How to Center feed Emphase IQ7+ from a soladeck?

[ChrisBarrett]

If I end up using conduit, I'd definitely prefer to use PVC over having to bend EMT up in an attic. That would be terrible lol.

[bob-n]

Again, I'm not an electrician, but I'm pretty sure that NM is not allowed and that conduit is required. An electrician will be able to quote the code on this. You can't stuff NM into conduit, either. You must use THHN or the appropriate spec for your situation. Most wire is multi-use specified (THHN & THWN or THWN-2).

Also, conduit must be marked every 8 feet and 1 foot from each bend and fitting. If interior, the markings must be reflective.

[ChrisBarrett]

I will check over 690 again, hopefully tomorrow. I feel like I read some caveats that *might* allow me to get away with it I'll have to prove that - but I'm not even sure if I want to prove it. It could be that installing conduit would be easier than running NM through the walls ...

[foo1bar]

You can't stuff NM into conduit, either. You must use THHN or the appropriate spec for your situation. Most wire is multi-use specified (THHN & THWN or THWN-2).

NM is allowed in conduit inside. It isn't wet-rated, so isn't supposed to be in outside or underground conduit (or outside junction boxes)
It usually is cheaper and easier to use THHN. (ex. can fit more wires into same size conduit, easier to pull, etc)

[SunEagle]

I will check over 690 again, hopefully tomorrow. I feel like I read some caveats that *might* allow me to get away with it I'll have to prove that - but I'm not even sure if I want to prove it. It could be that installing conduit would be easier than running NM through the walls ...

Just remember that your local code approvers or AHJ may have rules that are stricter then the latest NEC. So even though you meet the NEC you may still not meet the AHJ requirements.

[ChrisBarrett]

Okay it took me a little longer, but I did formulate an email to the inspector about this:

690.31(A) requires "Photovoltaic source and output circuits" over 30V to be in conduit or raceway, where "in a readily accessible location". However, 290.31(B) appears to apply only to PV System DC circuits.

Given that our attic is not readily accessible, and that 690.31(C)(3)(2) allows for multi-jacketed conductors to be secured to the building structure I feel like this might be permissible. However, I am using Soladeck roof mounted enclosures for each array and the NM 10/2 would terminate at terminals in the enclosure. The soladeck is flush with the roof surface, flashed, and waterproof. However - I am not sure if the codes permit NM to terminate in an external junction box.


Hopefully he has time to get back to me - if not I'll just apply for permits and get shot down if he doesnt like the wiring solution.

[ChrisBarrett]

The next question I have is about grounding the system - I'm not sure if the #10 romex ground will be sufficient for each array to ground it for lighting strikes? The Enphase IQ7+ says:

"The Enphase Microinverter models listed in this guide do not require grounding electrode conductors (GEC), equipment grounding conductors (EGC), or grounded conductor (neutral)."

However, I still need to ground the rest of the metal of the system. The wiring diagram calls for #8 bare copper along the racks, then tie into the romex.

[bob-n]

There are two issues for grounding. First is code. I can't recall the details. Others will have the facts, but it requires that the ground conductor be as large as other power conductors. I've heard of people using bare #6. There's nothing wrong with using green insulated wire (THHN), either. It would be marginally better protected at little difference in cost.

The second issue is related to the fundamental purpose of grounding the racks. If you get nearby lightning, it will induce current into your rack structure. It doesn't have to be a direct hit to cause very high current surges. You want that current to go right to the ground rod in the soil so any voltage drop along the way does not couple voltage drop into other systems. For that reason, I think you want one large ground wire separate from the wire in your conduit or NM, going right to (first choice) the ground rod or (second choice) to the ground bus in the service panel, which goes right to earth.

To me, lightning protection is one of those analog parameters. You can do more and give yourself more lightning protection, but there is no such thing as bulletproof. A direct hit from a serious storm could burn a building to the ground, even with protection. You can always make a case that more is better. For example, in lightning-prone areas, it might be wise to also have lightning rods on the roof, higher than any panel. Lightning arrestors are another form of protection. The better the arrestor (often more expensive), the more that they can divert safely. You may never need it, or may need it all the time. Lightning arrestors also wear out, and it's extremely hard to know when they need replacement.

[ChrisBarrett]

Yeah, thanks Bob. That's exactly what my concern was - and it seems like code was updated maybe in 2017 to not necessitate a second grounding electrode. I'll see what the AHJ has to say, I sent off an email yesterday.

Currently, the wiring diagram calls for solid #8 to NM 10. We'll see...

[ChrisBarrett]

I verbally spoke to the local inspector over the phone, and sent in all the docs for the permit. Over the phone, he confirmed that romex would be permissible in this installation, but suggested that MC #10 would be ideal - so that's what I'm going to use. He also confirmed the grounding strategy.

It does look like I might need a structural engineering review, so I'm working on that next as well as waiting for the permit to be approved. I received delivery of all the major components, and now I'm waiting for ancillaries like wiring, connectors, roofing safety stuff.

[J.P.M.]

I verbally spoke to the local inspector over the phone, and sent in all the docs for the permit. Over the phone, he confirmed that romex would be permissible in this installation, but suggested that MC #10 would be ideal - so that's what I'm going to use. He also confirmed the grounding strategy.

It does look like I might need a structural engineering review, so I'm working on that next as well as waiting for the permit to be approved. I received delivery of all the major components, and now I'm waiting for ancillaries like wiring, connectors, roofing safety stuff.

If a structural review/approval is required as part of the permitting process, like all other permitting requirements, the permit will be issued after a structural review shows all is per the building code.

[ChrisBarrett]

So it's not part of the permitting process, but it's part of a pre-inspection checklist. Actually found some interesting resources about this topic: https://www.osti.gov/servlets/purl/1326351

Basically lays out the case that code is more than 300% more conservative than actual performance of the structure.

Actual checklist item is:
Roof is designed and capable of supporting the weight of the PV system (IRC Section R324, IBC 1509.7)

[J.P.M.]

So it's not part of the permitting process, but it's part of a pre-inspection checklist. Actually found some interesting resources about this topic: https://www.osti.gov/servlets/purl/1326351

Basically lays out the case that code is more than 300% more conservative than actual performance of the structure.

Actual checklist item is:
Roof is designed and capable of supporting the weight of the PV system (IRC Section R324, IBC 1509.7)

The weight's not the usual controlling factor in strength calculations for PV arrays. The array weight (~ ~ 3 - 4 lbm/ft^2) is peanuts compared to external loadings, which, in the case of wind loadings may be an order of magnitude higher and in the uplift direction which most existing roof designs don't consider.

However, if such external loading considerations are not a mandated part of the permitting process, I'd not bring it up.

[ChrisBarrett]

right, yeah I think the guy running these experiments is just saying, code for roof strength in general is really really conservative - irrespective of solar or any other load.

[J.P.M.]

right, yeah I think the guy running these experiments is just saying, code for roof strength in general is really really conservative - irrespective of solar or any other load.

I'd probably agree that most reasonably constructed roofs are strong enough to support an array. I'd also think that most roofs are OK with respect to most external loadings such as wind with the caveat that most applications need checking for imposed loadings, usually per ASCE 7-16 or other recognized standard.