Solar panel fires!

[Mb190e]

Hadn't even heard of this one thought it was going to be the one in Latham last month. That one was an old BP solar account that was taken on by solar city and was chewed on by squirrels apparently.


Edit: never mind your pictures deceived me it was that one. Squirrels did it. Was installing on a fire department that responded to it at the time.

So I guess it was a good idea that I had bird and squirrel guard installed around my solar array. My installer was telling me they never had an issue with squirrels and birds but I insisted on having it installed.

[Panelmal]

So I guess it was a good idea that I had bird and squirrel guard installed around my solar array. My installer was telling me they never had an issue with squirrels and birds but I insisted on having it installed.

I've got hundreds of installs under my belt as well and laughed at the salesman trying to sell us squirrel guard in the warehouse one morning. That's just how it was reported, personally I think it was a failed connection in a bp panel which has happened a lot lately.

[SunEagle]

They are pretty much maintenance free and I don't think these fire are so much a maintenance problem but a problem of incorrect installation in the first place...most probably having to do with bad connections I would guess.

And you would know about bad connections done by the original installer. You found them and fixed them but I have no doubt that you would have had an issue if one of those wires you fixed had come loose and arced to the frame.

Maintenance may be nothing more then checking the integrity of an electrical system every so often to make sure nothing has changed or has been disturbed by outside sources like the weather and critters.

We called those walking PM plans where an electrician just did a walk around one of our machines to make sure there wasn't any broken connections or loose wires. Easy and cheap to do but priceless if a problem is found before it results in down time.

[thejq]

Can you find a single citation that supports the idea that <500 V DC lines are hazardous in the way you are suggesting?

Electric field = [(Resistivity) * (Current)] / [(pi/4) * (D^2)] (note that voltage doesn't matter, just current)

For 10 AWG copper carrying 20 A

E = [(1.68×10−8) * (20)] / [(pi/4) * (.00259^2)] = 0.064 V/m

Google is your friend. http://www.renewableenergyworld.com/...em-safety.html
It specifically talks about arcs caused by high voltage DC with more causes than what I thought of.

As I said before, the arc is not caused by magnetic field which's induced by electricity flowing in the wire (what you illustrated). I'm talking about static voltages. Think of it as a high voltage battery with +/- terminals that're pretty close. The + side attracts - particles, and - side, + particles. With enough build up and the right humidity, the two sides of particles (or ions) will some times touch and release the energy, in extreme case, produces an arc. If you enclose it in a metal conduit, the metal is a conductor that constantly neutralizes ions, so no charge build-up no arc. With AC, the polarity of the battery changes at 60Hz, so there's no time for charge to build-up.

[inetdog]

I'm not SE engineer, but from what I read, it detects both.

An arc fault detection circuit (very different from a ground fault detection circuit) will detect arcs. That means a radio frequency current and voltage signal which results from the arc repeatedly making and breaking or fluctuating in voltage.
It is possible to build an arc detector which is sensitive to both AC and DC arcs, but it is also much easier to build a circuit which detects one or the other best.
None of them detect constant electric fields unless those fields cause current to flow.

Be careful of the difference in implication between static electric fields (anytime the field values stay constant) and static electricity which commonly refers to fields generated by a mechanism that cannot sustain much current at all.

The current and duration profile of an arc produced by static electricity will have so little current associated with it that it will not trip most AFCIs.

In most if not all of the reported fires a ground fault detector would have alerted to the problem just as efficiently as an arc fault detector.
The only thing that a ground fault device will not detect is an arc or solid connection between + and - wires that does not involve any current flowing to the grounded panel/rack/conduit system.

[sensij]

Google is your friend. http://www.renewableenergyworld.com/...em-safety.html
It specifically talks about arcs caused by high voltage DC with more causes than what I thought of.

As I said before, the arc is not caused by magnetic field which's induced by electricity flowing in the wire (what you illustrated). I'm talking about static voltages. Think of it as a high voltage battery with +/- terminals that're pretty close. The + side attracts - particles, and - side, + particles. With enough build up and the right humidity, the two sides of particles (or ions) will some times touch and release the energy, in extreme case, produces an arc. If you enclose it in a metal conduit, the metal is a conductor that constantly neutralizes ions, so no charge build-up no arc. With AC, the polarity of the battery changes at 60Hz, so there's no time for charge to build-up.

OK, let's slow down. The equation I provided is not that of the magnetic field, it is the electric field. Large electric fields can be hazardous (corona discharge in extreme cases), but the electric field generated here is insignificant.

The hazards described in the article, or in the example of a battery terminal that arcs, are not due to the attraction of charged particles. They are due either to a termination failure or an insulation failure. In the case of insulation damage, dielectric breakdown can occur under the right conditions, and an arc can occur between + and -, or possibly to ground if it is not a floating system. However, just so it is clear, this has nothing to do with the accumulation of charged particles. Although grounded metal conduit serves many purposes, neutralization of ions is not one of them.

With undamaged insulation and proper terminations, I don't see how the risks of DC in PV systems are any greater than the risks of AC.

[SunEagle]

OK, let's slow down. The equation I provided is not that of the magnetic field, it is the electric field. Large electric fields can be hazardous (corona discharge in extreme cases), but the electric field generated here is insignificant.

The hazards described in the article, or in the example of a battery terminal that arcs, are not due to the attraction of charged particles. They are due either to a termination failure or an insulation failure. In the case of insulation damage, dielectric breakdown can occur under the right conditions, and an arc can occur between + and -, or possibly to ground if it is not a floating system. However, just so it is clear, this has nothing to do with the accumulation of charged particles. Although grounded metal conduit serves many purposes, neutralization of ions is not one of them.

With undamaged insulation and proper terminations, I don't see how the risks of DC in PV systems are any greater than the risks of AC.

Which is what I have also been saying.

Electricity in any version is dangerous to those that take it for granted.

Two rules to stay alive.

1. It ain't "de-energized" unless it is grounded.

2. Never become complacent while you are around it or we will be reading your epitaph.

[CodeSection]

Ok, as a newbie following this thread, I got lost somewhere on the middle. I never thought about the fire aspect.

I am having a very reputable company in my area install 81 LG280 panels with 81 M250 Enphase Microinverters in a desert inviroment with the racks on the desert floor as opposed to on the roof. It is my understanding that the Enphase Microinverters no longer require a ground wire.

So, with this thread about magnetic fields, arcing, grounding issues, installation issues, maintenance issues....do I need to rethink about even installing solar? I know that may sound silly to many, but to a newbie that has tried to read up on many items, this is the first I ever heard of the fire dangers. Are there regular routine maintenance items other than inspecting wires and the Microinverters? Are there electrical issues that could harm someone by the mere presence of standing next to the panels and perhaps building up static electricity that somehow could start a change reaction of some type? Please forgive me for my ignorance.

As always, I appreciate any and all comments!

Thanks!

[inetdog]

With undamaged insulation and proper terminations, I don't see how the risks of DC in PV systems are any greater than the risks of AC.

The risks in an undamaged system without adding somebody poking their fingers where they do not belong are vanishingly small for both AC and DC.
The question is how the two can be faulted and what the consequences are.

For fires, the most probable fault is damage to insulation causing either an arc or a glowing connection (local hot spot, not the result of arcing.)

I suspect the chances of both are similar for AC and DC of the same voltage.
But what happens after an arc starts is very different.
An AC arc will cause some damage when the arc strikes and will likely remove material to the point that eventually the arc will stop at the current zero crossing and will not restart.
A DC arc has no zero crossing, so the arc will not extinguish until so much material has been removed that the arc stops even though the voltage remains the same. By then you have a fire for sure.

The same difference applies to accidental shorts across the wires involving a person, except the human heart rhythm is an order of magnitude more sensitive to 50/60Hz AC than to DC.

[bcroe]

I never thought about the fire aspect.

That was one of the first things I thought about, and one of many reasons I have a ground
mounted array hundreds of feet from any building. I am going to add this which just might
relate to some fires.

After some 22 months of service, I have decided to join the "TIGHTEN ALL CONNECTIONS
REGULARLY" school. I have been running 15 KW out every sunny day not just at solar noon,
but for many of the daylight hours. None of the wiring is approaching its absolute limit,
but at the end of sun it can definitely be felt to be warmer than the surroundings.

There are a couple inverters, DC and AC disconnect boxes, several AC breaker boxes, and a
DC combiner box. All the connections were initially checked by me. About a year ago a
circuit breaker failed, apparently from heat at one of the line terminals. This month a DC
screw connection to a disconnect switch (carrying 21A @ 360 VDC) burned out. The
problem was detected & repaired in a couple hours, but the issue is, why did it happen
at all?

As others have suggested, perhaps the constant heat/cool cycles at significant power can
work these connections loose. Loose enough, the resistance will rise and they burn out.
I undertook (in the dark of course) to check them ALL for tightness. Sure enough, the
tightness of screws after 1.5 years varied A LOT. Since all were done by the same person,
often many the same day, some had apparently loosened up. My conclusion is that
these all need to be rechecked annually, at least until they ALL show some settling to
stability.

So far, no issues with split bolt or soldered connections. Bruce Roe

[Mike90250]

Ok, as a newbie following this thread, I got lost somewhere on the middle. I never thought about the fire aspect. ......

Thank user thejq for starting up a whole bunch of nothing.

In a modern Glass and silicon panel, there is a bit of EVA sealant, and a thin Tedlar backsheet. There is very little
in a PV panel to burn. It doesn't create mystery fields that will arc and torch your house. The AC/DC static field charge buildup bs is just nonsense.

But what can happen, is rats, squirrels, chewed insulation, nest materials, trapped leaves, palm fronds and that stuff can be lit off by damaged wires, and there goes the neighborhood.

And, yes, copper creep under pressure. Even bit me. Installer added a proper 40a conductor to a buss bar, and either forgot to torque and re-torque an hour later, or something, but the connection failed. Inside a UL box with UL wire. Scorched stuff, but no flames. I now have a gunsmiths torque driver to go through the ePanel with.

[thejq]

OK, let's slow down. The equation I provided is not that of the magnetic field, it is the electric field. Large electric fields can be hazardous (corona discharge in extreme cases), but the electric field generated here is insignificant.

The hazards described in the article, or in the example of a battery terminal that arcs, are not due to the attraction of charged particles. They are due either to a termination failure or an insulation failure. In the case of insulation damage, dielectric breakdown can occur under the right conditions, and an arc can occur between + and -, or possibly to ground if it is not a floating system. However, just so it is clear, this has nothing to do with the accumulation of charged particles. Although grounded metal conduit serves many purposes, neutralization of ions is not one of them.

With undamaged insulation and proper terminations, I don't see how the risks of DC in PV systems are any greater than the risks of AC.

I don't see the relevance of the electric field equation in this discussion, and the reason for keeping bringing it up, since both you and I agree it's not the cause of arcing. I also failed to see what your point is after all these. The original questions was whether high voltage DC is more prone to causing fire, and microinverter is safer in that aspect. My point from the beginning was most likely yes. What is your point? I gave one simple example of what can cause arcing that's unique to high voltage DC. I never said it's the only nor necessary-and-sufficient reason. By regurgitating the points in the article I linked, logical you agreed with my main point, but your wording sounded like disagreeing. Reading your reply reminded me of the days of high school debate where triumph by confusion is often the winning strategy.

[sensij]

...With DC, if the wire is not shield properly, charged particles will accumulate around the wire. With the right voltage and humidity, arc occurs to release the energy build-up...

...High voltage DC can create a static field around the wires. If not properly shielded, can cause arcs or sparks...

I'm talking about static voltages. Think of it as a high voltage battery with +/- terminals that're pretty close. The + side attracts - particles, and - side, + particles. With enough build up and the right humidity, the two sides of particles (or ions) will some times touch and release the energy, in extreme case, produces an arc. If you enclose it in a metal conduit, the metal is a conductor that constantly neutralizes ions, so no charge build-up no arc. With AC, the polarity of the battery changes at 60Hz, so there's no time for charge to build-up.

You said the above...

I don't see the relevance of the electric field equation in this discussion, and the reason for keeping bringing it up...

And you don't see how the strength of an electric field is relevant? Yeah, there is no point to this at all. Mike's got it right.

[SunEagle]

That was one of the first things I thought about, and one of many reasons I have a ground
mounted array hundreds of feet from any building. I am going to add this which just might
relate to some fires.

After some 22 months of service, I have decided to join the "TIGHTEN ALL CONNECTIONS
REGULARLY" school. I have been running 15 KW out every sunny day not just at solar noon,
but for many of the daylight hours. None of the wiring is approaching its absolute limit,
but at the end of sun it can definitely be felt to be warmer than the surroundings.

There are a couple inverters, DC and AC disconnect boxes, several AC breaker boxes, and a
DC combiner box. All the connections were initially checked by me. About a year ago a
circuit breaker failed, apparently from heat at one of the line terminals. This month a DC
screw connection to a disconnect switch (carrying 21A @ 360 VDC) burned out. The
problem was detected & repaired in a couple hours, but the issue is, why did it happen
at all?

As others have suggested, perhaps the constant heat/cool cycles at significant power can
work these connections loose. Loose enough, the resistance will rise and they burn out.
I undertook (in the dark of course) to check them ALL for tightness. Sure enough, the
tightness of screws after 1.5 years varied A LOT. Since all were done by the same person,
often many the same day, some had apparently loosened up. My conclusion is that
these all need to be rechecked annually, at least until they ALL show some settling to
stability.

So far, no issues with split bolt or soldered connections. Bruce Roe

Great to hear you have had no issues. Also good info that you found that the connections had changed over a period of time.

Yes I agree that a solar pv system is relatively "maintenance free", but all electrical wiring systems require spot checks every so often because they do change and usually not for the better.

As for the potential of a fire. Chances are higher that it will start inside the building from another source then from a failed pv wiring connection. And as Mike pointed out there really isn't much flammable material in a solar panel to burn.

So while this thread could be an eye opener for some, I wouldn't worry about your pv system going up in flames anytime soon.

[bcroe]

Great to hear you have had no issues. Also good info that you found that the connections had changed over a period of time.

Yes I agree that a solar pv system is relatively "maintenance free", but all electrical wiring systems require spot checks every so often because they do change and usually not for the better.

As for the potential of a fire. Chances are higher that it will start inside the building from another source then from a failed pv wiring connection. And as Mike pointed out there really isn't much flammable material in a solar panel to burn.

So while this thread could be an eye opener for some, I wouldn't worry about your pv system going up in flames anytime soon.

All my inside connections are inside steel boxes; not likely to ever flame. But some possible failures
could be difficult and expensive to fix, think inverter. All 15KW feeds through a couple split bolt taps,
which aren't so easy to check being taped up. But I can at least check for warmth by hand and with
my infrared thermometer. Bruce Roe