FSRI spent three years burning 15 EVs and 3 gasoline vehicles to figure out what happens when an EV battery joins the fire. The answer is mostly reassuring, right up until the battery starts doing battery stuff.
In some ways, we’re building the modern automobile while we’re driving it.
Occasionally, we’re also crashing it into stuff at highway speeds, catching it on fire, and then asking the local fire department to please figure out what to do next.
Today, we’re visiting the intersection of science and Jobtown, the magical place where professional firefighters go to earn their money, and where the volunteers go to earn… memorable stories, I guess.
Editor’s Note: Kay was a volunteer FF/EMT-B, about 9,000 years ago.
Researchers Burned 18 Cars… for Science!
The Fire Safety Research Institute just dropped a banger of a report on EV fires after a three-year research project that involved deliberately setting 18 full-size vehicles on fire.
Fifteen of them were EVs. Three were conventional gas-powered cars used for comparison.
Nine vehicles were simply allowed to burn, including six EVs and the three gas cars. Nine more EVs were subjected to actual firefighting experiments using handlines, water additives, fire blankets and under-vehicle nozzles.
This wasn’t somebody lighting a battery module on fire in a laboratory sink. These were actual cars, burning at full scale, with researchers deliberately waiting until roughly the point when a North American fire department would arrive before beginning suppression.
Science is a beautiful thing, sometimes.
First, Some Surprisingly Good News
One of the biggest findings should probably annoy both EV evangelists and EV doomers.
The EVs did not generally burn hotter or longer than gas cars – not during the critical phase that concerns passenger safety, at least.
Overall fire size, duration and total energy release were similar. Much of what makes a car fire an enormous toxic dumpster fire is still the passenger compartment, because cars are full of plastics, upholstery, wiring, adhesives and all of the other things you would never voluntarily put into a campfire.
And ordinary firefighting equipment still works remarkably well on the passenger compartment, where the critical life-and-death interventions happen.
FSRI found that firefighters could knock down the passenger compartment fire in under a minute using less than 200 gallons of water.
That’s great! They don’t mention it in the report, but those 60 seconds are generally VERY fun.
Unfortunately, even if the passenger cabin is extinguished, there is a giant electrochemical spaghetti-mess under the EV’s floorboards that may have its own ideas.
The Battery Has Entered the Chat
Once thermal runaway gets established inside the battery, repeatedly spraying the outside of the battery pack with water does surprisingly little.
The enclosure is literally designed to keep stuff out. This is normally desirable in something bolted underneath an automobile traveling through rain, snow, salt and the occasional lake.
That admirably tight seal becomes considerably less convenient when the stuff you’re trying to put inside the battery pack is several hundred gallons of fire-extinguishing water.
FSRI found that continuous water application to the outside of a burning battery could be ineffective and wasteful. A more successful tactic was to extinguish the ordinary vehicle fire, protect exposures and periodically keep the passenger compartment from reigniting while letting the battery finish doing whatever terrible chemistry experiment it had already committed itself to.
Ah yes, it’s starting to smell like Jobtown in here. What’s cooking? About a thousand pounds of lithium-ion batteries, that’s what.
The Smoke Is Also Trying to Kill You, Obviously
Here’s another fun characteristic of lithium-ion thermal runaway: the stuff coming out of the battery can be both chemically unpleasant to breathe and explosively flammable… because variety is the spice of life!
FSRI found many of the same nasty organic compounds in both EV and conventional car fires. EV fires also produced elevated levels of battery-associated metals including nickel, manganese and cobalt, along with elevated particulate fluoride.
My community-college-level chemistry knowledge is good enough to tell you none of that belongs in your lungs.
At the same time, the battery can vent flammable gases even after visible flames have been suppressed. NEAT!
The correct approach, really the only safe approach, is full turnout gear. Mask on, breathing from your SCBA tank. Approach from upwind when possible. Use the reach of the hose stream, from a 45 degree angle in case the flaming car shoots projectiles back. Those can be tires/struts, or flame jets in the case of EVs.
Turns out, the fundamentals they taught us on the fireground still work.
EV Battery Fire That Looks Totally Put Out: “I didn’t hear no bell.”
And then, there’s reignition.
Under FSRI’s formal definition, two partially burned EVs experienced reignition, out of the experimental pool of 9 EVs. In other words, about 22% of the EV fires appeared extinguished, but really they were just getting ready for a late-game rally.
One Tesla experiment reignited four times while firefighters were actively working on the battery, then experienced several additional reignitions after the experiment ended and the vehicle was moved.
In another test, researchers left a fire blanket on a Ford EV for about four hours. Roughly 30 minutes after removing it, partially burned material inside the car caught fire again.
The battery was still around 350°C, four and a half hours later.
In firefighter temperature units, this is in the general neighborhood of “don’t touch that, volly.”
The report goes so far as to recommend treating every fire-damaged EV as a reignition risk, considering a suppression unit escort for the tow truck, and storing the thing at least 16 feet from anything combustible once it reaches the tow yard.
Jobtown apparently has suburbs, and I tell you what, they absolutely effing suck. Miserable neighborhood, to be honest.
Fire Blankets Sometimes Help, But Can Cause Unpredictable Blowouts
Fire blankets are great, in theory. Take an enormous fireproof blanket, throw it over the burning automobile, starve the fire of oxygen. Quick and easy.
They actually can work very well at controlling the passenger-compartment fire.
The problem is that thermal runaway does not particularly care whether you can still see it.
In FSRI’s tests, battery temperatures remained high and thermal runaway continued underneath the blanket, producing large quantities of smoke and venting gas. That gas was observed to be flammable in every blanket experiment.
So you have successfully converted your highly visible automobile fire into a quieter automobile fire containing a growing collection of flammable gases underneath a fireproof tarp.
They’re Not Paying the Fire Blanket Tester Enough
In one experiment, researchers lifted the blanket a second time to take a peek. Researchers believe flammable products from the battery and incomplete combustion had accumulated beneath the blanket at concentrations too rich to burn. Lifting it introduced fresh oxygen, moved that mixture into the flammable range, and allowed a hot surface somewhere underneath to provide the final ingredient. Kaboom!
In another experiment, when the blanket was lifted, accumulated smoke and vapors ignited and produced flames approximately 20 feet high. Researchers then activated an under-vehicle nozzle, which produced a large fireball out the driver’s side of the car. Which, generally speaking, we prefer our firefighting methods to not “produce large fireballs,” that is a surefire way to agitate the Incident Commander.
Later came additional deflagration and reignition events, including some that occurred without anyone touching the blanket at all.
There were no consistent warning signs before these events.
Fire Safety Engineers Would Also Like More Instructions, Please
FSRI’s study lands particularly well alongside another new paper surveying 149 fire-safety engineers and related professionals about battery fires.
The researchers found an industry still wrestling with missing standards, incomplete or inaccessible datasets, limited full-scale testing and major uncertainty around thermal runaway, gas generation and fire spread.
Among the 86 respondents who had actually worked on battery explosion prevention or mitigation, 88.2% said they had encountered challenges like a lack of reliable test data and predictive modeling tools.
The broader group was almost exactly divided over whether the available battery-fire data was reliable. Which, uh, that’s not great. Remember how I said we’re building the car as we’re driving it and maybe lighting it on fire? The fire-safety engineers are just about evenly split on the question, ‘do we actually have reliable enough data about what these things do when they catch fire?’
Plus, the paper identifies missing standards and relevant fire-test data among the biggest practical challenges facing the profession.
Which brings us neatly back to those 18 burning cars.
We Are Building the Automobile While We’re Driving It
This new research does not tell us EVs are rolling incendiary bombs. Once they’re burning, FSRI found EV’s overall fire size and duration looked remarkably similar to gasoline cars.
In fact, FSRI explicitly found that EV and conventional car fires are much more similar in overall heat release and duration than a lot of dramatic coverage would suggest.
What EVs add is a different set of problems once the battery becomes involved: difficult-to-reach thermal runaway, prolonged off-gassing, flammable vapor, unpredictable deflagration hazards, reignition and a damaged battery whose internal condition may be nearly impossible for the people standing beside it to determine.
Right now, a lot of that burden lands on fire departments. And with this newly published science, the next generation of EV firefighter training can be built around a little less folklore and a lot more actual evidence.
