Contactor explosion - GE Limitamp IC 2814

[schwalbe262]

Need ideas as to how this contactor blew up. GE Limitamp IC 2814 circa 1977. No operating data (jogging etc) available other than weather was clear. Services 2400V 450 HP pump motor. Coil is original AC type and appears to be in good shape though resistance is low (0.1 ohm). Were explosions like this the reason GE converted their coils to DC operation later? As you can see it blew a hole through the door and not much left of the guts to pinpoint failure. Any ideas as to history of failures and/or reasons would be appreciated. Thanks.

 

[dpc]

It's going to be really hard to come up with a definitive cause.

Is the system solidly grounded or grounded via an impedance?

My best guess is that something (tracking, dirt, insulation failure, rodent) initiated an arcing ground fault that escalated to a three-phase arcing fault.

What cleared the fault?

 

[Mobius44]

If your looking for a responsible party to obtain compensation, you'll want to retain a good engineering consultant to review the evidence, and not use guesses from an internet forum.

That all said, I sincerely hope no-one was hurt by the incident, and here's my best guess of what happened:

First, I think you had a contactor meltdown, which is closer to burning than exploding.

One cause of contactor meltdown is it's contacts arcing excessively while making or breaking. The extra arcing can occur for several reasons, I list 3 of the most common reasons I have encountered below:

1) weak coil voltage
2) mechanically the contacts don't pull in and release quickly enough due to binding or dirt in the mechanism
3) contacts are extremely "pitted" and have exceeded useful life

I personally expect #3 in this application, because of the age of the contactor. However, if regular electrical maintainance has been performed which includes cleaning the assembly and replacing it's contacts, then I'd vote for #1.

I bet with all the experts on this forum we can find a hundred other possible causes, anybody else care to guess?

 

[itsmoked]

That little a space, that high a voltage, an arc formed and ran for a while before something tripped. Talk about arc flash!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Mobius44]

You can forget my reply about contacts arcing, I was thinking low voltage because I overlooked the extra 0 on the end of 2400V. Medium voltage vacuum contactors won't have contact arcing.

 

[Laplacian]

Can't see much of what's left of the contactor from that picture. The truck looks like from an air-break contactor; is it vacuum or air break? There are several failure mechanisms for each type; some the same and some different. Did this fail while closing, while opening, while closed, or while open? What was the duty cycle of the contactor; how many starts per month?

 

[dpc]

I don't know that this was a vacuum contactor. Looks like maybe the remains of arc chutes in the back of the cubicle.

When was the last time these were cleaned or tested?

 

[marks1080]

From looking at the picture, it looks like this is damage from an extended burn (most likely from an arc). If this were an explosion I would expect to see the panel walls and cover pushed and bent outwards. Looks like the hole in the panel door was melted through.

If this was an arc, there are million and one things that could have caused it.

 

[jraef]

Must have been one heck of an arc. I would not have wanted to be in front of that cubicle.

That was an air-break contactor on a roll out truck assembly. Here is what it was. Maunal link for GE HV contactors

You will likely never know the exact cause, but the most common mistake on air-break contactors is a lack of maintenance. Dust / dirt / contaminants build up, temperature changes allow condensation, tracking occurs then flashover.

JRaef.com
Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[ScottyUK]

Once that board is re-energised, consider performing a PD test on it before putting people at risk working on or near it. There are a few instruments available for this kind of test which are hand-held and simple enough for a non-specialist to get a broad good-bad indication of the state of the board. EA Technology in the UK make one called the UltraTEV. Ideally an offline PD test would be better because it is more sensitive, but it is often tricky getting a major HV board released for testing. This test will pick up insulation failure at an early stage before it develops into a major problem.

Presumably the whole board was stripped and cleaned? There will be a lot of carbon-rich deposits over the busbars, insulators and any bushings which could easily provoke a repeat incident.

If the fault was able to persist at a relatively low energy level for a long time, as seems likely from the photos, you might want to investigate your protection scheme a little further. Is the neutral impedance grounded? This would account for a relatively low current arc being able to persist, but would call into question the earth fault detection.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[schwalbe262]

Thanks for all the input!! Now let me answer the ones I can at this point. Grounding impedence unknown, buried grid construction. It is an air-break and was operated about once per day for up to 10 hours duration. It failed upon closing. Annual cleaning/visual inspection. Humidity was between 46 and 55% with temperature around 77 deg. inside climate controlled building. Nobody (including snakes and rats) was inside at the time, whew! Any thoughts on coil degradation due to being AC? They converted the remaining pump contactors to DC.

 

[rybrk]

generally power contactors will burn due to carbon,cobwebs and dust deposited on contacts tips,lackof preventive maintenance of contactors explosin will happen.


rybrk

 

[Laplacian]

One problem we experience with this vintage of Limitamp contactors with AC coils is that one of the shading coils will fall off the main coil. This will cause the contactor to chatter. One of our armatures chattered so bad and so long that every horizontal surface in the entire lineup of motor starters (10 units) was heavily coated in iron ferrite dust. The interlock linkage was worn from the vibrations to the point the disconnect switch would not operate. The contactor did not fail though.

The contacts on yours may have been degraded to a point where a poor connection was established on closure, consequently overheating and catastrophically failing.

 

[dpc]

I don't see much reason to suspect the coil. Have you talked with GE Service?

 

[swgrmfg]

I would start with the question: when was the last time PM was done, and where is the data?

 

[manning225]

I have come to this thread a bit late so I can't really add much but picture is spectacular and underlines the importance of regular condition assessment of any HV switch. The HSE publishes an excellent guide "Keeping electrical switchgear safe". Although it is aimed at larger supply switchgear the same principle would apply for one of these small 2400V motor contactors. The HSE doument is not law, but your best defence in case of an accident is to make sure you were following it 100%!

 

[swn1]

DC coils on limitamps give less maintenance trouble because, as mentioned, the shading coils on the AC units are not reliable. But it is very unlikely they coil caused this burndown.

A fairly common problem with these is degradation of the arc chute ceramic plates. I found one once with a chunk of ceramic wedged in the contact gap. It could easily have ended up looking like yours.

A contactor in good condition can fail if it is asked to interupt excessive current, like "bumping" a motor or having a trip during start, if the starting current exceeds the interrupting rating. You have not said what cleared the fault, but it obviously escallated past the contactor's fuses. Was there any trip indication on the motor controls?