appear more "normal" to the AFCI.

 

[meternerd]

Cool video...thanks. Seen that happen when station battery was dead or disabled. Usually also a huge fireball. Not saying that's what happened, but an arc that large for that long tells me something didn't trip that should have. No fire means the fault was probably ahead of the transformers, so there may not have been any high side relay protection at the sub. Most distribution substation relays or primary fuses are for transformer protection. Since work was being done during a shutdown, I'm voting for "Oops...forgot to hook that wire back up" or "I thought YOU took the grounds off!"

Let us know what happened....might save the rest of us relay techs a black eye. Pics also would be great.