8400w Grid Tie System - Kit or piece it together?

[Mr4btTahoe]

Thanks

I'm good on everything other then the potential distance between the 2 rows... Is there any specs for that? I would think if the UFO clamp provides enough gap.. that as long as I maintained that spacing between the rows, it would be enough (which I'm over that a bit...)

[NukeEngineer]

Thanks

I'm good on everything other then the potential distance between the 2 rows... Is there any specs for that? I would think if the UFO clamp provides enough gap.. that as long as I maintained that spacing between the rows, it would be enough (which I'm over that a bit...)

Actually, your UFO clamps technically provide zero gap mechanically between panels. One panel is butted up against the UFO bolt, followed by the next panel. If all your panels were flat, and the UFOs not tightened, and you pushed horizontally along a row, you would immediately slide the last panel on the row as there's no mechanical gap between them.

However, between rows, that's not the case.

Though, using the same gap that the UFO bolt puts between panels on one row for the space between rows is likely very sufficient.

[J.P.M.]

Just did a quick calc and seems like an almost 2m tall aluminum frame would expand only about 0.2-0.3mm over a temp range of -10C up to 55C, so perhaps as long as there's some minimal spacing between rows, then thermal expansion probably isn't a big concern.

Horizontally, though, the total expansion is probably on the order of centimeters. But, the good news there is that since your racking is also aluminum, it will expand with your frames and not stress the panels.

The induced stress from thermal expansion is f(structural rigidity, considered length) as well as temperature and other things. By way of example and simplicity, for the mostly academic situation of two fixed metal surfaces of, say, aluminum in contact at some temperature, an added compressive stress will be induced that is equal to the (positive) differential temp. induced thermal expansion times the elastic modulus of the material up to the yield point of the aluminum. If the elastic modulus of aluminum is something like 10EE7, it doesn't take much restrained thermal expansion in such situations to get to the yield point - something like 0.003 in/in. That's when structural design flexibility becomes important. Also one way and reason why bolted joints can loosen up over time in thermally cycled joints, particularly when the bolting and joined materials are dissimilar materials, which is why tightening and torqueing of bolting systems in a bolted joint is a bit of an art form if done correctly.

As for the similarity of materials you write of, that can work either for or against inducing stress brought on by thermal expansion.

As for the types of applications we're talking here, and for design considerations, because the frames and the racking will usually be at different temperatures, the temp. differential(s) are what will be driving any induced stress, and because of the dissimilar temps of the structural components, thermally induced stresses will be present under most conditions regardless of the similarity of materials.

[Mr4btTahoe]

First "home run" of cables done and in conduit.
IMG_20180504_174537402 by acidburn02zts, on Flickr

IMG_20180504_174542852 by acidburn02zts, on Flickr

IMG_20180504_175059685 by acidburn02zts, on Flickr

IMG_20180504_175049608_HDR by acidburn02zts, on Flickr

The next run will be fun as it will have 4 10awg pv conductors (which is nearly identical in O.D. to 8awg thhn), and 1 8awg EGC. The max fill for 3/4 sch. 40 is 5 8awg conductors... so based on size, it'll work.. just going to be tight fishing it through.

Everything is in conduit from the main box to the array. Each run of conduit has an EGC ran with the conductors sized for the 40a circuit (8awg copper).

The rest of the panels are going up today and finishing wiring. I'll get everything labeled and should be ready for AHJ inspection by mid next week.

Have 2 weeks until the turn-on date with the POCO so if the inspector needs anything changed, I'll have a bit of time to do it.

More to come!

[Mr4btTahoe]

More panels going up...
IMG_20180505_114032340 by acidburn02zts, on Flickr

All conduit finished and wiring ran (including the 6awg bare ground also ran in conduit to the grounding electrode)...
IMG_20180505_161132619 by acidburn02zts, on Flickr

IMG_20180505_155733828 by acidburn02zts, on Flickr

24 panels on the shop... (3 more will be mounted on the lean-to once built for 3 strings of 9)
IMG_20180505_144456558 by acidburn02zts, on Flickr

Wiring grouped by string... voltage tested and marked for polarity.
IMG_20180505_155723214 by acidburn02zts, on Flickr

All that is left is to button up the inverter wiring... pass inspection and flip the switch.

When voltage was tested, it was late in the afternoon.. cloudy and raining. Both strings of 9 were producing ~360v and the string of 6 was producing ~240v. Getting ready to go back out and hook up the DC side of the inverter now that it is dark and the array is not producing.

Getting close to being complete.

[Mike90250]

uh oh. You have a chafed green wire in the photo of the box. Some inspectors may not pass it, What abraded the wire, and is there other hidden damage to conductors ?

HiRes photos - a 2 edged sword.

[Mr4btTahoe]

I don't know what it is about the green wire... But I had issues with it everywhere. It has a paper thin plastic film on it that caught on everything. It peels off extremely easy (like with a fingernail). The PV conductors and other colors used for the AC side didn't have that issue.

There isn't any damage to the actual insulation so I'll clean that mess up before inspection.

Any idea what the coating is? Perhaps it's the coating that was supposed to make the wire easy to pull through the conduit....? It failed at doing it's job if that's the case.

[bcroe]

If the actual insulation in not injured, perhaps you could do a neat job of covering the exposed section
with green tape. That sort of thing happens here when green or white would be better but all my
wires are black. Bruce Roe

BiDirDistBox.JPG

[Mr4btTahoe]

If the actual insulation in not injured, perhaps you could do a neat job of covering the exposed section
with green tape. That sort of thing happens here when green or white would be better but all my
wires are black. Bruce Roe

Isn't that technically not up to code? I thought 250.119 stated anything 6awg and smaller could not be marked via tape/paint... had to be the actual insulation color.

Anyways... I got everything cleaned up and buttoned up at the inverter. Just waiting on the proper size parallel taps to show up and then its time for inspection.

IMG_20180506_182152933 by acidburn02zts, on Flickr

IMG_20180506_182200238 by acidburn02zts, on Flickr

I can't believe it's actually almost finished... has been a much simpler process then I originally had thought.

[Mr4btTahoe]

Well... everything is labeled as it should be (according to the diagram that came with my label package) and everything is buttoned up.

All that is left is the inspection. Today is election day in my state so I'll give the inspector a call tomorrow. Hopefully we can get it inspected by Friday so I have a bit of time to correct any issues he may point out.

More to come!

[Mr4btTahoe]

AHJ inspection passed without a hitch. Just waiting on the PoCo to tell me that I can flip the switches.

I'll post up some completed project pictures over the weekend.

Greatly appreciate the guidance and help!

[Mr4btTahoe]

PoCo just placed the net meter.

I'm at work but when I get off, the system finally goes online!

[NukeEngineer]

PoCo just placed the net meter.

I'm at work but when I get off, the system finally goes online!

Congrats!
goosey.gif

[TAZ427]

Nice job, I read most of the thread, but not 100%, did you have final tally of what it cost you. Not including your labor of course. Also, that was a straight string tied system, no power optimizer right?

[Mr4btTahoe]

No optimizers... 3 strings (SMA SB7.7 inverter has 3 mppt inputs). Very simple system and grand total of ~9200 including the mistakes I made ordering a few small (yet expensive) parts wrong.

So.. ~$1/w installed.

2 guys.. roughly 6hrs for 2 days to build the array (iron ridge racking made it very simple). A couple hours of wiring and labeling (after a couple weeks of studying the NEC and asking lots of questions).

I still need to take some finished pictures and post them up. Haven't had time yet...