I have a fantasy. I want a cool looking electric car, that gets 50+mpg and is not outrageously more expensive to buy and operate than a VW Golf when driven 12,000 miles per year. I was excited when I saw that the Union of Concerned Scientists released a much talked about report on the fuel efficiency of electric vehicles yesterday. There is much to like about this study. While electric vehicles do not release significant emissions of local or global pollutants when driving down the road, those electrons in the battery have to come from somewhere.
They come out of the wall plug at your house, place of work or those “free” charging stations at the airport. The UCS trace those electrons back to where they are produced. Electricity generated using coal has a higher carbon content than electricity produced using natural gas or carbon free sources of electricity such a hydropower. As a Nissan Leaf, one of the only fully electric vehicles on the market today, is advertised to draw 0.34 kWh/mile, one can calculate the CO2-equivalent miles per gallon. In other words, this is the miles per gallon that your gasoline-powered car would have to get if it emitted the same amount of CO2 as the Leaf.
The UCS was careful to use life-cycle emissions of CO2, which takes into account the full energy cost of that kWh delivered to your door (e.g. transmission losses). If you charged your Leaf with pure coal electrons, your gasoline-powered car would have to get 30 mpg, pure natural gas electrons 54 mpg, and if you could plug that solar panel straight into your Leaf, your gasoline-powered car would have to get a grandiose 500 mpg.
There is one potentially significant issue with the report, which the authors subtly acknowledge. It has to do with the use of average emissions instead of the marginal emissions in a given region. They argue that using marginal emissions matters when
“examining what power plants, or types of power plants, are likely to be deployed when new electricity demand is added to the electricity grid above and beyond the demand that already exists. [...] The concept of marginal electricity rates is important, especially when evaluating how electricity demand from thousands or millions of new EVs added to the grid over the coming decades will be met. [...] While a marginal emissions analysis of EV charging is important for forward looking studies of the policy implications of large-scale EV adoption, our goal in this analysis is to give consumers an idea of what the typical global warming emissions of the electricity used to charge their EV will be on today’s electricity grid. Therefore we use the average emissions intensity of the electricity”
Using average emissions ignores an important feature of how electricity markets function. Grid operators (such as California’s ISO) are in the business of balancing load. This means that they need to make sure that when you want to turn on your toaster at 7 am, you can. If demand is low, the cheapest units are deployed (usually coal) and as demand ramps up, more expensive sources of generation are brought online (e.g. natural gas). As a result, not only does it matter where you charge your EV, but it also matters when you charge because the marginal source of generation varies over time of the day. The marginal source of generation can also vary across time of the year (e.g., spring vs summer) even for the same time of day.
Demand for electricity is lowest at night. That is when you charge your electric vehicle. In most areas base load is predominately generated by coal. That means if you plug in your vehicle at night you are putting – yes you guessed it – mostly coal generated electrons into your Leaf’s battery. This is true for one Leaf, ten Leafs or a Berkeley full of Leafs.
Putting my ears to the ground of the econosphere, I hear that an all star team of environmental economists (Graff-Zivin, Kotchen and Mansur) is gearing up to calculate just these numbers. I bet that there are some areas of the US where you would do better in terms of CO2-equivalent miles per gallon driving a gasoline-powered Honda Fit. And that Fit would put you back $15k instead of the $35k for the Leaf.
(Yes. That is the price before tax breaks. Socially that is the right price to use. Privately you may use $27,000 if you must.)