Came across a very interesting situation this week. A customer had an old Square D Iso-Flex 5 kV rated starter, typical Class E2 style. When I first saw it there was about a 4 foot long soot/burn mark on the floor. Of course my very first response was a long sigh as I thought about all the fun I was going to have risking my life opening that door and saying, OK I'll go get my gloves and arc flash stuff. I didn't make it all the way to the truck though because the protective device for this switch is cutout fuses that blew and shut everything down. The fuse barrels were sitting nex to the pole, next to my truck, so securing them was trivial. So I promptly whipped out the camera/phone, flash light, and screwdriver, and went to work.

The disconnect is probably one of the worst designs I've ever seen. I've seen this one before but never bothered to pay much attention to the switch. The strangest part is that the disconnect handle moves up/down as you normally expect it to but to get the door open the entire disconnect handle rotates sideways 90 degrees and then the end of the handle folds in half so that the door can swing open past the handle mounted in the center of the starter instead of more conventionally off to one side. The disconnect itself sits behind 4 panels that I had to remove which sit behind the fuses and the starter itself. I had to remove the fuses, starter, and 4 panels before I could even expose enough of it to take a picture for a post-mortem. Most of what I was able to initially diagnose was via taking off the splice plate cover on the side. There was extensive damage from smoke, fumes, soot, and heat throughout the high voltage side. The motor fuses never tripped and if it wasn't for contamination and melted control wiring, I'll bet that the starter would still run. I'll never get it clean enough for medium voltage service though.

All indications to me is that one of the switch blades never quite closed in fully or lost tension or something that caused a high resistance connection at the switch that caused it to fail, most likely some kind of mechanical problem. The interesting twist here is that although I don't know how long before the incident the switch was closed in, the motor had been running for over 3 hours before the starter failed. This is the closest thing I've seen to a "spontaneous" arcing fault. Arc flash seems extremely unlikely here because after 3 hours of operation, personnel would have just had to have been walking by but they also don't routinely go into electrical rooms because none of the on site personnel are electricians.

Also of note is that the cabinet door did not open or blow off. The entire bottom 3" of the cabinet is vented with grills. Most of the damage was along the path between the starter and the vent. I can't supply pictures due to customer request but the entire fixed part of the left hand disconnect arm is completely vaporized and gone. There is no way that the "copper vaporization theory" holds any water based on the damage. To me it's just more confirmation of E-Hazard group's theory that arc blast is really all hot air (literally).

I'm sure that maintenance performed on the switch could very well have prevented this from happening. But I'll just say that there is nothing about this switch that makes it accessible for maintenance purposes. It wasn't maintained because it's not practical to do so. And by that I don't mean taking the time and money to shut it down...I mean it is kind of like the late 70's Pontiac Sunbird where the only way to change the spark plugs was to actually disconnect the motor mounts and pull the engine to reach the ones on the back right side. The disconnect on this starter can be serviced but nobody does it because it's not designed to be serviced.