South Korean lithium battery plant blaze kills at least 23 A fire at a lithium battery manufacturing plant in Hwaseong, South Korea, on Monday killed at least 23 people and injured eight others. Parts of the top floor of the facility – operated by battery manufacturer Aricell – collapsed, and chunks of concrete spilled into the road. The warehouse stored approximately 35,000 battery …
Indeed..
https://www.bbc.co.uk/news/articles/cjlllw0kj67o
People living near the scene of an fire in Renfrewshire have been told it is safe to go outside again.
Very impressive photos of the smoke plume. Sampling what's in that smoke may reveal how many people downwind have become involuntary heavy metal fans.
I'm intrigued to know what the root cause was.
Well, the smoke generator was from this company-
https://www.weee.co.uk/waste-batteries
So it's probable it was another battery fire. The Bbc's story was a bit coy on that aspect. It raises some important questions though given it's not the first battery fire, and unlikely to be the last. This-
https://notalotofpeopleknowthat.wordpress.com/2024/06/21/yet-another-e-scooter-fire-disaster-when-will-the-govt-act/
A spokesman for Hampshire and Isle of Wight Fire and Rescue Service said: “The initial reports and the preliminary fire investigation has found that an e-scooter is the most likely cause of the fire.
“The main danger occurs when e-bikes and e-scooters are left on charge and unattended in homes or in communal areas such as hallways and stairwells.
“Charging lithium batteries indoors increases the fire risk, especially if charged overnight when occupants are sleeping.”
destroyed 7 homes, and is one of many stories where recharging e-bikes or scooters has been the cause. With the e-everything policy, fire and safety risks are becoming far more common, and inevitably insurers are going to act. Which will be a problem, if they insist on charging outdoors in a safer area. That will be impossible for many existing properties, especially when it comes to EVs that will get charged in garages attached to many people's homes. Then there's the larger health & safety risks from all the grid-scale battery farms being built in close proximity to populations.
Which is also the serious point about this fire, ie sampling the plume and any contamination that might have caused to soil or ground water.
I have trouble comprehending how ebike and such batteries are so prone to going pop. Maybe they are made with crap cells.
I've designed some lithium ion packs as part of my job and also used some old 18650 cells to re-battery some old cordless tools and my crappy laptop. Admittedly these were all with good quality cells, Sanyo, Boston Power, Samsung, rather than complete random ones from China.
If you don't overvoltage them or try and charge them when they are too far discharged and don't over current them they should not go pop. And the video from the South Korean incident appears to show random spontaneous immolation. The only time I've seen a battery go pop was a NIMH that was just too far gone.
Saying all that I still don't leave batteries on charge overnight.
.. these machines are made down to a price, with safety being 3rd or 4th priority (if considered at all)
Yep. I guess this is a challenge for the fire service's forensic investigators, and maybe Trading Standards? Luis Rossman made some videos where he was building an e-bike, complete with smoke. Bigclive's YT videos also shows a lot of teardowns of dodgy battery powered kit. Problem seems to be a combination of cheap batteries and charging/battery management systems.
I think you're right about price though, but maybe also some pressures for things like fast charging, which may be pushing the charging too hard. I'm not expert, but I read that some batteries also suffer from dendrite formation that can lead to internal shorts inside batteries, and then potential fires. Thunderf00t also gave a good description that high capacity/energy density batteries are essentiall bombs, and all that energy just wants to break free.
I'm curious about the charging circuitry though. It seems that including thermistors or some other temperature sensing should be cheap and able to detect batteries getting too hot, then reduce or stop the charging. What I don't know is if when batteries get to that point, there's thermal runaway and it's too late. If I were evil, I'd acquire one of the e-scooters that often litters the pavements near me and strip it down. Their battery packs are in a cast aluminium(?) unit that looks like it's detachable and maybe helps with the cooling. I haven't seen any stories about those ones catching fire though.
(And on the subject of electrical fires.. Last week I had to replace a door camera/intercom system in a property. In which the lighting cicuit tripped when I disconnected the trasnsformer that should have been powering it. Then discovered someone had installed a second transformer tapped into a light fitting, and was buried under 2 layers of loft installation.. And was too hot to touch. So after some cursing whoever installed that, I removed it and added a smoke detector for the loft space as well.)
the design of batteries is flawed. The multi layered packs alternating between the cathode and anode is the real cause. Because there are lots of layers, this also means there is very little separating them, thus a minor fail such as a crack in the plastic means contact between the two parts and you have a self sustaining explosition. THe safety layer between the active components is also of course basically nothing because its all about weight so they cut down this part.
Other batteries more rigid and have two distinct or more separated parts, lithium batteries dont.
I've designed some lithium ion packs as part of my job and also used some old 18650 cells to re-battery some old cordless tools and my crappy laptop. Admittedly these were all with good quality cells, Sanyo, Boston Power, Samsung, rather than complete random ones from China.
From watching Bigclive.. there seems to be a problem with fake batteries. Including one where low capacity cells had been sleeved inside 18650-sized tubes and a fake lable applied. I guess for re-batterying stuff, the challenge is how to test cells to try and make sure they're safe.
Escooters (presumably, ebikes analogously) are often driven until they die. That means the batteries go through a lot of zero-full cycles. If the battery pack becomes unbalanced, that does bad things to the worst cell in the pack. Optimistically, the charge controller should detect this state and refuse to discharge once even one out of n cells is dangerously low. Realistically, they just look at total pack voltage.
If you google around, there's some fun papers about this with somewhat unsettling phraseology such as "negative cell voltage" and "copper bridges".
(Disclaimer: I've had an escooter battery cause an apartment fire.)
Optimistically, the charge controller should detect this state and refuse to discharge once even one out of n cells is dangerously low. Realistically, they just look at total pack voltage.
I guess that's a problem with cost and marketing. So having the battery management system able to detect a bad cell and generate an alert would cost more than just looking at the pack voltage. Deep-discharge risks could be avoided by making a hidden reserve. So just report the battery is flat when it hits 20% and call it good. But that wouldn't allow manufacturers to claim as high ranges. Tesla may have done this with their 80KWh models being 100KWh packs de-rated in software. No idea whether that was done to create a safety buffer, or just a commercial decision to just make 100KWh packs and create a cheaper Tesla just to meet a marketing price point.
I guess it's something that could be covered by safety standards and requirements, but all too often kit ends up being imported that just fakes compliance anyway.
I have read a few papers when I was curious about the dendrite issue. That seems like it's a mostly solved problem with newer battery formulations and anode/cathode materials. But I guess it's also a risk that old batteries may still end up being sold as fake newer generation versions. Something ISTR Bigclive has picked up on many times.
The batteries we're talking about here are Lithium Thionyl Chloride, they are Primary cells with zero electronics.
Most/all the comments above apply to Li-Ion rechargeable cells.
LiSOCl2 are a different construction, a different set of risks - and in many ways worse.
The energy density is higher, and they're fully charged of course, unlike Li-Ion in the factory (typ 20%).
SOCl2 is highly reactive, corrosive, poisonous - and reacts with water to produce mixed acids.
Nice stuff. It's what they use to burn-through the epoxy when de-capping packaged silicon chips.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
