Solar install turning out really bad - looking for some help/advice please.

[sensij]

The conduit runs out the bottom of the combiner box, then down into the attic so shouldn't need to derate the wires.

There will be a temperature correction for anything more than 30 deg, which is too low of a temp for an attic in AZ. The design temps used in the permit originally are probably still good choices, but if I recall correctly, even with those corrections 2 AWG is overkill. If HX_Guy can post the design temps from his permit, it will save digging through the threads and getting a magnifying glass out to figure it out.

Edit: And yes, although I think I would be capable of installing my own system, I think the money spent hiring an good installer is worth it.

[SunEagle]

There will be a temperature correction for anything more than 30 deg, which is too low of a temp for an attic in AZ. The design temps used in the permit originally are probably still good choices, but if I recall correctly, even with those corrections 2 AWG is overkill. If HX_Guy can post the design temps from his permit, it will save digging through the threads and getting a magnifying glass out to figure it out.

Edit: And yes, although I think I would be capable of installing my own system, I think the money spent hiring an good installer is worth it.

Yeah the #2 is overkill and now requires modifications to the termination points at the inverter as well as pulling that wire through the attic. Seems like a lot of work to satisfy an "over" design.

[HX_Guy]

I'll check on the permit when I get back home to see what the temp used in the design is.

[HX_Guy]

What a cluster &^%$! So I called the city to ask them what the procedure is if we need to make revision to the permit. $150 fee, resubmit new pages for only the pages that have changes, and 5 day turn around time...not too bad overall. That's the good news.

The bad news is that the city people are idiots and they don't understand how this SolarEdge system works, which is why the permit shows #2 wire. With a traditional system, once you take into account what the array amperage would be, and calculate for the 125% safety and then the temperature derate factor (.85 here), I guess a #2 wire would be necessary. That is only if you were to count each wire as "hot", which in a ungrounded system they are, but that's addressed in the next point. SolarEdge isn't a traditional system and the conductors will never have more than 34.5A running through them...the city for some reason doesn't understand or believe that.

So that means if we redo the permit, we need to also educate the city on how this system works, or appeal their decision until we get to someone who does understand.

Ok...so then stick to the permit right? Problem there again is that the inverter input will only accept #6 wire and SolarEdge is saying you absolutely cannot change out the terminal blocks as the whole assembly is UL listed in the configuration it came in. AWESOME! Honestly I'm not even sure how much that would come into play...there is no reason I can think of (but maybe some else can?) why it would be an issue to use a different terminal block aside that it wasn't tested by SolarEdge.

[inetdog]

... why it would be an issue to use a different terminal block aside that it wasn't tested by SolarEdge.

That is exactly why they cannot allow it. Any change to the components invalidates the testing done for UL listing. So SolarEdge could conceivably tell you that it is OK to change the terminal block as long as you deface the UL listing mark and remove the SolarEdge name from the assembly. Which would not make the inspector happy.

[HX_Guy]

That is exactly why they cannot allow it. Any change to the components invalidates the testing done for UL listing. So SolarEdge could conceivably tell you that it is OK to change the terminal block as long as you deface the UL listing mark and remove the SolarEdge name from the assembly. Which would not make the inspector happy.

In practicality though really, what in the world could the different terminal block do that's unsafe? I mean you could probably use a freaking Polaris wire connector block in there and it would be safe. The inspector doesn't know that the terminal block was swapped out, and it would match the permit without having to go through the whole hassle of resubmitting it.

[Ian S]

That is exactly why they cannot allow it. Any change to the components invalidates the testing done for UL listing. So SolarEdge could conceivably tell you that it is OK to change the terminal block as long as you deface the UL listing mark and remove the SolarEdge name from the assembly. Which would not make the inspector happy.

Exactly right. Once a component or system has been UL listed, you can't change it without going through the testing again. Maybe Solaredge could help educate the inspectors. After all, it's in their interest not to have their systems misunderstood by the powers that be.

[HX_Guy]

Now that I look at everything again more closely, I'm not 100% sure if the city needs to be educated, if it even matters. I think the problem may be in how the permit was designed.

The whole issue stems from this page of the permit..



Zooming in specifically on this area here...



Now I'm not sure how the city would calculate it but the way the permit designer calculated it was:

35.8A (maximum output amps) X 125% (NEC safety buffer) = 44.75A per conductor
44.75A x 2 conductors = 89.5A
89.5A / 85% (temperature derate factor) = 105.29A

#2 wire is rated to 115A @ 75º C so there you go. The big mistake here is the PER CONDUCTOR which is completely wrong because if if they are both "hot", the inverter will never have more than 34.5A sent through the wires (that's a different spec, but it's close enough to the 35.8A on the permit that the 35.8A could be left alone on there).

Now that's something the permit designer did wrong, not that the city asked him for I don't think. The two "hot" wires are no wired in parallel so I don't know why they doubled the amperage for the calculation...each wire should be carrying at most 35.8A and even after all the NEC safety and temperature derating, the total would be 52A which is way below the 65A rating of #6 wire.

[HX_Guy]

Also back on the AC side, it looks like #6 wire will not work. Well, it would work I'm sure, but I don't think it'll pass with the city.

The inverter is rated at 47.5A max AC output. You take that, multiply it by 125% and you get 59.38A...so far so good (it's under 65A for #6 wire) but then once you derate for temperature, 59.38A / 85% = 69.85A...oops.

Now I'm confident it would be fine because A) the inverter will never pump out 47.5A, that would mean all the panels are putting out 100% output and that won't happen because of different orientations (some facing south, some facing east) and B) that's already taking into consideration a 125% buffer, which is overkill anyway. The city though isn't going to buy any of that though, so looks like we do need to run #4 between the inverter and the meter.

Now to figure out if 1" conduit is really needed or if we can use the 3/4" already in place. Anyone know for sure what the conduit fill specs are? I seem to find different answers online.

[Alisobob]

Can you post up a photo of your "installer"...

I really need to see what this asswipe looks like....

[sensij]

Yes, your designer really screwed things up. In a SolarEdge system, the PV Circuit is only between the solar panel and the optimizer. Voc, Vmp, etc for the entire array is meaningless, and should not have been listed in the bubble the way that was done. From the optimizers forward, what matters is that the rated continuous amperage for each optimizer string is 15 A. After the combiner, with three strings, 45 A is the rated amps, at 350 V. 45 A * 1.25 = 56.25 A. Your THWN-2 base ampacity is 75 A. If 85% is the temp correction is correct, then 75 * .85 = 63.75 A. 63.75 A > 56.25 A, so you are ok. 85% can't be the right correction though, because that number is not in the 2011 NEC table.

If the temp correction is 76%, which cover you up to 55 deg C, #6 THWN-2 would just barely be acceptable.

It could be that in the original 4 string design, the combined amperage was 60 A (four strings * 15 A). 60 A * 1.25 = 75 A. With a 76% temp correction, 4 AWG THWN-2 would be too small (95*.76 = 72.2 A). Therefore, 2 AWG THWN-2 would have been required.

If the number of strings is what is screwing things up, a permit revision showing only three instead of four may solve the problem.

[sensij]

Also back on the AC side, it looks like #6 wire will not work. Well, it would work I'm sure, but I don't think it'll pass with the city.

The inverter is rated at 47.5A max AC output. You take that, multiply it by 125% and you get 59.38A...so far so good (it's under 65A for #6 wire) but then once you derate for temperature, 59.38A / 85% = 69.85A...oops.

Now I'm confident it would be fine because A) the inverter will never pump out 47.5A, that would mean all the panels are putting out 100% output and that won't happen because of different orientations (some facing south, some facing east) and B) that's already taking into consideration a 125% buffer, which is overkill anyway. The city though isn't going to buy any of that though, so looks like we do need to run #4 between the inverter and the meter.

Now to figure out if 1" conduit is really needed or if we can use the 3/4" already in place. Anyone know for sure what the conduit fill specs are? I seem to find different answers online.

85% is not a valid temp adjustment. Where did you get that number? 82% is the correction for up to 50 deg.

See this post. You are doing the math wrong.

[HX_Guy]

I spoke with the permit design company and they gave me that 85% temperature derate number, they said that's what my city (Peoria, AZ) requires to be used for conductors ran through metal conduit.

As for doing the calculation wrong, they actually said the city forced them to do that. In their first draft (on the right side), they actually showed the SolarEdge figures which were accurate and the city didn't accept them. They had to revise (the cloud bubble) to satisfy what the city asked for.



Your interpretation of SolarEdge though is also not right I don't think, at least not how it was explained by SolarEdge. While each string can have a maximum output of 15A, all strings combined will never have an output of more than 34.5A. This is regulated by the inverter and the communication back and forth with the optimizers. The optimizers will raise their voltage to drop their amps so that the total of all strings in parallel does not exceed 34.5ADC, so the correct figure to always use is 34.5A...which is why the inverter only allows max 6AWG wire. There isn't a scenario ever where you would need more than that because 34.5A x 1.25 x whatever temp derate you use is not going to exceed 65A.

http://www.solaredge.com/files/pdfs/products/inverters/se-single-phase-us-inverter-datasheet.pdf

[yocane]

Seams like quite the issue

haha get it?

http://www.howto.solar

[inetdog]

Now I'm not sure how the city would calculate it but the way the permit designer calculated it was:

35.8A (maximum output amps) X 125% (NEC safety buffer) = 44.75A per conductor
44.75A x 2 conductors = 89.5A
89.5A / 85% (temperature derate factor) = 105.29A

#2 wire is rated to 115A @ 75º C so there you go. The big mistake here is the PER CONDUCTOR which is completely wrong because if if they are both "hot", the inverter will never have more than 34.5A sent through the wires (that's a different spec, but it's close enough to the 35.8A on the permit that the 35.8A could be left alone on there).

Ouch! I suspect that the same designer would say that I have two 100A 120V loads, so I need a 200A 120/240 service.

Well, this will certainly be a good time to amend that part of the submission.

Although as you say, # 6 would probably not fly either.
I wonder just how the temperature adjustment calculation was done? Clearly only two current carrying conductors in the raceway, so it must all be pure temperature.
Some use the highest high ambient temperature and then take the conduit above roof adder on top of that. Did he figure in the correct height in inches for your conduit run? What temperature rating did he use for the insulation? You can probably justify using 75C, as long as the terminations at both ends are rated for 75C.
And if the run in conduit above the roof is short enough you may be able to ignore the whole calculation. The other place where you need the temperature calculation is inside the attic.
Quaint tip: If you have 90C wire and only 75C terminations at some point you can put in another junction box and go from large wire at the termination to smaller wire calculated at 90C for the body of the run. Just make sure you use 90C wire nuts or terminal strips to make the transition and run at least a foot of wire at the larger size.