230.95 GF protection of Equipment 480/277 only?

J

johnnydn

New User
Location
Oakland, CA
Occupation
Junior Electrical Design Engineer
I am having trouble interpreting the code which requires GF protection on service disconnects. I've read from different online resources that this is only a requirement for 480/277V but not 120/208V systems. My confusion is where does the code delineate which voltage systems are applicable and which are not? Does the Utility (CA in my case) have special exceptions which trigger this 480/277V only or is there a code section that articulates this delineation?

Taken from 2017 NEC:
230.95 Ground-Fault Protection of Equipment.
Ground-fault protection of equipment shall be provided for solidly grounded wye electric services of more than 150 volts to ground but not exceeding 1000 volts phase-to-phase for each service disconnect rated 1000 amperes or more. The grounded conductor for the solidly grounded wye system shall be connected directly to ground through a grounding electrode system, as specified in 250.50, without inserting any resistor or impedance device.

The rating of the service disconnect shall be considered to be the rating of the largest fuse that can be installed or the highest continuous current trip setting for which the actual overcurrent device installed in a circuit breaker is rated or can be adjusted.
 
johnnydn said:
I am having trouble interpreting the code which requires GF protection on service disconnects. I've read from different online resources that this is only a requirement for 480/277V but not 120/208V systems. My confusion is where does the code delineate which voltage systems are applicable and which are not? Does the Utility (CA in my case) have special exceptions which trigger this 480/277V only or is there a code section that articulates this delineation?

Taken from 2017 NEC:
230.95 Ground-Fault Protection of Equipment.
Ground-fault protection of equipment shall be provided for solidly grounded wye electric services of more than 150 volts to ground but not exceeding 1000 volts phase-to-phase for each service disconnect rated 1000 amperes or more. The grounded conductor for the solidly grounded wye system shall be connected directly to ground through a grounding electrode system, as specified in 250.50, without inserting any resistor or impedance device.

The rating of the service disconnect shall be considered to be the rating of the largest fuse that can be installed or the highest continuous current trip setting for which the actual overcurrent device installed in a circuit breaker is rated or can be adjusted.
Click to expand...
The only commonly used voltage system in the US that meets the requirement in 230.95 is a 480Y/277 volt system. There are also some 600Y/347 volt systems that would require this same ground fault protection but they are not near as common as the 480 volt system.

The 208Y/120 volts system does not qualify as the voltage to ground does not exceed 150 volts.

It also does not apply to the 240/120 volt 4 wire delta, even though the voltage to ground on one leg exceeds 150 volts to ground, because the rule only applies to wye systems.
 
The NEC does not explain the rules. For that you would need to research the documentation when the change was done, or understand electrical theory, or ask here.
277 ground faults won’t burn clear and the arc won’t self extinguish like a 120 fault will. Due to fires with 277 volt systems in the 70 ‘s gfpe rules were added
 
tom baker said:
The NEC does not explain the rules. For that you would need to research the documentation when the change was done, or understand electrical theory, or ask here.
277 ground faults won’t burn clear and the arc won’t self extinguish like a 120 fault will. Due to fires with 277 volt systems in the 70 ‘s gfpe rules were added
Click to expand...
And typically will not draw enough current, because of the impedance of the arc, to trip the OCPD.
 
tom baker said:
277 ground faults won’t burn clear and the arc won’t self extinguish like a 120 fault will. Due to fires with 277 volt systems in the 70 ‘s gfpe rules were added
Click to expand...
I've seen many incidents where 277 to ground either caused or nearly caused a fire even though on less than 1000 amp supply.

Then about a quarter century ago the code making geniuses decided we need AFCI protection on 120 volts to ground in many locations even though it does not sustain arcing without some external help:unsure:
 
kwired said:
I've seen many incidents where 277 to ground either caused or nearly caused a fire even though on less than 1000 amp supply.

Then about a quarter century ago the code making geniuses decided we need AFCI protection on 120 volts to ground in many locations even though it does not sustain arcing without some external help:unsure:
Click to expand...
I was going to submit a PI to delete 210.12 with the substantiation being that 230.95 tells us that self-sustaining arcs do not exist at 120 volts to ground, but since it would be rejected, I did not submit.
 
don_resqcapt19 said:
Probably two reasons...not a real common service and it is likely that the higher voltage to ground can flow enough current with a ground fault to open the OCPD.
Click to expand...
Line to line or line to ground are sort of the same thing with corner grounded delta when it comes to what you will have for fault current.

If you have 480 volt delta with a mid point ground and high leg - you only have 240 to ground (little less than 277) on two legs and 416 on the high leg.
 
kwired said:
Line to line or line to ground are sort of the same thing with corner grounded delta when it comes to what you will have for fault current.

If you have 480 volt delta with a mid point ground and high leg - you only have 240 to ground (little less than 277) on two legs and 416 on the high leg.
Click to expand...
But the impedance of the ground fault path will likely be much lower than the L-L impedance.
 
CoolWill said:
But the impedance of the ground fault path will likely be much lower than the L-L impedance.
Click to expand...
For a corner grounded full delta you kind of ignore the fact that one of them is grounded, the impedance is the same everywhere if all three coils that make up the delta and assuming there is reasonable balance of the primary side if the source is a transformer.

If you have an open delta there will be impedance differences
 
kwired said:
For a corner grounded full delta you kind of ignore the fact that one of them is grounded, the impedance is the same everywhere if all three coils that make up the delta and assuming there is reasonable balance of the primary side if the source is a transformer.

If you have an open delta there will be impedance differences
Click to expand...
No, I mean that a line to line fault involves two conductors of the same impedance , but a line to ground fault involves one conductor and the much lower impedance ground. So the overall impedance of the ground fault circuit is likely much lower just due to a lot more metal in the path.
 
CoolWill said:
No, I mean that a line to line fault involves two conductors of the same impedance , but a line to ground fault involves one conductor and the much lower impedance ground. So the overall impedance of the ground fault circuit is likely much lower just due to a lot more metal in the path.
Click to expand...
Why would the ground be lower impedance, for the general case?
 
CoolWill said:
No, I mean that a line to line fault involves two conductors of the same impedance , but a line to ground fault involves one conductor and the much lower impedance ground. So the overall impedance of the ground fault circuit is likely much lower just due to a lot more metal in the path.
Click to expand...
Delta with three identical coils has same impedance between all three corners, grounding one of them doesn't change that all it does is give that conductor a ground reference and ultimately extend that conductor to other grounded/bonded objects

Conditions of circuits you extend off the source may have varying impedance depending on conditions, and because that grounded conductor may have other possible parallel paths it could have less impedance at some particular points in the system, but at the source all three corners are essentially equal in impedance to one another.
 
jaggedben said:
Why would the ground be lower impedance, for the general case?
Click to expand...
Because there is very often multiple parallel paths for the ground. Especially in an install with a corner grounded delta. You will have the equipment ground wire, the conduit, the building steel, sprinkler and water pipes, etc. All of that adds up to a giant conductor that you might as well consider as zero ohms. It would be approaching half the impedance of a line to line short.
 
kwired said:
Delta with three identical coils has same impedance between all three corners, grounding one of them doesn't change that all it does is give that conductor a ground reference and ultimately extend that conductor to other grounded/bonded objects

Conditions of circuits you extend off the source may have varying impedance depending on conditions, and because that grounded conductor may have other possible parallel paths it could have less impedance at some particular points in the system, but at the source all three corners are essentially equal in impedance to one another.
Click to expand...
I'm only speaking of the fault circuit path itself, not the dynamics of the source. Two wires of equal length versus one wire and basically a zero ohm conductor. The ground fault circuit is almost half the impedance of the line to line short, everything else being equal.
 
CoolWill said:
Because there is very often multiple parallel paths for the ground. Especially in an install with a corner grounded delta. You will have the equipment ground wire, the conduit, the building steel, sprinkler and water pipes, etc. All of that adds up to a giant conductor that you might as well consider as zero ohms. It would be approaching half the impedance of a line to line short.
Click to expand...
Ok, maybe the probability of lower ground impedance is higher (or maybe not, irl) but it certainly won't be true every time. A line-ground fault that is only carried by the EGC (say through non-metallic conduit) will typically have a higher impedance, not lower, than line-line.
 
jaggedben said:
Ok, maybe the probability of lower ground impedance is higher (or maybe not, irl) but it certainly won't be true every time. A line-ground fault that is only carried by the EGC (say through non-metallic conduit) will typically have a higher impedance, not lower, than line-line.
Click to expand...
Go back and highlight the part where I said it will be lower every time.
 
You must log in or register to reply here.
Top