Recently I was invited to speak on a Princeton sustainability panel about my budding startup, Comet Motorcycles. Comet designs and produces small runs of electric motorcycles. Because it's based on a cruiser-style platform, it can carry an enormous 24kWh lithium-ion battery pack (the same size as one in a Nissan Leaf) mounted very low in the frame, giving it the longest range of any production electric motorcycle - about 200 miles of mixed riding, according to models based on research performed by MIT's Electric Vehicle Team.

Normally, Comet does not emphasize the eco-friendly side of electric power - we aim to produce the best motorcycles possible, which happen to be electric. However, seeing as this was a sustainability panel, I decided to bring up some points why gas motorcycles are so much more pollutive than people realize - even more pollutive than gas-powered cars! (at least in terms of non-CO2 output). Anyway, I was going over some of the common value propositions of EVs - zero engine vibrations, reduced heat output, and increased energy efficiency compared to an equivalent gas vehicle (which are typically about 17-21% efficient at converting the gas in their tanks into useful energy). At this point, I was interrupted by an audience member who started shaking his head and loudly repeating "That's not true, that's not true! Electricity from the grid is only 33% efficient!"

To flat-out cut off a panel speaker as an audience member, you must be pretty convinced you have a really important, undebatable point. As we shall see, the efficiency of EVs isn't quite that clear-cut, but it's definitely worth addressing if you're going to claim one way or the other. After all, you have to plug your EV into the grid, right?Nissan_Leaf And getting the energy from that grid - which generates its supply of power using ancient coal plants - to run your vehicle can't possibly be more efficient than burning gasoline in an onboard internal combustion engine, right?


To understand this question of EV efficiency, we need to decide what we mean by "efficiency". There are three such concepts that I'll deal with here:

"Tank-to-wheel efficiency" - as the name implies, this is how efficient the vehicle is at taking energy stored onboard and turning it into useable motion.

"Infrastructure efficiency" - this describes how efficiently we can take energy from the "source" (the ground, the well, a chunk of uranium, the sun, the tides, what have you) and put it into your tank.refinery-at-dusk Refineries, power plants, chargers, relays, ports, ships that carry barrels of crude oil across the Gulf - this is a big category for the refining and distribution of automotive-grade gasoline, as well as the generation and transfer of electric power using various source fuels (coal, LNG, solar, tidal, nuclear, etc.). This is also extremely difficult to estimate and you have to view the efficiency of different fuels in terms of opportunity cost and scale - that's a topic for a different post, so we'll just use worst-case estimates for now to get a lower-bound estimate of the efficiency of EVs over internal combustion vehicles.

"Well-to-wheel efficiency" - This is a measure of how much energy from the source fuel ends up as useful energy produced by your vehicle's power plant in the form of motion. By definition, then, this equals Tank-to-wheel efficiency x Infrastructure efficiency, and it's the most relevant, honest and all-encompassing measure of how energy-efficient our vehicle is. (note: To stay within appropriate scope for a blog post I restrict myself to discussing the well-to-wheel energy efficiency of EV vs. gas; I avoid making any claims with reference to the cost or emissions of EV vs. gas)

So first let's talk about the tank-to-wheel of a gas vehicle verses an EV: The EPA estimates that gas vehicles are about 17-21% "tank to wheel" efficient. A Tesla Roadster is about 79% efficient "tank to wheel" (88% chemical storage and conversion, 90% driveline efficient). Here's the point at which most Tesla drivers declare victory and zoom off. Well, we're not done here.

How much energy from the grid's source fuel actually ends up in your Tesla's "tank"? Let's go with the worst case scenario:coal-plant When you burn coal in an old-generation plant you are able to harvest 30% of the thermal energy originally stored in the coal into electricity (newer IGCC plants can bump this number up to over 60%. However, you then have to transmit and distribute that power to home users over the grid - there you lose about 6% of the energy that you generated. (100%-6%) x 30% = 28.2%. This is how much energy from the coal you burned that you actually get at your house where your Tesla is. The efficiency of Tesla chargers is about 92% according to the manufacturer. You multiply that by 28.2% and you get how much of the coal's original energy ends up in your Tesla's "tank" (battery pack): 92% x 28.2% = approx. 26%. Now you multiply that number by the Tesla's "tank to wheel" efficiency and you get 79% x 26% = 21%.

So what we've calculated is that a Tesla Roadster, running off of the least efficient coal plant you can find in America, is around 21% efficient well-to-wheel. That is - coincidentally - exactly the EPA's upper estimate for the tank-to-wheel efficiency of a gas car.

All the Tesla owners just rage-closed the window, but think about what I just said. A Tesla is as energy-efficient as a gas car, assuming that the refining and distribution of gasoline is 100% energy efficient. But I got news for you. It ain't.supplychain-1 Gasoline is actually really hard to refine and distribute. Building crude oil pipelines, lighting up gas stations, making everyone drive to those gas stations to fill up, carrying gas around in big gas trucks, etc. - the infrastructure required to distribute gas is titanic. The refining of gasoline from crude oil alone takes away about 15% of the energy. It's almost impossible to accurately account for all of the other energy-consuming hoops that gasoline has to jump through to get to your gas station and into your tank. So for the sake of being as generous as possible to gas-powered machines, we'll call that efficiency 100%.

Conclusions: We have calculated an extreme lower bound estimate for the added efficiency of an electric car over a highly efficient gas car - 15%. Are electric vehicles 100% efficient? No, not even close. But are gas vehicles as good or better? Definitely not. It takes some calculation and attention to accounting methods that aren't practical to present verbally, but make perfect sense when on paper. This is why you should probably leave the the more arduous arithmetic to the guy with the projector.