AC trunk cable transitions between different roof planes

[Cliffs]

Hi-

My AHJ does not allow any exposed roof wiring. I'm doing a microinverter system with multiple physically separated arrays which are on the same AC trunk cable. I need a way to transition the unprotected AC trunk cable (this stuff) to EMT conduit (required by my AHJ).

Does anyone know of a gland seal to EMT coupler? Like this, except EMT on one side and a gland seal, or some way to thread in the gland seal, on the other. This strategy would allow me to convert to EMT without using a box. Just pass the AC trunk cable through the conduit.

Another option would be to use a box, but all I can find are expensive ($50+) rooftop solar junction box. I just need something that mounts to solar racking (not a roof deck), small, waterproof, where I can attach an EMT coupler on one side and a gland seal on the other.

Advice appreciated!

 

[zanydroid]

I just finished a HoyMiles instal 1-2 months ago.

Does your AHJ require the EMT to be a complete conduit or just a sleeve?

You don’t need to use SolaDeck and the like for roof boxes. There are mounting systems for PVC boxes to rails. Unirac E-boss has one (I have a stack of extras that I didn’t end up using). E-boss also has EMT to rail and roof flashing clamps.

 

[zanydroid]

Complete conduit is probably better, it will require a larger diameter conduit to comply with the conduit fill rule.

You will very likely need a box actually because it is needed to break off the EGC from the trunk cable for bonding rails and panels. Unless you want to bond through the EMT.

If you insist to use a gland seal you will have two wire conduit fill limitation (low 30%) and it may not be possible to find a gland that is natively compatible for both a large trunk cable and a EGC. You also have to pull that #6 EGC the whole length of the conduit, OR do something janky like splicing the EGC inside a conduit body.

I’ve seen turnkey installs that transition conduit/EGC into a box, into THHN in the conduit.

As for roof deck vs conduit. It’s an aesthetic, potential penetration leak, and how good you are at bending EMT question. Note that securing EMT on roof deck will often anyway need a penetration for the EMT clamp (supporting EMT may not need one, but it will not be as robust).

 

[Hedges]

There are glands for 1 wire or several that thread into 1/2", 3/4" etc. trade size fittings. I use those into plastic or metal weatherproof boxes.

Or I run the wires in through non-sealing connectors on bottom of box, so any water drips away.

 

[zanydroid]

Another factor for deck box vs conduit is how hot your attic is. Some people would much rather bend EMT than go in there. Installers can typically knock out EMT way faster than using deck boxes.

 

[Cliffs]

Complete conduit is probably better, it will require a larger diameter conduit to comply with the conduit fill rule.

You will very likely need a box actually because it is needed to break off the EGC from the trunk cable for bonding rails and panels. Unless you want to bond through the EMT.

If you insist to use a gland seal you will have two wire conduit fill limitation (low 30%) and it may not be possible to find a gland that is natively compatible for both a large trunk cable and a EGC. You also have to pull that #6 EGC the whole length of the conduit, OR do something janky like splicing the EGC inside a conduit body.

I’ve seen turnkey installs that transition conduit/EGC into a box, into THHN in the conduit.

As for roof deck vs conduit. It’s an aesthetic, potential penetration leak, and how good you are at bending EMT question. Note that securing EMT on roof deck will often anyway need a penetration for the EMT clamp (supporting EMT may not need one, but it will not be as robust).

The point about the EGC & trunk cable in the same conduit is a good one, so I think I will need a small box to combine those two things into one conduit. Worth mentioning that I won't be breaking apart the trunk cable and transitioning to THHN, I'll just be passing it through. If I were not following AHJ rules I would just add some abrasion protection and fix the cable to prevent motion - I would not bother with a box/conduit.

Can you send the link for the E-boss part you have?

Is it allowed to run PV (<60V) and AC wires in the same conduit? How about through same junction box?

 

[zanydroid]

Worth mentioning that I won't be breaking apart the trunk cable and transitioning to THHN, I'll just be passing it through.

Ok then you need to run conduit fill calculator at 31%.

Box hanger: 00802JB-20

EMT clamp: https://www.solaris-shop.com/unirac-e-boss-00802cm-conduit-mount/

Note the EBoss catalog for the clamp will mention the bolt type you need to lock onto Unirac rails

Is it allowed to run PV (<60V) and AC wires in the same conduit? How about through same junction box?

By default no but there are ways around this/special cases. I’m not familiar with them. This shouldn’t come up with your install though unless you are doing something special like passing some DC conductors across arrays to group together panels in separate arrays onto one microinverter. I considered that but it’s not worth the code compliance pain.

 

[Cliffs]

Ok then you need to run conduit fill calculator at 31%.

Box hanger: 00802JB-20

EMT clamp: https://www.solaris-shop.com/unirac-e-boss-00802cm-conduit-mount/

Note the EBoss catalog for the clamp will mention the bolt type you need to lock onto Unirac rails



By default no but there are ways around this/special cases. I’m not familiar with them. This shouldn’t come up with your install though unless you are doing something special like passing some DC conductors across arrays to group together panels in separate arrays onto one microinverter. I considered that but it’s not worth the code compliance pain.

something special like passing some DC conductors across arrays to group together panels in separate arrays onto one microinverter

That's exactly what I'm doing. I'm using a microinverter with multiple MPPT's, so it works with multiple planes. Do I have any options? Is it acceptable to make a separate box for the PV conductors such that AC and PV never share box or conduit space? Or will PV in conduit have the inspector looking at me funny?

 

[zanydroid]

That's exactly what I'm doing. I'm using a microinverter with multiple MPPT's, so it works with multiple planes. Do I have any options? Is it acceptable to make a separate box for the PV conductors such that AC and PV never share box or conduit space? Or will PV in conduit have the inspector looking at me funny?

Well the AC/PV mixing is readily solvable by throwing conduits and boxes at the problem.

Satisfying RSD, not so much. It’s a science fair experiment , and you have no reference documentation on how to do it.

When I thought experimented it I figured it would take RSD on at least one , possibly both sides of the DC jumper, to satisfy RSD. And stacking two of them is far enough outside the standard install that it might not power up correctly.

The reason for both sides of the jumper is that some of the Hoymiles microinverters have inputs that are internally combined together. Not all microinverters have this problem, and not all config of the ones that do, will pull in this baggage. And note I only got to a theoretical analysis and didnt bench test to confirm behavior.

I regret having wasted 16 hours of my life researching and thinking about it. Negative ROI vs the cost savings

 

[Cliffs]

I just finished a HoyMiles instal 1-2 months ago.

Does your AHJ require the EMT to be a complete conduit or just a sleeve?

You don’t need to use SolaDeck and the like for roof boxes. There are mounting systems for PVC boxes to rails. Unirac E-boss has one (I have a stack of extras that I didn’t end up using). E-boss also has EMT to rail and roof flashing clamps.

Sorry one more question about:

There are mounting systems for PVC boxes to rails

The Unirac E-boss seems to be a conduit mount part. Not a PVC box mount. Are you thinking of this? https://unirac.com/product/j-box-mount/

If so can I attach something like this? https://www.homedepot.com/p/Commerc...-with-4-1-2-in-Holes-Gray-WSB450G/300845362Or is this really the wrong kind of box for what I'm trying to do?

 

[zanydroid]

Also RSD transmitter plus balance of system plus RSD disconnect is already rather pricy and the fixed costs/time need a lot of panels to amortize over to make rational sense

 

[Cliffs]

Well the AC/PV mixing is readily solvable by throwing conduits and boxes at the problem.

Satisfying RSD, not so much. It’s a science fair experiment , and you have no reference documentation on how to do it.

When I thought experimented it I figured it would take RSD on at least one , possibly both sides of the DC jumper, to satisfy RSD. And stacking two of them is far enough outside the standard install that it might not power up correctly.

The reason for both sides of the jumper is that some of the Hoymiles microinverters have inputs that are internally combined together. Not all microinverters have this problem, and not all config of the ones that do, will pull in this baggage. And note I only got to a theoretical analysis and didnt bench test to confirm behavior.

I regret having wasted 16 hours of my life researching and thinking about it. Negative ROI vs the cost savings

Hmmm - i'm using 4-input Hoymiles inverters all over my project - are you saying that the panels are connected in series, therefore voltage is too high, therefore RSD not satisfied?

 

[zanydroid]

Sorry one more question about:



The Unirac E-boss seems to be a conduit mount part. Not a PVC box mount. Are you thinking of this? https://unirac.com/product/j-box-mount/

If so can I attach something like this? https://www.homedepot.com/p/Commerc...-with-4-1-2-in-Holes-Gray-WSB450G/300845362Or is this really the wrong kind of box for what I'm trying to do?

Yes that is the one.

Your second link is busted but I assume it’s a metal bell box with pre drilled holes. I would not use that. It’s easier IMO to use a PVC box with no pre-drilled holes and then blast through it with a stepped bit as needed. It’s what the turnkey installs I’ve seen use. But you can probably post a top level question on that subject alone to confirm.

 

[zanydroid]

Hmmm - i'm using 4-input Hoymiles inverters all over my project - are you saying that the panels are connected in series, therefore voltage is too high, therefore RSD not satisfied?

No the panels are connected in parallel. The 2 MPPT 4 port definitely pairs connected in parallel which means one input will pull up another input unless both are shut down.

690.12 says (with some exceptions) conductors outside the array boundary are limited to 30V after RSD. The array bounding box is quite tight. Even the lowest voltage residential rooftop panels today (56/112 cell) are above 30V.

 

[Cliffs]

No the panels are connected in parallel. The 2 MPPT 4 port definitely pairs connected in parallel which means one input will pull up another input unless both are shut down.

690.12 says (with some exceptions) conductors outside the array boundary are limited to 30V after RSD. The array bounding box is quite tight. Even the lowest voltage residential rooftop panels today (56/112 cell) are above 30V.

So the inverter isn't doing anything wrong, me passing PV wires outside of the array boundary which are >30 V would be the problem ?

 

[zanydroid]

So the inverter isn't doing anything wrong, me passing PV wires outside of the array boundary which are >30 V would be the problem ?

Yes.

Apart from the inverter allowing >30V on one port to go to the other port. Both panels must shut down in sync, even if one panels is in the same array as the microinverter. OR you need to cutoff the DC from both sides (which has a high chance of messing up the RSD due to off label use).

 

[Cliffs]

Yes.

Apart from the inverter allowing >30V on one port to go to the other port. Both panels must shut down in sync, even if one panels is in the same array as the microinverter. OR you need to cutoff the DC from both sides (which has a high chance of messing up the RSD due to off label use).

Thanks for all this help, I'm going to go read 690.12 again.

I'm a bit confused about what you mean about panels shutting down. My understanding is the micros just turn off their AC inputs and do nothing to the panel inputs, and this is fine because it meets <80 V in 30 sec within the array boundary 690.12(B)(2)(2). Is that wrong?

 

[zanydroid]

I'm a bit confused about what you mean about panels shutting down. My understanding is the micros just turn off their AC inputs and do nothing to the panel inputs, and this is fine because it meets <80 V in 30 sec within the array boundary 690.12(B)(2)(2). Is that wrong?

Microinverters are intended to be placed within the array boundary. Most panels are under 80V, and microinverters generally do not stack panels in series (though there’s probably one or two intended to be used with two in series that adds up to under 80V. Hoymiles are not compatible with this usage). So if the microinverters shut down the AC trunk goes to 0V and maximum voltage within the array is VOC of your panels. You can’t go lower than the VOC of a single panel without maybe shorting the panel together in shutdown state.

When I say “panel shuts down” I refer to the combination of panel + its RSS.

 

[ncsolarelectric]

Thanks for all this help, I'm going to go read 690.12 again.

I'm a bit confused about what you mean about panels shutting down. My understanding is the micros just turn off their AC inputs and do nothing to the panel inputs, and this is fine because it meets <80 V in 30 sec within the array boundary 690.12(B)(2)(2). Is that wrong?

The panel wires extending outside of the array boundary are not compliant with 690.12.... because they do not drop to < 30V. If you're using a 4-in-1 microinverter, you need to arrange your panels in groups of 4 and keep all the wires tucked underneath except the AC trunk cable. I have HMS-1000's and 500's for situations where you can't group 4 together, or you can use three ports and cap the 4th.

 

[Cliffs]

The panel wires extending outside of the array boundary are not compliant with 690.12.... because they do not drop to < 30V. If you're using a 4-in-1 microinverter, you need to arrange your panels in groups of 4 and keep all the wires tucked underneath except the AC trunk cable. I have HMS-1000's and 500's for situations where you can't group 4 together, or you can use three ports and cap the 4th.

Yep we're all on the same page here.