Solar Array Equipment Grounding

[shortcircuit2]

Do all components need to be LISTED by a NRTL for use together to be used for grounding? Specifically, I?m referring to the PV modules with a mounting or ?racking system?...

Example:

1. The PV rail mounting system needs to be LISTED as suitable for grounding and not just mounting.
2. The PV module needs to be LISTED for all the different methods for grounding. IE: Grounding with nut, bolt and lug, or grounding with self tapping screw and lug, or grounding by WEEB system, or any other grounding method.

In reading the installation instructions of 2 different PV module manufacturers recently, 1 only describes the use of a nut & bolt with grounding lug to the panel for grounding, and the other describes 5 different ways to ground the module.

So, the PV module with only 1 method for grounding can only be grounded that way or the 1703 listing is void.

This 45-minute UL webinar seems to discuss this in detail...(Its long, but informative)

http://content.learnshare.com/courses/73/424588/UL 2703/lib/playback.html

After watching this, my conclusion is that all components do need to be tested and NRTL LISTED together.

Would this need to be done at both 600volts and 1000volts?

shortcircuit

 

[jaggedben]

I'm not following you.

Do all components need to be LISTED by a NRTL for use together to be used for grounding?

From where are you getting "for use together"?

Example:
1. The PV rail mounting system needs to be LISTED as suitable for grounding and not just mounting.
2. The PV module needs to be LISTED for all the different methods for grounding. IE: Grounding with nut, bolt and lug, or grounding with self tapping screw and lug, or grounding by WEEB system, or any other grounding method.

This is an example of what? (A code compliant installation?) I don't know where you are getting your items (1) and (2) that you list after that. They are not required by the code, in my opinion. The 2011 code says a bunch of stuff on this subject in 690.43(C) through (E). The section was heavily revised from the 2008 version. But none of what it says entails, to me, what you've written above.

So, the PV module with only 1 method for grounding can only be grounded that way or the 1703 listing is void.

True, if an AHJ wants it to be, because s/he can say it violates 110.1(B) to not follow instructions. But a "method" can include a lot of different potential components, such as any that are listed to such-and-such a UL standard.

After watching this, my conclusion is that all components do need to be tested and NRTL LISTED together.

I didn't watch the webinar, but I believe I disagree. All components need to be listed for their individual purpose. No component may be grounding by a method not included in its instructions, and no grounding hardware may be used in a manner outside the instructions. But there is no requirement that components used together be 'listed together.'

 

[shortcircuit2]

Yes, I guess I did present it in a confusing way. Typically we install the racking system to the roof and ground it with a equipment grounding conductor jumped from rail to rail, clamp the modules to the racking with weebs and presto?everything is grounded right.

Well, after watching that UL webinar, the racking must be listed as suitable for grounding (690.43(C) structure identified as an equipment grounding conductor), and 690.4(D) says the modules must also be tested and listed for the application. 690.43(D) would cover weeb type devices for grounding the modules to the structural racking.

Therefore each of the system parts would have been tested and listed for the application and could be used together as described above for the typical installation.

Maybe this is not what the UL 2703 standard says specifically right now, but from watching the webinar, this is how it will be soon.

 

[jaggedben]

the racking must be listed as suitable for grounding (690.43(C) structure identified as an equipment grounding conductor), and 690.4(D) says the modules must also be tested and listed for the application.

This is what I'm disagreeing with, as those sections of the code don't actually lay out such strict requirements. 690.4(C) permits 'devices listed and identified for grounding the metallic frames of PV modules" to bond modules to racking. (That basically permits WEEBS.) It then requires the mounting structures be 'identified as EGCs or shall have identified bonding jumpers or devices connected ....'

690.43(D) simply says that racking must be 'identified for the purpose of grounding modules' if it is used for that.

(Emphasis added).

So 690.43 does not require racking to be 'tested and listed', only 'identified', and requires 'listed' devices only when specifically bonding PV module frames. In other words, it's written so as to generally permit listed WEEBs.

The only remaining issue is whether bonding a PV module with a device that is not mentioned in its instructions is a violation of the module's UL listing and therefore a violation of 110.1(B). This depends on how the module's instructions are written, and manufacturers are all over the place with this. Some module instructions are wonderfully vague, saying things like "You must ground the modules. Use this device that we manufacturer, or use another method." A lot of them say to use the marked grounding holes, but the wording may or may not imply that you must use a method that requires a hole.

I don't bother arguing with AHJs who say I can't use WEEBs if it's not in the module's instructions; there's too little to go on. But the flip side is that unless I know that an AHJ cares, I generally don't bother checking if module instructions include WEEBs, because 690.43 permits them IMHO. And if the WEEBs were tested and listed by UL for bonding one type of aluminum framed module, then common sense says they will work with any such module because the frames are all the same material. If the AHJ doesn't object, then I certainly don't.

As for UL's evolving standards, I still haven't watched the webinar, but at one point I skimmed through these documents.

http://www.solarabcs.org/about/publications/reports/module-grounding/index.html

 

[shortcircuit2]

So then the issue is the interpretation of description "Identified"...well is this another "gray" representation in the NEC? or is the definition of ?Identified? in article 100 along with its informational note, give enough guidance to say that if the racking is ?listed? ,then it is ?identified?.

 

[ggunn]

So then the issue is the interpretation of description "Identified"...well is this another "gray" representation in the NEC? or is the definition of ?Identified? in article 100 along with its informational note, give enough guidance to say that if the racking is ?listed? ,then it is ?identified?.

Beyond that, check with your AHJ to make sure that WEEBs are allowed if you want to use them. I know of at least one where they are not.

 

[jaggedben]

So then the issue is the interpretation of description "Identified"...well is this another "gray" representation in the NEC?

I'd agree with that, unfortunately.

or is the definition of ?Identified? in article 100 along with its informational note, give enough guidance to say that if the racking is ?listed? ,then it is ?identified?.

If it is 'listed' then it is also 'identified', but it can be 'identified' without being 'listed'. Remember that being 'listed' means it's been tested to a certain NRTL standard. In the case in point, I don't believe there is any NRTL standard to which racking would be listed for carrying ground-fault current from PV modules. And would anyone say such a standard is needed, when we are talking about solid aluminum panel frames and racking that typically have a cross sectional area probably equal to something like a 1/0 conductor or larger? I'd say no such standard is needed. Instead, I think it's reasonable to say that if a WEEB is listed for bonding PV modules to a particular make of racking, then that racking has been 'identified for the purpose of grounding PV modules". If the racking manufacturer also approves of the use of such WEEBs in their instructions, then that's corroborating evidence that said racking is so 'identified.' After all the WEEB has been listed, and that's the likely point of failure. In any case, that's the argument I'd make to an AHJ who raises his eyebrows at WEEBs, recognizing that some will be swayed and some won't.

I find it paradoxical that some AHJs have problems with WEEBs, when the WEEB has evidently been tested by UL for precisely the purpose it is being used for (because there's no other type of connection UL could have tested it for). Whereas who knows if UL has also tested ISLCO direct-burial grounding lugs for how well they bond panels? All we know is those lugs have somehow been tested to UL 467, but we don't know if they tested them on PV panels. They probably didn't.

 

[shortcircuit2]

Thank jaggedben for bouncing this around with me. I have a somewhat clearer view of this grounding issue that I've been studying. I would agree that an aluminum rail would seem to be a good, low impedance ground-fault path as expected by 250.4, but I?m not in a position to identify it as such, and I think that is to be left to the experts of an approved NRTL. I?m sure they will all agree that they are best for identifying the rails for such use. I'll have to watch the webinar again, but the standard I made note of was UL 2703...

 

[SolarPro]

Do all components need to be LISTED by a NRTL for use together to be used for grounding? Specifically, I?m referring to the PV modules with a mounting or ?racking system?...

Check out slide 6 in Greg Ball presentation for the SolarABCs meeting at SPI in Orlando:

http://www.solarabcs.org/about/publ.../pdfs/10-Module_Grounding-Ball-14Sept2012.pdf

Industry stakeholders are trying to move the standards away from a high level of specificity to more general requirements and allowances.

I think that's a good plan. Limiting the industry to the specific grounding solution(s) that the modco calls for in its product installation manual isn't adequate in the long term. Every installation us different. New products come to market. Practices are improved slowly over time. We don't want to limit installers and designers to one solution just because that's what was en vogue when the installation manual template was written 5 years ago.

 

[marez]

we have the same problem with our ahj. last time it came up I asked him if #4 rebar building steel or copper pipe is listed with its components. bummer is he just said "good question. just do it my way."