For one thing, the electricity grid is getting cleaner every day.
A post I wrote last week on where electric vehicles cause more pollution than gas cars produced some unusually high reader emissions rates. Amid the more combustible reactions were a number of quite thoughtful ones, laying out some legitimate critiques to both the research study that prompted the post as well as our description of the work. So we reached out to several experts on electric cars and the environmental impact of transportation to offer a wider perspective.
First a quick recap: the NBER study I originally referenced measured pollution produced by EVs (via the electricity grid) and gas cars (via on-road emissions) across the U.S.—in effect, comparing power plant smokestack to vehicle tailpipe. In the West, EVs tended to be cleaner than gas cars; elsewhere they rated out worse. The economists leading the work priced this difference by geography to determine where electrics should be subsidized, and where they should actually be taxed.
The maps came down hardest on the worst environmental offenders (namely, gas cars in big cities, and coal power plants) and served as a reminder of the huge social costs of car reliance. But the researchers, concerned foremost with present-day policy, also meant them to “generate critical questions as to the merits of the federal subsidy” for EVs. They concluded that maybe it made more sense to set specific pollution fees or subsidies based on different locations.
In our follow-up discussions—a sort of informal peer-review for a working study yet to receive its official one—four key qualifications to that conclusion emerged.
1. You can draw different policy conclusions from similar findings
By far the biggest limitation of the study (acknowledged in both the original post and the paper) was that it didn’t perform what’s known as a lifecycle analysis. In addition to measuring tailpipe and smokestack emissions, a lifecycle analysis includes the other steps required to get an automobile in motion—often called “upstream” emissions. These are things like oil extraction, refinement, and distribution on the gas car side, and battery production on the EV side.
At least one recent lifecycle analysis, published late last year in PNAS, reached similar findings about the relative air quality impact of different types of vehicles. EVs running on electricity generated through coal produced more air pollution than conventional gasoline and hybrid cars, while those running on electricity generated by cleaner energy sources—such as natural gas or renewables (wind and solar power)—produced less.
That general story held true whether the researchers focused on tailpipe-versus-smokestack emissions or included the full lifecycle. Out of 11 vehicle types, coal EVs ranked last on both counts, conventional gas cars ranked toward the back (7th for lifecycle and 8th for tailpipe), and natural gas EVs ranked second best for both. If anything, “upstream” emissions accounted for a greater share of the air pollution impacts for EVs than for gas cars. (See Figure S3 for more.)
“I’ve sort of stopped referring to electric vehicles generically,” says study co-author Julian Marshall, an environmental engineering scholar at the University of Minnesota and co-director of the school’s Acara program, which develops business ideas to address environmental problems. “It depends where the electricity comes from.”
Marshall sees some strengths in the new NBER study; in particular, he says, the finding that some states end up exporting their pollution impact to other states, by drawing on a remote electricity grid, is “a really good point.” He also sees some big flaws, including that emissions rates are several years old and therefore don’t reflect recent improvements (more on this point below)—a shortcoming shared by many grid analyses, including his own, given data limitations.
But when it comes to EVs, Marshall interprets his own data in a more encouraging light than the authors of the NBER study interpret theirs. After all, the very best vehicles in his group’s analysis were electrics—albeit only those powered by natural gas or renewables. Instead of seeing that as a reason to tax coal-powered EVs, perhaps policymakers should follow Minnesota’s lead and subsidize programs that help these drivers buy cleaner electricity. It’s easier to clean up one grid, he says, than a million gasoline engines.
“When we looked for what are the solutions, what are the opportunities to significantly reduce the health impacts for transportation, they’re EV,” he says. “If you want a really clean transportation system it’s probably going to involve electric vehicles.”
2. Other lifecycle reports have given EVs the edge
David Reichmuth, senior engineer in the Clean Vehicles Program of the Union of Concerned Scientists, a science advocacy group with a climate focus, also questions whether or not it’s possible to make an honest comparison about vehicle emissions without considering a full lifecycle. “And both gasoline and electricity have these upstream emissions,” he says. “It just makes it hard to tell the full story when you leave out some of these emissions.”
In its own lifecycle analyses—one conducted in 2012, and a follow-up in 2014—UCS used a slightly different approach from the new NBER study. Instead of dividing the country into states or cities, UCS analysts broke it down by 26 electricity grid regions across the U.S. It also tracked the global warming emissions that come from the production of gas and electricity, as well as from driving.
In line with the NBER study’s findings, the UCS reports indeed found that some electric grids are cleaner than others. “Both our results and the results of others have shown that the emissions from an electric vehicle does depend on where you get your electricity,” says Reichmuth. “There are variations throughout the country.” Once again California fares well, with lifecycle EV emissions there comparable to driving a car that gets 95 miles a gallon.
But overall the UCS reports suggest there’s really nowhere in the country where EVs don’t stack up favorably to conventional cars when it comes to lifecycle pollution. The 2014 report finds that an EV charged in every grid area in the U.S. produces less global warming emissions than the average, 28-mpg gasoline car. And it shows that 60 percent of the country lives where EVs produce less global warming emissions than the best-performing gas car, a 50-mpg Prius.
That last figure is up from 45 percent in the 2012 report—a sign the grid has improved (and can improve) even in a brief time window.
“The key is that, from a global warming perspective, we see EVs today as being cleaner than gasoline cars, and definitely as we switch to cleaner sources we’re only going to see that comparison become more in favor of electric vehicles,” Reichmuth says. “Fundamentally, this is about making an investment in transforming to a cleaner transportation system.”
3. The electricity grid is already getting cleaner
As mentioned above, another caveat to the new NBER study is that it uses power plant emissions measures from 2010 to 2012. Again, to some extent that’s unavoidable, especially if you want accurate, fully vetted data. The latest version of EPA’s official eGRID data is from 2010. Scientific databases move slowly.
But the problem with using data from several years ago is that the electricity grid has been undergoing a series of extremely recent and rapid changes, says Marcus Alexander, an electric transportation analyst for the Electric Power Research Institute. Cap-and-trade programs, renewable energy portfolios, new (though legally challenged) federal mercury and air toxics standards, and the retirement of coal plants have all exerted “downward pressure” on emissions, he says. It’s no coincidence that coal represented half of America’s fuel generation a decade ago but today makes up under 40 percent.
Alexander has several more (admittedly highly technical) problems with the way electricity was measured in the NBER study. By rolling emissions into one monetized value, he says, the paper made it difficult to trace the source of any single pollutant. That’s critical because the work itself suggests non-carbon emissions are the big culprit for electrics, though it’s not clear which specific ones. And the new research estimated emissions with a “small-scale marginal” technique that minimizes the impact of significant policy changes, he says.
He prefers to measure the electricity grid as an “integrated system” rather than a series of individual power sources. This captures the intricacies of how the system turns up Technology X (say, renewables) and turns down Technology Y (say, coal), depending on things like changes in load and new regulations. Failure to look at the grid this way, he says, can result in an electricity model that doesn’t reflect adjustments made to remain in compliance with pollution laws—and therefore can produce unrealistically high emissions rates.
That doesn’t deny clear geographic differences in grid cleanliness. “There are some regions of the U.S. that are better than others,” says Alexander. “That will be a significant challenge going forward—trying to make sure everywhere is as good as it can be.” But he characterizes that variation as “between good and better” for electrics, and says EVs already have an advantage more or less across the board when compared against gas vehicles, at least according to EPRI’s own calculations.
“If we just focus on greenhouse gas emissions, we’re essentially already at the point where the grid is competitive or better than every conventional technology available,” he says.
4. The future is far brighter for EVs than for gas cars
One more point made against the new research is that in focusing on present-day policy it ignores the future. In acknowledging that point, study author Stephen Holland noted that conventional gas cars have made huge strides in fuel-economy. That’s true, but it also means people basically have to buy a new gas car to get a cleaner gas car. If you own an EV for a decade, on the other hand, it has the potential to get cleaner every year as the electricity grid gets cleaner—in a sense, updating itself in real-time.
“Only electric cars get cleaner as you drive them,” says Luke Tonachel, senior analyst and director of the Clean Vehicles and Fuels Project for NRDC. “We’re not going to meet our climate goals if our transportation continues to be monopolized by petroleum.”
One of the new pollution standards that Tonachel expects to have a big impact is the EPA’s Clean Power Plan, expected to be finalized this summer, which specifically targets carbon emissions from power plants. (The plan is not without its detractors.) But a problem Tonachel has with the new NBER research is that even before the power plan gets working, the maps struck him as “inconsistent” with current power plant emissions rates—specifically those in the Northeast, which he calls “actually fairly close to those in California.”
In support of that statement he points to an eGRID chart of 2010 greenhouse gas outputs. Indeed, the California regional grid fares well in this table, emitting roughly 610.8 pounds of carbon per megawatt hour. But the New England grid comes close (722 pounds) and the Upstate New York grid actually does better (545.8 pounds). NRDC’s own reports echo these trends. “We know from a carbon perspective, for example, that the Northeast does not have these kinds of emissions profiles that the map shows,” says Tonachel. “Typically, if they’re lower carbon, they’ll be lower in other emissions as well.”
New York recently announced a target to get half of its power from renewable energy sources by 2030. Moving forward, there’s reason to think states can do even better than getting half their energy from renewables. Some new research from Stanford’s Mark Jacobson and UC Berkeley’s Mark Delucchi argues that it’s “technically and economically feasible” for every U.S. state to convert its energy system to 100 percent clean wind, water, or solar energy by 2050—and outlines a plan to get each state there.
Even the new NBER research quietly recognizes this bright future for EVs. Buried in the paper’s final table is an analysis that estimates the potential environmental benefits for both gas and electric cars, assuming all coal power plants are replaced by natural gas power plants at some point in time. When that day does come, they report, environmental damages from both types of vehicles are lower—“and damages from electric vehicles are much lower.”