You Asked, We Answered: Response to 2013 Cars Report
By Daniel Yawitz, Alyson Kenward, and Eric Larson
In the month since we released our report, A Roadmap to Climate Friendly Cars: 2013 we have received many good questions and insightful comments from readers. Thank you to everyone who has taken time to give us feedback. We especially appreciate those who recognized the inherent challenges involved in carrying out and presenting lifecycle analyses. Our research team has reviewed all of the reader responses, and here we offer some additional thoughts and clarification to address those comments.
Several readers thought that for gasoline we considered only the emissions from combustion and not the full lifecycle emissions (including emissions during oil extraction, transportation, and refining). As we wrote in the report, we did consider the full lifecycle emissions for gasoline. To do otherwise would not be an apples-to-apples comparison.
One reader thought we had used a 20-year GWP when assessing electric vehicle emissions and a 100-yr GWP for assessing gasoline vehicles. In retrospect, our report is not clear on this point, so the confusion was understandable, but in fact we applied 20-yr GWP values for ALL vehicles analyzed. To do otherwise would have been an unfair analysis.
Several noted that for plug-in hybrids we assumed that these cars drive half their miles on gasoline and half on electricity, but you also believe that many plug-in owners don’t drive their cars this way. Driving behavior certainly does make a difference in total emissions generated while driving plug-in hybrids. However, we are not aware of good data that show on average what the mileage splits are on electricity and on gasoline for different model PHEVs, so we decided to keep it simple. But, in our 2012 report on climate-friendly cars, we did evaluate the impact on emissions of different driving behaviors with the Chevy Volt, and we referred readers to that report to get some idea of how the driving splits affect emissions.
One reader said our report was confusing and, apparently, self-contradictory. Aside from a couple of typos (which we have fixed), it is self-consistent, but admittedly we could have been more explicit when distinguishing between what we called emissions associated with “driving and charging” a car and the “lifecycle” driving emissions. The latter is the appropriate one for comparing the overall climate-friendliness of vehicles. We presented the former to help highlight the differences in electricity generation mix between states.
Several pointed out that as the grid mix changes in the future, so too will the emissions of electric vehicles. Absolutely. And we hope that the grid mix will evolve to lower and lower carbon sources. In our report we noted this was the trend from 2010 to 2012, but we also noted that coal use was on the rise again in early 2013. Since it is difficult to make predictions (especially about the future), we decided to present a “snapshot.” We will show another snapshot next year, and we hope that will show further improvement.
- Several mentioned a lack of uncertainty analysis. We acknowledge that we could have more clearly illustrated the sensitivity of lifecycle emissions to different input assumptions.
For example, some readers pointed out manufacturing emissions for the battery as a particularly important parameter. There is a wide disparity in the literature on estimates of battery manufacturing emissions. There are also not very many detailed and well-documented studies published on this. We chose to base our analysis on Hawkins' work (including a correction he published this year on his earlier, better known, work) because we found his analysis quite detailed and well documented. But we also did some internal calculations to check the sensitivity of our findings to the assumed battery manufacturing emissions. We found that they have a relatively small impact on the Leaf vs. Prius state-by-state rankings. Moreover, the impact of the battery manufacturing emissions decreases as the assumed lifetime driving miles increases, as one would expect.
Several noted that the time of day when a battery is recharged from the grid will have an impact on emissions. We acknowledged as much in our report, but we decided to use annual average emissions from electricity generation because we did not find data that would allow us to say with much confidence what the average charging behavior of the average electric car driver is.
A few readers noted that if charging is done from a solar panel, emissions would be very low. We agree, but chose not to evaluate the “off-grid” option, since there are insufficient data on how many electric car owners recharge in this manner. Also, as the number of electric cars on the road increases, an increasing fraction will be charging from the grid, so we felt grid-connected recharging was of broader interest and relevance.
There were additional comments about our not showing lifecycle emissions for lifetime car mileage above 100,000 miles. We are not aware of good data on what actual EV car lifetimes are, probably because there is not enough experience out there yet to say with much confidence. This is why we limited ourselves to 100,000 miles. However, we did note that emissions per mile for electric cars improve the further they are driven, and we also mentioned that if a car lasts for 200,000 miles without the need for a battery replacement, we find that the best EV (Honda Fit) is more climate friendly than the best hybrid (Toyota Prius) in 26 states (and, for those interested, at 200,000 miles, the Nissan Leaf is more climate friendly than the Prius in 23 states).
At least one reader pointed out that despite disagreements you have with some of our assumptions, an important underlying conclusion of our analysis is that for electric cars to improve their carbon footprint, lower-carbon grids are needed, and emissions associated with manufacturing the vehicles is needed. We agree.
- Another reader asked, “What is the right number for CO2 equivalents produced by the Prius?” and proceeded to calculate three different answers starting with numbers taken from different tables in our report. In fact, there is only a single number, which we can arrive at via all three of your calculation methods, if we correct the mistakes you made. Here are our corrections:
Emissions for Prius in Table 4 are 0.67 lbsCO2e/mi (not 0.73 that you used). This converts to 304 g/mile.
We agree with your calculation #2, which gives 306 g/mi.
- The Prius fuel economy according to EPA (combined city/highway drive cycle) is 50 mpg, which we show in Table 2 (you used 46.3 mpg for your calculation). Also, the vehicle manufacturing emissions (15,627 lbsCO2e) should be divided by 100,000 miles (not 120,000) to put it on the same footing as your calculations #1 and #2. With these corrections, calculation #3 gives 306 g/mi.
- Finally, it was noted that carbon offsets could be purchased from TerraPass to eliminate the “carbon debt” due to manufacturing of the battery of an electric vehicle. Presumably this could be done to offset a vehicle’s driving-related emissions, as well. However, if everyone just buys offsets and continues to drive carbon-emitting cars, it will not accomplish much toward solving the climate change problem. A low-carbon grid and less polluting manufacturing processes are needed for that.