Is conduit required in attic for microinverter system? And lots of other code Q's.

[Johann]

[QUOTE=foo1bar;153977]PV DC wires require metal conduit inside the building. (MC might satisfy that requirement - I'd have to look that up)


Doesn't MC stand or mean ''metal conduit'' even if it is flexible ?

[inetdog]

[QUOTE=Johann;154006]

PV DC wires require metal conduit inside the building. (MC might satisfy that requirement - I'd have to look that up)


Doesn't MC stand or mean ''metal conduit'' even if it is flexible ?

No, it means metal-clad cable. It includes the line and neutral conductors as part of the assembly.
There is no separate raceway, and for almost all MC the flexible metal jacket is not allowed to serve as the ground (EGC) conductor.

FMC is Flexible Metallic Conduit.
LFMC is Liquid-tight Flexible Metallic Conduit.
You supply the wires that go in it.

BX is an obsolete or proprietary name for a type of MC.

[giskard]

If the main is at the top (either right or left, since the breakers in both columns typically connect to the same common bus bars) then the PV backfeed need to be at the bottom. ...

The goal is to eliminate the possibility of load breakers being installed between the main and the PV breakers in electrical sense.

These 2 statements seem contradictory to me. If you don't want load breakers electrically between the main and PV breakers, then you want the PV breakers to be right next to the mains. <confused>

What would make sense to me is that placing the PV at the end opposite the mains, would minimize the current in the bus bars because the PV current subtracts from the loads.

[giskard]

1/2. When you say "flex conduit" are you referring to MC cable?

What I meant was flexible metal conduit. One of my existing branch circuits from the service panel is FMC. Pls see pics of wires coming out of my service panel.

Here you can see all the wires going up to the garage ceiling.
Copy of IMAG2995.jpg
Note there is 1 flexible metal conduit, 2 solid conduit, 4 Romex-looking cables with the color of a shiny gray snake, 1 yellow Romex, and 1 black Romex looking cable.

Here are the wires that come out the top. There are 2 more wires, coming out the bottom. (no closeups)
Copy of IMAG2992.jpg


My main Q is, why would flexible metal conduit be OK but no metal conduit at all, be OK?
And why can't I use Romex for the Enphase circuit? (6 of them, ~5.5A total)

[giskard]

5. If you have a single ground rod the new code is to have 2 rods, separated by at least 5' (I think, or maybe 6'?). The inspector will want to see this and it's a good idea, helps to kill stray voltage coming in on your neutral wire.

And this 2nd rod 5 ft away needs to be wired all the way back to the service panel same as the original rod?

6. Yes, for that small load connecting to a single 2 pole breaker is the way to do it, not a pair of 240V breakers!

You are correct, I meant a 2-pole 15A breaker.

[giskard]

I guess I should rank my questions based on hassle/expense of over-building:

1) Do I need to rip out the 1 flexible conduit circuit in the above photo, coming out of my service panel?

2) Can I use Romex in the attic space from the service panel to a spot in the attic just below the waterproof hole in the roof where the Enphase bus cable comes through? If not, why is Romex OK for 15A lighting circuits, but not for a 15A AC circuit going to Enphase inverters that will only put out 5.5A?

3) Do I need to drive in a 2nd ground rod 5 ft away from my existing one?

4) Do I need to upgrade my existing 1/2" ground rod to a 5/8" one?

5) San Jose, CA doesn't see many lightning storms compared to many other areas of the country. Would grounding the PV mounting rails by dropping a wire down the nearest side of the house to panels, be highly recommended? I just noticed that the satellite dish installed by a DirecTV tech is NOT grounded. Did he violate code? FWIW the house has vents for the heater, and those "chef hat" spinning turbine ventilators which are higher up than the panels will be, or the satellite dishes. Are those things grounded for lighting discharge?

[JFinch57]

And this 2nd rod 5 ft away needs to be wired all the way back to the service panel same as the original rod?

Yes, and I had the inspector in NJ tell me that it had to be a continuous wire, through the first clamp and ending at the second. Your electrician will know the requirements.

[JFinch57]

1) Do I need to rip out the 1 flexible conduit circuit in the above photo, coming out of my service panel?

2) Can I use Romex in the attic space from the service panel to a spot in the attic just below the waterproof hole in the roof where the Enphase bus cable comes through? If not, why is Romex OK for 15A lighting circuits, but not for a 15A AC circuit going to Enphase inverters that will only put out 5.5A?

3) Do I need to drive in a 2nd ground rod 5 ft away from my existing one?

4) Do I need to upgrade my existing 1/2" ground rod to a 5/8" one?

5) San Jose, CA doesn't see many lightning storms compared to many other areas of the country. Would grounding the PV mounting rails by dropping a wire down the nearest side of the house to panels, be highly recommended? I just noticed that the satellite dish installed by a DirecTV tech is NOT grounded. Did he violate code? FWIW the house has vents for the heater, and those "chef hat" spinning turbine ventilators which are higher up than the panels will be, or the satellite dishes. Are those things grounded for lighting discharge?

1- I'm not an electrician, but I'd say that it looks OK if properly fastened. I can't imagine the inspector flagging you on it.

2- Don't think you can use Romex, someone else posted that PV cables have to be in conduit. Not sure if you can run the rubber Enphase cable through the roof. I'd put a junction box fastened to the rail and run 1/2" watertight flexible conduit to a junction box in your attic.

3- Yes, you need another ground rod, I originally thought 5', maybe 6', someone else posted the requirement is 6'.

4- 5/8" ground rod does have more surface area and works better, but I think you can leave it. Check with your electrician.

5- Anywhere can be hit by lightning, especially once you fasten a ground wire, even though smaller than lightning protection wire, it's still ground. Lightning is attracted to ground and doesn't care that the feed wire may not carry the load. If lightning hits your panels and they have only equipment ground you will most likely burn out your panels, inverters and feed wire, basically a pile of toast. By providing a better ground the lightning will be led in that direction, POSSIBLY saving some or all of your equipment. Be sure to notify your insurance company of the added solar panels. If they're covered on your policy you should be covered for a lightning strike, verify it. DirecTV typically grounds their dishes to the house wiring ground with a small wire that's molded into the coax. The dishes are cheap, so no big deal if one gets fried. Roof vents and spinning turbine ventilators are not grounded to begin with and therefore do not attract lightning. That's not to say the the lightning couldn't go through one to pick up a ground wire in the attic, but unlikely.

[russ]

1- I'm not an electrician, but I'd say that it looks OK if properly fastened. I can't imagine the inspector flagging you on it.

Better to not comment then. You are posting WAGs and conjecture.

[giskard]

"Don't think you can use Romex, someone else posted that PV cables have to be in conduit."

Isn't that requirement for DC cabling? I have microinverters, so they're AC, and do not have any voltage when they do not sense the mains (i.e. when the breakers are opened).

[sdold]

And why can't I use Romex for the Enphase circuit? (6 of them, ~5.5A total)

I'm curious about this too. I looked through my 2008 NEC and 2010 CEC and can't find any reason not to use NM AKA Romex. The NM part of article 300 doesn't seem to have anything that would keep you from using it for this application. 690.31(A) requires "Photovoltaic source and output circuits" over 30V to be in conduit. Since "Photovoltaic source and output circuits" are shown to be the DC before the inverter in 690.1(A), I don't think that applies to the inverter outputs and we are free to use Romex in the attic for the AC back to the main panel.

Anyone know if this is correct? BTW check out the "Soladeck" for a transition through the roof, if you have comp shingles, a friend used it and it seemed like a good method.

[foo1bar]

"Don't think you can use Romex, someone else posted that PV cables have to be in conduit."

Isn't that requirement for DC cabling?

It is a requirement for DC cabling to the inverter (if inside the structure).
Whether it also applies to the microinverter output is a separate question.

My *guess* is it does - BUT you can check with enphase - they should have an opinion.
And you can present that opinion to the AHJ if you want to see if it'll convince them. The AHJ is really who matters - they're who makes the judgement.

[giskard]

And you can present that opinion to the AHJ if you want to see if it'll convince them. The AHJ is really who matters - they're who makes the judgement.

Isn't that like lining up at the DMV, waiting for service with a snarl?

[foo1bar]

Since "Photovoltaic source and output circuits" are shown to be the DC before the inverter in 690.1(A), I don't think that applies to the inverter outputs and we are free to use Romex in the attic for the AC back to the main panel.

There's at least some who say it is a "output circuit" and should be in conduit:

http://forums.mikeholt.com/showthread.php?t=129866

Isn't that like lining up at the DMV, waiting for service with a snarl?

The people working at my AHJ have been very polite and helpful.
But I'm sure it varies from town to town. (and probably depends a lot on how you ask too)

[giskard]

There's at least some who say it is a "output circuit" and should be in conduit:

http://forums.mikeholt.com/showthread.php?t=129866

They're treating NEC code like words from Mt. Olympus that you can't question and with which inspectors do a lot of hand-wringing interpreting the mysterious words of the gods. Seems to me from a safety point of view, conduit is necessary for DC that's always live, but unnecessary for micro-inverters because they go dead when the breakers are switched off. I mean, why would Romex be OK for 15A lighting circuits, but not for micro-inverters? It makes no sense.

In my particular field of engineering, we have to follow design safety regulations, and it's so much easier to get clarification from the (competing, multiple) Safety agencies, than the (monopolistic) NEC, interpreted by the (monoolistic) praetorian class of inspectors.

I just had a thought - does the wiring from a central inverter to a subpanel need to be inside conduit? If not, why would micro-inverter outputs need to be?

If indeed it needs to be in conduit, does the conduit have to be solid metal, or is flexible metal conduit (FMC) OK? Is there a reason FMC is "bad"? Is it because rats can chew through them? I heard something about FMC not being low impedance to fault currents, but that's silly if the conduit has a ground wire in it that will carry said fault currents... and again, silly that Romex is OK but FMC is not.

One guy in the mikeholt link wrote "I cannot make the logical distinction between the two types, unless those microconverters also have built in overload and short circuit protection. (Which I do not know if they do) "
YES THEY DO. I design power electronics for a living. Overload and short circuit protection is standard. And micro-inverters go completely dead in the absence of mains voltage to begin with.

AFAIK the "IG" versions of Enphase microinverters have internal double insulation, which means by design, (and the failure modes are tested by UL et al), they can't fail in a manner that will cause the PV panel to become live. Appliances and power supplies with double insulation do not need a ground wire to prevent the exposed metal from going live in the event of a failure. You can spot them by looking at their AC cables - they have 2 wires with no ground, and the plug has no 3rd ground prong. Cellphone chargers are like this. No ground prong. So IMO "IG" Enphase inverters do not need a ground wire for Safety - only the bracketry needs to be grounded, for lightning protection. If the code or the inspectors don't reflect this, IMO that is wrong.

Classic way that regulations are badly written or enforced, increasing costs all around. $10, here, $100 there, pretty soon we're talking real money.
</rant>

The people working at my AHJ have been very polite and helpful.
But I'm sure it varies from town to town. (and probably depends a lot on how you ask too)

So how does this work? Is there an office with friendly inspectors just sitting around waiting for common folk like me to ask stupid questions? Do you make an appointment and pay some hourly fee?