Solar power sales dudes and their scare tactics. How does a solar panel burn up?

[Ed.Kloonk]

Been getting a barrage (again) with solar power sellers. Latest one is warning of the danger of employing a company with little or low time spent in the business. The leaflet shows a picture of a shonky installation, a row of several solar panels, one of which is charred.

My question is how is this happening? What is the installer doing wrong that causes a panel to burn up?

Sydney. Aust.

[ataradov]

You may want to look into Tesla/Walmart lawsuit.

I think it is mostly down to shoddy workmanship and cheap parts outside of the actual panels, like wiring and distribution boxes.

[Rerouter]

generally it is the connections between panels, if they are done poorly, or the wire size is way off because they wanted to cheap out, then things get hot, its not uncommon for connections with a number of amps flowing to be able to heat a bad connection up to red hot, when this is surrounded by leaf litter, you have a fire,

[TerraHertz]

20190821
http://joannenova.com.au/2019/08/walmart-asks-telsa-to-remove-solar-panels-from-240-stores-and-pay-damages-after-7-fire/

20190913
https://www.naturalnews.com/2019-09-11-tesla-solar-panels-explode-elon-musk.html

20190924
https://www.zerohedge.com/technology/solarcity-panels-spontaneously-ignite-colorado-residential-rooftop

It seems that some genius decided to make panels cheaper, via all-plastic materials instead of glass-faced. So now the panels themselves are flammable.

As for how they catch fire... in addition to possibly poor quality connectors going resistive and catching fire, I'd suspect the reverse diodes in the junction boxes. These fail in three ways: open circuit (probably never happens, but no fire), short circuit (no fire), or resistive (possibly fire.) The diodes can often fail due to nearby lightning strikes and the EMP lighting produces.

I have a writeup of some examples here: http://everist.org/NobLog/20160504_solar_panels_vs_emp.htm

[NiHaoMike]

One solution is to add plastic water pipes to the backside, normally acting as a water preheater, but if a fire starts, the pipe melts and water sprays everywhere. The tradeoff is that ground fault protection would be even more important.

[rs20]

How about backwards installation of DC isolators:

[tooki]

Been getting a barrage (again) with solar power sellers. Latest one is warning of the danger of employing a company with little or low time spent in the business. The leaflet shows a picture of a shonky installation, a row of several solar panels, one of which is charred.

My question is how is this happening? What is the installer doing wrong that causes a panel to burn up?

Sydney. Aust.

In a nutshell, it’s that a) very high voltages are involved (600-1500V DC at up to 20A), b) cheap connectors have much higher contact resistance, and c) bad connectors, junction boxes, cable, and the like can allow water intrusion, or may fail due to weathering. So it’s critical that all the components be of top quality, and that they be installed correctly. High contact resistance, be it due to poorly made contacts, or due to corrosion from poor weatherproofing, results in overheating.

Remember also that because it’s DC, any arcs can be sustained. (It’s not like AC, where much arcing self-extinguishes twice per cycle, when it passes the 0V point.) This in turn informs connector and switch design, and the cheap chinesium clones often get it wrong.

[amyk]

Remember also that because it’s DC, any arcs can be sustained. (It’s not like AC, where much arcing self-extinguishes twice per cycle, when it passes the 0V point.) This in turn informs connector and switch design, and the cheap chinesium clones often get it wrong.

Not surprisingly, people have tried to make use of that too:

[Tomorokoshi]

Egads. Arc Burn City, Arizona!

[Sal Ammoniac]

It's not just installation issues that can cause problems, but the design of the panels themselves.

My 20KW solar installation originally used BP (British Petroleum) panels. They were recalled by the manufacturer because of design problems that could cause them to burn up. I actually had this happen to one of my panels. My installation is ground-mounted on racks in an area covered with dry grass in the summer--it's lucky that when this panel fried itself it didn't start a grass fire.

BP replaced all of my panels with new panels, which are more efficient that the previous panels. As a result, I have fewer panels that generate as much power overall as the old panels did.

http://www.bpsolarsettlement.com/

[orion242]

How about backwards installation of DC isolators:

I'll stand by my comments in the tesla fire thread.  The industry and regulations are a bit wild west yet.  The labeling is retarded on that device for starters.  Will note, the voltage at hand was not mentioned.   We just know its 10A.  Its more likely the output of that device will be miswired because of the labeling.  Still confusing in a case that leads to epic fail.

To have an disconnect labeled that way coupled with the cheapest install labor, disaster in the making.

[orion242]

Remember also that because it’s DC, any arcs can be sustained. (It’s not like AC, where much arcing self-extinguishes twice per cycle, when it passes the 0V point.) This in turn informs connector and switch design, and the cheap chinesium clones often get it wrong.

Not surprisingly, people have tried to make use of that too:

Bare foot, bare hand welding.

Ok, someone hasn't welded much or immune to burns and eye sight is optional.  This guy probably designed the disconnect that crosses poles front to back.

[rs20]

The labeling is retarded on that device for starters.

It's an interesting case, I thought about it for a minute and now the labeling makes perfect sense to me. They don't want to be overly prescriptive when labelling these things, because DC isolators are not always installed between a solar panel and an inverter. So they label each half of the breaker thusly: "when the circuit is broken, the voltage (and therefore arc current) appearing across the breaker must assume this polarity." That is the actual fact of the device that needs to be expressed, and the fact that enables other uses of the breaker.

It's a breaker-centric view, rather than a panel-and-inverter-centric view (which makes sense because ... it's labelling on a breaker).

To have an disconnect labeled that way coupled with the cheapest install labor, disaster in the making.

Agreed nevertheless. Makes you wonder if they should have additional labelling to help the color-by-numbers folks who won't want to think about what they're doing.

[amyk]

I think labeling with an arrow in the direction of (conventional) current flow would make more sense. Ditto for connections on solar panels themselves, the + should have an outward-pointing arrow and the - an inward-pointing one.

[rs20]

I think labeling with an arrow in the direction of (conventional) current flow would make more sense. Ditto for connections on solar panels themselves, the + should have an outward-pointing arrow and the - an inward-pointing one.

Arrows aren't necessarily the panacea they seem to be. This reminded me to run a poll, so here it is.

[Red Squirrel]

There's a lot of scammers in the solar industry as well so have to watch.   The best bet is finding a solar store that you can buy all the parts from and install yourself and bypass the middle man.  Worse are the ones where the panels are "free".  There is always a catch.   Unfortunately it's very hard to find stores that just sell the panels and other parts, at least here in Canada.  Some sites either don't have any kind of option to buy, or they charge a retarded shipping fee that seems to only be there to discourage you from buying.   Like 3x the price of the actual panels.