Sanity Check: GECs, subpanels, and inverters with transformers

[jaggedben]

In the past week I have failed three inspections and I feel like the major AHJ I work with has floated off into the deep end...

As a check on my sanity, I would like to ask folks here if they agree or disagree with the following statements:

1) Installing a circuit breaker for an isolated utility-interactive inverter (i.e. inverter with a transformer) into a distribution subpanel with other OCPDs presents no inherent safety danger or code compliance problem, including objectionable current.

2) A DC GEC for a utility-interactive inverter may be run through a subpanel without inherently creating a danger or code violation.

3) A combined GEC/EGC, as allowed by 690.47(C)(3) in the 2011 NEC, may be run through any number of subpanels or other enclosures to where it connects to the Grounding Electrode system, without inherently creating a danger or code violation.

4) An EGC for a subpanel may be used as an continuation or extension of a combined EGC/GEC for a utility-interactive inverter, so long as it complies with the provisions of 690.47(C)(3).

5) The presence of a transformer in a utility interactive inverter has no relevance to the safety or code compliance of a) the installation of the inverter output OCPD or b) the routing or manor of installation of a DC GEC, if required.

Assume that all inverter equipment is listed, that installations of GECs comply with the letter or 250.64, etc.

 

[Smart $]

FWIW, I agree with your statements.

 

[Zee]

1. yes
2. don't know.
(I land combined AC EGC/DC GEC in "associated equipment".
I interpret that term widely: wherever I land PV breaker.)
3. don't know
4. yes.
(Given 2. i must rely on (E) EGC from main to Subpanel to "continue" the GEC back to main.)
5. yes.
i would add that i don't believe a gec is even required for a TL inverter.

 

[SolarPro]

1. Correct.
3. Yes, otherwise you can't get there from here.
2. Must be okay, since a combined dc GEC and ac EGC is allowed.
4. Correct. Since the combined conductor needs to be unspliced, installers piggytail jumpers onto that conductor in order to ground equipment it passes through.
5. Correct.

 

[c_picard]

Perfectly sane.

Sent from my SCH-I535 using Tapatalk

 

[jaggedben]

Thanks for the votes of confidence.

I have yet to be able to talk to the AHJ personage who is responsible for this coming up....

 

[shortcircuit2]

I'm not a fan on the method in 690.47(C)(3) where we are allowed to land the combined GEC/EGC at the busbar of the associated equipment. I think the GEC should go all the way back to the grounding electrode.

1.What if the associated subpanel is fed with a 60amp feeder? It will only have a #10 EGC.

2.What if the associated subpanel feeder is installed in a ferrous raceway? Is the installer going to make the feeder comply with 250.64(E)

 

[SolarPro]

As I read this subsection, it is implied that the combined dc GEC and ac EGC has to be run to the busbar to which the main ac GEC is attached. It doesn't really make sense otherwise. Why run an unspliced combined GEC/EGC only as far as a groundbus bar in a subpanel? That clearly isn't consistent with the intent of options 1 and 2 or the explicit requirements in subsection (3). However, if you run an unspliced GEC/EGC all the way to the ground busbar to which the main ac GEC is attached, then what's the problem? That's basically as good if it were landed on the GE. That's the way I have always understood this option. (Of course, you typically do have to install pigtails on this unspliced combined conductor to pick up ground bus bars in all the equipment that you pass through en route to the main panel.)

 

[shortcircuit2]

Yes SolarPro...we understand the intent...but the language allows for different interpretations in the field. The language doesn't disallow a 690.47(C) GEC/EGC to land at a subpanel.

 

[SolarPro]

If there's a problem, I'd put in a proposal to fix the language. That option isn't likely to go away. But you can find room for improvement in many places in the Code, especially in Article 690.

 

[jaggedben]

shortcircuit,
I appreciate the comment. I have always taken it as a given that one has to somehow comply with the various GEC requirements all the way back to the grounding electrode system. And that's not precisely what's at issue with respect to why I started this thread and the problem I'm having with the AHJ. However, with that said, and with some risk of thread drift...

I am increasingly coming to doubt the underlying requirement that there even be a DC GEC. I'm admittedly not an electrical engineer or any other kind of credentialed expert on this subject. But let me describe my understanding of the subject and see where it goes.

...

The purpose of a premises GES is to dissipate surges and such from lightning or contact with high voltage sources. Also, regrounding the grounded conductor at the premises helps stabilize the voltage between the grounded and ungrounded conductors from the utility supply.

The more stringent requirements for installation of the GEC (as compared to EGCs) are there to :
a) ensure a reliable connection for fulfilling the purposes mentioned above. (The requirement for this is to be continuous.)
b) to prevent the 'choke effect' in case the GEC becomes a path for ground fault current or other return current to the utility. (The requirement for this is bonding at both ends.) The GEC may become a current carrying conductor for AC current in abnormal circumstances, such as an open neutral on the premises or at a neighboring premises. This isn't supposed to happen but needs to be accounted for.

So far so good? Maybe not. Anyway, now for the punchline...

None of the above relates in any serious way to the grounding or bonding of PV systems, (especially utility-interactive systems without stand-alone capabilities).

As far as surges from lightning or high voltage utility sources that come in along utility lines, the addition of a PV system to the premises does not predictably affect those issues. As far as stabilizing voltage, that does not apply to grounded DC PV conductors.

As far as a PV array being a potential target for a lightning strike, the NEC requirements for a DC GEC are not meant to address that, nor is that even generally within the scope of the NEC. (On a related note, see : https://www.youtube.com/watch?v=YuDqXFvRv94)

As far as ground-fault current traveling along the NEC-required DC GEC from a utility interactive inverter, that is not something that will come from the solar side of things.

As far as the operation of ground-fault detection circuits in inverters or charge controllers, that is handled by equipment grounding (aka "bonding"; as with OCPDs for AC circuits it does not require and should not rely on a connection to the earth).

In cases of stand-alone PV systems, especially those with DC utilization equipment, it makes sense to me that if a GES does not previously exist then one ought to be installed to earth metal parts for personal saftey, and also to fulfill some of the same purposes as an AC GES. However, on a premises with an existing GES, I do not see how any anything other than equipment grounding (bonding) is required to protect persons or property.

Am I wrong? Do solar panels themselves create the potential for static discharges that would be mitigated by connection to the earth? If so, is ordinary equipment bonding (i.e. 250.122) to an existing GEC not adequate to address this?

 

[shortcircuit2]

Hmm...you have put a lot of thought into this jaggedben...I'll need some time to sort your comments out.

Here is my kneejerk answer...

Without a DC Grounding Electrode Conductor on PV systems with isolation thru a transformer, there would be no system connection to the Grounding Electrode System...which on an AC system performs one function as protection for lightning strikes. Now we say that the NEC isn't intended to be a guide for lightning mitigation and protection, but silently it has a few rules that are in place for that purpose.

Man still doesn't fully understand lightning. We put into effect some rules that we think will protect from lighting. But lightning is a phenomena that can pass through air, so how do you protect from that?