Finally installing PV on my roof in LA: day 1

[sensij]

So, engineering drawings and calculations are great and all, but how did the mounts hold up in the wind this weekend? Still got all your panels?

[J.P.M.]

The installer sent me engineering drawings which seem to show wind load calculations have been done by a registered professional engineer.
(I can share them privately upon request if anyone's curious.)
So in theory at least that checkbox has been ticked off.

Here's what one of the lower reverse tilt mount attachments looks like from below:

https://goo.gl/photos/VXwbkr9Zp5in2jDu6

I hope that plywood holds, and the bolt holds on well to the (dimensional, 100 year old, possibly termite-addled) 2x4.

I think I'm going to see if I can dig up a mechanical engineer alum friend from college to have a look at all this.

NOMB, but that flick shows something that would make me nervous if it was my stuff.
Calcs to estimate wind loadings, and a design and material specs are one set of criteria. Buiilding to the material specs and drawings are also required. Might be that the designer didn't know/consider the condition of the roof ?

[DanKegel]

Still there

I'm going to send those pictures (and maybe a walk around video) to the engineer and see if he gulps.

[inetdog]

Still there

I'm going to send those pictures (and maybe a walk around video) to the engineer and see if he gulps.

+1

Glancing through your album I see three things of concern:
1. The patched circular hole (wrong place?)
2. The locations where the sheathing does not sit tight against the support member and so has been destroyed by the lag screwed mounting point which is still not tight against the structural member.
3. The knee in the tall braces, which is adding a bending moment to the pipe part of the support instead of just a compression (tension in wind) load. ]
Like others, I do not think that the roof end of the mounts is designed for a rotational moment load. At the short mounts the moment arm is small and so probably of no consequence. Some diagonal bracing would be nice for strength but would also probably shadow the back sides somewhat more than the current arrangement. If the long braces were straight the mounting rails themselves would triangulate the structure nicely.

[DanKegel]

The patched circular damage isn't as bad as it looks from the album - the hole was somewhere else, the damage is just someone standing a cut conduit upright on a hot roof, that hasn't been patched yet.

I still wonder if guy wires could add some stiffening without too much shading.

[inetdog]

The patched circular damage isn't as bad as it looks from the album - the hole was somewhere else, the damage is just someone standing a cut conduit upright on a hot roof, that hasn't been patched yet.

I still wonder if guy wires could add some stiffening without too much shading.

Since you have two non collinear compression members in the brace with a knee, adding a tension member is not going to help much if at all.
A strut to a clamp at the base of the pole might help though.

[DanKegel]

I was thinking one set of guys from the knee to the roof, and another from the knee to the panel rail, so there would be triangles. But IANAE (I am not an expert).

Tell me more about the strut idea.

[SunEagle]

What about a see through panel on the back side to deflect the wind from under the solar panel to as not to create a "sail".

I think I have seen that type of installation by the manufacturer of those double sided panels.

[inetdog]

I was thinking one set of guys from the knee to the roof, and another from the knee to the panel rail, so there would be triangles. But IANAE (I am not an expert).

Tell me more about the strut idea.

Simple: You now have the flat metal running from A to B (the knee) and the pipe running from B to C. Add another compression member directly from A to C.

Or else run a tension member from A to C and add another tension member from the middle of that wire to B.

Your idea of tension from A to C and tension from B to the panel rail will not help since both of those tension members will be trying to flex the knee rather than extend it. The downward load (weight) of the panel system is already trying to flex the knee.

Your tension members would come into play in counteracting upward forces from wind, but only that.
My tension members would be counteracting gravity and downward wind forces only.
A solid rod from A to C would resist both flexing and extending of the knee.

[DanKegel]

I'm going to send those pictures (and maybe a walk around video) to the engineer and see if he gulps.

He gulped. He's calling a meeting with the installer and me on site

[inetdog]

Just realized that there is also absolutely no lateral triangulation for forces parallel to the width of the panels. Those all come to bear on the short upper supports only as both linear and angular forces.

[DanKegel]

Had a conference call with the PE and the installer just now.
The installer was still unclear on the problem, but the PE and I talked him into using a long enough adjustable tilt leg from IronRidge.
The PE also suggested using an extra brace from the uphill mount to the middle of the adjustable tilt leg for extra stiffness.
When I mentioned the issue of shading on the underside of the panel, the PE said "What do you mean bifacial? I wasn't told this would be a bifacial panel. I've never heard of a bifacial panel. How can you mount a panel without obscuring its underside? Madness!"
He was very skeptical that a frameless panel would be strong enough (despite the manufacturer's assurances on the datasheet).

Good times. Hopefully they'll straighten it out shortly.

(Also saw a scary video from Unirac saying "for the love of god, don't do reverse tilt mounts"

http://solpowerpeople.com/solar-pv-racking-with-unirac/

It's about half an hour in.)

[DanKegel]

The replacement reverse tilt mount is still being designed, but in the meantime, the installer sent someone out to turn on the system for testing (and install the missing performance meter, and plug in the ethernet). The old analog meter is spinning merrily backwards now

First day output was 40.0 kWh.
Faceplate capacity is 25*305 + 4*300 = 8825 watts, so that's 4.5 Wh/W. Not too bad for winter.

[SunEagle]

The replacement reverse tilt mount is still being designed, but in the meantime, the installer sent someone out to turn on the system for testing (and install the missing performance meter, and plug in the ethernet). The old analog meter is spinning merrily backwards now

That is great news Dan. Hopefully the reverse tilt mount can be resolved so you can get data on those panels.

[DanKegel]

2nd day output was 42.8 kWh, or 4.8 Wh/W. (I guess that first day was partial.)