Is GFCI needed for this interconnection?

[Matt Donovan]

I have a 480Vac PV system connected to a 208V service, so we are using a step-up transformer between the PV disconnect and the existing main switchgear. This is a load-side interconnection, and the existing gear has a GFCI main, so normally I would need to use a GFCI breaker for the interconenction, per 705.32. But does the transformer prevent ground fault current from flowing from secondary to primary? If so, then my PV system does count as a ground fault current source to the existing switchgear, and I should not need a GFCI on the PV breaker. The transformer is is delta primary (utility side) :: wye-G secondary (PV-side).

 

[ggunn]

I have a 480Vac PV system connected to a 208V service, so we are using a step-up transformer between the PV disconnect and the existing main switchgear. This is a load-side interconnection, and the existing gear has a GFCI main, so normally I would need to use a GFCI breaker for the interconenction, per 705.32. But does the transformer prevent ground fault current from flowing from secondary to primary? If so, then my PV system does count as a ground fault current source to the existing switchgear, and I should not need a GFCI on the PV breaker. The transformer is is delta primary (utility side) :: wye-G secondary (PV-side).

To begin with, your transformer is step down, not step up, under normal operation. Some transformers are OK in both directions and some are not. Be sure to get it clear with your transformer manufacturer what your application is. The 480 wye side of the transformer is separately derived; even though it is connected on the load side it should be treated as a new service with its own neutral to ground bond, fused disco, and grounding electrode system. I'm not sure how that applies to your question, though.

 

[Matt Donovan]

The question is can ground fault current from the PV system pass through the transformer to the main switchgear? Or does the transformer prevent ground fault current from the inverters from reaching the equipment on the primary? Yes, the separately derived system will have a neutral-to-ground bond, as usual. Also, the grounding electrode will be shared with the primary (through building steel).

 

[ggunn]

The question is can ground fault current from the PV system pass through the transformer to the main switchgear? Or does the transformer prevent ground fault current from the inverters from reaching the equipment on the primary? Yes, the separately derived system will have a neutral-to-ground bond, as usual. Also, the grounding electrode will be shared with the primary (through building steel).

In the event of a ground fault on the AC side the current comes from the service, not from the inverter.

 

[Engineer#4827502]

The question is can ground fault current from the PV system pass through the transformer to the main switchgear? Or does the transformer prevent ground fault current from the inverters from reaching the equipment on the primary? Yes, the separately derived system will have a neutral-to-ground bond, as usual. Also, the grounding electrode will be shared with the primary (through building steel).

It depends on the transformer. Some will and some won't. The way to think about it is: will a ground fault on one side of the transformer be detected as a ground fault on the other side. On a wye:wye transformer, a ground fault on one side will look like a ground fault on the other. On a wye:delta, a ground fault on the wye side will not read as a ground fault on the delta; it will look like a phase fault.

 

[jaggedben]

The question is can ground fault current from the PV system pass through the transformer to the main switchgear? Or does the transformer prevent ground fault current from the inverters from reaching the equipment on the primary? Yes, the separately derived system will have a neutral-to-ground bond, as usual. Also, the grounding electrode will be shared with the primary (through building steel).

I believe you understand that the idea behind 705.32 is that, with the inverter powering loads on the load-side of the main switch gear, a ground fault on that load side might not be detected by the ground fault protection device. Therefore the inverter must be connected on the supply side of the GFP device, or provide some protection of its own

Consider, then, that the ground fault may be on the low voltage side of the transformer, and that in that case the existence of the transformer is moot. Whether a ground fault on one side of the transformer will be detected on the other side is moot, because we're still worried about a ground fault that is in the same system as the switchgear with the primary GFP.

In your original post, where you say "normally I would need to use a GFCI breaker for the interconenction", others would say 'normally I would need to a connection on the supply side of the ground-fault protection". This is because GFP breakers that are suitable for backfeed are rarely encountered. That's a whole side discussion, but either way, I don't think the presence of a transformer in your design changes the reasoning.

I'm not an engineer, just repeating what I believe other engineers have said on this forum.

 

[pv_n00b]

I have a 480Vac PV system connected to a 208V service, so we are using a step-up transformer between the PV disconnect and the existing main switchgear. This is a load-side interconnection, and the existing gear has a GFCI main, so normally I would need to use a GFCI breaker for the interconenction, per 705.32. But does the transformer prevent ground fault current from flowing from secondary to primary? If so, then my PV system does count as a ground fault current source to the existing switchgear, and I should not need a GFCI on the PV breaker. The transformer is is delta primary (utility side) :: wye-G secondary (PV-side).

1) The ungrounded delta side of the transformer can't supply ground fault current, unless it's a multi-line to ground fault. But then it's a line to line fault through the grounded metal which is not exactly a ground fault even though it flows through ground. :huh:

2) There's a big difference between GFCI and GFP. NEC 215.10 and 705.32 are connected and are talking about GFP, and would not apply to a 208V service. If you have a main breaker that has GFCI then 705.32 does not apply. But I've not heard of a 3 phase 208V GFCI breaker. So maybe there is some confusion over what you have.

 

[electrofelon]

It depends on the transformer. Some will and some won't. The way to think about it is: will a ground fault on one side of the transformer be detected as a ground fault on the other side. On a wye:wye transformer, a ground fault on one side will look like a ground fault on the other. On a wye:delta, a ground fault on the wye side will not read as a ground fault on the delta; it will look like a phase fault.

I am not sure I agree. I think it depends on what you mean by "look like." For a wye:wye, a ground fault on one side would result in a turns ratio proportionate line to neutral load. I guess we could call that like a ground fault, however this will not result in the GFP tripping, even if the Y points are bonded together (which they would in essentially every installation).

 

[iwire]

I am not sure I agree. I think it depends on what you mean by "look like." For a wye:wye, a ground fault on one side would result in a turns ratio proportionate line to neutral load. I guess we could call that like a ground fault, however this will not result in the GFP tripping, even if the Y points are bonded together (which they would in essentially every installation).

I agree.

 

[jaggedben]

...
2) There's a big difference between GFCI and GFP. NEC 215.10 and 705.32 are connected and are talking about GFP, and would not apply to a 208V service. If you have a main breaker that has GFCI then 705.32 does not apply. But I've not heard of a 3 phase 208V GFCI breaker. So maybe there is some confusion over what you have.

Don't forget about 230.95.

 

[pv_n00b]

Don't forget about 230.95.

Yes, 230.95 goes into more detail on GFP setting. But since a 208V service does not qualify for any of these provisions it's a moot point. If someone did put a GFP main breaker in this service it was to address an issue outside of the NEC and it would be worthwhile to find out why it is there.