An interesting rule of thumb seems to be that a lithium based battery contains about the same additional potential energy available in a thermal runaway as is contained in its rated capacity.
Which places some bounds on what is needed to control one. 100kWh is enough energy to boil about 160kg of water. So double that if you have a fully charged 100kWh battery. 300 litres of water.
But clearly this is a lower bound and difficult to achieve at best. Maybe there is the germ of a chance if a battery pack was designed from the outset to accept a water feed in a safe manner.
The car is going to be toast no matter what. But a controlled release of energy might be possible in a useful way.
Doing so might bridge the gap from the method linked to above that requires mechanical breaching of the battery container. One needs to control the output as well, which is going to include a hydrogen fire initially, and spew steam once the fire is out. So that is going to be an interesting challenge.
The other fun thing is that the result of burning hydrogen is water. Which the lithium can then react to, generating hydrogen again. Which burns, creating water. Which then… well, you get the idea.
Lithium is quite greedy. Atmospheric oxygen might get the process going but the resulting fire generates a lot of heat/energy.
The generation of hydrogen fluoride in a lithium battery fire is another fun side effect I’d rather not have to deal with.
This. If you have a large solid thing that produces heat throughout its internal volume, that heat has a difficult time getting to the exterior surface by itself. You need to circulate coolant through it and then run that coolant through a heat exchanger to get rid of the heat, same as an elephant does with its giant ears. And that’s pretty much what is done with EV battery packs. Here are pics and schematics of the cooling system inside Tesla’s battery packs.
I was curious as to the actual chemistry of a thermal runaway, and had a bit of a search. There is surprising little real data. But I found the following paper which provides some interesting insights. Perhaps the most interesting takeaway is that hydrogen is only a trace output, the most flammable are ethylene and the evaporated organic solvents from the cell. This perhaps helps explain the ferocious orange flame one sees - hydrogen does not burn like that. Lots of carbon dioxide as well. The lithium metal in the battery is swiftly oxidized out of contention. It does play a small part in evolving hydrogen, but that is a minor player in the runaway. The tl;dr is that most of the popular accounts of what goes on are wrong.
I messed around with building high voltage battery packs utilizing multiple flashlight cells in various series and parallel configurations for reproducing obsolete radio power supplies.
They will work fine for the application but it is important that every cell be of the same brand & even lot # to try ensure they are all roughly the same charge level, age, internal resistance, etc. A single cell that goes wonky in series or parallel can cause big trouble for the battery as a whole. I was surprised to find the rechargeable cars are essentially running on a metric crapton of flashlight cells. I wouldn’t park an EV in my garage, I’ll say that.
Nope. Are essentially running on a large number cells of about the same case size as flashlight cells, but with nearly zero of the same internals. The concerns you raise are real. And they’re engineered for. Not perfectly, as nothing is perfect.
It’s unclear to me that gasoline cars in garages are any safer than EVs in garages.
But the usual reaction of water with lithium is for the lithium to combine with water and make lithium hydroxide, while liberating hydrogen and heat, so if you throw lithium into water it will float (lithium is lighter than water), generate hydrogen gas and heat, which will ignite in the oxygen atmosphere.
I don’t know of a lithium-water reaction that liberates free oxygen. I would’ve thought that lithium under water would not burn because there was no oxygen. It sure as hell would react, though. So either (as suggested above) the description of “burn” isn’t quite correct, and you’re getting a reaction that gives off hydrogen gas (which could catch fire when it breaks free of the water, if there’s a heat source), or else there’s some lithium-water reaction that does generate oxygen that I’m unaware of.
Incidentally, reactivity goes up with atomic number. Sodium is a LOT more reactive when you toss it into water than lithium, and so is potassium. Cesium and rubidium ought to be more impressive. I’ve worked with metallic lithium, sodium, and potassium, and had to put them in water for my work (don’t ask), and speak from experience.
By the way, if you want to get a really slow reaction from the alkali metal and liquid – as when you wan to clean the surface of it – use a heavy alcohol, like pentyl alcohol, instead of water (water is arguably the lightest alcohol). The heavier the alcohol, the slower the reaction. In pentyl alcohol, sodium merely fizzes, like alka-seltzer in water, rather than skipping around and prictically exploding.
Lithium burns in water in the sense that there is a reaction taking oxygen and releasing energy. Water is the oxidant, because lithium binds more tightly to oxygen than hydrogen does. That is, the oxygen in water is “free” from the point of view of the lithium.
IANA expert, but there have been cites in other EV-related threads that on a per-vehicle basis, more gasoline cars burn in garages than EVs.
Society is learning at its own pace about this novel hazard. What’s unclear to me is whether you or I are taking the right lesson from that growing experience. The “good” news about this novel hazard is that most novel hazards are incremental additions to the total. This one is not.
e.g. …
40 years ago nobody had to worry abut their purse or pocket bursting into flame since nobody was carrying a mongo potentially incendiary battery with them every waking hour of every day. Now everyone has that worry (or at least that hazard exposure whether they’re aware of it or not). That hazard is both novel and purely incremental.
But every EV that goes into a garage displaces an ICE that used to sit in the same garage. The hazard nature is novel, but the number of hazards isn’t increasing. And as said above, the jury is still out on whether the hazardousness of any single instance is increasing or decreasing as we exchange ICEs for EVs.
“EV fire burns down garage” is highly newsworthy. ICE doing the same is not. Beware reporting bias.
Gasoline and diesel vehicles can certainly catch fire but rarely do so when they are just sitting unused (although there was a recall on Fords a few years ago because the ignition system would spontaneously catch fire); most car fires are due to fluids leaking from the head gasket or oil system dripping on the engine or exhaust manifold. BEV fires, on the other hand, can definitely occur during charging. A number of commercial insurance companies require additional riders or just don’t allow charging stations in underground garages for that reason.
Although to the extent that they are reactive (and they are, a little), it does increase with increasing atomic number. Xenon hexafluoride, etc. The outer shell electrons aren’t as tightly bound with high atomic number as low.
Stranger already covered this, but I’ll just add that I’m not convinced that this is generally true, and even if it is, I would argue that it frequently involves external causes that can be controlled for – like someone being careless with a blowtorch – rather than a spontaneous function of the vehicle itself.
Conceptually, an ICE car in a garage is a passive, static thing that is very unlikely to do anything spontaneously, and in the very rare circumstance when it does, the trigger is almost always electrical. Whereas an EV sitting in a garage not only has a lot more points of electrical vulnerability, it’s also anything but passive; it will spend a good deal of its time charging from a high-voltage connection, and all that energy is actively pouring into around a thousand pounds of potentially dangerous material.
That said, I wouldn’t hesitate to store and charge an EV in my garage, provided only that it had a reasonable track record of safety.
Quite true, and I’m curious how this works with EVs using very large numbers of cylindrical cells, since even if they’re manufactured to high quality standards, they’re not all going to deteriorate in exactly the same way.
The UPS I use with my computer runs on a 24-volt system powered by two 12-volt batteries in series. If one of those batteries gets significantly weaker than the other, I don’t just lose capacity – the whole battery assembly is flagged as failed. Batteries connected in parallel are probably less sensitive, but EVs need to have both serial and parallel connections in their cells.
The vehicle’s anti-lock brake system module could leak brake fluid internally and cause an electrical short. An electrical short could result in significant overcurrent in the ABS module, increasing the risk of an engine compartment fire while driving or parked.
And yes, it was electrical in nature, but that just drives home the point–ICE cars are electrical, too. They may have smaller batteries, but still have many points of failure and flammable fluids that can accelerate any fault.