How did the main trip?

[ralpha494]

New industrial manufacturing building in South Milwaukee, WI. We're still adding and changing things, working on punch list, finishing phase two. Cutler-Hammer gear, two switchboards, don't know what gfci was set at.
The word I got from our foreman:
5000 amp main for switchboard B tripped.
1200 amp feeder for panel that feeds the crane tripped. 480 volts, ground wire, but no neutral.
500 amp breaker for the crane rail locked off and tagged off. Also 480 volts, a ground wire, but no neutral.
Crane guys were working on batteries for magnets that can lift 10 tons.
They said 239 volts dc negative ground.
Accidentally grounded positive lead up on the overhead crane, a flash, and half of the building goes down, generator started, management begins chanting "When in danger,fear, or doubt; run in circles scream and shout."

How does grounding that battery positive take out the main?

What is the gfci sensing?

How lucky was I to be in the job trailer (fed from swbd A and unaware of any problems) for lunch when it happened?

 

[vhay]

ralph,
I would start by determining what the ground fault settings are on all the relative feeder breakers. They should be coordinated to prevent such occurances. I would check to see if there was a flashover from the DC system to the AC system on the crane. It could be that the ground fault relay settings are still set at the minimum values. If all are set at minimum, it is just a race to determine which will trip first.
Know What I Mean, Vern

 

[mdshunk]

Somebody tell me that you all did a some testing first, before slapping that 5000A breaker back in? I don't take anybody's word for anything they said happened, until I test things myself.

 

[LarryFine]

You may have caused a voltage gradient across structure metal, due to the high current, and created a voltage difference between neutral and earth, which was detected by ground-fault equipment. Does this make any sense?

 

[RRelec]

You may have caused a voltage gradient across structure metal, due to the high current, and created a voltage difference between neutral and earth, which was detected by ground-fault equipment. Does this make any sense?

Makes perfect sense, the ground fault sensor could be on the bonding jumper in the main sensing any current on the ground, normal reading, this should be zero. OP indicates no neutral present at the load.

Did anyone check to see what conductors the sensor is monitoring?

R

 

[kingpb]

It sounds like a common lack of protective device coordination. I would imagine, as I have seen many, many times, that no coordination was done at all. The factory default settings have probably never been changed, and therefore as Vhay stated, its a race as to which one trips first.

Somebody needs to do a fault current analysis using minimum fault current, then based on that, go through and coordinate all the relay settings. Do arc falut analysis, place stickers on equipment, and finally, test equipment and cable to make sure it is still functioning properly.

 

[Pierre C Belarge]

Were any of the GFP sensors tested and documented before the tenants occupied the building?

I would have Cutler Hammer send a tech to work on the setting of the GFP(s)...after some type of coordination study is performed...should not all of this have been done already?

 

[jim dungar]

My guess is that somebody saved a few dollars by not paying for a coordination study and the testing of the ground fault system (not just the relay). I will bet that the 5000A main is the only device with ground fault sensing.

But the description is filled with too much second-hand information to make any other guess as to the cause of the 5000A tripping. Someone needs to make a one-line sketch of the affected electrical system (power and grounding).

 

[frenchelectrican]

My guess is that somebody saved a few dollars by not paying for a coordination study and the testing of the ground fault system (not just the relay). I will bet that the 5000A main is the only device with ground fault sensing.

But the description is filled with too much second-hand information to make any other guess as to the cause of the 5000A tripping. Someone needs to make a one-line sketch of the affected electrical system (power and grounding).

I going to say say the same thing as well but i am supecting that the setup is not right at all and some reason why the battery for the crane is not fused so i am supecting something else going on as well.

I am not sure if this fall in the NEC code or other area that the large storage battery like they used in crane it should have some sort of disconnection switch that can disconnect the postive and negtive source of battery system [ point it to me if i am wrong on this ]

Merci , Marc