eevBLAB #64 - Tesla Solar Panels Are CATCHING ON FIRE!

[Marco]

I realized in a big way, Tesla's solar-shingle electrical connectors are going to have great difficulty not burning up.
It's high heat, water, cold on 100's of connections.

I think you could make them contactless. Put down a sheet with an insulated electrical backplane and then glue the shingles to it, with capacitive coupling between them.

Sounds like something which is patented though.

[tautech]

I don't typically see isolators per panel.  If they are only installing them at the end of a series string....that's a fair number of connectors with little way to kill the power.

Once the string has the load removed there is zero current through the connector.
With the current (excuse pun) range of panels sizes and ratings, connectors need only see 10A but granted with a 10 panel string some 300V.

So with 100's of panels on a typical WalMart installation there only needs be some installer wally with the wrong crimper or limp wristed termination into the uInverters and you have a recipe for flame outs, especially with HV DC.

In a tiny 3 panel LV system we're working on has been wired by a registered sparky and every termination has bootlace ferrules.

[maginnovision]

I think they've deployed some solar shingle type roofs. They're not all of the solar type though, only some. Even if a section of the roof is optimal it's only partially the solar shingles. I don't know how many real paying customers have had them installed vrs massive supporters/employees. Even still number of installations seems very low, I can't find concrete numbers though.

[floobydust]

FF past 1:30
Finally found more on the Tesla Solar Roof Tiles- I'm wrong about the connectors, they look like MC4 underneath. The tiles sit up high raised up off the roof, I thought they were flush.
Safety features are having the array floating wrt earth ground, "16V"? diode trunk and loss of AC automatic disconnect.
edit: added pic and I like the hinge but don't like the overlap on each tile for shade and the fact that windy rain will blow water in there.

[orion242]

Nice find.

They sure look nice over standard panels.  What an utter nightmare to find a half dozen dead shingles once you notice something isn't right.

[orion242]

added pic and I like the hinge but don't like the overlap on each tile for shade and the fact that windy rain will blow water in there.

Doesn't look great for a roofing material.

https://ca.brockwhite.com/userfiles/documents/info/2_0800377-firestone-clad-gard-metal-roofing-underlayment_info.pdf

The crap they put below sounds fairly solid, but its certainly not meant to be the final water proofing.  If all those connectors are exposed to water regularly, seems bound to have failures just do the shear number of connections required.  The air gap behind the shingles should allow for quick drying, bonus.  Troubleshooting a failure still looks insanely expensive.  How the hell would you narrow it down?  The "diode trunk harness" and all the wiring is inaccessible without removing the shingles.  Would like to see what this diode harness looks like.  At 14:11 in that video, it might be the vertical thingies at the left side of the video.  Broken up, but those guy pretty much destroyed every part of that install as firefighters seem to do.

[orion242]

With all the BS OSHA safety crap is around arc flash hazards today, I'm bewildered how a connector like this can be rated for 1KV @ 30A without requiring permanently embossed warnings, high viz colors and LOTO provisions.  1kv AC has an absolute $hitload of safety requirements at any moderate power level.  Internally in a building, your going to have to accommodate that with special construction in most cases.

Unlike the AC power line, solar panel outputs are effectively impedance protected by the voltage to current curve of the solar array which limits voltage and current to values not much higher than the typical operating conditions.  So you can get hazardous voltages and currents but power and energy are limited preventing anything like arc flash.

A short just reduces the voltage to zero and raises the current 10s of percent.  The power ends up dissipated in the panels which have plenty of surface area.  If that increase in current causes a connector to fail, then it was defective already.

I understand the flash part is certainly not an issue as it is with mains.  The available fault currents are very different.  The arc potential though if they are running high voltage DC setups certainly is there.

High viz vest, safety glasses, hard hat, work boots just to be in a 99% complete building, yet this kind of arc might be possible opening a non-descript connector?  Working on anything live, holy crap...its a PPE nightmare these days.  Flame retardant undies, everything must have labeled fault currents, etc.  Nearly impossible to comply and still operate.  I look like an astronaut to check low voltage controls on high power circuits and barley have any fine movement control at that point, which IMO is more dangerous than all the BS one has to wear..  There are certainly cases the village idiot is on a roof near these installs and may decide to move them out of the way for a minute.  How about the Tesla moron walking on the panels decides to inspect the connectors.  As a former non-solar sparkie, I would have never expected that kind of power on something so simple to access and disconnect by hand with free air cabling.  Will it blow up in your face like mains, doubtful.  Could it burn the hell out of your hands and pose a serious shock hazard, looks possible.

Personally, I don't think the NEC/AHJ is aware of the potential issues with these installs.  Admittedly, I don't know enough what common practice is on these installs either.   Anything close to 1kv @ 30a Dc on these connectors is insanity without full LOTO.

[floobydust]

Heat with full sun at 800-1000W/m^2 while generating electricity into a short-circuit- that is extra heat for the cells.
I think the connectors and module internal interconnects can see much more than ISC if there is an arc, as other panels in the system contribute, they are generating current in parallel connection and we're all too cheap to use a fuse or extra diode barrier there.

Walmart and Tesla are now "actively negotiating to resolve the lawsuit" to contain the hull breach. The fact that Walmart has sue Tesla to get them to act and fix the substandard installations, is really terrible.  Ma and Pa would be totally ignored if they had any troubles with their solar system, including the fire.
I can't tell if the company is selling unicorns and rainbows but the conflicts of interest around the purchase of SolarCity really says something creepy about the business priorities.

Heat a mated connector and wires to say 70C and dunk in cold water and see what happens. Similar to a hot connector suddenly get cold rain from a thunderstorm- water gets sucked inside. I've seen many products fail because the seals malfunctioned during the cold shock, or the tiny pressure change internally makes a low pressure, or the housing did not contract fast enough to keep seal tolerances.
So I am not a fan of outdoor connectors being immersed or getting wet at all. The Weidmuller PV stick is only IP65, but Stäubli MC4, Amphenol H4 are IP68.

[SilverSolder]

We are still in the early days of the solar industry...  -  What's the betting that in time,  regulations will require solar panel wiring to be completely enclosed...   -  which, after reading this thread, is what I would be insisting on with any installation I have anything to do with!

[Marco]

Instead of corded connectors, why don't solar panels simply have sockets on the junction box? It would vastly reduce the exposure of connectors to water.

[floobydust]

Amphenol UTX was subject to recalls in late 2016, vendor contaminated plastic which cracked and allowed moisture ingress and shock hazard. Tesla is right to implicate them.

The industry has too many different proprietary connectors that come with the panels. They all just might plug into each other... and there's no safety agency approvals or tests on how they do when cross-mated. It's a mess.

There are some junction boxes with MC4 on there, I think it's convenience to have the connector on flying leads.
The panel junction boxes seem to run hotter than everything else?
Here are thermal pics taken from photovoltaikbuero which show a shorted bypass diode, and an open string within a three string panel.

The open string runs hotter- which is counter-intuitive:
"In order to understand the image, keep in mind that solar cells in an open-circuit state are about 2-3K warmer than cells operating based on maximum power point (MPP). This is because, in an open circuit, all of the solar energy that’s not reflected from the panel is turned into heat in the panel; in the case of MPP, on the other hand, some of the energy is drawn from the panel in the form of electricity. As paradoxical as it may seem, this current leads to cooling. Unlike in the case of a short-circuited bypass diode, for this fault you won’t see the typical image with some much warmer cells, like in the thermogram above. Here, all cells in the affected area are equally warm, and the panel socket looks much warmer than its surroundings."

[orion242]

Amphenol UTX was subject to recalls in late 2016, vendor contaminated plastic which cracked and allowed moisture ingress and shock hazard. Tesla is right to implicate them.

"Sold At:
Solar panel installers nationwide from November 2016 through December 2016 for about $330 per solar panel with the connectors."

Ok Tesla can blame how many installs on a specific connector only available for two months?  They were installing these over a several year period, most before Nov '16 off memory.  Site after site with issues, doesn't sound like this is going far to explain things.

we're all too cheap to use a fuse or extra diode barrier there.

Maybe there is a recall on the bolts used in the fuse holder as well.  lol.

Ma and Pa would be totally ignored if they had any troubles with their solar system, including the fire.

Agree, it sure sounds like the wild west yet on these installs, codes, etc.  Resi setup whatever.  Smaller number of panels, the crews are likely a bit better and actually give a rip.  Almost always local and don't want to piss in their pond.  Hundreds of panels installed across the country as cheap and quick as possible, bit more scary.  Step above traveling carnies I suspect.  Install and never seen again.  Only part they brag about is how quick they did it.

Have a new customer now with a commercial scale solar install.  Haven't had a reason to be on the roof yet.  Next visit there, I'm checking it out.

[orion242]

I've seen many products fail because the seals malfunctioned during the cold shock, or the tiny pressure change internally makes a low pressure, or the housing did not contract fast enough to keep seal tolerances.

Water is nearly impossible to keep out 100% if exposed outdoors.  Seal it perfectly in an outdoor enclosure, hard problem.  Condensation is a big problem.  Hot and humid when you perfectly seal, ambient temps fall at any point you have condensation to deal with on top of the pressure change.   If any air can enter, condensation is a compounding issue.  Underground conduits can collect water in the driest places if air can get in.  Its almost guaranteed to have water to some degree with wide temp swings and exposed to weather.

The longest lasting connections in outdoor enclosures I have seen, have tight sealing covers/openings but make provisions for water to escape.  Aka, specifically designed to be open to atmosphere and using gravity to remove the little water that does get in.  Seen numerous enclosures with robust sealing that just trap in water and fill up if they didn't allow for water to escape.  Open the cover and it pours out, connections submerged in water and completely nackered.

Underwater, aka water proof doesn't need to deal with the wide temp swings as something outdoor.  The pressure/condensation problems are minor compared to sitting on a roof.

[floobydust]

The point is Amphenol had a plastics quality problem that caused the connectors to fail, albeit in late 2016 but recalled due to an installation company encountering the problem.
Would Amphenol be forthright over dangerous quality issues otherwise, or just sweep it under the rug? Let's replay their official PR response, oh wait it been scrubbed from the Internet as their stock takes a big hit over this. Just amazing.
You need connectors with batch/serial numbers it seems. How does Ma and Pa know if their connectors are affected?

Weather-Pack/GM/Delphi/Aptiv and Deutsch automotive connectors use triple-lip silicone rubber seals for both the housing and wires. Under the hood, you get extreme temperature excursions and water splash- they are proven to work very good. They have no agency approvals or HV design of course.
Why are PV connectors sealed with only one tiny o-ring? Expected to last 20 years? They could be built much better and not as cheap.


I think a drone fly-over with thermal imaging cam would be well worth it to spot problems on the panels, but the wiring is underneath so we need a small rover with thermal cam to drive underneath. Musk should postpone his Mars work, and maybe make an inspection robot.

[G7PSK]

This is a humorous look at this.

https://youtu.be/f58ohDLw6Tw

[Marco]

There are some junction boxes with MC4 on there, I think it's convenience to have the connector on flying leads.

You could also have the default be one socket and one flying lead on the junction box, for simply attaching neighbouring panels you would still have everything at hand with the panels.

For every string there would be only one connection which the (amateur) installer would be tempted to simply leave dangling on the roof surface, instead of one every panel.

[floobydust]

We want the convenience of plug-in connectors, so a plumber can wire a rooftop installation within minutes.

Any electrical connector located outdoors has a really hard life. It's getting attacked by UV, heat, water, ice, oxidization, vibration from wind etc. I guess all we can do is protect them from as much as possible.

A lot of mechanical engineering seems to have gone into the panel's mounting struts and extrusions, with no thought about the wiring, just use low cost tie-wraps that age and easily break. A simple cable-tray might be better and could be part of the extrusions as a U-channel, to keep the connectors and wiring from laying on the roof.

The industry is kind of screwed, with over a dozen different panel connectors that are implicitly "MC4" compatible. The connectors need to be open-source - which could never happen due to certification costs or quality control.
Stäubli (Multi-Contact) surely has the tight tolerances and quality German manufacturers are renowned for. But nobody wants to pay extra for that, so Cheap Panel Co. makes their own connectors which come already assembled to their panels.

[SparkyFX]

The point is Amphenol had a plastics quality problem that caused the connectors to fail, albeit in late 2016 but recalled due to an installation company encountering the problem.
Would Amphenol be forthright over dangerous quality issues otherwise, or just sweep it under the rug? Let's replay their official PR response, oh wait it been scrubbed from the Internet as their stock takes a big hit over this. Just amazing.
You need connectors with batch/serial numbers it seems. How does Ma and Pa know if their connectors are affected?

Incompatible to "Freedoming", but we´re getting political here.

Weather-Pack/GM/Delphi/Aptiv and Deutsch automotive connectors use triple-lip silicone rubber seals for both the housing and wires. Under the hood, you get extreme temperature excursions and water splash- they are proven to work very good. They have no agency approvals or HV design of course.

There are known cases in which engine oil was pulled through an oil pressure sensor wiring into the engine controller - which does have a pressure equalization valve(?! No idea how it would even be possible). Utilizing the insulation as a hose. So yeah, it is kind of airtight.

But when it comes to high power, it is all crimped or crimped cable lugs on zinc coated bolts.

Why are PV connectors sealed with only one tiny o-ring? Expected to last 20 years? They could be built much better and not as cheap.

Cost-optimized to death... seriously, if it was up to me i´d go with a hardwired install, crimped cable lugs mounted to a busbar, in metal enclosure with added pressure equalization membrane, ideally under a shroud. But yes, it is probably more expensive to install.

I think a drone fly-over with thermal imaging cam would be well worth it to spot problems on the panels, but the wiring is underneath so we need a small rover with thermal cam to drive underneath. Musk should postpone his Mars work, and maybe make an inspection robot.

Speaking of reinventing the wheel:

[Marco]

Another stupid question, why don't solar panels come standard with a peel off cover? Given how frequent high voltage strings are used it would seem a convenient protection.

[SparkyFX]

At first they are usually isolated high voltage DC, which means safe to the touch unless there is a ground fault.

As far as i understood there is a glass-like front over them already, mostly for mechanical and chemical reasons, as they need to withstand hail and ice to some degree and that also serves as electrical insulation. Another cover would block/filter a certain amount of sunlight and any peel off film does usually not withstand the weather for a long time.

[Marco]

It needs to last a couple of days at most, blocking sunlight was kind of the point ... a vinyl peel off film will last a lot longer than that.

[SparkyFX]

Ohh! You mean during installation or maintenance to block them from generating voltage? I guess so you can inspect them for damage before you haul it on the roof. Installers might have cloth and clamps for that case.

Electrically the safety depends on the order in which things are connected, I don´t see a problem if the isolator switch is toggled last. The connectors probably burn/spark only while conducting high currents, so as long as the circuit is open at the isolator switch, there is no tendency for arcing.

[tautech]

Another stupid question, why don't solar panels come standard with a peel off cover?

Considering all the plastic waste that would arise those with a real green bent that install PV's might be offended by such a consideration. 

Given how frequent high voltage strings are used it would seem a convenient protection.

We did a 3 panel 300W jobbie the other day and just left one connector open until everything else was finished.
Easy with a tilted array but not so if they are mounted flat.

[Marco]

Considering all the plastic waste that would arise those with a real green bent that install PV's might be offended by such a consideration. 

At a couple 10s of microns it's really not all that relevant, doesn't make a mess either unlike the foam packing.

[SparkyFX]

It depends on what the users/handlers think about such solutions and if the amount of linked accidents during installation is really so high that it justifies demanding this for every panel produced. If it electrostatically charges the panel while peeling, it might damage other parts of the installation and the manufacturer would have a hard time to work out the claims, so it can´t be cheap plastics, it needs to be defined conductivity or special properties, non stick/weld/adhesive type that is guaranteed to not stain the surface, would need additional chemical cleaning or attract dirt electrostatically. The devil is in the details for this special sauce.

I guess if one installs dozens of these, it might be more annoying than it helps with safety, as this means you need to collect large volumes of trash on the roof and need to get it down safely. Plus the unnecessary waste disposal takes time and if these go flying you annoy the neighbors too, forget to remove them and you need additional service. If it prolongs the process, requires more tools or creates new safety risks, raises additional costs it might not be worth doing.

I personally think as this would be a one time use only thing, it´s probably not considered eco friendly. Of course it is a ridiculous low amount per panel... but even apart from that, i think it would be a waste of resources and money. With guaranteed recycling some energy of it might be preserved, but thats not mandatory everywhere. In the end you´d pay triple: higher purchase price, additional process time on the roof, disposal cost just to avoid doing things in the right sequence.