Wednesday, September 01, 2010

Grading Cars on the Curve

By now you may have seen some prototypes of the new-car fuel economy stickers on which the EPA is seeking public comment. The versions that prominently display letter grades for overall fuel economy performance are certainly eye-catching, rising above the potentially confusing mix of numbers and graphics in the body of the sticker. Yet although the current stickers are clearly inadequate to illuminate the choices and consequences associated with buying vehicles powered by an increasingly diverse array of fuels, devising a similarly simple summary page may be beyond the skills of even the cleverest engineers and graphic designers. And in the hyper-connected world in which we now live, the necessity of presenting all this information in one place deserves at least as much thought as the proposed new stickers themselves.

I don't envy the EPA its assigned task of coming up with a useful replacement for the venerable fuel economy stickers that adorn the windows of all new cars at every dealership in America. It should be obvious that the current stickers, displaying city, highway and assumed average fuel economy--a subject for discussion in itself--along with estimated annual fuel expenditures, are not up to the task of informing consumers faced with a choice of vehicles running on gasoline, diesel, natural gas, ethanol, electricity, or a mix of several of these. Whether you consider it necessary to attempt to do so probably depends as much on your personal philosophy as on the inherent complexity of the situation, so for the purposes of this posting, I will accept it as a given and focus on evaluating the range of sticker options furnished by the EPA.

Start with the prototype shown above, for a plug-in hybrid car (PHEV) capable of running on both gasoline and electricity. A larger image of this sticker is available on page 4 of the PDF file on EPA's site. The two features that stand out in this design are the letter grade, which compares the fuel economy of the stickered vehicle to all others on a scale running from A+ to D (further conveyed in a green to amber color range) and the estimated 5-year fuel cost savings compared to the average new car. I've seen comments elsewhere suggesting that the graded comparison should be based only on comparable vehicles, rather than all cars, and there's some merit to that. There might be a few folks out there looking to replace a Chevy Suburban with a Nissan Leaf, but I'll bet they're in the minority. At the same time, the goal of the program is to reduce fuel consumption and emissions, and people are generally smart enough to figure out that if the vehicles that meet their needs only range from B- to a D, the B- choice will probably cost less to run and be better for the environment.

Of course that raises questions about whether the letter-grade system dumbs-down the whole process and diverts attention from details that actually matter a great deal, including the assumptions underlying the system, many of which are displayed in the fine print. One of the biggest of those is that only emissions from the tailpipe count. That sounds like a technicality, but when the result is that vehicles powered directly by electricity are guaranteed to get most of the A's in the class, it has serious consequences. For example, it would skew the comparison between an only-moderately efficient PHEV and an extremely efficient conventional hybrid (non-plug-in.) The latter might produce fewer lifecycle GHG emissions than a plug-in running on the electricity mix of the applicable regional grid, but because its emissions mainly come out of the tailpipe, it would be at a disadvantage. This kind of comparison is only one aspect of the emerging transportation energy market for which any static, national-level representation such as a sticker plastered on a car window seems likely to be wrong more than right.

The new stickers also introduce several new concepts to motorists, including the MPGe, or mile per gallon equivalent--a worthy evolution of mpg. This provides a handy way to compare the energy content of different fuels, including electricity, to the standard energy content of a gallon of petroleum gasoline, approximately 115,000 BTUs. The problem is that in the case of electricity, the stated conversion rate of 33.7 kWh per gallon-equivalent ("eGallons in the stickers' parlance) grossly understates the energy required to produce most of the kWhs on the grid. It's only accurate for the 31% of our national electricity mix attributable to nuclear, hydro or other renewables. In the case of electricity from natural gas turbines, it can understate the fuel requirement by much more than half--and thus overstate electricity-based fuel efficiency by more than double. In other words, an accurate comparison of the equivalent fuel economy of an electric vehicle requires more information about the source of electricity for each consumer than any sticker could conceivably collect. I'd prefer to see a more conservative conversion rate for electricity, such as 14.4 kWh/gallon (based on a typical gas turbine heat rate), but at a minimum the CO2 grams per mile figure on all such stickers should be asterisked along the lines of, "Your emissions will vary depending on your local electricity source."

Another new concept incorporated in the sticker for PHEVs is the notion of "Blended Electric + Gas" fuel economy. Unfortunately, while it provides a basis of comparison among similar vehicles, its accuracy is limited to the precise combination of electric and non-electric miles that went into the EPA's calculation, accentuated by the kWh/eGallon problem described above. Drive mostly electric miles and just a few on gasoline, and you'll get a result like what I saw when I drove the Chevrolet Volt last winter. I much prefer the alternative sticker shown on page 12 of the PDF, in which consumers are given the fuel economy on gas only and electricity only and allowed to work out the likely result for their specific circumstances. I also like the range graphic on this PHEV sticker and the electric vehicle sticker on page 11, allowing an easy comparison between those two vehicle types.

Yet while the EPA is clearly working hard to provide consumers with more information about the performance of vehicles that can use combinations of electricity and liquid fuels, I find it inexplicable that their proposed sticker (page 15) for flexible fuel vehicles (FFVs) would display the fuel economy only for gasoline, rather than for both gasoline and E85 ethanol, since the latter tends to be a quarter to a third less, based on the EPA's own results for actual FFVs.

The last issue I want to raise relates to the assumptions underlying the annual fuel costs and savings shown on all the stickers. When the first stickers were introduced in the 1970s, there was no easy way to convey to consumers up-to-date information on current and expected future fuel prices. That's certainly no longer true, and posting cost estimates relying on the assumption that we all pay the same price for gasoline and electricity and will do so for the life of a car makes little sense. Why not omit this information and replace it with a link to an interactive website that, with the input of just a zip code, could determine local fuel and electricity prices and calculate future savings based on those and the latest forecasts from the Department of Energy?

While I commend the EPA for its effort to make alternative fuel vehicle characteristics more understandable and for making these proposed stickers public now, I believe the agency is attempting to over-simplify a truly complex set of parameters and relationships, at the risk of inadvertently misleading a significant number of purchasers. Consumers would get more reliable value from stickers that provided them with just a few clear metrics, plus access to the information needed to work out how the vehicles among which they are choosing would be likely to perform in their circumstances of where and how they drive. And unless the basis of the letter grades can be expanded to include lifecycle emissions, rather than just those from the tailpipe, they should be jettisoned as fundamentally flawed. It will be very interesting to see what emerges from the next 60 days of public comment, and I encourage my readers to put in their two-cents worth.

No comments: