Ground Fault Selective Coordination

[brother]

I always thought it was a NEC requirement that a 'BRANCH circuit' overcurrent device was required to 'trip' before the Feeder device on even on a ground fault condition (basically the settings on the feeder is suppose to be delayed enuff to give the branch time to open) when the 'fault' is on the branch circuit end of course.

But after reading NEC 2008 240.13, 700.27, and 517.17 which deals with 'feeder and services(no branch). Even though it would be 'sensible' to be sure that the branch opens before a feeder, am I missing the 'language here?? This is not just dealing with emergency systems even though I listed some of those articles pertaining to that. Any help would be appreciated.

 

[charlie b]

I don't keep a copy of the NEC at home. But from what I remember the ground fault protection concept relates to a high level fault. It is a service and feeder thing. The GFP requirement applies to services above a certain rating. The hospital article requires one additional level downstream of the service to have GFP as well.

As to other types of faults, such as when a wire inside an enclosure touches the case, the NEC rule for fault coordination was not manditory until 2005, and even then it applied only to emergency and legally required standby systems.

 

[brother]

I don't keep a copy of the NEC at home. But from what I remember the ground fault protection concept relates to a high level fault. It is a service and feeder thing. The GFP requirement applies to services above a certain rating. The hospital article requires one additional level downstream of the service to have GFP as well.

As to other types of faults, such as when a wire inside an enclosure touches the case, the NEC rule for fault coordination was not manditory until 2005, and even then it applied only to emergency and legally required standby systems.

They have the NEC online too. Thanks for responding.

 

[templdl]

Coordination is always desirable but doesn't always work, as we would expect it to.
We always expect the device closes to the fault to trip and clear it before an upstream device trips. But that does not always happen when it comes to the instantaneous trip element.
I all devices that are in series with the fault see a fault level of current that is within their instantaneous trip calibration at the same instance it will be pot luck as to which one will trip.
It is not uncommon for the main breaker of a home distribution panel to trip when there is a bolted fault on one of the branch circuits where you would expect the branch breaker to have tripped.
Arc faults may not trip a breaker magnetically unless the arc draws enough instantaneous current to trip the breaker. If they don't the arcing will either ignite a fire or will trip the breaker thermally should the arc draw enough current over a given length of time that would be required to trip the breaker thermally

On the larger industrial circuit breakers the magnetic pickups often times are adjustable from 5-10 times the rating of the breaker which can provide some coordination. Even still one can not expect coordination should there be a bolted fault.
Going one step farther electronic trip units can also be equipped with a time delay for their instantaneous pickup which can give the upstream breaker a brief hiccup to allow a down stream breaker to clear a fault first. There is the short time pickup and delay for additional coordination. In addition there are breakers that communicate. A down stream breaker can indicate to an upstream breaker that it has seen the fault and will trip so that the upstream breaker doesn't need to.
Ground fault elements can be coordinated in a similar fashion. The reason that I used instantaneous and short time delays as an example is because the ground faults pick up and delays are usually required options before you can even get the ground fault features.

And all of these extra bells and whistles cost more money and are often omitted because of cost.