Just how green is your fleet of Teslas, really?
Just how green is your fleet of Teslas, really?
A story of 10 cars
I’m going to lay my biases bare. The Wall Street Journal’s Dan Neil is one of the most respected auto journalists alive. He’s an absolutely top notch writer, with a penchant for flair. He engages the reader and transports them into the cockpit of the car he’s driving. Absolutely world class scribe, no questions or doubts here.
But his absolutely bizarre bias towards anything and everything Tesla is truly mind boggling. To Dan, anything the not-at-all-controversial company produces is perfect and beautiful, a technological and environmentally friendly marvel. While his appreciation of EVs as the best of both worlds (drive better than gas guzzlers! good for planet!) is perhaps questionable on many fronts, it’s at least a somewhat defensible position. But his fondness for EVs appears to be specific to a single brand alone. His review of the Jaguar I-Pace (one of the first luxury EVs aimed directly at Tesla’s customer base) was truly striking in its shallow, petty criticisms. Especially given the car was positively reviewed by every other major auto journalist and publication. This behavior has continued unabated by Mr. Neil for years. I don’t know what his angle is, but it’s truly bizarre to behold.
So when Dan’s latest for the WSJ came out, titled: Next to Tesla, Plug-In Hybrids Are an Illusion of Eco-Consciousness, I absolutely could not let this stand any more. Dan is absolutely wrong about PHEVs, which I will prove, empirically, below. The fact that he further compares PHEVs to not EVs as a whole, but specifically Teslas, is an abject embarrassment. The environmental benefits of one EV versus another, if both use roughly the same battery technology and drive train types (hint: they do) are identical. And to be frank, giant battery packs, for many vehicles are a waste if our goal is to minimize impact on the environment.
Here’s an example of Neil’s dishonesty:
In November the environmental pressure group Transport & Environment published a study of the emissions of the popular BMW X5, Mitsubishi Outlander, and Volvo XC60 plugins. The study observed that, even with a fully charged battery and in optimal conditions, the emissions of these vehicles were 28-89% higher than the official value.
Citing and nitpicking individual problems with single vehicles (especially given Tesla’s own issues with accurate mileage estimates), completely misses the forest for the trees. That climate change is a global problem and, if we’re not switching to public transport as fast as we can (and we’re not, sadly), resources should be dedicated to creating the maximum net benefit. So let me show you how to approach this problem holistically.
I got great feedback from my Work From Home post, so it's time for another model!
The 100 Kilowatt-hour Battery
Let’s say for modeling purposes we’ve got ~100 kilowatt Hours of automotive grade lithium ion cells/packs/pouches and we need to buy 10 cars for a fleet. These numbers were selected for even division since it’s hard to make fractional cars. My goal is to minimize Greenhouse Gas emissions from the fleet as a whole. This is, more or less, how tax refund/HOV lane/parking incentives are designed (or at least justified) on the government level. “We want to minimize global carbon impact and local pollution” is the stated mantra. Companies are graded on their fleets as a whole and climate change metrics are accomplished on a macro (not micro or per-unit) scale.
My fleet can be economy or luxury and the cars must be more or less fungible in features and performance. So here are the contenders:
Luxury Choices
2021 Tesla Model S - EV
2021 BMW M5 - ICE
2021 BMW 530e xDrive - PHEV
Economy Choices
2021 Tesla Model 3 (SR) - EV
2021 Toyota Corolla - ICE
2021 Toyota Prius Prime - PHEV
And here are the battery sizes for each of the Non-ICE vehicles:
Now, we only have limited supply of batteries to go around, so we have to construct our fleets in a way to only use ~100 kilowatt-hours of cells, so here they are:
So, when looking at these fleets, we shouldn’t look at blended MPGe (Miles Per Gallon equivalent) from the EPA ratings. Why? Well, different usage scenarios for PHEVs lead to drastically different emissions profiles.
Assuming all of these cars are charged in California (a very green power grid and an aspirational goal for other States and Nations to strive for), at 202 gCO2e/kWh, this is about as good as you’ll get for getting the most bang for your buck on EVs. The cars are also filled with CARB gasoline.
Instead of dividing total MPGe for the fleets, we’ll look at driving activity by day
A normal 25 mile commuter day, where the model PHEVs (BMW 530e and Toyota Prius Prime) will do most of their driving on pure battery
A 100 mile day, perhaps for a longer commute or for running errands around town or driving to the beach
A 500 mile road trip day, where the PHEVs will spend the majority of time in Hybrid Mode
Here are the daily CO2 emissions for these days, using manufacturer provided kWh/mile and MPG estimates.
As expected, the pure EVs do best on all three types of days. Great! Notably, however, on a normal “commute to work and then back home” kind of day, there’s not really much difference between the PHEVs and EVs. These days represent the bulk of American driving habits.
Next, we take these driving days and divide them into separate buckets:
Scenario 1: a commuter who drives to work 5 days a week, 25 miles per day. This person also does a 100 mile day twice a month and drives 6 days on road trips. At ~12k miles, per year this is a very standard American Driving profile per the IIHS.
Scenario 2: This person does more 100 mile driving days per year (80), with perhaps a longer commute but occasional work from home. This also includes 6 road trip days per year, totaling again ~12k miles per year
Scenario 3: This person has two cars for their family and only uses the car for commuter purposes, 5 days per week, 25 miles. Once a month, they use 100 miles in a day. No road trips. Annual mileage is ~7k
You’ll notice that the performance ICE car (BMW M5) really suffers across the board here, while the two plug in hybrids again do a great job when used primarily for commuting.
Putting it all together
Once we take the above data and look at emissions from a fleet level, the results turn common perceptions around. Remember that we have a limited supply of batteries (and this reflects real world supply limitations). Once the battery from a large EV pack is repurposed or divided into multiple PHEVs and applied to the standard American driving pattern a clear trend emerges:
I’ve long railed against the so-called green cred of the Tesla Model S, and I think the above chart is a good representation why. The Model S, when first released was an expensive car pitched to people who drove luxury vehicles. Most every early Tesla sale used an insane amount of battery cells to replace a single car for someone who would have bought a different status vehicle. One Model S battery pack, divided and put into 10 PHEVs, like the Prius Prime, effectively nullifies 35 *thousand* kilograms of CO2 emissions, per year! So you can imagine the frustration from those (like myself) who think about conservation from a macro standpoint seeing these giant cars, pitched to the upwardly wealthy, getting the largest tax breaks. It’s asinine.
jUsT mAkE MoaR BaTTeRieS!!!!!!!!
So when I talk about opportunity costs for deploying batteries into the global fleet, I often get some pushback. “More batteries, now!” they cry. There are some problems here. Rare earth mineral extraction, in addition to having a large initial carbon footprint and being a source for troubling toxic waste and water quality issues, requires lots of planning and infrastructure and time to get rolling. Its geographic footprint is small and even with the huge forward EV demand, many operations still require co-products such as cobalt and copper to make unit efficiencies work.
I’ve posted below an average estimate of (very optimistic) current and forward lithium ion cell projections out until 2030:
To give you a sense of scale, roughly 75 million cars are made globally every year. If we took all of the EV cell production from near term capacity (500 gigawatt*hr) and allocated it to car type, assuming the global market stays flat, we could make the following:
5 million Tesla Model S (6.6% of the global fleet) or,
9.3 million Tesla Model 3 (12.3% of the global fleet) or,
55 million Toyota Prius Prime (75% of the global fleet)
Now do you see the problem? When looking at *fleet* emissions each 54 kWhr (Model 3 pack size) of cells has an opportunity cost associated with it, that opportunity being: use subsidizes and incentives to move those batteries into small EVs or PHEVs instead.
When approaching it from this framework, on a fleet level, my Fleet #3 vs Fleet #4 makes sense as follows:
Every 100 kWhr of capacity shifted from Model 3s into Prius Primes saves 11 thousand kilograms of CO2 emissions per year. So, now take that assumption and put it into the global capacity chart above.
Between 2022 and 2030, this crude model projects that, if viewed as opportunity cost:
🚨3.5 billion metric tons fewer CO2e emissions total (Swap from 100% EV to 100% PHEV fleet)🚨
Even if we drop that opportunity cost delta assumption by 75%, that’s *still* nearly a billion metric tons of CO2!
Look, this model is really rough and I understand that global transportation economics and environmental impacts are, uh, complicated. But when we take a step back, look at the big picture and view Lithium Ion cells as the limiting supply for EV production, the “PHEV CARS SUCK” argument becomes, well, pretty friggin weak.
Until Next time
I'm extremely sympathetic to the idea that hauling around an extra 500lb of batteries to every grocery store and school drop off when you'll only use them once a month is non-ideal, and that there's a lot of CO2 to be saved by putting those batteries to work on more daily commutes instead, but I wonder (and don't know of good data about) the real-world usage statistics, and whether such a trade is actually real.
For instance: my understanding is that, in practice, most PHEVs never get charged. People buy them because they get killer gas mileage, aren't much more expensive, and come with HOV lane stickers. Of course, I hear plenty of "I get 1000mi/tank average" anecdotes, but there's almost certainly hefty selection bias at play there, given, you know, I'm commenting on a substack called "ESG Hound." Anyway, if PHEVs amount to "better fuel economy" or "A jeep with HOV stickers" as I think the trend will lean, I wonder how much theoretical CO2 there actually is to recover.
And of course PHEVs present plenty of practical usage limitations favoring this behavior. Small batteries mean (much) slower charging in terms of real power, i.e. it takes you "approximately overnight" or "all day at work" in real-world terms to charge your Prius Prime for 25mi usable, where those same periods can also get you a full BEV charge. Of COURSE this isn't strictly true, but given the similarity of user experience, I think it's practically so. You could fully charge your Prime at L2 while you're in the grocery store, while the same in a tesla would require DC fast charging. But especially as PHEVs proliferate, I think a smaller and smaller percentage of drivers WILL. In other words, getting most of your miles from the wall in a PHEV is nearly impossible in an apartment, where it's "a little extra time spent DC fast charging" in a BEV.
Add to that that only half or less of a PHEV's power is electric - EV mode is an objectively stunted experience compared to hybrid mode. Again, I don't know how real-world usage shakes out, but I'd be willing to bet most PHEV owners (and even more future-owners) aren't so eco-motivated as to make that sacrifice consistently.
In other words, yes, PHEVs used optimally stand to remove a lot more CO2 per kWh of battery capacity vs BEVs, but BEVs come with a *guarantee* of grid-sourced miles, where PHEVs come with a substantial raft of incentives against electric miles. On top of which, PHEVs probably delay the collapse and price-increase of the gasoline market (to the extent that is even likely)
I think you've made a compelling case (as Toyota does) that PHEVs COULD be a better use of resources than full BEVs, but I think there's more to whether they are likely to be in practice.
You are far too generous to the Tesla scenarios. The manufacturing emissions of big batteries are enormous (especially those made in China via coal power). It is unclear if a 100kWh car (especially an unreliable Tesla) will live long enough to have lower cradle-to-grave emissions than an efficient PHEV or HEV. Many Teslas need their batteries replaced more frequently that people realize, which is a huge emissions event. This (not behind paywall) has links to some of the studies. https://seekingalpha.com/article/4351863-electric-vehicle-subsidies-are-unsustainable