John Voelcker’s article ”How Green Is My Plug-In?” delves into the carbon impact of plug-in hybrid electric vehicles (PHEVs), and it generated quite a bit of discussion. After that article appeared in March 2009,IEEE Spectrum’s David Schneider spoke with Jeremy Michalek, an assistant professor of mechanical engineering and of engineering and public policy at Carnegie Mellon University (CMU). Michalek’s upcoming study in Energy Policy looks at the sizing of batteries for such cars; in it, he calculates that plug-in hybrids with large battery packs—like the 2011 Chevrolet Volt—may never save consumers any money.
The Electric Power Research Institute (EPRI) has a different view of the future than Michalek does. Here, in an interview with Voelcker, Mark Duvall, director of electric transportation at EPRI, explains the results of his organization’s detailed analyses of the environmental and energy impacts of PHEVs—which were conducted with the Natural Resources Defense Council (NRDC)—and relates some of the experiences developing models of the impact of plug-in hybrid technology.
IEEE Spectrum: What was your goal in doing the EPRI-NRDC study?
Mark Duvall: We wanted to conduct the first comprehensive study of the environmental impacts of plug-in hybrid vehicles—both for greenhouse-gas emissions and air quality. To really analyze these issues, you need sophisticated and detailed models of the electricity sector to understand which power plants would be used to generate the electricity to charge PHEVs.
Spectrum: Did that study assess the costs of cutting carbon using plug-in hybrids?
MD: That’s very difficult to do for vehicle technologies. While transportation as a whole is a large emitter of greenhouse gases, each vehicle saves a very small amount of carbon [one liter of gasoline produces about 3 kilograms of CO 2 ; 1 gallon, about 11 kg]. If you don’t get all of the costs exactly right—especially the purchase cost of the vehicle—you get really large variations in cost-effectiveness, all the way from saving money by cutting carbon to costs of US$ 1000 or more per ton of carbon. Our work indicates that PHEVs, once the technology matures and they attain volume production, can be manufactured and sold for a reasonable price and that owners can expect to save money over the useful life of the vehicle.
Spectrum: One area where you differed with the Carnegie Mellon study was the cost of batteries, which it pegged at $1000 per kilowatt-hour. How do you view that estimate?
MD: We agree that until battery costs reach some more reasonable level (for example, $500/kWh), no plug-in hybrid can save enough fuel to cover its higher production cost. However, we’ve done quite a lot of cost modeling. Once batteries drop to the $500/kWh range, our studies indicate that plug-in hybrids with 10, 20, and 40 miles of range will all save money over their useful life—and this was with gasoline at about $3 per gallon. If gasoline costs more, as many analysts seem to expect, then the savings increase.
Spectrum: When do you see batteries dropping to that level?
MD: We are seeing some indications that battery costs for high-volume orders may be below $1000/kWh already. With the large number of production PHEVs now planned, and the current focus on developing cell-manufacturing capacity, I expect that we will see these price levels by the time the second generation of PHEVs arrives, in 2015 or so.
Spectrum: Are there other cost components to plug-in hybrids besides the battery pack?
MD: Absolutely. The electric motors, power electronics, and electric accessories are also pricey, and every one of those components is needed in some form for any type of hybrid or plug-in hybrid. This is why our cost studies found that shorter-range PHEVs didn’t have the best economics. The ”sweet spot” appears to be about 20 miles for typical driving patterns. But if you habitually drive longer distances, you might want more electric range, and vice versa.
Still, battery cost is the primary bogey. The CMU study looked at variations in cell cost and state-of-charge range (the amount of a pack’s total energy capacity used, which is generally maintained within limits—30 percent to 80 percent for example—to avoid overstressing the battery), but not at the combination of those changes. That study’s assumptions—$1000/kWh, and a 50 percent state-of-charge range—correspond fairly well with what we expect first-generation plug-in vehicles to offer. That cost is almost entirely due to the limited volume and conservative approaches used by the automakers when launching new technologies.
Once production volumes increase and the industry feels more certain about the real-world performance of the batteries, we will likely see cost fall to $500/kWh and the state-of-charge range rise to as much as 80 percent. Those changes will cut vehicle mass to less than the most lightweight case in the CMU study. These factors really have to be considered together, not separately, to get an idea of where things are headed.
Spectrum: What effect does the larger battery pack’s higher mass have on a plug-in with greater electric range?
MD: Adding mass to a vehicle clearly reduces its efficiency somewhat. However, with plug-in hybrids, a larger battery comes with a more powerful, more capable electric drive system. Among the improvements accompanying the larger battery are higher system efficiency, greater regenerative braking capacity, and higher average engine thermal efficiency—all of which raise the car’s overall efficiency. We have a lot of detailed modeling work to show that even in ”charge-sustaining mode” (where gasoline provides all the energy), plug-in hybrids hold their own or do better against today’s hybrids. And to some extent, plug-ins with larger batteries can hold their own against systems with smaller batteries. So to assume that the larger battery of a PHEV or EREV (extended-range electric vehicle) will significantly impact electric energy efficiency is simply not the case.
Spectrum: The CMU study seems to indicate that even with gasoline at $6 a gallon, a plug-in hybrid with 7 miles of electric range will save its owner money, but the 40-mile range of a car like the 2011 Chevrolet Volt never will. How do the conclusions of your study differ, and why?
MD: We disagree on the relative costs and attributes of a ”mild” (low-electric range) PHEV against a vehicle like the Chevrolet Volt. EPRI has a detailed analysis on PHEVs with 10 miles (PHEV-10) and 20 miles (PHEV-20) of electric range; for most drivers, the PHEV-20 wins on cost at gasoline prices of $3 a gallon and higher. This is because the PHEV-10 has all the costs of the PHEV-20—a hybrid drive system, electric accessories, regenerative braking system, et cetera—but its battery is slightly more expensive and heavier per kilowatt-hour because it still has to produce the same power level.
Spectrum: To jump-start the market for plug-ins, the U.S. government will grant billions of dollars in tax credits to consumers who buy vehicles with packs in the 4-kWh to 16-kWh range. When will these plug-in vehicles be able to compete without subsidies against liquid-fuel cars?
MD: We anticipate that by 2015 or so, around the time the government subsidies expire, there will be a robust and mature market for all types of PHEVs, EREVs, and even battery-only electric vehicles.
Spectrum: What do you consider to be the most cost-effective way to displace gasoline in U.S. transportation?
MD: There are numerous studies in this area. The answer is probably somewhat boring, like driving less, driving slower, or introducing minimum tire-efficiency standards.
But reducing petroleum consumption to any reasonable level will require us to use every technology we have that can be made cost-effective. Hybrids, plug-in hybrids, and electric vehicles are just a few of the options we need to develop if we are to have any hope of making progress on this issue. But we don’t see them as competing with other options—biofuels, public transit, et cetera—because we’ll need all of those as well.
Spectrum: Michalek’s advice to consumers was ”Buy small, charge often.” What’s your thought there?
MD: This is a terrible idea. Frequent charging of a smaller battery will wear it out very quickly. Market research shows that vehicle owners expect the battery to last the life of the car. This makes sense, considering that we’re now used to 100 000 to 150 000 relatively trouble-free miles from a new car. The more frequently you charge and discharge a battery, the shorter the time until you have to replace it.
For example, let’s say I have a battery that lasts for 4000 deep cycles—a challenging goal, but one that our research and testing in this area leads us to believe today’s battery technology can reach. If I buy a 7-mile PHEV and drive it exactly 7 miles between charge cycles, I will need to replace the battery in just 28 000 miles. No one will be very happy with that. In fact, we have market studies that show consumers would be unhappy.
But if I buy a 40-mile PHEV and drive it exactly 40 miles between charges, I’ll have 160 000 miles on it when the battery wears out. And the difference in battery life is even greater, since a 40-mile battery will make a lot of trips of fewer than 40 miles, with correspondingly less discharge, whereas most drivers are likely to completely discharge a 7-mile battery on almost every trip.
One question I would add: Is it worthwhile to go past 40 miles of electric range? A little less than 80 percent of U.S. drivers travel fewer than 40 miles per day. Also, if you charge a Chevy Volt at work, it is roughly equivalent to a battery-only electric vehicle with an 80-mile range. Several years ago, we analyzed a PHEV with 60 miles of range and found the savings from 60 miles is not that much higher than 40 miles—unless you have a really long commute.
In real life, most of these vehicles will run on a mix of electricity and gasoline. Otherwise, they’d be electric vehicles! But it’s vital to note that the smaller the battery is, the harder it becomes to make it last the lifetime of the car. That’s why a plug-in with 7 miles of range probably won’t offer a lot of electric-only range. It will operate in hybrid mode most of the time, blending gasoline and electricity. So it will drive a lot like your current hybrid, only with better fuel economy. If you want real electric range, even for around-town driving, you realistically need the equivalent of at least a 15- or 20-mile battery.
Spectrum: In less than two years, we will see first-generation plug-in hybrids available for sale in the United States. What are the biggest unknowns for future generations of plug-in technology?
MD: We are primarily concerned with how fast the market will grow. I would be happy with a growth rate similar to hybrid vehicles, which have reached a total of 1 million sales in the United States after eight years—though I would prefer we get there quicker. This will require lots of manufacturers entering the market, strong customer demand, and a variety of vehicles to choose from. I’m not worried about the vehicles themselves; they will be very, very good.
Spectrum: Any last comments?
MD: The real strength of the plug-in hybrid concept is that the battery can be sized to accommodate the daily driving patterns of most users. The data tell us that more than 70 percent of U.S. drivers cover fewer than 40 miles a day. There is no one ”right amount” of electric range, because everyone has different needs and price sensitivities. Remember that when hybrids were first introduced, most analysts complained that none of them made economic sense. Now nearly all of them do. We expect the same thing to happen for plug-in hybrids with 10 to 40 miles of range.
In the end, people will buy plug-ins because they want them, and the most important thing automakers can do is give them reasons to buy them. It is more important to get well-engineered plug-in vehicles into production as soon as possible and get them out there than to worry about what amount of range is best.