Pat and I have just hired a car in the UK for 10 days: a Peugeot 308, a proper 4-seater with a non-trivial boot and a turbo-diesel engine, generally a nice car. The onboard computer told us the fuel consumption, for a mixed cycle of town, country and motorway driving, was 54 Imperial mpg, or 45 US mpg. It hadn’t been zeroed between hires, so the value covered 2000 miles. Not bad; and comparable to what you can expect from a Toyota Prius hybrid. Peugeot, along with other European carmakers, are working on diesel hybrids: they promise 83 Imperial mpg for the comparable model due in 2010, or 69 US mpg.
It’s a puzzle why US drivers are so set against diesels. They are now as quiet and flexible as the petrol engines of 20 years ago, have low particle emissions and better consumption.
The main thought this experience has triggered is the insight that miles per gallon is now an obsolete and misleading measure.
What we should worry about is carbon emissions: grams per kilometre (or mile). For traditional cars, mpg is a perfectly good proxy. But not for hybrids. Once you enable recharging the battery from the mains, a car is no longer an isolated system. A hybrid used as an urban runabout in pure electric mode could have an infinite mpg; but it has positive carbon emissions, because electricity isn’t carbon-neutral.
The carbon intensity of electricity, and hence of any partially electric vehicle, varies with the generating mix and the daily and seasonal load curves. Nuclear, solar and wind energy have negligible short-run marginal costs and so are the preferred base-load forms of generation. Night-time electricity (forgetting about solar) is normally less carbon-intensive than daytime. Furthermore, Americans can look forward to dramatic changes in electricity generation and transmission over the lifetime of a new car. Any benchmark indicator will be misleading; but a bad one is better than none, as Dr. Johnson said of dictionaries.
My proposal is that the EPA:
* abandon miles per gallon for grams of carbon (equivalent) per 100 miles as the basis for fuel economy standards;
* for hybrid cars, calculate both the values for an autonomous cycle without recharging, and a typical urban cycle with maximum recharging;
* apply the latter metric to all-electric cars.
Working assumptions: recharging is done at night using off-peak generation; the electricity generating mix is uniform across the United States and optimally managed – an efficient interstate grid would soon make this a plausible approximation; the carbon intensity of electricity reflects that of the generating capacity currently installed and under construction. The basis should be revised every three years.
The same goes in kilometres for the EU.