Irreversible splice requirement in PV junction box.

H

hex4def6

Member
Location
San Jose, CA
Occupation
HW Engr
Hi all,

Doing a solar PV install in California, and have a question about grounding (GEC + EGC).

The house is a typical 1980s style build. I'm not sure where the existing ground ties to, but I'm assuming it's the water line.

The PV install in question consists of Enphase microinverters (DC -> 240VAC at a per-panel level), on aluminum racking. The proprietary 2-conductor Enphase 240VAC cable lands in a metal roof-mounted junction box (Soladeck), and feeds down to the Enphase Combiner through EMT.

The microinverters are ungrounded / double-insulated. Therefore the only grounding of note is the railing (on the roof).

This is done with #6 between rails, with ironridge grounding lugs, to the Jbox. Inside the Jbox, there is a ground bar, on which the #6 from the rails, and a #10 THHN which feeds back to the combiner, and a wire from a grounding bushing on the EMT all land. This EMT feeds into the Enphase Combiner box on the side of the house.

1. According to the AHJ inspector, using a ground terminal block inside the jbox on the roof is not acceptable, and they need to be irreversibly spliced together. Is this true? Any code I can quote to the contrary? Isn't this a GEC requirement rather than a EGC requirement?

2. My plan is to add two grounding rods (obv, irreversibly spliced here), 6ft+ apart, to satisfy current code, and land that in the existing SE panel, which has a ground terminal block. The Enphase Combiner (physically next to the service entrance), then has a grounding conductor also land on that terminal block. Do I need two, or can the existing water pipe bond count as one of the two (assuming I'm not ohming it out for the <25-ohm single rod req)?



Ref:

PV enphase microinverter

Does IQ7-60-2-US micro inverter 240V single phase require neutral brought to them? Anyone have manufacturer datasheet or instruction that says neutral is not required?
forums.mikeholt.com
 
1) 250.102(B) bonding conductors can be attached using 250.8 (terminal bars). I don't know if there is something I am missing though. I always thought the #6 is for bonding metal parts of the cells ahead of the inverters. I have never really gotten clarification on it.

2) The way the code is written, when using a ground rod as a grounding electrode it always requires a second rod (supplemental electrode) be installed, unless you meet the less than 25 ohms test.

Bonding for Roof Top Solar Array

Roof top solar array: Enphase microinverters kicking out AC (two conductor). Solid copper EGC bonded to racks. EMT sleeve runs from East to West side of roof and between strings on west side. EMT does not terminate in a box (i.e., open ends). Obviously the EMT runs must be bonded to the EGC...
forums.mikeholt.com
 
Elect117 said:
2) The way the code is written, when using a ground rod as a grounding electrode it always requires a second rod (supplemental electrode) be installed, unless you meet the less than 25 ohms test.
Click to expand...
More precisely, when the ground rod is a required electrode. If you are using the ground rod itself as a supplemental electrode the 25 ohm requirement does not apply.
 
GoldDigger said:
More precisely, when the ground rod is a required electrode. If you are using the ground rod itself as a supplemental electrode the 25 ohm requirement does not apply.
Click to expand...
If you're implying that a water pipe electrode, plus a single ground round as the supplemental electrode, always suffices, that is incorrect. The supplemental ground rod for the water pipe must itself meet the 25 ohm requirement or you must add a second ground rod (or other non-water pipe electrode) to it.

(2017) NEC 250.53(D)(2) has the language requiring a supplemental electrode for a metal underground water pipe. It says "If the supplemental
electrode is of the rod, pipe, or plate type, it shall comply with 250.53(A)." And 250.53(A)(2) has the language requiring a supplemental electrode for a rod, pipe or plate type. It says that the supplemental electrode shall be "of a type specified in 250.52(A)(2) through (A)(8)." That list notably omits 250.52(A)(1), which is the metal underground water pipe.

So the metal underground water pipe may not serve as the 250.53(A)(2) supplemental electrode for a rod, pipe or plate electrode.

Cheers, Wayne
 
Thanks!
So unless I want to do the actual GND test, I should just suck it up and install two rods , >6 ft apart.


Just to be clear on the first point (and the bigger issue tbh, since it requires me lifting panels again... ):

I see I'm not the only person with this issue, and it seems to stem from whether it's considered an EGC or a GEC:

Irreversible Splice+Junction Box+solar

Hi guys, I am new here so bare with me. I had an inspection today for a ground mount PV system. We failed because we did not use An irreversible splice inside the junction box. We have a continuous Bare copper run connecting the modules, rails, the structure and two ground rods located at...
forums.mikeholt.com

It is not a GEC, correct? Therefore, irreversible splices are not relevant.
 
What rock has this inspector been living under for the last 10 years? Yes there used to be a lot of belief that you needed to treat the ground from the roof as a GEC and for early Enphase micro-inverters that was actually true because there was a bona-fide GEC. But Enphase changed those inverters around 2012 and the 2014 NEC clarified that an ungrounded DC system only needs an EGC. You do not have a GEC from the roof.

I met one AHJ who cited 250.4 for wanting a GEC to the array. I gave up on him. Good luck.
 
Hmm,

If it were a GEC, this is the thing that would trigger (250.64 (C) I assume? ):
Except as provided in 250.30(A)(5) and (A)(6), 250.30(B)(1), and 250.68(C), grounding electrode conductor(s) shall be installed in one continuous length without a splice or joint. If necessary, splices or connections shall be made as permitted in the following:
(yada yada, irreversible connections)


But you're right. I think I'll try quoting this from the enphase install manual:

The microinverter models listed in this guide do not require grounding electrode conductors (GEC), equipment grounding conductors (EGC), or grounded conductors (neutral). Your Authority Having Jurisdiction (AHJ) may require you to bond the mounting bracket to the racking. If so, use UL2703 hardware or star washers. The microinverter itself has a Class II double-insulated rating, which includes ground fault protection (GFP). To support GFP, use only PV modules equipped with DC cables labeled PV Wire or PV Cable.


Thus, the only thing needing grounding is the racking, and that's through an EGC.

How to convince the dude, on the other hand... Sigh.
 
Unfortunately it did not do the trick...

The guy mumbled something about irreversible splices being required in rooftop junction boxes according to the California code but couldn't give me the actual code reference (he said he'd get back to me on that).

(I'm betting it's going to be 250.64)
 
Joe.B said:
2022 CA Electric Code Section 690. Specifically 690.41 System Grounding, and 690.47 Grounding Electrode System.
Click to expand...
Neither of which requires a GEC or any irreversible splices. In particular, 690.47(A)(1) says all you typically need is an EGC:

"For PV systems that are not solidly grounded, the equipment grounding conductor for the output of the PV system, where connected to associated distribution equipment connected to a grounding electrode system, shall be permitted to be the only connection to ground for the system."


Cheers, Wayne
 
Agreed. I was just providing the code sections in case someone needed to look it up. Say, maybe the inspector? Just saying...
 
It is a huge misconception. Been using wire nuts or split bolts in J-box for 20 years.
In CA.

It is an EGC.
There is no such irreversible requirement.
 
Also, #10 bare copper under panels, rail to rail. As long as covered and protected from physical damage. One guy who was having a literal meltdown demanded #6 but I don't work there any longer. Small city so I can afford to skip it, and I don't need the stress from the local AHJ Hitler.
 
Pass one or two certifications and call it good. Met the minimum job requirements. Those tests are hard, must know everything.

In my opinion one certification a year is bare minimum. Two a year is very reasonable and shows you're continuing to grow and learn. Any inspector who's out there acting like they're the ultimate expert on everything is most likely over-compensating.
 
Joe.B said:
Pass one or two certifications and call it good. Met the minimum job requirements. Those tests are hard, must know everything.

In my opinion one certification a year is bare minimum. Two a year is very reasonable and shows you're continuing to grow and learn. Any inspector who's out there acting like they're the ultimate expert on everything is most likely over-compensating.
Click to expand...
Yeah. They should also not act like the code never changes after they become an inspector.
 
You must log in or register to reply here.
Top