The EV lifecycle conundrum
Image credit: Lambourghini
How well do electric vehicles stack up against their counterparts when you take all the inputs into account?
A seminal analysis of the environmental cost of cars turns 20 this autumn. In ‘On The Road In 2020: A life-cycle analysis of new automobile technologies’, the late Malcolm Weiss and colleagues from the Energy Laboratory at the Massachusetts Institute of Technology calculated how much it takes not just to keep a car on the road but to put it there in the first place. Written as Al Gore turned his attention from a failed presidential campaign to warning of climate change, the report focused on the greenhouse gas emissions from conventional and future automotive technologies.
The major step taken in ‘On The Road In 2020’ was to look not just at how much carbon dioxide a vehicle emits while running – a challenge pretty much any electric vehicle (EV) can win hands-down against anything with a conventional petrol or diesel engine – but over its entire life-cycle, and to take into account the resources it consumes.
Another competition the battery EV wins over traditional technology is complexity – this plays a big role in life-cycle costs. If there is one advantage that an internal combustion engine has, even one armed with a forest of sensors and electronic control units, it is that it is relatively simple and cheap to make. An EV needs a lot more exotic materials, from the lithium and cobalt that goes into the batteries to the many rare earths now needed for the control electronics and motors. It uses many more semiconductors produced by highly sophisticated fabs. This relative complexity puts much greater emphasis on the energy intensity of the manufacturing process. The tricky part is estimating the net effect of the two main competing forces.
A major problem with any life-cycle analysis is balancing one-off production costs against usage. Because the emissions from EVs are more heavily loaded towards production, working out a realistic lifetime mileage is crucial. A low-mileage environment pushes the advantage in the direction of traditional diesel- and petrol-driven vehicles because they, at least, did not take as much effort to make.
Luckily, for a comparison of environmental costs, the lifetime distance points to a relatively large total. In the UK, and the country is no outlier, a study using MOT data from the early 2010s by engineering company Ricardo-AEA found the lifetime usage for today’s vehicles tends to cluster around 150,000km. Some vehicles never do more than 5,000km but these cars – probably used just for a weekly shop and a little holiday travel – represent just 1 per cent of the total. More than 80 per cent of petrol-engine vehicles ran for more than 100,000km over their useful lifetime; the ratio increases for diesel. Ultimately, more vehicles run up 200,000km on the odometer than those that are driven less than 100,000km before they are sent to the scrapheap.
Trends in the future may change lifetime mileage. A strong possibility with the rise of EVs is that of autonomous driving. With full autonomy, it should be easy to order a vehicle to come to your door instead of having to go to a car-sharing parking spot. Once you are finished, the vehicle takes itself to the next user. In this scenario, it becomes an easier decision even for people in rural areas with little access to public transport to forgo car ownership. From a lifetime-mileage perspective, it is possible fleet owners will run vehicles for longer to maximise capital usage though they may instead look at reliability as a factor and decide to retire vehicles early. Without the large aftermarket that exists for former rental vehicles there is the possibility they head to the scrapheap or recycling centre sooner than with heavy private usage.
The battery pack is arguably the biggest weakness of the conventional EV > < environmentally. It is big. It is heavy, which depresses fuel efficiency. And making each one consumes a lot of energy. Sustainability issues with battery technology may lead to a bigger push towards fuel-cell designs that run primarily on hydrogen or, if that is impractical, synthetic carbon-based fuels, though these may have other problems that make them impractical.
Even with conventional chemistries and common materials, estimates of the carbon dioxide produced per kilowatt-hour vary by an order of magnitude in the different analyses researchers have performed so far. In a 2018 report, the International Council on Clean Transportation (ICCT) found battery production ranged from just 56kg of carbon dioxide per kilowatt-hour of storage up to almost 500kg. Most cluster around 150 to 200kg. The high number came from a 2016 study by researchers at UC Davis that was designed to take into account how electricity was generated. The high end of the range assumes coal-fired generation, which remains typical of production in China. At the other end of the scale, though, are factories like Tesla’s GigaFactory battery production plant in Nevada, which is designed to run entirely from solar power. Such a plant could be expected to cut carbon-dioxide emissions for battery production by more than a third.
If you scale up the average emissions per kilowatt-hour to a small family car such as a Ford Focus or Nissan Leaf with a battery capacity in the 25 to 30kWh range, the result is close to five tonnes of carbon dioxide before the vehicle has even been delivered. Assessed on a per-kilometre basis, though, the numbers do not look quite as bad. Assuming the car is run for the average lifetime of 150,000km, it amounts to 33g of carbon dioxide per kilometre. The body, motor and other electronics would most likely account for another 30 to 35g/km.
In an ideal scenario for EVs, the emissions stop there. This is where the question of the degree to which the local grid has been decarbonised comes into play again. Maps produced in 2016 by Jeremy Michalek of Carnegie Mellon University and colleagues showed how this would change across US states. Solar-rich southern states naturally favour EVs while the coal-rich northern Midwest means high-efficiency petrol-driven vehicles could wind up with an environmental advantage.
Trying to answer the question on a larger scale, last year an international group of researchers led by Florian Knobloch at the University of Nijmegen modelled the situation for countries around the world, taking into account how green initiatives might change the picture. Under the most likely trajectory used by the Nijmegen team, coal will still remain responsible for close to half of the global total even in 2050: an estimated 20PWh out of 45PWh.
Under the more optimistic 2°C maximum-warming scenario, coal’s contribution drops to a tiny fraction of that, although natural gas is likely to remain on its current trajectory of generating roughly 10PWh by 2050. Total energy consumption worldwide drops a little in this scenario to just over 40PWh: the 2°C scenario assumes greater emphasis on energy efficiency in industrial production and home heating.
For the most likely scenario, renewables account for 20 per cent of overall energy use. The share of renewables climbs to just over half of generating capacity under the 2°C scenario, which improves the picture for EVs significantly.
When it comes to whether or not EVs are better than fossil-fuel cars, the breakeven is when the grid is producing 700g of carbon dioxide per kilowatt-hour. To put that into perspective, the average for electricity generation in the US in 2018 was 1kg per kilowatt-hour. As a result, the benefits will not be distributed evenly. Based on current estimates of electricity generation, an EV-heavy fleet is good news in France.
Only around 25 per cent of the emissions of an EV built in France, which relies for the most part on nuclear energy, come from its time on the road even using 2015’s numbers for grid generation. Similarly, Iceland, Sweden and Switzerland benefit from nuclear, renewables or both. Germany, which largely switched nuclear for coal while it tries to build up generation using wind and solar, means an EV entirely built there would produce twice as much carbon dioxide during manufacturing as its entirely French-constructed equivalent. Were German drivers to import their EVs from France, the fuel cycle would account for nearly three-quarters of the total in the average vehicle. The UK is not far behind, though recent events have shown that coal generation is relatively easy to push out of the picture. The European country with the worst reason to switch to EVs right now is Estonia: its heavy use of gas from shale extraction makes EVs 40 per cent more emission-intensive.
Although France can more than halve its vehicle emissions by switching to EVs, the Nijmegen team expects the US to realise savings nearly as dramatic despite operating a far more carbon-intensive grid than France. The seemingly counterintuitive reason for its reduction of more than a third overall by switching to EVs comes from the American appetite for gas-guzzlers: Tesla has a relatively easy argument to support its long-term plans in the US. Conversely, the Japanese consumer’s preference for smaller, highly efficient petrol vehicles combined with a grid that no longer uses a large nuclear component makes the trade-off less obvious than you might expect.
The many factors that determine the environmental impact of motor vehicles make the question as to whether battery EVs help or hinder far from straightforward to answer, though the results of studies point to an overall advantage over petrol-driven vehicles. Changes in usage are likely to yield bigger savings in carbon dioxide, especially if car sharing becomes the norm and reduces the size of the national fleet. The sensitivity of the numbers to changes in technology may lead to hybrid designs becoming better options overall, just as long as their own inherent complexity does not push up their life-cycle costs.
How to keep life-cycle costs down
The core assumption that underpins many of the estimates of lifetime emissions production from electric vehicles (EVs) is not just that the average vehicle will travel at least 150,000km during its lifetime but that the battery will survive at least that long. But as numerous laptop and smartphone owners have found, lithium-ion batteries degrade over time and with usage.
As the battery is a major component of the environmental cost of an EV, responsible for half of it during manufacture in even a medium-sized car, if it needs to be replaced at some point the potential savings fall away.
The experience of Tesla, which has specialised in EVs with longer driving ranges to suit US business users, suggests the issue of capacity falling off over time is not as great as feared. With a single-charge driving range of 400km, an operational life of 200,000km suggests an absolute minimum of 500 charging cycles. In practice there will be many more cycles as few drivers are going to risk going from full to almost empty on each cycle. But studies suggest a reasonable minimum of 1,000 full cycles before the battery pack begins to degrade significantly.
Some of the pressure on the battery packs could be relieved by moving some of the workload to alternative technologies. R&D is progressing on supercapacitors: electrochemical versions of conventional electrical capacitors that offer much capacity. They lack the energy density of batteries but have the same ability to capture and release charge without degradation. Lamborghini has built supercapacitors into the frame of its Sián car to beef up the transmission system so that it copes better with high acceleration.
Recycling of batteries into fixed installations for grid storage would also change the arithmetic of emissions. Even when batteries fail to provide sufficient capacity to be usable in an EV, they still have the required density to prove useful over a number of years. That avoids the full burden of the life-cycle cost falling on EVs directly. But it is a calculation that requires electricity generation to move more quickly to a system based on renewables.
On a mission to reduce emissions
According to most lifecycle-cost reports, the lion’s share of battery-production emissions arises when the components are assembled into cells and packs before they go into the vehicle. In lithium-battery production, mining and materials purification account for just 20 per cent of emissions. Of that 20 per cent, aluminium extraction is the biggest single contributor at around 40 per cent of the total for most lithium-based chemistries.
This figure for raw materials cost remains pretty stable across the different battery chemistries despite their differences. However, recycling may improve the situation particularly with aluminium as it is possible to cut its emissions by more than 90 per cent for batteries that do not need to use any freshly extracted metal. Recycling is far from unusual in automotive batteries: almost all lead-acid batteries in the developed world use recycled components. The European Union has mandated that half the mass of EV batteries should be recycled. Although the instruction does not specify which components, the balance of mass in a battery and the relative costs suggests that structural materials like aluminium and steel will be favoured by manufacturers. This indicates a relatively straightforward saving of 10 per cent in energy cost before improvements in construction are taken into account.
Changes to sulphur and air-based chemistries may reduce the demand for some of the rarer elements that conventional lithium-ion batteries needed. But these are some way from commercialisation so for the near term, comparisons using existing chemistries seems a more accurate option.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.