The humble Mitsubishi Mirage has none of the hallmarks of a futuristic, environmentally friendly car. It is fuelled by petrol, runs on an internal combustion engine and spews exhaust emissions through a tailpipe.

But when the Mirage is assessed for carbon emissions throughout its entire lifecycle — from procuring the components and fuel, to recycling its parts — it can actually be a greener car than a model by Tesla, the US electric vehicle pioneer, in regions with particularly high carbon emissions from electricity.

According to data from the Trancik Lab at the Massachusetts Institute of Technology, a Tesla Model S P100D saloon driven in the US Midwest produces 226 grams of carbon dioxide (or equivalent) per kilometer over its life-cycle — a significant reduction to the 385g for a luxury 7-series BMW. But the Mirage emits even less, at just 192 g.

Although multiple studies show that electric vehicles to be greener than comparable combustion engine cars, many components of the electric car life-sycle are left out. To capture electric cars’ full environmental impact, regulators need to embrace life-cycle analysis that takes into account car production, including the sourcing of rare earth metals that are part of the battery, plus the electricity that powers it, and the recycling of its components. Life-cycle studies show that the idea of “zero emissions” is misleading. Too much energy is consumed in the manufacturing process of lithium-ion batteries, and to recharge them, for the environmental impact to be nil.

Also, the lack of regulation differentiating between electric vehicles encourages car makers to sell cars with bigger batteries and longer ranges — features that sound great but are at odds with electric vehicles’ green image, given the amount of lithium and cobalt used in the batteries.

However, the problem for makers of electric vehicles is that their efforts to limit emissions in the supply chain can only go so far. The uncomfortable reality is that battery manufacturing plays a bigger role in life-cycle emissions than anything else the car maker does.

A decade ago, this was not such a problem. Researchers could assume electric vehicles were small cars such as the Smart fortwo, which weighs less than a tonne. But Tesla upended these assumptions with the Model S, its roomy saloon which can weigh up to 2,250kg because of a massive battery that powers its impressive range.  These bigger batteries could damage the green credentials of electric vehicles, even if power grids are fueled by less coal and more renewables, given the poor environmental and ethical standards involved in procuring metals such as cobalt, 60 per cent of which comes from the Democratic Republic of Congo.

Tesla has been credited with accelerating a broader shift into battery-powered cars, but one result of its appeal is that average electric vehicle batteries will double from 20 kilowatt hours today to 40 kWh by 2025, according to UK investment bank Liberum. Peter Mock, managing director for Europe at the International Council on Clean Transportation, says many electric vehicles produced today feature a range that is too high, and the trend is towards even bigger batteries.

The average electric vehicle sold today offers a range of less than 250km, according to EV Volumes, a data provider. But the Renault-Nissan-Mitsubishi alliance announced plans in September to create 12 electric vehicles with at least 600km of range by 2022.

“For 90% of the vehicles it just doesn’t make sense to have such a big battery,” Mr Mock says. “Maybe it’s useful now in the transition phase . . . But rationally it doesn’t make sense. Most of us drive less than 100 km a day.”

Mr Meilhan points out that petrol-engine cars weighing just 500kg — such as the French Ligier microcar or some popular “kei cars” in Japan — emit less lifecycle emissions than a mid-sized electric vehicle even when driven in France, where carbon-free nuclear power generates three-quarters of electricity.

“If we really cared about CO2,” he adds, “we’d reduce car size and weight.”

My rebuttal of the authors rebuttal in the FT (can be found here).

The original paper is here.  Miotti, M., et al. 2016. Personal Vehicles Evaluated against Climate Change Mitigation Targets. Environ. Sci. Technol 50:10795-10804.

What a waste of time.  This paper evaluates fossiled versus EV using the wrong evaluation criteria.

• Vehicle cost shouldn’t be in a scientific study of energy efficiency.
• Only energy use per mile and energy used over the life cycle.
• Nor should EV subsidies be used in the cost, the Republicans or next financial crash will end or lower them.

Secondly, the paper assumes that the electric grid will continually use more renewable power.  But since conventional oil, the master resource that makes all others possible, including solar and wind contraptions over their entire life cycles, peaked in global production in 2005, the electric grid cannot long outlast declining fossil fuels, and their life expectancy is only 20 years (wind turbines) to 30 years (solar). So even if we doubled and doubled and doubled the solar and wind generation of today, we’d have to stop making these contraptions at some point of oil decline – oil will be rationed to agriculture and other essential services, like heavy-duty trucks of all kinds, since they can’t run on electricity (see posts here).  And they aren’t doubling every year as you’ll see below, in fact at current rates of increase they won’t even be 10% of generation by 2030, and likely less since the best, most profitable sites are already taken (NREL 2013).

Solar and wind contribute a tiny fraction of electricity. So far in 2017, solar provided 0.95%, in 2016, 0.7 %, 2015 0.48%, 2014 0.38%, 2013 0.27%, and so on. In all cases the difference in increases was from one-tenth of 1% to a quarter of 1%. Not the doubling required.  But these low percentages belie the fact that for half of the year solar (fall and winter) the percent declines by half those months (see posts here). Most of the U.S. has very little solar power, even though it’s subsidized, because they are too far north.  Solar power is highly skewed:  California has 60% of solar generation, Arizona 24% (see post here). It can’t grow much more, the best spots are already built out. Sure, there are a lot of sunny places left, but it costs tens of millions to build the transmission lines to them (EIA 2017). And that cost isn’t ever included in the cost of solar electricity or solar plant construction. It is a huge free subsidy and substantially lowers the true cost of solar and wind electricity.

Wind contributed more than solar, but still a trivial amount of overall electricity generation: 1.9% in 2014, 2.0% in 2015, 2.5% in 2016, and 2.6% so far in 2017 (through August) (EIA 2017).  But the best sites for wind are already built out as well.  Wind is also highly seasonal, doesn’t blow at commercial scales across most of America in the summer, and never at commercial scale year-round in the South East (see posts here).  We are far from having a national grid, so you’ll need a horse or bicycle to get your groceries half of the year just about anywhere you live.

Another consequence of peak oil the study doesn’t acknowledge is that declining oil (coal, and natural gas) will reduce CO2 far more than EVs ever could.

Just at it is absurd to assume the grid will be 30% renewable power by 2030, it is also absurd to assume that batteries will grow more and more energy dense.  The energy density of batteries has only increased 5-fold over the past 210 years, but need to be 50-100 times more energy dense to come close to matching gasoline. That means that batteries will always be too heavy for diesel vehicles, and too expensive for 90% of consumers. The reason it is so damned hard to make batteries more energy dense are here and here.

Worse yet, comparisons are made on the basis of vehicle costs, with the assumption that EV costs will grow lower over time. Ridiculous! We’ve reached nearly all limits to growth, and lithium is certainly finite (lots of it but mostly embedded in minerals for which the chemical and/or heat energy is too high). Vehicle cost shouldn’t be in a scientific study of energy efficiency. Only energy use per mile and the energy used over the life cycle to make and operate the car, from mining to recycling.  And especially not subsidies, the Republicans or next financial crash will end or lower them.