NEC 690.12 - RSID's for Solar Carport with Inverter AC Output Conductors Run Over a Building to Electrical Room - What Solution will Satisfy 690.12?

[jaggedben]

...There's nothing vague or ambiguous about the definition of "building" in the NEC.

The issue is just that the definition doesn't match the (presumed) expectations the writers of 690.12 had, nor the expectations of the AHJs applying 690.12. ...

In the real world, those statements amount to a contradiction.

 

[ggunn]

I'm more or less with you on your comments except on this point as it relates to 690.12. There's nothing vague or ambiguous about the definition of "building" in the NEC.

The issue is just that the definition doesn't match the (presumed) expectations the writers of 690.12 had, nor the expectations of the AHJs applying 690.12. So that's fine, they can collectively decide that the definition is wrong, or 690.12 is written wrong given the definition, and apply their collective judgement to enforce a more reasonable rule. Or perhaps most often, never even look at the NEC definition of "building" and just use the definition they're used to.

Basically this is just a case of everyone collectively agreeing (perhaps unknowingly) to ignore Charlie's rule. Until someone made a PI for the 2023 NEC for 690.12 that says (I'm guessing, I didn't read it) "hey this doesn't actually say what we've all been thinking it says, so maybe we should fix that."

Given all that, I would say that you're correct in your advice to the OP to just assert in the permit application that the carport is not a building and so 690.12 doesn't apply. And then only if the AHJ pushes back, look to convince them in fact it's not a building because the support structure is all equipment, or to allow the use of not-yet-adopted 690.12 Exception 2, or move the inverter outside the footprint of the carport, or whatever.

Cheers, Wayne

Whatever. We were building PV AC combiners that conform to what is now 705.12(B)(3) before the code allowing them was added to the NEC because requiring them to conform to the "120% rule" was just plain stoopid and everyone knew it. Charlie's Rule notwithstanding, a modicum of common sense is necessary in interpreting the NEC.

Rapid shutdown was brought into the NEC to protect firefighters who are on a roof chopping holes in it to ventilate the building below it. Obviously that won't happen with a parking shade irrespective of whether or not one might interpret the NEC as defining it as a building. Exception 2 merely codifies what we already knew.

 

[wwhitney]

In the real world, those statements amount to a contradiction.

No. "Vague" and "ambiguous" still aren't the correct words to use when there is a mismatch between the NEC definition and the user's expectations. The NEC definition is clear enough. Probably "unexpected" is the best description of the NEC definition of "building". Or depending on your attitude, maybe "stupid" : -)

But anyway, I don't think there's much substantive left to discuss about the use of the word "building" in 690.12. Good to remember that the 2023 NEC added Exception 2.

Cheers, Wayne

 

[ggunn]

There's nothing vague or ambiguous about the definition of "building" in the NEC.

You called it "silly". I will go with that.

 

[pv_n00b]

When I read this, "Solar carports are generally considered ground mounts, and not subject to the rapid shutdown requirements of NEC 690.12." I knew that the discussion would quickly descend into, is the carport a structure or not, and the OP's question would be forgotten.

"My concern is that the AHJ may require RSID's on the DC side of the system, because these are specifically "listed" as RSID's."
This is a valid concern but one that will only be alleviated by talking to the AHJ. I would say most would not require it but you need to know before you design and build it. Go talk to the AHJ.

 

[jaggedben]

When I read this, "Solar carports are generally considered ground mounts, and not subject to the rapid shutdown requirements of NEC 690.12." I knew that the discussion would quickly descend into, is the carport a structure or not, and the OP's question would be forgotten.

"My concern is that the AHJ may require RSID's on the DC side of the system, because these are specifically "listed" as RSID's."
This is a valid concern but one that will only be alleviated by talking to the AHJ. I would say most would not require it but you need to know before you design and build it. Go talk to the AHJ.

I mean, your latter paragraph is certainly apt if the prior issue can't be resolved.

 

[ggunn]

When I read this, "Solar carports are generally considered ground mounts, and not subject to the rapid shutdown requirements of NEC 690.12." I knew that the discussion would quickly descend into, is the carport a structure or not, and the OP's question would be forgotten.

"My concern is that the AHJ may require RSID's on the DC side of the system, because these are specifically "listed" as RSID's."
This is a valid concern but one that will only be alleviated by talking to the AHJ. I would say most would not require it but you need to know before you design and build it. Go talk to the AHJ.

Of course. It is always best to confer with the AHJ before building anything that is the slightest bit controversial. It is far better to resolve it beforehand than to fight about it at final inspection.

 

[Elect117]

I had a couple of follow up questions, and I could be totally out of place here. I am not the best with 690. I think it's the language of the section.

Are the AC conductors from the carports, up the wall, on the roof, and then entering the building to the electrical room circuits considered outside of the array? Or is that only pertaining to the DC output conductors of the system? 690.12(B)(1)

If they are supply side tapped, is a disconnect required within the electrical room or could they theoretically have 6 or less breakers at a combiner panel at the carport?

If it is supply side tapped, would they need to encase the conduit in 2in of concrete to remain outside of the building? Like a fire pump?

 

[ggunn]

I don't think that any part of 690.12 applies, but as we have been saying, you'll want to confirm that with the AHJ.

If it is supply side connected, then yes, you will probably need a disco in the electrical room and most likely another on the outside wall. Only the one closest to the Point Of Interconnection (POI) needs to be fused.

The underground (UG) conductors from the inverter(s) to the building need only be buried in compliance with 300.5.

 

[solarken]

2) Moving the inverters outside the array boundary to escape 690.12(A)(2) is a great idea. Now I need to evaluate the space availability and cost difference between a separate fenced off inverter area versus adding RSID's to the system.

690.12 Rapid shutdown has nothing to do with your inverter output circuits when inverters are mounted on columns supporting a solar carport, regardless if the AHJ considers that 690.12 rapid shutdown applies to the structure mounted array or not. The inverters mounted on the columns are not in the boundary of the array, they are at a lower level and may also happen to be under the array, which is different. The controlled conductors described in 690.12(A)(2) only apply to inverter output circuits when using microinverters or string inverters that are mounted in the boundary (and in the plane) of the array. There is no need to move the inverter mounting locations off the columns.

 

[wwhitney]

The inverters mounted on the columns are not in the boundary of the array, they are at a lower level and may also happen to be under the array, which is different.

Well, that's the question isn't, are the inverters under the array, or within the array?

2020 NEC 690.12(B) says "The use of the term array boundary in this section is defined as 305 mm (1 ft) from the array in all directions." All directions clearly includes vertical, so if we have an elevation at which all of the array is above that elevation, and the inverter is at least 1 ft below that elevation, the inverter is outside the array boundary.

But the definition of Array is "A mechanically and electrically integrated grouping of modules with support structure, including any attached system
components such as inverter(s) or dc-to-dc converter(s) and attached associated wiring." That says the "support structure" is part of the array, as is any inverter attached to that "support structure".

So if the column is part of the array "support structure," an inverter that is attached to the column is part of the array. If you mount the inverter on a separate post that is 1 ft away from the array and its support structure in all directions, then the inverter would be outside the array boundary, even if the array encloses the inverter.

Here it would actually be useful to argue that the carport solar structure is in fact not just a ground mount array (where all the supports are clearly part of the array support structure) but is in fact a carport that has an array mounted on it. And that therefore the columns are not supports specific to the array but are carport supports.

Which is fine as long as you're under the 2023 NEC and have Exception 2 to 690.12, or you and your AHJ are interpreting "building" as "enclosed building" in applying an earlier version of 690.12.

Cheers, Wayne

 

[solarken]

Here it would actually be useful to argue that the carport solar structure is in fact not just a ground mount array (where all the supports are clearly part of the array support structure) but is in fact a carport that has an array mounted on it. And that therefore the columns are not supports specific to the array but are carport supports.

Good points. In my view this is the case. A carport solar gen system is a carport, meant to provide a place to park and protect cars, that just happens to have a roof that is made out of a solar array instead of conventional roofing. The PV array is the modules, the rails, the clamps, the PV string wiring, and any MLPE. The rest of the structure would exist if it were a regular carport with a conventional roof and not a solar carport. I suppose it is possible to mount a string inverter at a height of less than a foot below the rails of the array, but I don't know why anyone would ever want to do that, and most inverters require a clearance above them that would tend to discourage that anyhow.

 

[jaggedben]

If a building (one with walls and a roof) with a rooftop solar system were part of the array by virtue of being part of the panels 'support structure' then the array boundary definition - indeed the whole purpose of defining an array bounday - would make no sense. So I think it's obvious that the building is not part of the array, and similarly one can distinguish between the columns supporting a carport array and the array itself. If I had to draw a firm line I would probably draw it at the supports (rails or whatever) that directly support the modules.

 

[wwhitney]

If a building (one with walls and a roof) with a rooftop solar system were part of the array by virtue of being part of the panels 'support structure' then the array boundary definition - indeed the whole purpose of defining an array bounday - would make no sense.

Sure, but what do you have if you remove the panels from the carport array? Something that nobody would have any reason to build. Not true of most buildings (in the usual, non-NEC sense).

So I think it's obvious that the building is not part of the array, and similarly one can distinguish between the columns supporting a carport array and the array itself.

I think there's a progression of possibilities:

1) Ground mount array
2) Higher ground mount array that a car can drive under
3) Pre-existing carport to which panels have been added (e.g. there's a metal roof under the array)

For (1), the support structure is obviously all part of the array, per the definition. I have trouble with the idea that just raising up the panels by lengthening the columns means that now none of the support structure is part of the array.

Cheers, Wayne

 

[jaggedben]

Sure, but what do you have if you remove the panels from the carport array?

Um ... A carport?

I think there's a progression of possibilities:

1) Ground mount array
2) Higher ground mount array that a car can drive under
3) Pre-existing carport to which panels have been added (e.g. there's a metal roof under the array)
....

Yeah, i don't buy that 'pre-existing' has any significance; there is nothing in any relevant code to support that. I think an AHJ can make a reasonable determination whether the entire structure is the support structure for the array or not.

 

[pv_n00b]

Of course. It is always best to confer with the AHJ before building anything that is the slightest bit controversial. It is far better to resolve it beforehand than to fight about it at final inspection.

And yet time and time again I see people's plan is; I'll just ask for forgiveness and the AHJ will accept what I have done because it's too hard to fix and I'm pretty sure the AHJ will not be that mean.

 

[wwhitney]

Um ... A carport?

I've been assuming in this "carport array" that there is no roof underneath the solar panels, and the solar panels are the only "roof". If that's the case, you don't have a carport after removing the panels, you just have some columns and some beams.

If you do have a roof under the panels, then I agree it's a carport, and none of that carport is the array support structure, same as case (3).

Cheers, Wayne

 

[pv_n00b]

Sure, but what do you have if you remove the panels from the carport array? Something that nobody would have any reason to build. Not true of most buildings (in the usual, non-NEC sense).


I think there's a progression of possibilities:

1) Ground mount array
2) Higher ground mount array that a car can drive under
3) Pre-existing carport to which panels have been added (e.g. there's a metal roof under the array)

For (1), the support structure is obviously all part of the array, per the definition. I have trouble with the idea that just raising up the panels by lengthening the columns means that now none of the support structure is part of the array.

Cheers, Wayne

Added a few more for spice.

1) Ground mount array
2) Ground mount array with sheep under it
3) Higher ground mount array with farming under it
4) Slightly higher ground mount array with asphalt surface underneath that a car can drive under
5) Pre-existing carport to which panels have been added (e.g. there's a metal roof under the array)
6) Parking garage with PV on top level. All the parking garage does is hold up the PV and provide a place for cars to drive under it.

 

[pv_n00b]

I spent 20 minutes a week or so ago arguing with a room full of AHJs who were sure that a transformer, panelboard , and EV charger pedestal in a parking lot were a "structure" and needed a local disconnect on the transformer primary. That's how bad the NEC definition of structure is. In 2017 "other than equipment." had to be added to the definition of structure because of interpretations like this and I still have to argue with AHJs.

 

[ggunn]

And yet time and time again I see people's plan is; I'll just ask for forgiveness and the AHJ will accept what I have done because it's too hard to fix and I'm pretty sure the AHJ will not be that mean.

You won't find any plan of mine like that. Fortunately, most of the commercial PV systems I have designed had to go through a plan review by the AHJ where any issues were hashed out before any equipment was bought or anything was built. There is one notable exception to that, but the AHJ was amenable to having preconstruction meetings to work out any questions.