preferable grounding method

System · December 2006

What's the prefered grounding method?

1) No ground cable from inverter/disconnect to array. Negative PV lug wired to combiner box grounding lug with panel frames, rack and combiner box ground lug wired to grounding rod.

or

2) Ground cable from inverter/disconnect ground lug. Panel frames, rack, and combiner box ground lug wired to grounding rod. Negative PV lead not wired to combiner box ground lug.

It just seems kinda weird to run a ground from the array back to inverter/disconnect ground lug when the negative line is already bonded with ground at the disconnect.

Is one of these a better option?

Roderick · December 2006

Re: preferable grounding method

I wrestled with this, too. If you haven't come across the document yet, Google for "John Wiles Photovoltaic NEC" and there should be some helpful documents.

This may have been mentioned in another thread, but where are you located, again? I hear European systems are commonly ungrounded.

Also, even if you're in the US, see if your inverter implements Ground Fault Interruption on the DC side. If so, then I think bonding array negative to ground before the inverter would defeat the GFI protection. If array positive had a weak short to ground (say, the panel frame or racks), it would just go back into array negative instead of tripping the GFI. I think, at least. It's worth checking with the inverter manufacturer for their recommended grounding. It may not be in their manual, but in my case, their website had a tech support form, and they were very helpful.

If the inverter does not have GFI, then I'd say yes, connect array negative to ground close to the array. I have no qualifications in NEC code, but my ad hoc impression was that inspectors were looking for a separate ground wire rather than just using the negative. This could be for lightning protection or just arbitrary custom, but I suggest running a ground wire of at least the same gauge as your supply conductors down from the array, if you don't want to do a lot of explaining, begging, or arguing.

mike95490 · December 2006

Re: preferable grounding method

http://www.sandia.gov/pv/docs/John_Wiles_Code_Corner.htm

Is there a FAQ page here, maybe this link should live there?

System2 · December 2006

Re: preferable grounding method

Thanks.

I already looked thru his documents. In a 2000 doc (http://www.nmsu.edu/%7Etdi/pdf-resources/cc74.pdf), he basically states these two methods, but doesn't really say if there is a better one. In 2004 (http://www.nmsu.edu/%7Etdi/pdf-resources/CC103.pdf), he only covers method #2 and doesn't offer #1 as solution. It would appear that running a disconnect-array ground is now required.

The question of running the ground cable is turning out to be moot as the cost per foot of 2/0, 2/0 copper is basically on par with the cost of 4/0, 4/0, 2/0 aluminum. Its only about $20 difference in total cost.

mike95490 · December 2006

Re: preferable grounding method

Are you prone to lighting strikes in your area ? (Thunderstorms)

think of all that nice metal up there, wired to your $4,000 inverter.

Will a strike fry all the panels in an array, or just the few it hits? (or do they all vaporize)

preferable grounding method

Comments

Categories