From the EMailbag:

I've invited RH to comment if necessary

Bill

The short circuit he described probably did cause a ground fault from hot to armor.

Yes, I would like for RH to explain why he thinks a GFCI is too sensitive to a ground fault.

Am I to believe that a ground fault anywhere in that part of the building will kill the whole section of the building?

The only solution I see is to use successively more sensitive GFCI protection throughout the branch.


[This message has been edited by Dspark (edited 05-03-2001).]

These are those adjustable types right? We are onto ma setting, instead of trip curve ?

[This message has been edited by sparky (edited 05-03-2001).]

A short circuit ground fault is going to be way beyond milliamps.

We are not talking GFCI, it is GFPE. GFCI is Ground Fault Circuit Interruptor and is for protection of personnel. GFPE is Ground Fault Protection of Equipment as covered in 215-10 and 230-95. I can't imagine a branch circuit that is protected at 20 amps causing a GFPE set at 1600 amp to trip. For that to happen the GFPE would have to already be on the verge of tripping. That would mean that close to 1600 amps is constantly being fed to ground, the GFPE is defective or the feeder is very unbalanced. 230-95 gives the maximum setting to be 1200 amps. The trip time at 3000 amps has to be less than or equal to one second. A 20 amp circuit breaker should be instantaneous at around 100 amps. The GFPE requirements are for feeders and services because of the high short circuit available to that types of equipment and the conductors. Have the breaker trip units and sensors(CTs) been tested lately? Is the 20 amp somehow causing a very large load to start?

Okay, that is a different story.
Are you sure whether 1600 A was the GFPE setting or in fact the total OPD limit?

Sounds like the typical thing happening on GFPE protected services.

The trip setting is something like 5 amps to 100 amps of leakage to ground before tripping.

Scott SET

For the types of breakers the the original poster was talking about I think the setting would be closer to 100 amps minimum. The trip units usually use multiples of 100 amps to ground. That 5 amps would allow the service GFPE to go out before the branch circuit. A small motor shorting out mid-winding would bring the whole system down because the short circuit protection may be set so much higher than 5 amps and the overload device would probably be too slow. The overload device is to protect the motor, not for safety to prevent elevated voltages at ground. Services going down because of a branch circuit fault is definitely poor selective protective coordination.