Fact-checking the fact-checkers

From Climate Scepticism

BY MARK HODGSON

The Guardian has taken it upon itself to explore “the myths and realities surrounding EVs”. The heading to the article asks “Do electric cars pose a greater fire risk than petrol or diesel vehicles?” and I think we know the answer the Guardian will come up with. They start by telling us that the Luton Airport car park fire resulted in lots of speculation that the fire might have started in an electric vehicle. (Indeed, we at Cliscep speculated about it too, though we did – and continue to – keep open minds on the question). However, we are told that this “theory was quickly doused by the Bedfordshire fire service, which said the blaze appeared to have started in a diesel car.” And so we see how this is going to pan out. The claim has been doused, but has it? The support for that claim is the fire service told us that the fire appeared to have started in a diesel car. So far as I am aware, nobody has yet definitively established that this is the case, even though the blaze occurred almost six weeks ago. The most recent discussion on the internet seems to be an article in the Daily Sceptic, which makes the following observations:

Was the Luton airport fire caused by an electric vehicle? The official line is that it was a diesel-powered vehicle. However, a video of the fire in its early stages has been noted by many on social media to appear to show a Range Rover Evoque ablaze on the passenger side at the front.

The Range Rover Evoque has mild hybrid EV (MHEV) models which have been recalled in at least one country because of a fault that caused the lithium ion battery located under the passenger seat to short and catch fire.

The Telegraph‘s Allison Pearson notes there “was none of the thick black smoke you would expect with a diesel fire”.

It seems odd that no further details – such as confirmation of the make and model of the car – have yet been put into the public domain. Is that because they would confirm it was a hybrid and detract from the official ‘diesel vehicle’ line?

It may be that it will ultimately be shown to have started in a diesel (or hybrid) car, but as of the time of writing, that isn’t the case. There are no hard facts yest in the public domain. We simply don’t yet know for sure one way or the other. So at this early stage, the lie is given to the Guardian’s claim that it “has spoken to experts and looked for hard data where possible to address some of the most common criticisms of electric vehicles…”. It might have looked for hard data, but it relies on a statement that something appears to be the case to rubbish a negative story about EVs.

Still, the issues to be considered are fairly set out:

The claims about electric car fires fall into two broad categories. The first is that fires are more common in electric cars, while the second is that when fires break out, they are more damaging.

If electric cars do pose more of a fire risk than petrol or diesel, that would have a host of consequences. One could be a requirement for larger car park spaces to stop fires spreading, while the Conservative MP Greg Smith, who serves on the transport committee, said in July that EV owners should pay higher insurance premiums to cover the extra costs to firefighters.

And so it’s obviously crucial to establish that EVs are less prone to spontaneous combustion than the internal combustion engine (ICE) alternative. In order to achieve this, we are treated to a claim by Colin Walker, the head of transport at the Energy and Climate Intelligence Unit think tank, to the effect that EVs are much, much less likely to set on fire than their petrol equivalent. He tells us (without evidence – or at least none that is cited in the article) that many, many fires in petrol and diesel cars “just aren’t reported”. If they aren’t reported, how does he know about them? How does he know how many such fires there are? By definition, if his claim is correct, we lack hard data. The Energy and Climate Intelligence Unit, by the way, isn’t simply a “think tank”, as the Guardian would have it. Rather, it’s a climate alarmist campaigning organisation, as its “Who we are” web page makes clear, with funding being provided by, inter alia, the European Climate Foundation, the Quadrature Climate Foundation, and (previously) the Grantham Foundation for the Protection of the Environment, the Climate Change Collaboration, the Oak Foundation and the Tellus Mater Foundation. Mr Walker is no doubt sincere, his claim may be correct, but the claims made on his behalf by the Guardian provide absolutely no hard data to back up the assertion that ICE vehicles are far more likely to burst into flames than are EVs.

The next paragraph rather seems to make the case against Evs, rather than to offer a defence:

Fires can start in several ways. Car batteries store energy by moving lithium ions inside a battery cell but if cells are penetrated or if impurities from manufacturing errors cause short-circuits, then unwanted chemical reactions can start “thermal runaway”, where cells heat up rapidly, releasing toxic and flammable gas. In petrol cars, fires can start via electrical faults causing sparks or if the engine overheats because of a fault in the cooling systems, potentially igniting flammable fuel.

So a battery fire in an EV can start “thermal runaway” “releasing toxic and flammable gas” (sounds pretty alarming) while an electrical fault in a petrol car might “potentially” ignite flammable fuel.

We turn to Norway for the next round in the defence of EVs. As is well known, Norway is a country where new EV sales represent a very high proportion of all new car sales. And so the Guardian assertion that “[i]n Norway, which has the world’s highest proportion of electric car sales, there are between four and five times more fires in petrol and diesel cars, according to the directorate for social security and emergency preparedness” seems pretty compelling. At least it does at first sight. It’s difficult to find detailed information from the internet (so intent are search engines on pushing pre-EV information at searchers) as to what proportion of cars in use are EVs and what proportion are ICEs. Statista offers some useful information , however. They suggest that in Norway in 2022 there were 1,135,538 diesel cars on the road; 822,133 petrol cars; 599,169 EVs; 348,969 “other” and just over 200 running on gas or paraffin. At face value, then, the Guardian is on to something (assuming their statistics are correct – they provide no link to assist curious readers in checking it out), since the number of ICE fires are said to exceed EV fires by four or five to one, while the number of ICE vehicles on the road exceeds EVs by just three or four to one. However, even then, the comparison isn’t direct, since most EVs are likely to be new or relatively new, while most ICE vehicles are likely to be relatively or very old. Old cars are more likely to develop faults (including faults leading to fires) than new ones are, so unless and until we have data covering older EVs, these statistics are interesting, but far from conclusive. At this stage, we are simply not comparing like with like. Incidentally, I have searched the website of the Norwegian Directorate for social security and emergency preparedness without success for the report to which the Guardian refers. It occurs to me that if the report dates from 2021 or earlier, then ICE cars would then have outnumbered EVs on Norway’s roads by four or five to one (2021), by between six and seven to one (2020), by around nine to one (2019) or by as much as twelve to one (2018), in which case, far from supporting the Guardian’s claims, the report would actually undermine them. But the Guardian doesn’t provide a link to the report, doesn’t tell us its date (crucial information), and I can’t find it. So much for hard data.

Next we are treated to a statistic from Sweden, which even the Guardian more or less acknowledges doesn’t count for much:

The Swedish Civil Contingencies Agency this year found that there were 3.8 fires per 100,000 electric or hybrid cars in 2022, compared with 68 fires per 100,000 cars when taking all fuel types into account. However, the latter figures include arson, making comparisons tricky.

I would go further. If the former number excludes arson and the latter number includes it (as appears to be the case), then citing the numbers is simply pointless. They establish nothing.

Next up are some numbers from Australia (the Guardian has apparently searched the world in its quest to prove that EVs are safe):

Australia’s Department of Defence funded EV FireSafe to look into the question. It found there was a 0.0012% chance of a passenger electric vehicle battery catching fire, compared with a 0.1% chance for internal combustion engine cars.

I followed the link since one was provided in this case. While the findings are broadly in line with the claims made by the Guardian, they are considerably more tentative than the Guardian’s phraseology suggests. Here is what EV FireSafe’s website actually says:

Our intial [sic] research findings, based on global EV battery fires from 2010-2020, indicate a 0.0012% of a passenger electric vehicle battery catching fire.

While it’s difficult to find a similar stat for internal combustion engine (ICE) passenger vehicles globally, a range of country-based reports we found suggest there is a 0.1% chance of an ICE vehicle catching fire.

That’s potentially useful indicative information, but it hardly represents compelling hard data, and doesn’t, so far as I can see, begin to justify the weight put on it by the Guardian. Furthermore, the website in question also stresses the problems with thermal runaway when EV batteries ignite; that an EV lithium traction battery burns hotter than an ICE vehicle; that EV fire behaviour is different and presents new challenges; that it’s not smoke, it’s a vapour cloud of highly flammable gases; that EV traction battery fires may require more resources; that The location of an EV battery makes fire harder to extinguish; that best practice is to allow a traction battery to burn out; and that EV traction battery fires can reignite, hours or days later.

Finally, the case for the defence of EVs concludes by telling us that Tesla “says the number of fires on US roads involving Teslas from 2012 to 2021 was 11 times lower per mile than the figure for all cars,, the vast majority of which have petrol or diesel engines.”

Again, a link was provided, so I followed it. And while the claim made in the Guardian is broadly correct, it does misrepresent the facts in one possibly important respect – it wrongly claims to be comparing Tesla car fires with ICE car fires per mile travelled. In fact, Tesla compares Tesla car fires with “data from the National Fire Protection Association (NFPA) and U.S. Department of Transportation shows that in the United States there is a vehicle fire for every 19 million miles traveled.” That’s vehicles, not cars, so presumably includes buses, lorries and all other vehicles using the roads. It also, by definition, presumably includes EV fires, if it’s a database of all vehicle fires, so that will skew the figures somewhat, and will reduce the ratio from 1:11 to a smaller number, though admittedly perhaps not much smaller. The inclusion of vehciles other than cars in the federal data may or may not be relevant – I simply don’t know. Also of potential relevance is how the fires occurred. Tesla make the point that:

In order to provide an apt comparison to NFPA data, Tesla’s data set includes instances of vehicle fires caused by structure fires, arson, and other things unrelated to the vehicle, which account for some of the Tesla vehicle fires over this time period.

I concede that that is a pretty compelling statement. But we don’t know how many of the ICE vehicle fires were the result of, say, multi-vehicle crashes, which might have led to fires. We also don’t know how serious the ICE vehicle fires were compared to the Tesla fires. When a Tesla burns, it burns in a big way.

Then there is the fact that not every EV on the road is a Tesla, and increasingly it seems we can expect to see more and more Chinese EVs on the roads. Are their safety standards as high as Tesla’s? I don’t know. The Guardian doesn’t follow up by considering these questions.

At the end of the Guardian’s case for the defence, the reality is that we have been presented with very little meaningful hard data despite their claim to have looked for it. Perhaps there isn’t much.

In fairness to the Guardian, they do also offer some of the case for the prosecution with a section of the article headed “Any caveats?”. Here they acknowledge that EV fires “can be furious infernos” and acknowledge “the ominous risks of “vapour cloud explosions and rocket flames” when the gases burst out of cells.” And there is a belated recognition of some of the negative findings of EV FireSafe, namely “battery fires require more water to put out, can burn almost three times hotter, and are more likely to reignite.”

The article concludes with a verdict, which was rather more balanced than I expected, albeit it’s much kinder to EVs than my own verdict:

Despite the increased danger once a battery fire is burning, the probability of being caught in an EV fire appears overall to be much lower than for petrol or diesel cars, based on currently available data – although this could change as more people get electric cars.

Walker said it was possible that the prevalence of fires in EVs could increase as the average age of batteries on the roads increases. However, at this point, it appears that they would have to multiply by many times to be worse than internal combustion engines.

I am pleased to see the acknowledgement that as EVs age, the data could well become much more negative with regard to EV fires. However, the claim that ICE fire numbers are worse (by many times) than ICE fire numbers simply isn’t borne out by the data provided in the article.