I’m a fan of electric vehicles. As I’ve noted, they can be a smart and flexible option to help the grid accommodate variable energy resources like wind and solar. And EVs are the option likely to have the most success at decarbonizing transportation.
I’ve noted that the claim “The only way to do X is Y” is probably incorrect. I expect EVs to become the best way to get transportation CO2 emissions to zero for most applications, but not the only way. Hydrogen fuel cells, cellulosic biofuels, and liquid fuels produced by renewable electricity also exist. Maybe we will use some of these technologies for specific applications. But it is my estimate that EVs will achieve the greatest reductions in transportation carbon emissions.
UCS has shown under the current electricity system, EVs reduce emissions compared to the average gasoline car in all parts of the country, even when considering manufacturing impacts. In most of the country they beat the best gasoline cars. That analysis uses the 2015 version of EPA’s eGRID database, which actually only includes data through 2012. The grid has gotten considerably cleaner even in the past few years.
The grid impact
So, theoretically, what would happen if all light-duty vehicles were EVs?
As a back-of-the-envelope calculation, that would be about a 25% increase in electricity demand. Light-duty vehicles account for about 3 trillion vehicle-miles per year in the United States. EVs get roughly 3 miles per kilowatt-hour. We would need an additional 1 trillion kWh of electricity. The U.S. currently uses about 4 trillion kilowatt-hours per year; an additional 1 trillion kWh represents a 25% increase.
US cars and light trucks produced just over one billion metric tons of CO2 in 2014, or about 20% of all U.S. energy-related CO2 emissions. In 2015, US power plants produced 1,925 million metric tons of CO2 while producing 3,931 billion kWh, for an emissions rate of about 0.5 tons per thousand kWh. The additional trillion kWh for the EVs would result in 500 million tons of emissions, or about half of what light-duty vehicles emit today, for a 10% reduction in energy-related CO2 emissions. That estimate applies if EVs are charged by the grid of today. It does not account for the fact that the grid is getting cleaner through existing market forces and policy actions, nor the ways that EVs can specifically help the grid accommodate greater use of variable renewables, nor actions such as buying green power or using a “green charging” algorithm.
Scenarios of EV market growth
It will be a while before EVs represent even 10% of vehicles, let alone 100%. EV sales were about 0.9% of new car sales in the US in 2016, up from about 0.7% in 2015. With somewhat over half a million sold during the period 2012-2016, EVs are currently about 0.3% of cars on the road.
Sounds small? Well, compound growth is an amazing thing. Continuing the 2012-2016 growth rate of 32% per year would put EVs at 10% of all new car sales by 2025, and about 40% by 2030.
Now, that might be a stretch. It’s too short a time to extrapolate from, and technology diffusion tends to follow an S-shaped “logistic function” rather than a constant growth rate. Still, even with lower growth rates, existing targets are achievable.
The eight states that have signed the Zero Emission Vehicle Memorandum have a combined goal of 3.3 million vehicles on the road by 2025 (and all new vehicles being zero-emission by 2050). I expect that most of these will be EVs, although some may be other technologies. These states represent about a quarter of the US population. If the ZEV states meet their goal, and the other states with three times as many people deploy only half as many EVs, that would be 5 million EVs by 2025. This would represent roughly 2% of all vehicles on the road in the US, and possibly 5% of new car sales in that year.
US EV sales would need to grow at a rate of 25% per year over 2016-2025 to hit 5 million cumulative sales in 2025 (sales in that year would be 1 million). As shown below, 2016 represented a 37% increase in EV sales over 2015 (the 2012-2016 period saw 32% annual growth).
Assuming we hit 2025’s targets, what would happen next? One of two things.
Maybe the market keeps growing. In UCS’s “Half the Oil” report, we present a scenario in which California meets its Zero Emission Vehicle (ZEV) requirements, with 16% of new vehicles being ZEV by 2025. With further growth beyond that, ZEVs would reach 28-36% of light-duty vehicles in California in 2030. This Hawaii-specific study similarly forecasts a “high” case of about 33% of new car sales being EVs by 2030. Other studies feature even faster diffusion.
Or, vehicle sales might fall dramatically. If shared autonomous vehicles become the norm, far fewer vehicles would be needed. At any one time, no more than about 13% of cars are on the road. You might think you still need one car per person for rush hour, but 1) our number of commuters is less than half our number of vehicles; and, 2) these commutes are spread around a fairly broad time and are typically under half an hour. So one vehicle can support several commutes, even before you consider ridesharing. EVs could be grabbing a larger share of a shrinking market – at least in the countries that currently feature widespread car ownership.
Considering the progress from Google, Tesla, Apple, and others, full autonomy may be technologically possible by 2020. Regulations would take another few years to catch up. Granted, autonomous vehicles could be privately owned; the degree to which people choose to switch from owning vehicles to buying “transportation as a service” is uncertain. Still, the prospect of shared autonomous vehicles is why I think projections beyond 2030 are especially difficult.
Our transportation system might look much, much different.