home

A Hobbyist's Take at How Sensible are EVs, CO2-wise

June 2023

Previously I was wondering just how sensible are plug-ins in terms of lowering CO2 emissions (you can read about that here). The next logical step after that would have been to check the same for EVs. Except that I didn't. Well, I once ran a half-marathon and then everybody was saying "Now you should go for a marathon next year" and I was reasoning like "Hell no, just because it seems the most logical next thing or everybody says so, I am not gonna do it." And so I wasn't. Like seriously. Until one day, when flipping through a magazine while in some checkout line, I read that Jack Gylenhall was running like almost a half-marathon distance every day in preparation for a movie (or sth like that) and I said to myself that if that skinny Brokeback cowboy can do a half marathon day after day, well heck, then I should surely be able to survive one marathon. And so I did. Plenty of 60- or 70-somethings crossed the finish line before me, but I wasn't going to complain about it. (I was in my late thirties then but was never a runner, on both occasions I trained for a couple of months prior to the event and then stashed the sneakers away.)

So, how about EVs? Just because EVs don't have a tailpipe, it doesn't follow that driving them doesn't cause CO2 emissions. It does, it is just that they occur someplace else and at a different time, ie. they are not being generated in the moment the EV is spinning its wheels and in the hardware where its driver is seated as is the case with internal combustion engine vehicles (ICE).

CO2 emissions from electric vehicles come from 3 directions:

  1. from CO2 footprint of electricity generation (characteristic of the power network into which an EV is plugged-in to charge its traction battery),
  2. the emissions created during the production of the EV's traction battery (most often than not a lithium ion battery), amortized during the battery's lifetime (for the purpose of this text, 250,000 km driven by the vehicle).
  3. the energy losses during charging, estimated to 15% (it seems that most sources put it between 10 and 25%)

Let's calculate these components.

Calculations - theory

CO2 emissions are expressed in g CO2 per km driven. Here's how to calculate them for EVs:

(1) CO2 emissions from electricity consumption

Average electricity consumption of the vehicle (in kWh/100 km);

TIMES

CO2 emission from electricity generation, characteristic for the network from which the EV gets its glow juice from (in g CO2 /kWh of electricity produced);

TIMES

1,15 to account for the 15% losses at charging point (from point 3 above);

DIVIDED BY

100 to express the value per km instead of per 100 kms.


(2) CO2 emissions related to lithium battery

Vehicle's battery size (in kWh);

TIMES

g CO2 emitted to produce 1 kWh of lithium battery1;

DIVIDED BY

250.000 (to get a value of g CO2 due to battery production per each km driven by the vehicle during its lifetime, which in this text is, rather arbitrarily, set to a generous 250k km).


Wait, why have add losses here to the calculation at point 1 when noone talks about losses when calculating consumption for gas guzzlers? Don't the trucks that haul petrol across the country to get it to gas stations add up to losses then too? Well that would rather be similar in content (and probably in size too) to losses in electricity transmission over long distances. But in addition to that, charging EVs also incurs losses right at the charging point (the 15% added in the calculation) while usually all the petrol that passes through the hose at the petrol station also makes it into my car, I don't tend to sprinkle it around water gun style. (There are some losses in the form of vapours when filling up cars that run on dinosaurs and ferns, but those don't affect consumption calculations.)

Re: one other factor at calculation 1, the values for emissions from electricity generation vary a lot, for the purpose of this text the following reference values will be used:

To get the CO2 emissions for the vehicle, the 2 components from the 2 calculations above have to be summed!

Selection of cars

One small (VW e-up, 32 kWh battery), one bigger (Tesla 3 Long Range, 75 kWh battery)

Data rules

The 2 pieces of data about vehicle consumption mentioned below (can that be called a sample?) are real-world consumption data logged by users from a popular site for logging fuel consumption. Requirements: cars year of make 2018 or newer; data from at least 100 users; for the average consumptions below I took the mean value between the average consumption of all cars and the consumption of the median car, just in case the average is perhaps distorted due to any extreme value on any of the extremes.

Calculations

Car Average
consum-
ption		 Battery
capacity		 Elect. prod.
emissions		 Emissions due
to consumpt.		 Emissions
due to
battery		 Car
emissions
total		 Equivalent
consumpt.
in petrol car		 Equivalent
consumpt.
in diesel car
kWh/
 100 km		 kWh g CO2/kWh g CO2/km g CO2/km g CO2/km l/100 km l/100 km
Tesla 3 440 (Wrld) 95 39 133.6 5.8 5.0
18.7 75 79 (France) 17 39 56 2.4 2.1
40 (PV) 9 39 47.6 2.1 1.8
VW e‑UP 440 73 17 90.4 3.9 3.4
14.5 32 79 13 17 30.2 1.3 1.1
40 7 17 23.7 1.0 0.9

Sorta analysis

Charging the Tesla 3 in an average corner of the world (where during production of 1 kWh of electricity 440 g CO2 are emitted) gives already a good result, but still one that is nowhere spectacular. Modern medium sized diesel cars are easily satisfied with less than 5.0 l fuel per 100 km (which would give the same CO2 emissions figures as from a Tesla 3 in an average region as regards electricity production) and 5.8 l gasoline per 100 km offers no challenge for modern gasoline self-charging hybrids. (I drive an older spec self-charging hybrid, ie. a hybrid that is not a plug-in hybrid, I am definitely not able to hypermile it but my average consumption, calculated at the gas station from fuel-ups over the last 4 years, is 5.1 l/100 km or 46 US mpg.)

But still, that is comparing the consumption of the average Tesla 3 with the consumption of a car from the lower part of the distribution of data about consumption of ICEs. I would estimate the average real-world consumption figures of ICE cars, at least in Europe, to be 7-8 l/100 km for gasoline and 6-7 l/100 km for diesel cars, so at least some 20% worse than an average Tesla 3. If the same Tesla is driven in France or Norway, where the CO2 footprint of electricity production is low, then no fossil fuel car can realistically have emissions so low (2-2.5 l fuel per 100 km driven).

Conversely, if you charge your Tesla 3 in a corner of the world where CO2 emissions from electricity generation are high, as in where sourced mainly from coal, when the figure for CO2 emissions climbs over 900g/kWh generated, then driving your EV doesn't make much sense. (Yes, electric cars are fun, yes, Teslas are cool, yes, EVs can be way cheaper to run than ICEs and all that, but here the angle on things is exclusively from the perspective of CO2, so that's where the absence of sense comes from.) Namely, in this case you would realistically be nearing a drive where 200 g of CO2 are emitted per each km driven, a lousy result.

In this sense I find it pretty stupid, or better deceptive and offensive, when car producers for their EVs proudly wave the banner of "CO2 emissions: 0 g" in every ad or catalogue. Well, sorry, that's crap. There should be two numbers there, one for on the spot emissions (which are nil) and one for the driving imputed emissions which would take into account some average numbers for battery production pertaining to that car and a range for electricity generation numbers pertaining to the country where this ad is being shown or this vehicle is being sold.

As regards batteries, it seems that with current battery production technology the size of the lithium ion battery can have quite an impact on the overall emissions of an electric car - see the 39 g of CO2 emitted per kilometer driven for the Tesla 3 vs the 17 for the Volkswagen minicar. The small VW e-up (which is, I think, discontinued as of 2023 but there are plenty of other electric cars with smallish batteries available), when driven in a part of the world with average carbon intensity of electricity production, offers CO2 emissions figures that can be reached only by serious hypermilers, while when driven in low carbon intensity countries (Vive la France) is completely out of reach of the most extreme of the sparsamest of drivers. And conversely, big electric cars with batteries over 90 kWh (a Mercedes-Benz EQS packs a 110 kWh Li-ion battery) and then also probably higher average consumption (say 25 kWh/100 km) would come out at 180 g CO2/km (same as 8 l gasoline or 6.7 l diesel/100km) and so offer no benefit whatsoever in terms of CO2 emissions reductions. Note that for each kilometer that a car with 100 kWh of batteries covers on the road, it emits - from an accounting perspective that is, for the CO2 has already been emitted during production of its batteries - an amount of CO2 equivalent to what comes out of the tailpipe of a gasoline car that consumes 2 l/100 km, solely for the fact that it sports a battery of 100 kWh.

But we humans are where we came to be - or basically we are what we are - because we excel at one thing: standing on the shoulders of giants (or actually anyone) that came before us. And so there is no question that all of the constituents of the equation - battery, drivetrain and electricity production technology - will eventually be improved to the point where every and any electric car will be a no-brainer in terms of CO2 emissions plausibility.




1 To best of knowledge set to 130,000 g CO2 for each kWh of battery capacity (sources: https://www.transportenvironment.org/wp-content/uploads/2021/07/2019_11_Analysis_CO2_footprint_lithium-ion_batteries.pdf , Effects of battery manufacturing on electric vehicle life-cycle greenhouse gas emissions (theicct.org) , both accessed Feb 2023)

2 Source: Carbon intensity of electricity, 2000 to 2021 (ourworldindata.org) (accessed Feb 2023)

3 The average of the two: Carbon intensity of electricity, 2000 to 2021 (ourworldindata.org) and https://app.electricitymaps.com/map (past data for 5 years) (Both accessed Feb 2023)

4 Best shot from sources: Life Cycle Greenhouse Gas Emissions from Solar Photovoltaics (nrel.gov) , Re-assessment of net energy production and greenhouse gas emissions avoidance after 40 years of photovoltaics development (nature.com) (Both accessed Feb 2023)


Ivo Makuc, 2023
byguesswork@gmail.com