Do electric cars emit half the CO2 as diesel from a life-cycle perspective?

Question

From The Guardian:

"On average, electric vehicles will emit half the CO2 emissions of a diesel car by 2030, including the manufacturing emissions,” said Yoann Le Petit, a spokesman for the T&E think tank, which commissioned the study.

-"Electric cars emit 50% less greenhouse gas than diesel, study finds" (2017-10-25)

Is it true that electric cars pollute 50% less than diesel from a life-cycle perspective?

Also I can't find a link to the report from https://www.transportenvironment.org (not a bad question just bad journalism) — daniel, Oct 25 '17 at 23:08
Around here you wold need to include CO2 emissions from the coal-fired power plants that produce the electricity. And of course take into account the loss in electricity transmission from the power plant to my house. — GEdgar, Oct 26 '17 at 00:24
There's a lot of complicated factors going into this one, as there're efficiency loses and costs all over the place in the manufacturing chain and in on-the-road energy consumption, both for electric cars and diesel engines. I suspect that this isn't a claim that could be easily fact-checked by off-hand knowledge; a good study'll be required. — Nat, Oct 26 '17 at 00:27
Here's a relevant part from [a study](http://www.sciencedirect.com/science/article/pii/S1361920916307933) linked by the claimant article: **"_Using [electric vehicles], in [the European Union], saves on average 50–60% of [green house gas] emissions compared to internal combustion engines._"**. — Nat, Oct 26 '17 at 00:30
This is confusing, because the [headline](https://skeptics.meta.stackexchange.com/questions/4074/how-should-we-deal-with-claims-that-only-appear-in-article-headlines) makes the claim, but the quote doesn't make the claim until 2030. — Oddthinking, Oct 26 '17 at 02:02
@Oddthinking the question could've been about the projected estimates for 2030 or the electric cars in Poland today which I see as falsifiable. But, no sign of the report with their working, sensational headline, the clearly biased group making the claim, I would just say this is fake news and move on. Even the Guardian hints there is some shenanigans with their working, but probably aren't allowed to call it out for fear of their greenwashing management. — daniel, Oct 26 '17 at 06:41
@daniel: is that helping us improve the question, or just speculating the motives of the journalists? — Oddthinking, Oct 26 '17 at 09:38
Short comment: it depends entirely on how you produce the electricity. Produce the electricity with a coal plant: no. Then it is just as bad. Produce the electricity with hydro power, wind power, solar power or nuclear power, and yes, then you put a big dent in the per km (or per mi) emissions. You need approximately 2 kWh per 10 km/6 mi. 1 kWh of Swedish nuclear power results in 5 g of CO2 (lifecycle assessment). The corresponding kWh of the German electricity mix is ~500 g of CO2. — , Oct 26 '17 at 11:17
@GEdgar, and Michael - I'm fairly certain that, even using coal, the efficiencies of industrial-scale energy production plants make even coal-powered electricity more efficient than individual sized fossil-fuel plants under the hood. Also, if anything happens to your grid to upgrade the "green-ness" of your energy source, then your electric vehicle magically becomes greener without any changes. The "better" might be more in relation to gasoline, though, since, in the USA, diesel is in a huge minority as a choice. — PoloHoleSet, Oct 26 '17 at 18:44
@GEdgar: But around here (western US) most of the electricity might come from hydro, with some geothermal and wind. Or possibly from PV solar on the owner's roof. That's why it's so hard to come up with a definitive answer. — jamesqf, Oct 26 '17 at 18:51
@PoloHoleSet: take a look at the study linked by Nat: there are some EU countries with high carbon use for electricity production and there the e-vehicles are not better than comparable combustion engines. But yes, less CO2 emissions for electricity production (France, Sweden: nuclear power) change the picture a lot. — cbeleites unhappy with SX, Oct 26 '17 at 21:08
This is clearly going to depend on where the car is used, as the article actually points out. — dont_shog_me_bro, Oct 27 '17 at 09:30
@PoloHoleSet In Sweden, a car that classifies as "Environmental Car" by law (literal translation) may not release more than 95 g/km. This puts it on par with an electrical car run on coal generated electricity. A more "normal" car will end up at around 150 g/km. Close enough for me to say a coal powered electrical car is no improvement, while then an electrical car on Swedish nuclear/hydro/wind power would release 0.1, 0.2 and 0.3g per km respectively Fossil fuels must go away **everywhere**, not just on the road but in our power plants as well. — , Oct 27 '17 at 10:58
@MichaelKarnerfors - I don't disagree with your statement about what needs to happen. I would also not disagree with stating that a reduction to 95 g/km improvement is "not enough." I would disagree that it is "close enough to say a coal powered car is ***no*** improvement." Having one be 50% more than the other, or cutting by over one third is a pretty significant difference. — PoloHoleSet, Oct 27 '17 at 15:03
@PoloHoleSet Falling from a 5th floor window is a huge reduction in falling compared to falling from the 10th floor. Does not really help you now does it. It is not survivable until we are down on the ground floor. And such is the case with carbon emissions. We need a **huge** reduction in order to solve the problem. Coal needs to go. — , Oct 27 '17 at 18:31
IMO, calling CO₂ emissions "pollution" is deceptive. Usually, when we talk about "pollution", we are talking about local problems that that can be solved---maybe for a price---without disrupting industries and people's ways of life. CO₂, on the other hand, is a global problem, and it won't go away if we don't completely stop using fossil fuels. — Solomon Slow, Oct 28 '17 at 17:21
@Michael Karnerfors: A point that you and the study authors perhaps are missing is that an EV that runs off coal-generated electricity today need not do so in the future, if whoever is in charge of the electric grid decides to change the generation mix. Or the owners decide to install PV panels on their roofs... — jamesqf, Oct 29 '17 at 03:27
@MichaelKarnerfors - interesting analogy, except cutting emissions by a third ***does*** help vs. the status quo, so not really applicable. — PoloHoleSet, Oct 30 '17 at 14:20
@PoloHoleSet Switching to electric cars is a **necessary** but not a **sufficient** step towards getting rid of fossil fuels. — , Oct 30 '17 at 14:23
@MichaelKarnerfors - agreed, as I very specifically said in my previous comments ("I would also not disagree with stating that a reduction to 95 g/km improvement is 'not enough.'"). That's not what "is **no** improvement" says. Looks like we're going in circles now, so I'll bow out. — PoloHoleSet, Oct 30 '17 at 14:30

score 3 · Accepted Answer · edited Oct 28 '17 at 18:43

Summary: I found numbers from University of Cologne (2010) which together with production-related emission numbers spread out over a number of German newspaper articles are in agreement with the claim.

There is large uncertainty caused by uncertainty about the development of renewable energies in relation to increasing e-mobility and also somewhat on conditions for battery production, though.

Sorry: all but the book chapter link go to German language sites

http://www.zeit.de/mobilitaet/2014-01/elektroauto-energiebilanz compares a VW Golf (gasoline, 169 g CO2/km*) with a Nissan Leaf (e-vehicle, 2014 German e-mix 106g CO2/km; both taking numbers from real-world gasoline tanking/charging data collections) saying that production of battery takes about 125 kg CO2/kWh, for the Leaf about 3t CO2 for battery production. OTOH, the rest of the car needs less, so production of the car incl. battery 2,74 t CO2 more for the e-car. That is, after 28000 km they are about equal in their CO2 emission. Unfortunately no absolute values for the production of the whole cars are given (only the difference).

However, lifetime of an e-car battery is cited as 80000 / 100 000 - 160 000 km So we may say: in addition to the 106g CO2/km for electricity we have another 19 - 38 g CO2/km (18 - 36 %) for battery production.**

So with these numbers, the current (2014ish) state is the Leaf producing 74 - 85% of the CO2 emissions of the gasoline car according to the numbers from the Zeit, with the CO2 emissions from the fuel consumption collections below, e-vehicle and diesel car cause about the same CO2 emission.

(The Geo article refers to a "Swiss study" (no further details given) saying that assuming a lifetime of 150.000 km, a maximum of 15 % of the total environmental pollution for an e-vehicle (production, use, dismantling) is for the battery.)

Richter & Lindenberger: Elektromobilität als CO2-Vermeidungsoption - Ergebnisse einer aktuellen EWI-Studie, e|m|w 4, 2010, p. 20-22 (pdf available at University of Cologne) which refers to this book chapter(which I don't have access to) give current e-vehicle as emitting 75 % of the CO2 compared to a similar diesel (2010). And give a graph of projected development which for 2030 projects the diesel to emit ca. 10 % less CO2, and the e-vehicle improving by about 50 %, leading to in 2030 the e-vehicle having emissions in the range of 40 - 50 % of those of the diesel - which even if battery production needs to be added would be rougly around half of the diesel's CO2 emissions. I note though, that their explanation of well-to-wheel calculation does neither list production of the vehicle nor the battery for the e-vehicle.

An important point in their discussion is that depending on how the electricity market, the CO2 emission certificate market and electricity production develop, there may be no CO2 savings at all or any CO2 not emitted by conventional cars is truly saved. After all, if you have enough renewable energy, you can use that also for battery production - note that a former Tesla manager is about to build a battery factory in Sweden where comparably low amounts of fossil fuel are used - while LG Chem is building another battery factory in Poland who at the moment have coal being 80 % of primary energy production.

This recent newspaper article says the projections for battery production are particularly difficult as the CO2 balance depends largely on a lot of factors such as

where and how the raw materials and battery are produced
battery size and lifetime
possible secondary use as stationary buffer for photovoltaics and recycling

* https://www.verbrauchsrechner.de/ (https://www.spritmonitor.de) list Golf VI gasoline with an average of 6,93 (7.22) l/100 km (161 - 168 g CO2/km) and diesel with 5,19 (5,55) l/100 km (137 - 147 g CO2).

Note also that for combustion engines, the fuel consumption varies easily by more than a factor 2 depending on driving style and type of route. E-vehicles have different behaviour in terms of acceleration/deceleration, so that these influences have different effect on energy consumption compared to combustion engines. E.g. e-vehicles can be much better in start-and-stop situations, but also there recuperation depends on driving style.

This puts natural limitations on conclusions - though less on a car-to-car comparison than on projections including population behaviour (which are needed for the CO2 emission for energy mix and renewable energy projections).

** These numbers do not account for possible savings due to recycling. And they cannot, as they are guesstimates (according to the Zeit article, by IFEU not numbers by battery producers) for the current state of the art in battery production. AFAIK, currently no medium- or large-scale recycling facilities for e-vehicle batteries are in operation, so real-world numbers for current state of the art are not available. I found only some older newspaper articles referring to a pilot plant for e-vehicle battery recycling that was to be built by last year IIRC. However, I did not find news relating to an operational pilot plant. The FAZ article linked above says that there's currently too much uncertainty in use and recycling scenarios to get useful (reliable) predictions. There's a number of things that can and may be done, but already what would be sensible (as possibly opposed to what is actually done) seems to depend rather sensitively on technical circumstances (for the whole e-vehicle concepts).

E.g. (my thoughts/questions)

Would it be better to have easily exchangeable car batteries, allowing
- super-fast "recharging" by exchange for a full battery, and thus
- be able to charge most car batteries on surplus renewable energy when that is available?
- (on a per-household basis or at "battery stations" like we do it when buying gas bottles)
- That would also allow e-vehicles with usually comparably low battery capacity (50 or 100 km for typical daily commute use etc.) that for long distance drives plug in a high-capacity battery and exchange it say, every 300+ km - or charge it on a e-highway), i.e. from a macro-economic perspective, less total battery capacity may be sufficient and/or better surplus renewable charging may be achieved.
- While that sounds appealing, does it make sense given the cost (also energy-wise and raw material-wise) of producing more batteries?
There are concepts under discussion to "downgrade" old car batteries (< 70 % of original capacity) and use them as stationary buffers for a while before recycling the batteries.
But that's probably going to be a competing technology to the battery exchange concept above.

Great writeup! Thanks a lot, Would like to point out that it's not tesla but a previous employee of tesla that is building the battery factory in Sweden, and it's not yet funded. Tesla on the other hand only run battery-factory on renewble energy wich kind of makes the production-cost argument voided all together. Theese calculations also assume zero recycling right? — Kristoffer Nolgren, Oct 28 '17 at 13:07
@KristofferNolgren: good point with the (possible) Swedish battery factory. Production cost and "renewable energy used" argument: that totally depends (the Uni Cologne paper discusses this in detail, in case you read German) and even if the battery factory (and the charging car owner) use contracts that deliver 100% renewable energy, the crucial question for net CO2 savings is whether this renewable energy is taken from the total energy mix leaving that mix a bit less renewable or whether additional renewable electricity is produced and in turn less gasoline/diesel(coal) is used... — cbeleites unhappy with SX, Oct 28 '17 at 16:07
@KristofferNolgren: I updated the answer for the recycling. I also realized (seeing that you are from Sweden) that different regions have very different renewable mixes. E.g. you have lots of water power, whereas in Germany there isn't that much (basically we've been using for >100 years what is available, we have more of the more volatile wind and solar power) - so using surplus renewable energy is more important here (in 2016, 645 M€ compensations were paid for wind power that has *not* been used. My guesstimate: that's ≈10 TWh ≈ 1.5 % total electricity production wasted). — cbeleites unhappy with SX, Oct 28 '17 at 18:05
How reliable are the CO2 numbers for the VWs since they predate the general awareness of VWs altering their behavior during emission tests? — Eric, Oct 29 '17 at 23:54
@Eric: those numbers are from databases where real use consumption is collected. So not directly affected by altered test stand behaviour - though real use consumption may be affected by changed software settings for road behaviour in possible updates of the software as a consequence of dieselgate (although VW claims the software update does not lead to higher fuel consumption). My guesstimate is that a larger contributor to uncertainty in those numbers may be that people who are interested in and/or good at fuel saving driving behaviour may be more likely to submit their data than people who — cbeleites unhappy with SX, Oct 30 '17 at 20:57
aren't. Another group of people who I can think of as being more likely to submit their data are people/cars (combinations) who need a lot of fuel and want to make public how upset they are about the much higher than expected fuel consumption. Those numbers are probably more relevant than prospectus numbers, but the distribution may not reflect full driver/car/road conditions population. Still, for the same model, e.g. Golf VI 1.4l gasoline, you find 95% of the entries ranging 5.75 - 8.69 l/100km which gives you an idea of uncertainty for actual fuel consumption (CO2 production). Though one.. — cbeleites unhappy with SX, Oct 30 '17 at 21:12
... can make a point for standardized road tests (prospectus numbers) because they normalize driver and road condition influence and reflect (both for combustion and e-vehicles) roughly the maximum you can get. — cbeleites unhappy with SX, Oct 30 '17 at 21:25

Do electric cars emit half the CO2 as diesel from a life-cycle perspective?

1 Answers1