Yet another grounding question.

[msheets]

I am assisting with some design questions on a largish ground-mount system. A question has come up regarding the size of the GEC at each row. We are using NEC 2020.

The support structure is a single-axis tracker with driven piles with ~11' embedment. Most rows hold 72 modules with 10 piles per tracker row. The piles in each row are bonded so that the piles are electrically continuous, and each row is bonded to a neighboring row through a CAB messenger wire. They have #8 CU wire from the inverter to the array, but that PV output circuit combines multiple strings which may not all be on the same row.

The Engineer of Record originally specified that a ground rod be driven at each tracker row, with #6 CU bonding the tracker pile to the ground rod. The installer is questioning this since each driven pile is already a grounding electrode, complying with 250.52(2), making the ground rod superfluous.

Per 690.47(B), the GEC should be sized per 250.66, but I'm not sure exactly how to interpret that for this type of system. None of the conductors are truly parallel, so do we size it off a single #8 CU, making the GEC #8 CU (probably #6 since it's exposed)?

 

[jaggedben]

I personally agree with installer that the engineer's spec is pointless, (but I'm not an engineer). 690.47(B) says that the rods are permitted though. Your last paragraph about the size of your DC conductors is unclear, but I would say that the GEC should be sized to 250.66 based on the largest DC conductor present at the given array. It sounds like "making the GEC #8 CU (probably #6 since it's exposed)" is probably correct, from the limited info you've given. It might vary per array depending on how those combined output circuit conductors are arranged.

 

[msheets]

690.47(B) says that the rods are permitted though.

Permitted, yes. Would they be required without increasing the size of the GEC, though? I suspect that the ground rods were included so they could use a #6 GEC, per 250.66(A), without having to think about it too much (as I apparently might be doing now). This might be getting a little too much into the weeds, but what is it about rod, pipe, and plate electrodes that allow the smaller GEC? Why can't we use that rule for other types of grounding electrodes?

Your last paragraph about the size of your DC conductors is unclear, but I would say that the GEC should be sized to 250.66 based on the largest DC conductor present at the given array. It sounds like "making the GEC #8 CU (probably #6 since it's exposed)" is probably correct, from the limited info you've given. It might vary per array depending on how those combined output circuit conductors are arranged.

For clarification, the modules are First Solar (high voltage, low current), with 6 modules per string. They are using wire harnesses to combine up to 5 strings at the array. Each home run is #8 CU. The inverters are in a centralized arrangement near the AC gear, so the largest DC conductor anywhere in the system is the #8 CU PV output circuit described above.

 

[jaggedben]

If one or more of the piles complies with 250.52(2) like you said, the rods are not NEC required. I don't understand what you mean by "Would they be required without increasing the size of the GEC?". Whether or not they are required, if you decide to put them in then 690.47(B) has some requirements about GEC size.

I think the reason single rods and such have a limit on the max GEC size is some theory about the max current a single electrode can carry. Not that it is very scientific or has any relevance in a situation like this, where the electrode probably does nothing. ( I think they just don't want to make the code longer and longer with more granular requirements and exceptions.) Probably in this case a #8 does nothing that a #14 doesn't accomplish. But 690.47(B) says 250.166, so #8. But now I'm noticing that 690.47(B) does NOT refer to 250.64 so arguably the GEC does not need to be #6 even if is 'exposed'. Regardless, since according to your info this is all basically job spec required by the engineer rather than the NEC, better ask the engineer what they want.

 

[msheets]

But now I'm noticing that 690.47(B) does NOT refer to 250.64 so arguably the GEC does not need to be #6 even if is 'exposed'.

690.47(A), which also applies to this installation, indicates that the grounding electrode system should be installed in accordance with Part III of Article 250 (250.50 through 250.70).

Regardless, since according to your info this is all basically job spec required by the engineer rather than the NEC, better ask the engineer what they want.

The installer is going to submit an RFI to the EOR. We'll see what they say.

Thank you for your insights.

 

[jaggedben]

690.47(A), which also applies to this installation, indicates that the grounding electrode system should be installed in accordance with Part III of Article 250 (250.50 through 250.70).

Yes but if the structural steel qualifies as an electrode that takes care of that. And that section only requires you to bond it to the PV EGC.

 

[asif.k]

I would agree with the Installer.
Moreover, in my view, there is no need of making every pile of each row as Grounding Electrode. Take one pile of each row as Grounding Electrode. However, this single pile shold be connected to the main Grounding System or Electrode of the project.

In regard to wire size, wire from the array to the inverter is probably inside raceway so he used it #8, otherwise it should be #6.
Similarly, for GEC it shall be at least #6 if subject to physical damage otherwise according to Table 250.66.