photo for aspiring solar forensics

[tankfullofsun18]

Howdy All,

so I'm new to the forum, and have an interesting *ground level* photo of a failed system that I often drive past. It's my seventh year in the solar industry (I got my nabcep sticker a while back, and have done a fair amount of rooftop residential and some commercial as well.) But this is a sort of a brain-twister for me. I'll let the group take a crack at some armchair solar forensics.

DSC02116-1050.jpgDSC02119-1050.jpg

(hopefully these render in the thread...)

Following what was by all accounts an unusually high-snowfall winter here in new england, there appears to have been a structural failure of the frames on the PV modules-- the bottom rows look like little dimples!

Any thoughts? My working hypothesis is that the array mounting design was chosen for aesthetics rather than strength (perhaps some kind of "rail-less" design), and there was another factor that compromised the strength of the PV modules. It might be significant that this array is located at a plant where concrete is processed..

will

[J.P.M.]

Howdy All,

so I'm new to the forum, and have an interesting *ground level* photo of a failed system that I often drive past. It's my seventh year in the solar industry (I got my nabcep sticker a while back, and have done a fair amount of rooftop residential and some commercial as well.) But this is a sort of a brain-twister for me. I'll let the group take a crack at some armchair solar forensics.

[ATTACH=CONFIG]6526[/ATTACH][ATTACH=CONFIG]6527[/ATTACH]

(hopefully these render in the thread...)

Following what was by all accounts an unusually high-snowfall winter here in new england, there appears to have been a structural failure of the frames on the PV modules-- the bottom rows look like little dimples!

Any thoughts? My working hypothesis is that the array mounting design was chosen for aesthetics rather than strength (perhaps some kind of "rail-less" design), and there was another factor that compromised the strength of the PV modules. It might be significant that this array is located at a plant where concrete is processed..

will

Looks sort of like an ice dam to me.

[J.P.M.]

Howdy All,

so I'm new to the forum, and have an interesting *ground level* photo of a failed system that I often drive past. It's my seventh year in the solar industry (I got my nabcep sticker a while back, and have done a fair amount of rooftop residential and some commercial as well.) But this is a sort of a brain-twister for me. I'll let the group take a crack at some armchair solar forensics.

[ATTACH=CONFIG]6526[/ATTACH][ATTACH=CONFIG]6527[/ATTACH]

(hopefully these render in the thread...)

Following what was by all accounts an unusually high-snowfall winter here in new england, there appears to have been a structural failure of the frames on the PV modules-- the bottom rows look like little dimples!

Any thoughts? My working hypothesis is that the array mounting design was chosen for aesthetics rather than strength (perhaps some kind of "rail-less" design), and there was another factor that compromised the strength of the PV modules. It might be significant that this array is located at a plant where concrete is processed..

will

Looks sort of like the result of an ice dam situation to me.

[tankfullofsun18]

Looks sort of like the result of an ice dam situation to me.

JPM - Could you elaborate? By ice dam do you mean heat lost from the building causing snow to melt and then refreeze on the roof, and that the forming ice caused it? Or are you talking about the shifting potential of large amounts of snow on a PV array, building up somewhere towards the bottom row?

I have seen a situation where the dynamic load of an ice dam has pulled railings or pulled loose poorly installed fasteners. never heard of an ice dam crushing or dimpling a PV module. The faces and frames must be built to withstand dead loads of many, many p.s.f. Am I wrong?

[Panelmal]

Looks sort of like the result of an ice dam situation to me.

Bet you're right about that, this winter was bad for those here in ny on arrays with the constant weather swings between sunny and 30s then below zero again for the next week. Guess railless isn't all it's cracked up to be.

[J.P.M.]

JPM - Could you elaborate? By ice dam do you mean heat lost from the building causing snow to melt and then refreeze on the roof, and that the forming ice caused it? Or are you talking about the shifting potential of large amounts of snow on a PV array, building up somewhere towards the bottom row?

I have seen a situation where the dynamic load of an ice dam has pulled railings or pulled loose poorly installed fasteners. never heard of an ice dam crushing or dimpling a PV module. The faces and frames must be built to withstand dead loads of many, many p.s.f. Am I wrong?

Perhaps the local dead load was severe, perhaps that, coupled with some additional wind loading from the increased area of the ice. Perhaps the connections at the roof got kicked out/moved by ice expansion and rattled around in the wind causing a fatigue failure. Could be a lot of stuff, but most of my money's on the ice dam scenario.

Snow will accumulate on an array. During the day, if temps. get above freezing and the sun come out for awhile, the array will heat up some, a bit of the snow will melt and run downslope. Most will run off. Some will not, trapped at the array frame . At night, the water will turn to ice, leaving an ice "lip" at the bottom of the array. Next day: same scenario. This time however, the ice "lip" holds back a bit more water that again freezes. Over several freeze/thaw cycles the ice will become quite thick, turning from a lip to a dam holding back progressively more water. - perhaps 30 or 40 cm. or more. The weight of that ice plus perhaps the added wind profile area can cause failure via several mechanisms.

What you describe with respect to fasteners, etc. is usually separation due to expansion or fracture of connecting members likely due to expansion when water turns to ice. The same way big rocks crack when water gets in the small cracks and makes them big(ger) cracks.

What the photos show is more likely due to a dead load of ice that exceeded the panels' loading capacity, perhaps coupled with some wind induced fatigue stress or rupture at other points. Hard to tell without more data, some witnesses and/or a site visit.

Anyone who's grown up in a cold climate knows the ice dam scenario well as it pertains to roofs when snow warmed by poorly insulated roofs, or when sun shines on a dark roof ridge made clear by wind melts some snow, then it refreezes when it hits the colder eaves and starts the damming process. After a few day or weeks, a 1,000 lbm chunk of ice will drop from a roof, often, but not always at an inside corner, and literally shake the ground when it hits. Been there, seen it.

The heat source is a bit different here for arrays, but there are a lot of similarities and the basic mechanism of freeze/thaw is pretty much the same.

Another reason why I now live in San Diego.

[tankfullofsun18]

Snow will accumulate on an array. During the day, if temps. get above freezing and the sun come out for awhile, the array will heat up some, a bit of the snow will melt and run downslope. Most will run off. Some will not, trapped at the array frame . At night, the water will turn to ice, leaving an ice "lip" at the bottom of the array. Next day: same scenario. This time however, the ice "lip" holds back a bit more water that again freezes. Over several freeze/thaw cycles the ice will become quite thick, turning from a lip to a dam holding back progressively more water. - perhaps 30 or 40 cm. or more. The weight of that ice plus perhaps the added wind profile area can cause failure via several mechanisms.

So it seems like the solution for a cold-climate moderately-pitched rooftop is to incorporate a horizontal alley, to decrease the snow-dam watershed.

Another reason why I now live in San Diego.

i envy you're in that 5-6 kW/m2 country. Must be a ton of solar going up where you are. Plus fish tacos.

Thanks for the feedback!

[LETitROLL]

Bet you're right about that, this winter was bad for those here in ny on arrays with the constant weather swings between sunny and 30s then below zero again for the next week. Guess railless isn't all it's cracked up to be.

Ice/snow dam, enhanced by the reduction in roof pitch (transition) immediately below the area of damage, ice easily breaks pipes, and rocks, definitely looks like movement due to dam effect in that transition area.

[inetdog]

Ice/snow dam, enhanced by the reduction in roof pitch (transition) immediately below the area of damage, ice easily breaks pipes, and rocks, definitely looks like movement due to dam effect in that transition area.

Your very own glacier on the roof.

[Mike90250]

Crushed by weight

Either raw snow build up (doubtful, as the entire array would show damage)
or the eves never thawed, and warm snow slid downhill till the eves blocked the slide, froze, repeat the next day till enough weight built up to crush the panels