Electric vehicles, emissions and thermal runaway

Complex problems may not have one-size-fits-all solutions and even Very Good Things can have downsides and risks

Value for money is multi-dimensional. Decision-makers and consumers invest in value propositions (ideas about some preferred future) for many reasons beyond satisfying self-interested preferences, two examples being the intent of creating value for natural ecosystems and future generations. Value propositions focus on the benefits of investing in a product or service. However, it’s also important to consider whether an investment has side-effects, downsides or risks, and how it compares to its next-best alternatives.

Today’s example: electric vehicles (EVs). For the few percent of humanity wealthy enough to afford them, electric cars are a tangible and visible way to reduce one’s transport-related carbon dioxide (CO₂ ) and particulate emissions. Not only that, they’re low maintenance, quiet, powerful, awesome to drive and you can refuel them at home while you sleep. All while attracting the admiration and envy of your friends and neighbours. Have your cake and eat it with this emblem of growth and de-growth, complete with vegan leather and autopilot. I foresee a profitable future for consumer goods that can pull off all these virtues simultaneously.

Electric vehicles and carbon dioxide emissions

There are of course many factors involved in choosing a car. But let’s take a moment to consider one benefit that’s almost synonymous with electric cars: reduced carbon dioxide emissions. EVs are often labelled zero-emission vehicles, but there’s no such thing. Even my bicycle added CO₂ to the atmosphere as a by-product of its manufacture.

Manufacturing batteries generates a lot of CO₂ - so much, in fact, that if you compare two new cars that are similar aside from their means of propulsion (say, electric and petrol versions of Hyundai’s Kona), the electric version starts off life having released significantly more CO₂ than the petrol one (12.9 vs 5.5 tonnes, according to this estimate). In other words, there’s a fixed ‘cost’ in CO₂ emissions for every vehicle manufactured, and the electric Kona starts out on the back foot before the two vehicles are driven anywhere.

Once you start driving them, it’s a different story. On the road, the battery electric car can produce lower CO₂ emissions per km driven than the liquid hydrocarbon version: its variable (per km) carbon costs can be lower. Potentially, variable CO₂ could even be zero if you recharge your EV entirely from renewable sources. However, it is also true that some EVs are 70% coal-powered (see the graph below). Two identical EVs can have very different carbon footprints, depending what power sources they charge from.

Australian electricity generation by fuel mix, 2021. Source: energy.gov.au

Assuming we’re plugging into a green enough grid, once our electric and petrol cars have driven far enough, their cumulative CO₂ emissions curves will cross and the lifetime emissions of the electric version should ultimately be lower. How long would it take you to break even on CO₂ emissions in an EV, compared to an equivalent sized car running on fossil fuel? It depends on contextual factors like which vehicles you’re comparing, how much driving you do and what fuel mix is used in your location to generate electricity.

In Aotearoa New Zealand, more than 80% of our electricity comes from renewables. I compared electric and petrol Konas and estimated that in NZ, the curves would cross at around 50,000km (about 8-10 years of driving for me).¹ This is based on the small battery (39 kWh) model; the long-range (64 kWh) one would take longer to break even because its fixed carbon emissions are higher.

Break-even analysis on carbon dioxide emissions for petrol (2.0 Series II) and electric (39 kWh Series II) Hyundai Kona based on Hyundai NZ claimed efficiency statistics, average electricity emissions data for NZ 2016-2020 (statista.com), and manufacturing-related emissions from this source, which I have accepted at face value for the purpose of these approximations.

For Australia’s three most populous states (NSW, Vic & Qld), I estimated that the curves for our 2-litre petrol and small-battery electric Konas would cross after 200,000km (assuming both cars last that long).

My numbers are approximations. You’ll find studies online that take a more detailed and rigorous approach than my simple y=mx+c analysis, taking more factors into account such as CO₂ emissions from end-of-life recycling and disposal. I’ve seen a range of break-even estimates from 20,000km to 160,000km and this study suggesting that in Australia, EVs have higher lifetime emissions than petrol cars (i.e., the curves would never cross).

Despite the wide variation, I am not questioning (nor have I checked) their veracity. One study doesn’t have to be wrong in order for another to be right, because they’re based on location and vehicle-specific scenarios. You’d be lucky to find analysis based on a scenario that applies to your individual circumstances.

You could spreadsheet a personalised estimate for your locality and vehicle preferences. If you do, remember to factor in how often you intend to replace your car. As Rowan Atkinson pointed out, many of those privileged enough to afford an EV also upgrade to a new car every few years, and their personal CO₂ footprints may be more heavily impacted by the fixed carbon outputs of car manufacturing than the variable carbon outputs of car usage. If you care about minimising transport-related emissions, keep your car longer and drive it less.

It’s also worth comparing hybrid options with full battery EVs. Kelly Senecal and colleagues published a comprehensive analysis in the US, modelling and comparing scenarios for different states. They concluded that in some states currently, hybrid cars are expected to have lower lifetime CO₂ emissions than full electric vehicles, because hybrids have smaller batteries with lower fixed carbon footprints, and the fuel mix used to generate electricity in many states remains quite carbon-intensive. As Senecal says, “the future is eclectic”. The dynamics will evolve as more renewables come on line.

Source: Burton et al. (2023).

Digital watches don’t always tell the right time, and EVs are not zero emissions. Sometimes they’re lower emissions than other options. Sometimes they’re not. It depends. As my favourite YouTuber says, “they’re called facts, liking them is optional”.

Speaking of whom. What caught my attention this week was a couple of videos from Australia’s Auto Expert, John Cadogan. With a degree in engineering and long experience as a freelance automotive journalist, he’s as well-informed as he is entertaining. He does great work, in his inimitably independent, forthright and not especially politically correct way. Just so you know what to expect.

And some of his recent videos gave me a long overdue education on catastrophic thermal runaway battery fires - and just how catastrophic they can be under the wrong circumstances. Let’s take a look at that now.

Lithium-ion battery thermal runaway

A few salient facts.

• Lithium-ion batteries (like the ones in your phone, e-scooter and electric car) can start fires and explode (remember the Samsung Galaxy Note 7 fires?)

• This doesn’t happen very often - but it does happen, and when it does you really don’t want to be anywhere nearby. This isn’t about whether EVs are more likely to catch fire than internal combustion engine cars (they’re not) but about the severity of the consequences when they do. EV fires are more toxic and, once a lithium-ion battery goes into thermal runaway, it can’t be extinguished.

• Inside your battery are lots of little energy storage cells (that’s literally what a battery is). If any one of those cells is compromised, it can ignite. And when that happens, the ones next to it ignite, and then the ones next to those, exponentially. And if there’s another EV parked alongside, that will ignite too. And so on, and so on, and so on. Just like the shampoo ad from the 1970s, only more dangerous.

• These thermal runaway fires produce their own oxygen as an inherent part of the chemical reaction, meaning they are self-sustaining, self-accelerating, and can’t be smothered. You simply have to wait for them to run out of fuel - and that can take a long time.

• It’s the cobalt in Lithium-ion batteries that is particularly toxic and unstable (as if unethical cobalt mining practices weren’t problematic enough already).

• Lithium-ion battery fires are profoundly toxic - releasing nasties like hydrogen cyanide, hydrogen fluoride and cobalt. The latter goes straight through protective clothing and is absorbed through the skin, causing serious harm including brain, lung and heart damage. In Victoria, two firefighters have been permanently disabled from cobalt poisoning.

Yikes!!

I can count at least 30 lithium-ion batteries in my house and garage, in computers, phones, mophies, e-bikes, scooters, gardening tools, security cameras, a vacuum cleaner that works while we sleep then plugs itself in to recharge and, ironically, smoke alarms. Nothing as big as an electric car battery, but still big enough to start a ferocious unstoppable fire.

EV Fire severity

In the following video, Cadogan unpacks some lessons we really need to learn from a recent disaster involving the Fremantle Highway, a car-carrying ship that caught fire off the Dutch coast on 25 July this year causing one death and severe injuries to crew members. It was transporting nearly 3,000 cars, including 500 EVs. The cause of the fire on this giant floating multi-storey car park was officially yet to be determined at the time of recording but, whatever the cause, the only reason it kept burning for days, and could not be put out, was the presence of EV batteries. And it’s not the first time this has happened.

A large EV battery has over 70,000 individual cells. It takes just one cell in one car to malfunction and ignite, and every battery and every car on board is “a done deal”. Cadogan estimates that the EVs on board this ship collectively had around 200,000kg of batteries, closely packed, making the Fremantle Highway an extreme example.

Nonetheless, this is a real-life demonstration of a low-frequency, high severity event that can happen in a built up area. The risks apply to all of us, not just those who drive EVs, because we could be nearby when it happens.

What if a row of EVs, parked together in a charging zone, were to ignite in a covered car park near you? What if that car park was in a shopping mall, or underneath an office building? What if that car park vents to the street, with pedestrians walking by? When this kind of thermal runaway event happens in a city centre, Cadogan says, there is no plan to deal with it. At least not in Australia.

The video also includes footage of EV thermal runaway fires. Because sometimes you have to see things to fully appreciate them.

This👆should be an autoexpert.com.au video but in some people’s feeds it is showing as a repeat of the Waynes World vid. Here’s a link to the intended video.

In another video, Cadogan argues that this shouldn’t really be called a fire. It doesn’t behave like the things we think of as fires. It’s an “unstoppable hellscape chemical reaction”. If this sounds melodramatic, at least the melodrama is proportionate to the level of danger in an established multi-EV fire. It’s impossible to extinguish. You can’t fight it “and the Dutch coastguard hasn’t been fighting it - they’ve been waiting for it to go out because that’s all you can do under these circumstances”.

The number of EVs in circulation is (I presume) only going to increase, and that’s a good thing in many ways. But even good things have downsides and dangers, and this one is worth being prepared for.“Appropriate pre-emptive safeguards are needed”.

The United Firefighter’s Union Australia has urged the federal government to provide better regulation and public education for battery fires, including vehicles and battery energy storage systems that are used in solar-powered buildings.

Public health and safety messages

A bit like a plane crash, maybe it’ll never happen to you, but they do happen (I survived my first and don’t wish to have another). The consequences of battery fires are sufficiently severe that it’s worth being prepared and doing what you can to prevent a catastrophe.

If you only remember one thing from this post, make it the following:

If you encounter a thermal runaway EV fire, get upwind immediately - i.e., away from the smoke and as far away as possible. Teach your loved ones so they know what to do. It should be as automatic as knowing what to do if a mask drops from the overhead panel in flight.

If you charge your EV at home, don’t do it in the garage. Charge it outside, on the apron, well away from buildings and other vehicles. Similarly, don’t charge bikes or scooters inside.

Don’t charge phones etc overnight and don’t leave them charging for prolonged periods of time. Don’t charge them on combustible surfaces like beds, clothing or paper. Or in direct sunlight. Or in parked vehicles. Do not charge them unattended.

If your portable device starts to get very hot or starts smoking, switch it off. If you get to it soon enough, you can prevent a thermal runaway situation. Move it away from anything that could catch fire, and away from exit doors.

If your battery is already on fire, keep away from it and call the fire department.

More safety info here and here.

TVNZ’s Seven Sharp program recently screened a controlled house burn live on TV, demonstrating “the dangers of incorrectly charging lithium-ion battery products, how fast fire can spread and the importance of escape plans and interconnected smoke alarms”. A scooter charging in one room caught alight, and within minutes the whole house was ablaze. Fire and Emergency NZ has posted a link to the story here.

Slightly off-topic for Evaluation and Value for Money, but worth the resources if it saves a life. Happy to be corrected on any errors or omissions. I’m not anti-EV btw, I have several and I love them, though none of them are cars. When it comes to traffic congestion, all cars are the same.

1. I also ran a financial break-even analysis and found the cumulative cost curves cross at around 120,000km, based on NZ prices for short-range electric and petrol Konas (net of clean car rebates), fuel and electricity. EVs can save carbon and money, if you keep them long enough.