Ground Fault Protection Feeders/SDS

[RonShap]

In most areas of the USA, the NFPA 70 (NEC) is code.
Section 215.10 requires ground fault protection for feeders that are "1000 amperes or more and installed on solidly grounded wye electrical systems of more than 150 volts to ground, but not exceeding 600 volts phase-to-phase" unless "ground-fault protection of equipment is provided on the supply side of the feeder" etc.

The following examples are not services, but feeders:
If there was GFP on the primary of a delta-wye solidly grounded transformer (1000kVA, 480/277V secondary, separately derived source), would you consider the feeder on the secondary side of the transformer GF protected on it's supply side?

Is your answer the same if the input feeder to a 1000kVA dual conversion type UPS module, 480/277V output, separately derived source, is protected upstream by GFP? Is the output feeder of the module GF protected on it's supply side?

 

[davidbeach]

A separately derived source cannot be protected by a ground fault sensing device in another system. A ground fault on the secondary of a delta-wye transformer shows up as a phase to phase overcurrent on the primary; a ground fault device on the primary would never see the secondary ground fault.

 

[RonShap]

David,
If Power in = Power out + losses, then the primary would have to "see" the fault.

 

[davidbeach]

The secondary phase-to-ground fault shows up on the primary as a phase-to-phase fault (or overload depending on fault magnitude) and will not be seen by a ground fault sensing device. There is not a zero sequence connection between the primary and secondary, so no zero sequence current flows in the primary while zero sequence current flows in the secondary.

Draw yourself a diagram of the windings of a delta-wye transformer and trace the current flow for a phase-to-ground fault on the wye side. You will see that all of the current is in one or the three windings on the secondary. With current in only one winding on the secondary there will only be current on one winding in the primary and in the delta that current is phase-to-phase. No unbalanced current for a ground fault sensor to detect. Conservation of kVA is maintained.

 

[waross]

Ron
What davidbeach says. A ground fault on a transformer secondary will be seen on the primary as either a load or an overcurrent, not as a ground fault.
respectfully

 

[RonShap]

If a ground fault on the secondary would be reflected as a single phase or phase to phase short (or overload), why wouldn't the GFP on the primary trip, if set to a feasible value considering fault current availability.
Ground fault current is sometimes calculated as large as bolted phase fault depending on the impedance of the ground conductor.

 

[jghrist]

The ground fault on the secondary would not be reflected as a single phase short or overload. It will be reflected as a phase to phase fault. There is no ground current to detect in a phase to phase fault.

 

[davidbeach]

Current out on one phase and back on an other. The vectorial sum of the phase currents would be zero and thus a ground fault device would see zero current. You could see a secondary ground fault from the primary by looking at negative sequence current in the primary but you would not be able to distinguish a very badly unbalanced load from a ground fault. That's why ground fault detection has to be done at the proper location.

 

[RonShap]

If the loads are three phase, and the GFP is set properly, then is there a big problem not being able to distinguish a very badly unbalanced load from a ground fault? There would not be an unbalanced load (substantial anyway) unless it was a ground fault.
In a critical environment, it is typical to have to make coordination/selectivity decisions based on reliability vs. equipment protection.

 

[davidbeach]

The negative sequence current of a bolted ground fault (fault on secondary, current measured on the primary) would probably be enough to distinguish it from a highly unbalanced load, who is to say that the fault will be bolted. Enough fault resistance and you can't tell. Put the ground fault detection on the secondary side of the transformer and the difference is clear, but not on the primary. Time to dig out your Power System Analysis text from college and brush up on sequence components and what happens to to the sequence networks as you go through transformers.

 

[RonShap]

I agree that the most accurate place for the GFP is on the secondary of the transformer and the output of a UPS module.
With that said, when dealing with certain occupancies, and certain design priorities, sometimes I need to abide by the absolute minimum code.
That is why I was trying to get opinions regarding the varied interpretation of the NEC article 215.10 Exception#3 "ground-fault protection of equipment is provided on the supply side of the feeder".

 

[davidbeach]

A feeder does not pass from one system to another. A separately derived system will have its own feeders and the protection at the supply side of the feeder will be as close to the separately derived source as practicable.

 

[RonShap]

The definition that I acknowledge I can't get away from....
A feeder is:
All circuit conductors between the service equipment, the source of a separately derived system, or other power supply source and the final branch-circuit overcurrent device.