HELP!!! No EGC between array racking & combiner box???? This feels very wrong

[SunFish]

I was recently talking with a solar installer in my area who informed me they were NOT installing an EGC between the array racking/module frames and the combiner box. They are running an EGC from the combiner box down to the inverter and bonding all equipment in-between but they have no EGC bonding the racking/module frames to this EGC in the combiner. The local AHJ is operating under the 2008 NEC and they ARE installing the "additional array grounding" per 690.47(D). They had considered this acceptable as the racking is "grounded" and there was no need to run a ground connection between the array and the combiner box.

I had been taught that there needed to be a DC EGC between the racking and combiner box continuing all the way down to the inverter ALONG with the additional array grounding requirements of 690.47(D). I opened my code book later and found 690.47(D) states "Additional electrodes are not permitted to be used as a substitute for equipment bonding or equipment grounding conductor requirements". This along with 690.43 tells me there MUST be a DC EGC between the racking & module frames connected to the combiner box and continuing down to the inverter.

I'm 99% sure I'm correct and by code there must be an EGC installed between the array and combiner box continuing down to the inverter ALONG WITH the direct connection to ground per 690.47(D), are there any additional code items or documentation that can reinforce my claim?

My major concern is that if there is a ground fault at the array, say some squirrel damage, the current will find a path to ground through the additional electrode required by 690.47(D) and will possibly bypass the GFDI at the inverter, allowing current to continue to flow to ground without popping the GFDI as long as there is light on the PV modules.

Can anyone help me out with this question? I feel like this could be a major safety hazard and they will need to go back to ALL of their jobs to install the EGC between the array and combiner box? Note they are using weebs and lugs at the array properly, it's just this bond between the array and combiner that is missing.

 

[SolarPro]

Yes, it sounds like there is a reason to be concerned.

Here's a relevant quote from an article we ran in SolarPro within the last year:

David Del Vecchio is a senior engineer with Strata Solar, a project developer and EPC. Del Vecchio encourages installers to exceed minimum equipment-grounding requirements to ensure system safety over time. He cautions: ?Your ground-fault protection is only as effective as your return ground path. Let?s assume you are using a UL 2703?listed racking system, and the installation requirements call for a single EGC connection. What happens if this single connection fails over time, perhaps because it was under- or over-torqued or otherwise improperly installed? Now you have a section of the array that is effectively ungrounded.?

Del Vecchio concludes: ?If a ground fault occurs at a location without an effective return ground path, there is no ground-fault detection. The inverter will continue to run and the consequences can be disastrous. Therefore, when grounding rooftop racks, I always recommend multiple EGC attachments beyond the minimum requirements, because redundant grounding points in a bonded section of an array ensures a return path for ground-fault current and GFP detection.?

Here's a link to the whole article, which is long and dense?but also comprehensive.

Note that 690.47 describes system grounding requirements, which have changed somewhat over time. But your main question is about equipment grounding requirements, which have not really changed over time (though the equipment itself has).

 

[jaggedben]

You are correct about the issue although you are a bit off about the reason...

...My major concern is that if there is a ground fault at the array, say some squirrel damage, the current will find a path to ground through the additional electrode required by 690.47(D) and will possibly bypass the GFDI at the inverter, allowing current to continue to flow to ground without popping the GFDI as long as there is light on the PV modules.

You are correct that not having an adequate EGC path could result in the GFDI failing to operate properly. However, this is not because current will bypass the inverter through the ground, but rather because the additional electrode may not pass enough current to be detected by the GFDI. The current does not want to go to ground; it wants to find a path back to the other side of the solar panels. The earth is not a sufficiently reliable low-resistance conductor to serve as this path.

The arrays you are looking at are just waiting to experience two ground faults on each side of the circuit; the first will be undetected by the GFDI, and the second will be detected when it starts a fire.

...I'm 99% sure I'm correct and by code there must be an EGC installed between the array and combiner box continuing down to the inverter ALONG WITH the direct connection to ground per 690.47(D), are there any additional code items or documentation that can reinforce my claim?

Besides 690.43, there is 690.5. If the equipment grounding is not installed in a manner sufficient for the GFDI to operate correctly then you are violating that section. Further, if the inverter's instructions specify how equipment grounding is required for the GFDI to operate (they should, although they may not), then it would be violating 110.3(B) not to follow them.

I feel like this could be a major safety hazard and they will need to go back to ALL of their jobs to install the EGC between the array and combiner box? Note they are using weebs and lugs at the array properly, it's just this bond between the array and combiner that is missing.

I agree completely. Especially if the conduit between the array and combiner is PVC or something else that doesn't have a metallic path. Note that a properly installed metal conduit could meet the code requirements of an EGC, but industry best-practice is to install a wire EGC regardless.

 

[jaggedben]

I was looking for something else and stumbled back upon 250.4, which may also help, as it states in a couple places that 'The earth shall not considered an effective ground-fault current path."

 

[MWh_Pro]

+1 for Jaggedben's response. Although the WEEB clips or whatever they are using create a nice bond from panel to rack, without the EGC bonding the racking to the combiner then it seems the earth is the only ground fault current path. It is possible that a ground fault could occur within the array, the GFDI in the inverter would never detect it, and the equipment would become energized, posing a serious shock hazard, in addition to risk of fire.