In the wake of the EU Parliament's controversial decision to ban sales of combustion-engine vehicles by 2035, the harshest criticism essentially revolves around the purported “idiocy” of adopting electric engines in vehicles when most electricity is still produced from fossil fuel (often supported by the memetic news of diesel generators being used to charge fully-electrict Tesla cars).

While there is little doubt that such a setup is, shall we say, “less green” than it would be if electric vehicles were charged with ”green” energy (i.e. energy generated by renewable and/or less polluting sources such as wind, solar, water, or even nuclear), objecting to a wider adoption of EVs on that basis1 not only completely misses the point (a “green” transition can be gradual, it doesn't have to be all-or-none), but it's particularly stupid in the sense of the perfect being enemy of the good.

The fact that a diesel-charged fully-electric vehicle is still “greener” than a combustion-engine equivalent has been remarked by many when discussing Tesla's “loss of face”, but I don't have any particular appreciation for the articles I've found on a quick search on topic, so I've decided to present my own take on the subject. Note that this take only mostly looks at the “finished products”, so for a wider discussion on the topic of the “greenery” of vehicles (electric or not) including manufacturing and whatnot you'll have to look elsewhere. What you'll find here is some a few key advantages of the fossil-fuel-backed EV transition that I deem often overlooked.

Pollution delocalization

This particular advantage is in fact the first I thought of, the one that triggered my desire to write this post, and even if it was the only one (it is not), it would be —for me— sufficient.

Replacing all internal-combustion-engine vehicles (ICVs henceforth) with full-electric vehicles (EVs henceforth) displaces the pollution source. ICVs usage is concentrated in the same areas where people live, with each of them being a moving point-source of pollution right under our nose (and eyes, and skin, etc). Switching to EV would concentrate and move all those individual point-sources into a pollution sources located elsewhere, most typically farther from densely inhabited places.

Pollution from ICVs has well-known effects on health problems ranging from birth defects to premature death. While delocalizing the pollution source doesn't eliminate the effect altogether, it does improve things, and the farther the power plants are, the better.

In addition to displacing (if not reducing, see below) good ol' air pollution, replacing ICVs with EVs also massively decreases noise pollution. In fact, EVs are so much quieter that there are worries about the safety implications of the lack of noise.


This may come as a bit of a surprise, but EVs can actually be more efficient than ICVs in energy usage from the same amount of fuel. This depends on multiple factors ranging from the vehicle use-case to the power plant generation.

Current vehicle combustion engines have peak efficiencies ranging from 35% (gasoline) to 45% (diesel). Of course, in practical usage this only happens under ideal conditions (full load during acceleration with the engine at peak efficiency RPMs, which is usually between 2K and 3K RPMs) and one of the big ironies of ICVs is that maximum fuel economy is instead achieved at cruising speeds at the highest possible gear with the lowest possible RPMs (typically around 1K) which is actually not very efficient, in terms of fuel energy extraction. Combined with the more or less frequent (and immensely wasteful) stop-and-go (most typical in urban usage, yet less uncommon than most people would believe in extra-urban and highway usages), the effective efficiency of most ICVs is between 12% and 30%, with worst cases dropping as low as 6%, and best cases at around 37%.

{ Verify if drivetrain losses are accounted for in these figures. }

{ Add considerations on the cost of refining crude oil into vehicle fuel. }

Electric vehicles are more efficient in using the energy from their batteries (at least 60% considering all losses) and much less affected by idling or other low-efficiency usages. Of course, this has to be compounded with the efficiency of power plant energy production and grid distribution losses. The latter amount to between 10% and 30% (depending on distance, quality of the grid and a number of other factors), leading to an overall efficiency between 40% and 55% from power generation to motion.

The crux is, unsurprisingly, at the power generation step. Even though energy efficiency of nearly 60% are possible with modern tech, most power plants do not reach such levels (typical efficiency is 35% for coal, 38% for oil, 45% for natural gas, with the most efficient ones reaching resp. 42%, 45%, 52%), resulting in an effective efficiency for EVs (from fuel to wheel) between 14% and 29%.

So even without additional considerations emerges that switching to EVs from ICVs would typically result in comparable efficiency in fuel usage. However, the argument doesn't end here: there's more to consider both on the EV side and on the power generation side.

One of the significant advantage of EVs over ICVs is regenerative braking, i.e. the capability to recover some of the energy spent to put (and keep) the vehicle in motion when braking. Although similar systems (particularly KERS, Kinetic Energy Recovery Systems) have been explored for ICVs (particularly in racing cars), they have not seen any meaningful adoption for civilian transport, in contrast to the widespread use of regenerative braking in electric and hybrid vehicles. Taking brake energy recovery into account, the efficiency of EVs rises to the 75%-90% range, for an effective efficiency (fuel to wheel, including power generation and grid losses) between 18% and 42%.

The other aspect to consider is that while most of the inefficiency of ICVs goes into wasted heat, power plants can co-generate electric power and heat, with overall efficiencies as high as 88%.

Although this doesn't improve the fuel-to-wheel efficiency of EVs per se, it does improve the overall fuel consumption efficiency, thus reducing waste.

Efficiency improvements

Technology improves (in fact, it's interesting to note that technological progress seems to have had a higher impact on the efficiency of ICVs than in the efficiency of EVs, although this is largely due to the fact that the EV efficiency is already much higher, and that considerably less R&D has gone into improving EVs until recently).

One question that is interesting to pose is: how long does it take for a technological improvement to have an actual measurable effect (e.g. leading to lower pollution or higher efficiency).

In a largely saturated market like that of civilian vehicles, even if all new cars were to adopt the better technology, the replacement of the existing cars with the new ones would take decades if not for government incentives to switch to lower-emission vehicles.

With the highest sources of inefficiencies for EVs being located outside of the vehicles themselves (distribution grid, and most importantly power plants), many technological improvements would lead to indirect benefits to the effective efficiency of EVs without any intervention on the user side.

This doesn't hold true for all improvements (e.g. a better battery technology leading to higher density and thus lighter batteries for the same capacity would still require physical maintenance on the vehicles, although still less problematic than buying a whole new car as needed for most technological progress on ICVs), but e.g. a 5% reduction in power grid losses or a 5% improvements in efficiency for power plants would automatically lead to the corresponding gains in the overall effective efficiency (fuel to wheel) for all EVs recharging on said power grid.

Smoother transition

The previous point naturally segues into the final and (for some perspective) most important point: an early transition from ICVs to EVs will lead to smoother transition to other power sources.

Road transportation accounts for nearly 50% of oil consumption in the EU and constitutes the main emitting source for a number of pollutants responsible for the low air quality (and related health issues) in urban areas. Even if there was a full switch to “green energy” generation today, the around 250M vehicles currently in circulation in Europe would remain responsible for this massive consumption of oil and the associated pollution and health issues.

Even though, as discussed above, an accelerated transition from ICVs to EVs would neither eliminate our dependency on oil nor reduce the associated pollution (although it would reduce the health problems associated with the emitting sources being concentrated in highly populated areas), it would make the transition away from fossil fuel more effective, any subsequent increase in the percentage of energy produced from “green” sources would automatically (albeit indirectly) make road transportation more “green”.

Addendum: could it happen or not?

With the sales of electric vehicles doubling worldwide every 2 years or less (average growth rate 50% or more), one might even wonder if the EU initiative is even needed: if the trend continues and the percentage of EV car sales (to total car sales) were to keep doubling every 2 or 3 years, starting at an 8% of car sales being for electric vehicles in 2021, we would approach 100% of car sales worldwide being EVs in 10 to 15 years (thus with the target of the EU parliament proposal). The growing prices of car fuel (largely a consequence of the 2022 Russian invasion of Ukraine) is also likely to support such a trend.

(It should be noted however that these statistics include, so-called plug-in hybrid vehicles, that have both an internal combustion engine and a battery-backed electric motor. These are the ones that have seen the fastest growth in the most recent years, yet they would be affected by the ban proposed by the EU Parliament.)

What is missing on the other side is the infrastructure to support such a transition: charging stations are still few and far apart, and mostly concentrated in the higher-density, higher-traffic regions. Massive infrastructural upgrades are needed to support the target of the EV transition, and not just in terms of power distribution: large increases in the number of circulating EVs will also require an adequate growth in power generation. And between the looming energy crisis, the impact of climate change on “green” energy production, and the long times, increasing costs and general resistance to nuclear, that is something that might not ramp fast enough in the envisioned time frame.

It makes one wonder if investment in infrastructure (power generation, better grid, more charging stations), a differentiation of power sources, and support for local power production (“solar panels everywhere!”) to bring down electricity costs would be a more effective (albeit indirect) strategy to incentivize adoption of EVs. One thing for sure is that these things have to happen anyway for the “only EVs after 2035” to be sustainable.

Post Scriptum: a bet that I'm sure to win

Assuming the proposal (or some equivalent initiative to accelerate the adoption of EVs) passes, you can bet that 50 yeas from now, when we will be enjoying the benefits of the widespread use of EVs over ICVs, libertoloids (libertarians, ancap et similia) that are now so vocally against the EU plan will boast how the free market led to the resulting quality of life improvements, conveniently forgetting about the massive impact that regulations and incentives have had in directing such market.

How can I be so sure? Because that's exactly what they are doing about the improved energy consumption and reduced pollution that were driven by large scale government initiatives, particularly from the 1970s onwards.

Claiming “engineers did that, not government regulations” is a platitude inasfar as it's true, and is otherwise false. Yes, engineers were essential to achieve the technological progress that improved energy consumption and reduced pollution, but the main incentive to move in that direction came from the government regulation. We'd still be dying by smog in large numbers if private entrepreneurship profit had remained the driving motive for technological progress.

Until and unless the markets finally manage to incorporate the true cost of large scale externalities such as the environmental damage, it will never be able to lead to such improvements in quality of life. It's not by chance that the industrial revolution actually led to a decrease in life expectancy, in highly industrialized cities, with respiratory issues becoming the dominant cause of death (PDF warning).

(And yes, a similar discussion holds for epidemics, but that would be way off topic.)

  1. of course there are other objections, such as “this shouldn't be forced by law, but a decision made by the market”, which I'm not discussing here, and not just because “the market” not accounting for externalities —which are key in this discussion— makes such an objection irrelevant. ↩