Possible arc flash ???

[hlfireinspector]

Can someone enlighten me on what could have caused this arc on the door to this switchgear. This was found at a fire scene this morning. The machine that this switchgear goes to was not suspect to the fire cause but the machine was damaged in the fire. The fuses are not blown. If you zoom in the metal bar over the left fuse is damaged. The switch gear was still on but the main supplying it was off when I arrived. When I tried to move the lever to the off postion it was stuck.

 

[GoldDigger]

I think it is pretty clear from the picture that the arc was located either at the left hand switch contact or the upper left hand fuse clip.
Based on the handle being jammed and looking at the carbon pattern, I think it is slightly more likely that the arc was in the switch, but the initiating event might have been at least in part at the fuse clip. I think I see bright metal in the fuse area, where clean unoxidized copper is exposed.

It probably started with overheating because of a bad contact rather than an overload, followed by loss of pressure on the switch contact surfaces and a series arc across the contacts. The current would still be limited by the load on the panel at the time, but with a decent arc distance developing a voltage the energy given off by the sustained arc could be quite substantial.
The internal melting is probably what has jammed the switch in the closed position, but I bet that a continuity tester would see an open circuit across the left switch contacts.
If the fuse is welded to the upper clip, then there probably was arcing there too.

 

[big john]

What do the hinges on the of that drive-arm look like? Is there evidence of really high current flow through the operator? I imagine the paint would be burned on the exterior of the disconnect where the drive-arm rivets attach.

Those old ITEs are funny, they don't have knives, they use a linear plug that moves horizontally to bridge two contacts.

I almost wonder if there wasn't a breakdown between the phenolic guard that insulates the bridge-plug from the grounded metal drive-arm.

 

[don_resqcapt19]

I am not sure there was even an arc...looks more like heat damage to me. I would guess a poor connection issue, either at the fuse clip or at the actual switch contact.

 

[GoldDigger]

I am not sure there was even an arc...looks more like heat damage to me. I would guess a poor connection issue, either at the fuse clip or at the actual switch contact.

I think that to visibly melt the metal of the door (at least it looks like that to me from the photo) you would need an arc and not just a hot spot.
Never mind, that is just blistered paint.
I do find it interesting that there is a clean spot in the middle of the deposits on the inside of the door. Possibly a good indication that the material was thrown out around the circumference of contact area but the direct path was blocked by the contact cover.

 

[big john]

I am not sure there was even an arc...looks more like heat damage to me....

I imagine that for that kind of discoloration to be caused by heat alone, there would be massive deterioration of the conductor insulation, and that doesn't look to be the case.

Looks like splatter all over the door and bottom of that enclosure. And it appears that the vapor cloud blew out of the open KO on the lower right side and discolored the neighboring disconnect.

 

[meternerd]

Sure looks like a phase to ground arc to me. Poor connections or overheated fuse clips are a likely culprit. 480 is 277 to ground, so any arcing will most likely flash over to ground eventually. Seen it quite a few times at pump MCC's here.

 

[under8ed]

That is most likely in my thoughts. The machine had some insulation breakdown from the surrounding fire. When it shorted to ground, a weak spring or loose connection reacted faster to the overcurrent than the fuse. Didn't something upstream trip? something that reacted faster to further explosive arching than what must be time delay fuses.

 

[ron]

Was there an burned rodent anywhere in the vicinity of the fused switch?

 

[hlfireinspector]

Electrical is my weakness as a fire investigator so I always try to learn something at every fire. There were no rodents found in the disco. The power was shut down prior to my arrival. I have not been able to find the person the shut the power down. If the main was killed prior to operation of these disconnects then I do not think this was caused when the person tried to shut this disco down. If a firefighter was shutting down the discos in a limited visibility then he could have received a flash from this. I cannot understand why the fuse was not blown ( Slowblow fuses react to load not arc)?? If the arc went to the bar above the fuse this would have sent a shock to anyone operating the disco???


Thank you so much for your input: I'm sending you one more picture.

 

[cornbread]

I would be curious as to what the Arc flash label / calc was?

 

[GoldDigger]

Electrical is my weakness as a fire investigator so I always try to learn something at every fire. There were no rodents found in the disco. The power was shut down prior to my arrival. I have not been able to find the person the shut the power down. If the main was killed prior to operation of these disconnects then I do not think this was caused when the person tried to shut this disco down. If a firefighter was shutting down the discos in a limited visibility then he could have received a flash from this. I cannot understand why the fuse was not blown ( Slowblow fuses react to load not arc)?? If the arc went to the bar above the fuse this would have sent a shock to anyone operating the disco???


Thank you so much for your input: I'm sending you one more picture.

If this was a series arc across the contacts there would be no excess current to blow the fuse.
If it was a parallel arc to ground, it was ahead of the fuses in the circuit, so again they would not blow. It would have to be an upstream device that opened.

 

[ron]

There were no rodents found in the disco. The power was shut down prior to my arrival.

When you find out who saw the disconnect first, see if the door was blown open initially. If the door was open, then the rodent could be just about anywhere (and maybe not in one piece)

 

[Rick Christopherson]

Just thinking outside the box a bit, could it be possible that this was not the cause of the fire but instead the result of the fire? The smoke discoloration in the first picture doesn't seem to point to a direct source. Also, since the disconnect was fused (stuck) in the On-position, it was not in a partial disconnect state when the arcing took place.

Therefore, is it possible that there was enough smoke buildup in the disconnect enclosure to have a dielectric (air) breakdown that resulted in the arc flash, which was secondary to the actual fire?

Whatever caused the arcing is no longer present. So that leaves things like smoke or rodents as the cause for the arc.

 

[big john]

...If the arc went to the bar above the fuse this would have sent a shock to anyone operating the disco?

Maybe not. If the grounding was good enough to clear the fault quickly and prevent a high voltage from developing on the operating mechanism, the person touching it my never have felt it.

I'm sending you one more picture.
View attachment 11801

What is this from?

I gather from this discussion nobody knows for sure whether someone was interacting with this disconnect at the time of the flash?

 

[electricguy61]

Any chance something was pushed into the door of the disconnect, allowing it to slam into the left fuse holder? I don't see bends in the door, but we only have the one picture of the exterior of the door.

 

[hlfireinspector]

As stated before the machine that this disco goes to was not suspect to the fire cause. The machine had a lot of internal melting due to the heat in the building. I have never worked a fire in which I had so many breakers thrown on circuits that were not involved in fire. I think we lost a large amount of the electrical in the building. The fire was very isolated but we lost all lighting and most of the 110 outlets. This is a picture of a subpanel for the 110 in the building. The smoke and heat was very intense and if smoke can carry an electrical arc that could be what happen inside the disco. The disco cover was closed until I opened and it received no mechanical damage from an exterior source. The pop rivets on the disco handled were not damaged. I am hope the insurance investigator will want to arc map the building when they see the massive amount of electrical damage the building sustained.

Between the 220 and 110 I had over 20 breakers thrown. One outlet was involved in the fire area.

 

[big john]

What indications of heat do you have in the areas near the 120V subpanel? I've never seen completely unloaded breakers trip because of a hot environment, but thermal-magnetic breakers like what's pictured do have to be derated if the ambient temperate gets above 40?C or they will trip prematurely.

 

[jim dungar]

I cannot understand why the fuse was not blown...

The arcing in the photos occurred on the line side of the fuse. Fuses rarely blow simply due to heating, even if excessive, at their ferrules/end caps.
Or, are you asking why upstream devices did not open?

As to the arc flash; PPE requirements are based on cal/cm^2 not the 'size of the flash'.
A low level fault that lasts for a long time can melt a lot of metal and produce a lot of heat (think of arc welding). This arcing is why GF protection is required on some 480Y/277 volt systems.
Some faults can be so great that they 'blow the doors open', but the current is so high the protective device operates instantly, preventing further arcing damage.

 

[electric_cal]

I agree with what everyone has said. The visual indications are pointing to an arc flash event across the line side connections of the disconnect switch. There is also a very good indicator of the event in photo 3. Take a close up view on the bottom right side of the enclosure. There is a great deal of debries in the bottom of the switch box, and if you look you can see that there is an open KO that shows that the flash also hit the box to the right through the open KO.