Author Topic: Solar panel fires in the news in the UK after an alleged 60% increase in 2025.  (Read 5309 times)

0 Members and 1 Guest are viewing this topic.

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 6004
  • Country: gb
https://roofingtoday.co.uk/uk-fire-service-tackles-solar-fire-every-two-days/

This seems a fair article.  There are lots of "Anti-Renewable" lobbyists reporting it too of course.

The note that the largest number of fires occur in the inverters and the next in the "panel itself".  The trend change in the past 5 years here has been away from <450DC panel strings and a outdoor wall mounted inverter to micro-inverters per panel and a single 240VAC drop straight to the consumer unit.

So now the inverters are in the panels and on the roof.  Does that account for the majority of the "Panel itself" fires too.

DC arch'ing and high current from soaked or rotted connections exposed to too much weather is still going to effect 400VDC panel strings.  However, installing twice as many strings is expensive.  So is parallel strings in cabling.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline BadeBhaiya

  • Frequent Contributor
  • **
  • Posts: 319
  • Country: in
We have been seeing a similar 70% increase since 2024 in domestic solar installations but micro inverters are not very popular. I guess it's because they cost more, but also for domestic settings they don't make much sense. Large wall mounted inverters dominate here, the 10kW ones are massive.

Never heard of any solar related fires. Only sometimes about shoddy installations coming undone during storms. We mostly use made in China panels and inverters.
 

Offline indeterminate

  • Regular Contributor
  • *
  • Posts: 197
  • Country: au
In Aus the biggest problem we had was with water ingress into DC isolators mounted on the roof  & shoddy workmanship.
Now that these isolators are no longer required and the regs around instillation and equipment have been tightened up its not so much of a problem.
Your gov will read our reg book and update your reg book in due cores.


 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 8748
  • Country: gb
The article appears to make no real attempt to account for aging of installations. It also misrepresents installations as panels, suggesting a far higher incidence of fires than is the case. Lacking a sense of scale, too - one solar installation fire every other day.. out of the 150 structure fires and 400 overall fires every day.

Your gov will read our reg book and update your reg book in due cores.

The government has absolutely nothing to do with the wiring regulations (thankfully).
« Last Edit: November 07, 2025, 04:24:46 am by Monkeh »
 
The following users thanked this post: Someone

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: nl
I'd expect panel cost and increase in batteries to wipe out microinverters in the near future, except for the <800W wall socket stuff.

Panel cost is making even optimizers a hard sell if you DIY.
 

Online brucehoult

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: nz
I'd expect panel cost and increase in batteries to wipe out microinverters in the near future, except for the <800W wall socket stuff.

Panel cost is making even optimizers a hard sell if you DIY.

Microinverters and optimisers that go on the back of every panel seem like a clear incentive to skimp on quality because buying so many of them is very expensive. Also, their power handling doesn't seem to be increasing as quickly as panel sizes are growing.

On the other hand, wiring from different strings going to equipment inside the house is becoming a major cost. I spent $200 on the panels for each of my 1320W arrays, plus $110 for the 25m long 6mm^2 cable into the house. (Admittedly with only around 12-13 amps at MPP I could have used 4mm^2)

It would cut costs a lot if there were optimisers that could handle groups of say 4 modern panels (400-650W) with the same orientation and shading and output 450V or so to minimise the DC current. You could build a good quality unit for much less than the cost of 4 small ones.
 
The following users thanked this post: mikerj

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: nl
At that point you might as well make the optimizers constant voltage and connect them in parallel too. Micro-inverters, without the inverting, just MPP and DC-DC.
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 9530
  • Country: nl
  • Current job: ATEX product design
I really wish the utility companies would spend half the effort on grid improvements instead of this anti-rooftop-solar FUD and scaremongering.

My favorite one:
"The grid cannot handle the pressure"
OK, please tell me WTF is the pressure in a wire, I'll wait. Here is a textbook of Electrical 101, please find all references of "pressure" in an electrical system.
 

Online brucehoult

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: nz
At that point you might as well make the optimizers constant voltage and connect them in parallel too.

My point exactly.
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 6004
  • Country: gb
The first thing that passed my mind was:  A panel has a 30 year life expectancy, where an inverter is about 10 years.  So putting them in an inaccessible location, on the roof under each panel seems a rather expensive way to replace it twice.

All sounds like it works well in the installers favour.  Especially if they are charging "full DC install" money for a micro-inverter install.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: nl
I really wish the utility companies would spend half the effort on grid improvements

There are few things they spend more effort on, but convincing politicians to make the trillion room in the budget for them is hard.

I suspect that simply shifting load for say 6 hours for half the users on the grid is cheaper than upgrading the grid. The cost is ridiculous, especially in a country like mine.

 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 9530
  • Country: nl
  • Current job: ATEX product design
I really wish the utility companies would spend half the effort on grid improvements

There are few things they spend more effort on, but convincing politicians to make the trillion room in the budget for them is hard.

I suspect that simply shifting load for say 6 hours for half the users on the grid is cheaper than upgrading the grid. The cost is ridiculous, especially in a country like mine.
That number is way off. They were talking about spending 200B in the next 25 years total.
But the incentive structure is way off. A company, making a total profit of 1B creates 35B damage yearly by not connecting new customers fast enough. The damage is enormous, new houses are not built fast enough, and whichever are finished, they cannot move in.
New companies are looking for 18 months of delay. You want to open a factory, it's delayed by 18 months. And they don't care at all about it, it's about their bottom line. There should be no more than 3 months necessary to do this, and heads should roll, and fines going out if it does.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: nl
They could allow new connections in all congested areas right now by selling contracts for connections which are simply disconnected during two blocks of time during congested hours. Add some peak consumption based pricing if you want to incentivize them not to max out their connections recharging their batteries after congested hours. This would solve congestion overnight.

Of course Tennet and friends (individuals, not companies) stand to make a lot more money when they get to spend 200 Billion.
 

Offline richard.cs

  • Super Contributor
  • ***
  • Posts: 1224
  • Country: gb
  • Electronics engineer from Southampton, UK.
    • Random stuff I've built (mostly non-electronic and fairly dated).
I am kind of curious what the consequences of a "panel fire" on a typical UK installation are, it feels like they unlikely to spread to the structure. I assume that most panel fires are probably really connector fires, and the connector housings and cable insulation will be self extinguishing once the arc goes out - at some point the copper will burn back far enough or the sun will go down. Even if it manages to ignite the plastics in a panel it's still mounted to steel brackets on the outside of what's most often 20mm thick concrete tiles. If this idle speculation is true it probably means two things, firstly that such fires probably don't matter very much in the grand scheme of things, and secondly that they're probably under-reported - with most fires leading to the install not working and a bit of charred and melted grot that's only found on repair.

Hunting around on google images I could only find three photos of panel fires that weren't AI slop or building fires that had spread to the roof/panels. I'm sure PV related roof fires happen, though I would certainly worry more about loft mounted inverters than about panels.
 

Offline chasnc

  • Contributor
  • Posts: 30
  • Country: us
Here in the US, some insurance companies are refusing to insure houses with solar panels and grid tied inverters. My friend does fire investigations and he told me that the solar fires are being caused by many factors, not just one or two. First off, the installers buy components based upon price almost universally, so they are getting connectors that are incompatible, cheap inverters (can't handled grid line surges and the thermals needed to survive), and the splicing devices that they are allowed to use in the electrical panels require good technique, i.e., are easy to make a mistake.

And then there is lightning, with these solar panels being mounted on roofs attracting lightning strikes. Add in lazy installers, and you get a witches brew. I don't know why the electrical code here allows the attachment of the inverter grid connection - literally in front of the main breaker. The smaller spliced wires are routed to another breaker panel that goes to the solar inverters. But they are spliced into the mains cable coming directly into the home before the main breaker. The splice connector is an interesting compression style insulation displacement connector. And electricians say that it requires expertise and care to use it correctly.

I would be interested if that is the same in Europe and other areas.
 

Offline richard.cs

  • Super Contributor
  • ***
  • Posts: 1224
  • Country: gb
  • Electronics engineer from Southampton, UK.
    • Random stuff I've built (mostly non-electronic and fairly dated).
I don't know why the electrical code here allows the attachment of the inverter grid connection - literally in front of the main breaker.
...
I would be interested if that is the same in Europe and other areas.
In the UK it is common to add a separate small consumer unit (breaker box) on solar retrofit for several reasons. It allows the new install to meet current standards even if the rest of the property is stuck in the 60s and it makes it easier for the installer who doesn't have to find or deal with any existing wiring faults. We don't use insulation displacement connectors for it though, we have "Henley Blocks" which are just big screw and tunnel terminals.

This photo is very typical, meter just visible on the left, isolator (only on recent installs, older installs you have to pull the fuse before the meter for isolation) Henley blocks, existing consumer unit at the bottom (with a mix of old and new wire colours), new small consumer unit at the top right.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: nl
Over here everything has to be modernized and behind a master switch now, solar is almost always just another group with its own dual function circuit breaker (not mandatory, but since it has a specific GFI class it doesn't make much sense to use separate ones).
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 6004
  • Country: gb
Can't find the video right now, but a fella posted his experience of a house fire in a bedroom which resulted in abandoning the house and phoning the fire brigade.

As they started to tackle the fire, a guy went up the ladder and sprayed the solar panels in thick black rubbery coating.  I believe you can just peel it back off again if the panels didn't otherwise get damaged.

Apparently they don't take 450DC and water hoses lightly.

I might recall incorrectly, but I believe in his case the fire folks then crowbarred most of the panels off while checking the roof for fires.  Dumping them all in the back garden, destroyed, cables cut.
« Last Edit: November 13, 2025, 01:01:31 pm by paulca »
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 6004
  • Country: gb
I don't know why the electrical code here allows the attachment of the inverter grid connection - literally in front of the main breaker.
...
I would be interested if that is the same in Europe and other areas.
In the UK it is common to add a separate small consumer unit (breaker box) on solar retrofit for several reasons. It allows the new install to meet current standards even if the rest of the property is stuck in the 60s and it makes it easier for the installer who doesn't have to find or deal with any existing wiring faults. We don't use insulation displacement connectors for it though, we have "Henley Blocks" which are just big screw and tunnel terminals.

This photo is very typical, meter just visible on the left, isolator (only on recent installs, older installs you have to pull the fuse before the meter for isolation) Henley blocks, existing consumer unit at the bottom (with a mix of old and new wire colours), new small consumer unit at the top right.

Either I am miss reading that wiring or the consumer unit is wired "pre-meter" and will not be billed.  The tails from the consumer unit do NOT pass the meter.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline richard.cs

  • Super Contributor
  • ***
  • Posts: 1224
  • Country: gb
  • Electronics engineer from Southampton, UK.
    • Random stuff I've built (mostly non-electronic and fairly dated).
Either I am miss reading that wiring or the consumer unit is wired "pre-meter" and will not be billed.  The tails from the consumer unit do NOT pass the meter.
The meter is just visible at the centre left, and the two grey cables marked "3N" and "4L" are the metered outputs. These cables go up to the isolator with the red switch and then to the two consumer units so both consumer units are metered.
 

Offline Stray Electron

  • Super Contributor
  • ***
  • Posts: 2803
Here in the US, some insurance companies are refusing to insure houses with solar panels and grid tied inverters. My friend does fire investigations and he told me that the solar fires are being caused by many factors, not just one or two. First off, the installers buy components based upon price almost universally, so they are getting connectors that are incompatible, cheap inverters (can't handled grid line surges and the thermals needed to survive), and the splicing devices that they are allowed to use in the electrical panels require good technique, i.e., are easy to make a mistake.

And then there is lightning, with these solar panels being mounted on roofs attracting lightning strikes. Add in lazy installers, and you get a witches brew. I don't know why the electrical code here allows the attachment of the inverter grid connection - literally in front of the main breaker. The smaller spliced wires are routed to another breaker panel that goes to the solar inverters. But they are spliced into the mains cable coming directly into the home before the main breaker. The splice connector is an interesting compression style insulation displacement connector. And electricians say that it requires expertise and care to use it correctly.

I would be interested if that is the same in Europe and other areas.

   I would be interested to know what state and what insurance companies you're talking about. I've had several sales companies come around and trying to sell or lease a system like that to me and I'm very interested in getting one. However I don't want one on my roof for exactly that reason (and others).  Right now, I'm in central Florida which is the lightning capital of the world and lighting is major concern for me. I've already had a tree in my front yard blown up by lightning and a tree behind my property set on fire by lightning and lightning completely destroyed all of the wiring and electrical equipment in my neighbors house so it is a very real threat in this area. I want a ground mount system away from my house but the local solar companies REFUSE to do any kind of installation other than on the roof of the home.

    A friend of mine is installing a large home solar collection system in central Florida and the county is putting him through Permitting Hell.  EVERYTHING he does has to have reams of schematics, parts lists, and other documentation submitted to them and everything has to be certified by a Professional Engineer, even though PEs know absolutely zero about electrical or solar system.  After two years, he's finally been allowed to install the system (but not to connect it to the grid, that's a whole other fiasco.) Just a couple of weeks ago the county sent an inspector out to look at the installation (the Final Electrical Inspection) and the inspector CLEARLY didn't know what he was looking at. When the owner offered to open up the battery cabinets so that the inspector could check inside the inspector said "OH, Hell No! and practically run from the building! 

  The latest:  My friend owns a good size farm and technically it's two pieces of property (adjoining of course) with his house on one property and the solar system on the other. Now after installing the system and testing it and having it approved by the county, he's been told that they won't allow him to connect it to the grid unless he legally combines the two pieces of property into one!  No one, not even the county can offer any rational for that but that's their latest demand!  So right now his system is up and running and has been approved and his house is connected to it but it's not connected to the grid. 
 

Offline JPortici

  • Super Contributor
  • ***
  • Posts: 3888
  • Country: it
I would be interested if that is the same in Europe and other areas.

some insurance companies are now refusing or increasing the premium astronomically if you have battery storage, they fear battery related fires getting more widespread (incidentally, the same if you also have a chimney, we get maaaaany rooftop fires because people do not clean their chimneys every year)
 

Offline chasnc

  • Contributor
  • Posts: 30
  • Country: us
I don't know why the electrical code here allows the attachment of the inverter grid connection - literally in front of the main breaker.
...
I would be interested if that is the same in Europe and other areas.
In the UK it is common to add a separate small consumer unit (breaker box) on solar retrofit for several reasons. It allows the new install to meet current standards even if the rest of the property is stuck in the 60s and it makes it easier for the installer who doesn't have to find or deal with any existing wiring faults. We don't use insulation displacement connectors for it though, we have "Henley Blocks" which are just big screw and tunnel terminals.

This photo is very typical, meter just visible on the left, isolator (only on recent installs, older installs you have to pull the fuse before the meter for isolation) Henley blocks, existing consumer unit at the bottom (with a mix of old and new wire colours), new small consumer unit at the top right.

Either I am miss reading that wiring or the consumer unit is wired "pre-meter" and will not be billed.  The tails from the consumer unit do NOT pass the meter.


For some reason, my reply to your post did not get published. So let me try again.  :-+

They are allowing the installers to attach the new inverter cabling between the meter and the main breaker, so the meter does see the solar output going to the grid. But the horror to me is the connection method being allowed which is an insulation displacement / piercing connector that is starting to make news here due to fires that it causes. This helps retrofit installations, but it is dangerous in my opinion as seasoned a power electronics guy.
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 6004
  • Country: gb
Either I am miss reading that wiring or the consumer unit is wired "pre-meter" and will not be billed.  The tails from the consumer unit do NOT pass the meter.
The meter is just visible at the centre left, and the two grey cables marked "3N" and "4L" are the metered outputs. These cables go up to the isolator with the red switch and then to the two consumer units so both consumer units are metered.

I was assuming the isolator of the right with the massive black cable was the "presentation point".  Is it the solar incomer?

EDIT as an aside, I lived in a building with an unmetered connection.  There were 3 flats.  In the meter box the incoming tails split 4 ways.  One to each meter for each flat and another directly to a small consumer unit that ran the shared hall lights and the fire alarm system.  No meter.  I believe in such "shared" cases for small buildings the utility retailer just pays a fixed monthly estimate.  Saves having someone read the meter.  They can estimate a fire alarm and 2 lights on time out push switches.  Besides, if it is "commercially billed", then they don't typically bill by consumption at all, but by peek load per 15min interval.  The peek load in this case is both lights on and that's like 200W.  So barely billable.
« Last Edit: November 13, 2025, 04:40:01 pm by paulca »
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline chasnc

  • Contributor
  • Posts: 30
  • Country: us
@Stray Electron, Florida is the state that started the insurance trend due to lightning at first, but then installation fires started occurring. Here in NC we have the #2 spot for lightning at an average of 9 strikes per square mile. Florida has 13 strikes per square mile. I remember reading that there is an area in Australia that is slightly worse, where they can easily view lightning sprites above the clouds (which is fascinating on its own).

My friend here that added solar to his home, he just directly asked his insurance company about his solar system. He wishes he asked prior, because the added insurance cost almost ruined the financial investment. They told him it would have been better to not have it part of the residence and to use an outbuilding for the batteries and power connections. He could have even used a separate transformer there in the outbuilding and provide a separately derived ground right at the building to help attenuate the common mode voltage a lightning strike. Another plan is to use lightning rods, but the H fields during a strike then become an interesting problem. Keep the rods and ground cables away from the house or the electronics in the house can be damaged by induced voltages (parallel wiring to the lightning ground wiring).

And it is becoming very common for installations here in NC to be faulty with poor energy production and outright damaged components and homes. So it is a buyer beware situation.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf