Finally installing PV on my roof in LA: day 1

[DanKegel]

Hey, I'm a babe in the woods when it comes to mounting solar panels. Do you know of a list of good steep-roof reverse-tilt mount kits I can point them to and say "If you don't have an engineer on staff to design mounts properly, you should be using one of these instead of improvising"?

And I'm not trying to get away with anything here. I'm just a guy who trusted an installer when they said they could do reverse-tilt mounts. I'd like to have some good ammo Monday when I talk to them about this.

[J.P.M.]

I don't think the supports were designed, they were improvised.

More NOMB, but it reads to me as if you're going to allow that improvisation to be changed without some knowledgeable professional at least looking at it ? What are you thinking ?

If the inspector is myopic or doesn't know/care much about structural design, you might get away with it.

Not my roof and still NOMB, but I'd not use what I think I see. I'm not a fear monger, but simply put, I'd be concerned about safety.

Similar to the standoffs in a line that looks to be non parallel to the roof ridge that Sensij found, this is another example of barn job work.

[DanKegel]

I don't think the supports were designed, they were improvised. The lower part of the long supports were designed for tilt mounts on a flat roof, not for reverse tilt on a sloped roof. The resulting install does seem kludgy... struts with knees? Really?
I'm not even sure how one would check to see if the screws were coming loose - they're not visible (except maybe from the bottom; I think at least one didn't go directly into a rafter.)
I guess I can go up and wiggle it by hand once a year and see if it feels loose.

Anyone have a favorite brand of reverse tilt mount kit, for future reference?

I hope the installer is planning to install rails parallel to the panels to clamp them to. I'll be watching, with a copy of the sunpreme data sheet handy.

The strut a couple inches below the panel will "shade" it from diffuse light. But I think only one of the panels has this problem; the other seems like it might be unobstructed. If so, this'll be a natural experiment to see how much difference it makes.

[sensij]

Yeah, it won't be shadowing any beam irradiance. Since there will be LG panels installed on the reverse tilt stand as well, those LG panels should be located over the unistrut support members, which look like they will be just a couple inches under the panel. It might not make much difference, but having empty space behind the Sunpreme's will give any diffuse light the best chance to get to a cell.

I was thinking that the panels would be installed portrait orientation, with clamps on the short ends. On further review of the Sunpreme installation manual, I'm not sure that the panel can mechanically tolerate end mounting without support in the center. The manual recommends that in a high albedo environment, the portrait orientation is good but the rails should run along the long sides, not the short sides (see notes in rightmost column of page 1, and figure 2 on page 3).

[J.P.M.]

[QUOTE=sensij;n300608]the bifacials will have their best chance to outperform if they can be mounted so the support brackets are not shadowing the backside.

QUOTE]
How would that be done ? Anyway, assuming you're thinking about shadows, and particularly after looking at the surroundings from the photos, since it seems likely that little of the insolation that gets to the reverse side of the panels will be specular in nature, and therefore little if any shadow will be cast, except probably a bit early and late in summer, how much of a problem will it be ?

[J.P.M.]

Dan: Absolutely NOMB, and I thought I'd stay out of your parade and not rain on it, but I'd have and do have a concern about the longer reverse tilt support arrangement. Who designed them ? Were they built to the drawings ? As built, unless the connection between the bracket and the panel frame is designed to take out that moment (bending) load, which is unlikely, or the bracket at the roof is a pin type connection, wind or other external loads will exert a moment load at the roof connection. That type of loading and of the magnitude the longer arm will induce is usually not something a screwed in connection to the roof of that type is designed to withstand, particularly when the loading is cyclic, or occasional and reversing, or recurring. Over time, sooner or later, the alternating stresses will cause rocking motion as things loosen up, as they will with this arrangement, and that will cause problems.

Having the longer supports oriented normal to the panel plane with a pin type connection at or close to the roof roof might have been one of several other ways to do it.

Just my $0.02.

[sensij]

Nice adjustment. You've probably already planned for it, but the bifacials will have their best chance to outperform if they can be mounted so the support brackets are not shadowing the backside.

Stick to your guns on the hardwire. SolarEdge Zigbee is like early wireless routers... even if it only needs to be reset a couple times a year, it will still be a couple times more than you'd like.

[DanKegel]

Yep. You can see how they fixed it in today's pictures ( https://goo.gl/photos/s6s1n69k1EBuSejE8 ), glad I caught it early.

They're a little flummoxed by my request to use hardwired ethernet. I made it easy for them by running the ethernet right to the solaredge box, and giving them a link to the how-to video ( https://www.youtube.com/watch?v=AyL67b7JMGk ), all I'm asking for them is to put it in a conduit and plug it in. Sounds like everybody else goes for Zigbee.

The LG310s look real pretty. The bifacials do, too, but they're not mounted yet -- the edge clamps they need aren't here yet. So no generation until those come in, maybe Monday.

Did anybody notice the identical triplets in one picture? https://goo.gl/photos/oSmhoKiqELrkS3e28

[SunEagle]

Wow. I guess it was easy to make the mistake until they add the rails and then it became easy to see the low angle for those reverse tilt panels.

[DanKegel]

oh noes! reverse tilt mount was supposed to be tilted towards south 20 degrees, but it's more like 2 degrees.
Trying to explain problem to my official contact at the installer now.
Pics show one of the installers measuring current tilt, and demonstrating what 20 degrees would look like.

Installer says they'll fix it. Glad I was paying attention - wish I'd caught it sooner, but at least I got it before panels were mounted.

[SunEagle]

Looking good Dan. Thanks for the photos.

[DanKegel]

They're installing the rails and conduits today. The crew that installed the penetrations never did quite finish, so this crew is having to flash one mount and caulk another. They ran out of LG 305's, so they're substituting same number of LG 310's, so that's a bonus 125 watts, woot Panels themselves go on in a couple days.

Pics as usual at https://goo.gl/photos/s6s1n69k1EBuSejE8

[DanKegel]

Correct, I signed the contract and put down my deposit back in June; I was probably very nearly the last house to get the last drips of the LADWP subsidy.

The roofers have put shingles down around most of the mounts... one mounting point lacks flashing, tsk. They'll probably fix that Monday.
Pictures at https://goo.gl/photos/s6s1n69k1EBuSejE8 scroll down for latest.

[samotlietuvis]

But the LADWP subsidy evaporates in about a month, so it better be before then!

Nice progress. What LADWP subsidy are you referring to? I thought all subsidies are gone from 2014 September, only NEM 1.0 remains.

[J.P.M.]

Nope, they're 32" to 33" apart... but they're dimensional 2"x4"s, that may be what fooled you. (The house was built in 1912, before quarts of liquor turned into fifths, and 2x4's turned into 1.75x3.5's. Heck, I bet even Almond Joys actually *had* an almond in every bite back then

No snow load in Los Angeles. Worst stress might be wind on the solar panels (or maybe earthquakes).

And, FWIW, that wind loading may be uplift, and therefore probably not planned for in 1912. But back then, not much in the way of dynamic loading was considered. In any event, The positive (downward) load on the roof will be about the same.