Has anyone seen, or heard of any solar panels burning??

I have a few pics of a roof fire, upwards of 200+ panels affected. The root cause is not determined as of now.

Keep us informed.

When you look at the currents involved in an array that big, a fire would not be totally unreasonable.
I think it will be a connection or a device in the system more than a panel but that is a guess.

Greg:
I put a pic up in mod photo area, I can't seem to move it here.

This one.

I would really like to be in on figuring out where that fire started. If the poly tarp is any indication of the damaged area finding the origin may not be that easy.
I imagine it will just be called "electrical"

As fire departments around the nation weigh in...

More and more will find that residential rooftops are hardly ideal for PV arrays.

For the fireman have to fight a structure that can't really be de-energized while the Sun is out.

I expect someone to come out with foam blocker for just such a purpose.

BTW, silicon does burn -- just like charcoal -- once you get it hot enough. Instead of carbon dioxide -- you get white sand. (Silicon dioxide)

The reaction throws off plenty of heat, too. At least it's not a stinky reaction -- but the rest of the stuff is.

========

I'm inclined towards PV sourced power. I'm wary of it being installed 'at retail.'

I'm of the opinion that roof falls, fire hazards, and violation of roofer warranties will all cause us to regret ever putting them up.

What seemed so good in the science fiction novels is in reality highly impractical and excessively dangerous.

It's the roof falls. Have you seen the Workmans Comp rates for roofers?

They're in orbit.

Solar oriented ECs are grossly underpricing their risk.

The number one reason that roofers go out of business is roof falls.

At some point, the boys get too casual -- violate safety rules -- and everyone else picks up the pieces.

Roof mount is popular because it is seen as wasted space but it does bring up it's own problems, as you noted.

I suppose the fix for shutting down the collector output would be a contactor in each collector, controlled by a master EPO.
If these have grid tie inverters in each collector, simply dropping utility power would do it.

Greg:
This is the combiner box closest to the assumed point of origin of the fire. No definitive cause or location was available as of this afternoon.

There was certainly a lot of magic smoke in that box.

Tesla:

Thank you for this:
"BTW, silicon does burn -- just like charcoal -- once you get it hot enough. Instead of carbon dioxide -- you get white sand. (Silicon dioxide)

The reaction throws off plenty of heat, too. At least it's not a stinky reaction -- but the rest of the stuff is."

I hope to have time to stop at the site tommorrow and check the removal progress. Also to find out if any definitive cause, or location of point of origin has been determined.

That is (was) a 16 string combiner box. Another adjacent to it has issues also, although not as 'well done'.

Ironically, this system was to go 'grid tie' next Saturday with a scheduled utility shutdown. Oh, well!!

Hotline...there was a fire on a roof in Bakersfield California similar to what you are showing. The cause of that fire was determined to be a groundfault. Here is a link to info on that fire...

http://nfpa.typepad.com/files/target-fire-report-09apr29.pdf

Add here is a good article by Bill Brooks...

http://www.solarabcs.org/about/publications/reports/blindspot/pdfs/BlindSpot.pdf

Your fire should be documented for the solar industry.

Where, when, and what do you know about this fire?

shortcircuit

That appears to be a nonmetalic combiner box with EMT. Were bonding jumpers installed?

shortcircuit

Looks like PVC to me, I see the barcode on the bottom-most conduit.

John,

Yes please keep us informed. I would like to see the outcome of this problem was. I am still on the fence about the safety of this things.

I know I would prefer the PV systems that have the mini inverter on each PV panel. This way the DC current stops right there on the roof.

Shortcircuit:
Thanks for the links!!! Appreciate it!!!

Combiner boxes are metalic, 3R, etc. conduit is PVC with grd conductor.

Documentation is being compiled on my end.

Here is another angle pic.

That looks like a Solectria combiner box.

As of Fri. there has been no root cause determined. The area of the roof has been set into a grid, the remains being tagged and boxed and removed from the roof. The remains will be reviewed further to see if a cause can be determined.

Assuming a number of the collector pans look like this.

I guess the first question I have is what happens to those collectors if they feed a bolted fault. Is there collector level over current protection?
You might be looking at a symptom there and you had a bigger problem on the load side.

Greg:
The combiner boxes like these are 16 circuit, finger safe fuses (midget). I'll venture a guess at 100K minimum for the fuses.
I'll get the rest of the specs tommorrow.

http://www.cooperindustries.com/con...ntal/pv-fuses/solar-pv-fuse-14x51mm.html

Bussman shows 10,000 Amps as the interrupting rating.

Never forget: PV cells saturate current flow with a bolted short at much lower levels than rotary induction devices.

15, 20, 25 & 32A

Notably, these fuses are designed towards the European standard; hence the 32A type.

I have a little question.
Has anything in the box actually ignited?
Sure it looks pretty black in there, but I'm wondering if that isn't because of transmitted smoke/heat that may have come in through one of the conduits.

Trumpy:
I have no details/info that is 'firm'.

It appeared that there was some damage in one combiner box that was not soot/heat, but meltdown. An adjacent box looked like transmitted heat/smoke/soot. Keep in mind I am not the forensic party that is looking into this.

Most of the damaged area has been removed, grid identified, boxed up & transported to an interior location to be investigated. Could ba a long haul for an answer.

Tesla:

Littlefuse KLKD-15 on another jobsite is 100K AIC for AC and 50K AIC for DC at 600 volts.

I have not been back to the problem site yet, but probably by middle of next week.

Tesla:
Check out Bussman PVM fuses, they are 50K AIC-DC also.

http://www.cooperindustries.com/con...cts/supplemental/pv-fuses/solar-pvm.html

I'm getting the gut feeling that either it's going to be a long time, or kept hush-hush as to a cause.

Can't share what I don't know.

Just a bump & update.

This array is now replaced and requested a inspection be scheduled.

No cause has been mentioned to me as of today, although there have been a lot of comments.

Coincidently, there was a large fire in an adjacent Twp yesterday (Mon), ao a large whse bldg, that also has a solar panel install. The cause is not determined, either a roof fire that ignited the panels, or a panel fire that ignited the roof.

The fire chief of that Twp only heard of the fire in this thread about an hour after his fire.

Thanks for the follow up Hotline. Any determination of cause should be shared with the industry for all to learn I would hope. I know NJ had a solar boom for awhile there and I would think that some installations were not done properly. We have a solar boom here in Mass now.