New Data Show Electric Vehicles Continue to Get Cleaner

, senior engineer, Clean Vehicles | March 8, 2018, 10:48 am EDT
Bookmark and Share

New data from the US EPA on power plant greenhouse gas emissions are in, and electric vehicles (EV) in the US are even cleaner than they were before. The climate change emissions created by driving on electricity depend on where you live, but on average, an EV driving on electricity in the U.S. today is equivalent to a conventional gasoline car that gets 80 MPG, up from 73 MPG in our 2017 update.

Cleaner electricity means cleaner EVs

Based on data on power plant emissions released in February 2018, driving on electricity is cleaner than gasoline for most drivers in the US. Seventy-five percent of people now live in places where driving on electricity is cleaner than a 50 MPG gasoline car. And based on where people have already bought EVs, electric vehicles now have greenhouse gas emissions equal to an 80 MPG car, much lower than any gasoline-only car available.

To compare the climate-changing emissions from electric vehicles to gasoline-powered cars, we analyzed all of the emissions from fueling and driving both types of vehicles. For a gasoline car, that means looking at emissions from extracting crude oil from the ground, getting the oil to a refinery and making gasoline, and transporting gasoline to filling stations, in addition to combustion emissions from the tailpipe.

For electric vehicles, the calculation includes both power plant emissions and emissions from the production of coal, natural gas and other fuels power plants use. Our analysis relies on emissions estimates for gasoline and fuels production from Argonne National Laboratory and power plants emissions data recently released by the US EPA.

EVs getting cleaner over time

An important difference between EVs and conventional cars is that existing EVs can get cleaner—and, over time, they are getting cleaner. It’s difficult to make burning gasoline cleaner, and electricity is trending cleaner over time as we shift away from coal and add more renewables. This means that EVs that were sold years ago can run much cleaner than when they were purchased. Our initial analysis of EV emissions used data from 2009, while this update incorporates 2016 data. By switching between these two maps, you can see the improvement made in many regions of the US.



More efficient EVs now available too

The maps shown above are based the efficiency of the average EV. However, there are now options on the market that are even more efficient. Using one of these more efficient EVs (Hyundai Ioniq BEV, Prius Prime, and Tesla Model 3) means lower emissions. With these cleaner EVs, 99 percent of the country is in a region where electricity emissions would be lower than a 50 MPG gasoline vehicle.

How do other EVs compare? Use our EV emissions tool to estimate the emissions from a specific EV in your area.

The most efficient EVs are much cleaner than even the best gasoline cars in many regions of the US. Currently the most efficient EVs are the Hyundai Ioniq BEV, Tesla Model 3, and the Toyota Prius Prime (while operating on electricity).

A trend that’s likely to continue

Electric vehicles produce less emissions now because the electric grid is getting cleaner. Over the last ten years, the fraction of power from coal has fallen from nearly 50 percent to 30 percent. Over the same time, utility-scale renewable power like solar and wind power have grown to make up 10 percent of electricity generation.

This analysis relies on data from power plants for 2016, the most current data that includes details on the geographic location of emissions. However, based on the overall data on from 2017, it looks like emissions will continue to fall, with both coal and natural gas declining while renewable power continues to increase.

The falling emissions from electric power over the last decade also highlights the need to work to clean up electricity generation and transportation now. While we are moving in the right direction with renewable power and growing numbers of EV models, it takes time to replace existing power plants and gasoline cars. It’s vital that we accelerate the adoption of EVs, even if all power is not yet from renewable or low-carbon sources.

Utility-scale electric power generation. Power from coal has dropped over the last decade and clean renewable power has increased. Data Source: US Department of Energy, Energy Information Agency.

Posted in: Vehicles Tags: , , ,

Support from UCS members make work like this possible. Will you join us? Help UCS advance independent science for a healthy environment and a safer world.

Show Comments

Comment Policy

UCS welcomes comments that foster civil conversation and debate. To help maintain a healthy, respectful discussion, please focus comments on the issues, topics, and facts at hand, and refrain from personal attacks. Posts that are commercial, self-promotional, obscene, rude, or disruptive will be removed.

Please note that comments are open for two weeks following each blog post. UCS respects your privacy and will not display, lend, or sell your email address for any reason.

  • Scott Williams

    Interesting results in the Midwest – Eastern Wisconsin is actually worse in 2016 than in 2009. I suspect this is because of the closure of Kewaunee nuclear plant (which basically offset all the carbon emissions reductions from the renewable portfolio standard). Similarly, I was scratching my head as to why West Virginia and Indiana would be better than Wisconsin, since they are more coal dependent. But they get lumped in with Chicago/northern Illinois, which also has a lot of nuclear power.

  • Cosmo Kramer

    This is really a great article and it seems very well vetted. Thank you. I’m wondering if you have done any analysis of the vehicle “fuel system” cradle to grave GHGs as well. I realize it was not part of this analysis, but I see different reports and articles on the carbon emissions of Li-ion battery production as potentially offsetting years of operation with lower GHG in-use emissions. Articles and reports seem to be all over the place on this topic and your fair approach to the information you’ve provided here makes me interested in your thoughts on battery production effects. Thanks again. Great article.

  • Yucca

    Small error: SRMW should be purple, not blue. Or the number shown is incorrect.

    • Thanks for catching that! The MPG rating is correct, so the color should be purple. The map should be updated soon.

  • Jerome Carman

    Excellent article. One criticism – this could be made more of an apples-to-apples comparison by correcting for the following electricity losses and estimating the following additional sources that are not included in the eGRID estimates:

    “Plant level emissions in eGRID reflect a combination of monitored and estimated data. Emissions and emission rates in eGRID represent emissions and rates at the point(s) of generation. While they do account for losses within the generating plants (net generation), they do not take into account any power purchases, imports, or exports of electricity into a specific state or any other grouping of plants, and they do not account for any transmission and distribution losses between the points of generation and the points of consumption. Also, eGRID does not account for any pre-combustion emissions associated with the extraction, processing, and transportation of fuels and other materials used at the plants or any emissions associated with the construction of the plants.” (Pg. 15 of the eGRID Technical Support Document:

    Because GREET considers emissions from the well to the wheels I think it would be fairer to attempt to apply the same inventory boundary to electricity as well.

    If I assume an additional 20% of emissions associated with the above sources/losses for the electricity sector (unsupported rule of thumb I use based on my own experience), and assume that I can directly scale your MPGe estimates by 1/1.2 = 0.833, your overall message still applies.

    Just a thought. Overall this is a great analysis, and I love the approach of presenting in a common unit of MPGe.

    • Jerome Carman

      Ignoring emissions from power plant construction, as I believe GREET also does not consider these emissions.

    • Hi Jerome-
      Thanks for the comments. For our analysis, we did consider transmission losses, as reported by eGRID, as well as GREET 2017 upstream factors for coal, NG, biomass, fuel oil, and nuclear power. The factors range from 9 to 167 g/kWh and make up about 6-17% of the total g/kWh emissions (depending on mix in the region). Transmission losses vary regionally, around 4-5%.

      Our methods are listed in our last full report:

  • Ed

    Excellent report. We need to somehow reduced this to a single statement to summarize for the non-believers who frequent the comment sections. Is it “Today’s electric vehicles already achieve the emissions equivalent of 80 miles per gallon, and this is increasing as more renewable energy joins the grid.”?

    I also hope to see Tesla advertising like this someday, taking advantage of their combined businesses:

    • neroden

      Hmm. “An EV is now always cleaner than a gasoline car, and it gets cleaner every year”?

      • Ed

        A master of simplicity, neroden!

  • timclaes

    Hi, did you publish a methodology page? I assume you took the average generation mix to calculate the mpg equivalent. I wonder if you took into account the fact that most EV’s get most of their energy from charging at home, at night. At that point, mostly or only base load is running, which might have a different carbon footprint (more nuclear, coal and perhaps wind?)

    • neroden

      It’s completely impossible to figure out marginal generation. Average generation for the grid subregions is the only thing which makes any sense. A bunch of the subregions are approaching 100% renewable energy anyway — the marginal generation is renewable.

      If you’re using “base load generators” like nuclear, then that power is essentially free, you’re using no energy at all: “base load generators” have to operate whether or not there is any demand, and they just have to turn the energy into heat if there’s no demand.

      • guest9

        Time of charging is incredibly important. And I suspect the above analysis did not consider it. You CAN determine what type of plant is operating at the margin. It’s been done by eGRID region in several studies. To the extent that EVs charge at night in regions where coal base load is operating, e.g., they are creating a source of demand for coal-fired power. This is important. EVs are promising—yes. But charging during peak renewables periods is the key.

      • “they just have to turn the energy into heat if there’s no demand” is completely false.

        Google “Pumped-storage hydroelectricity”, a technology that’s been in use for _decades_ to store excess energy from power plants that are uneconomic to throttle down so that the energy can be time-shifted to when it is needed. The Wikipedia page has a beautiful description for example stating “Pumped-storage hydroelectricity allows energy from intermittent sources (such as solar, wind) and other renewables, or excess electricity from continuous base-load sources (such as nuclear) to be saved for periods of higher demand.”

        The up and coming complement to that is battery storage. Google “south australia tesla battery” for an example.

    • timclaes

      could the author comment please?

    • Hi- The methods in this update are the same from our most recent report, ‘Cleaner cars from cradle to grave’ ( ), with the following changes: Upstream emissions now use GREET 2017 data, electric emissions are from eGRID2016 (Feb 2018 release), and the average EV electric drive efficiency for vehicles sold (2011-2017) is now 0.338 kWh/mi (slightly worse than last year).
      We use the ‘total’ (average) emissions rates for our analysis for several reasons. First, there is no direct data available on true marginal emissions at the national level. There are several methods to try to infer what was marginal, though they are imperfect. As demand response and energy storage become more prevalent, this will become even more difficult. There are also issues regarding the use of short term or small scale marginal estimates vs long term and/or large scale marginal analysis ( . This is especially true if renewables and/or storage capacity is added to meet increased demand. In that case, increased load over the long run drives additional low carbon power, though a load-order or other short term marginal analysis might lead to the conclusion that only fossil fuel can be on the margin.
      We have chosen to instead use the historical aggregate generation and emissions for a subregion, as it can be measured and does reflect the variation in generation sources across the country. While this approach does not capture the short-term marginal impact of EV charging, it does incorporate the larger shifts occurring in electricity generation. This usage also conforms with the EPA’s guidance on using eGRID data:
      “While nonbaseload rates can be used to estimate the emissions reductions associated with projects that displace electricity generation, such as energy efficiency and/or renewable energy, these rates should not be used for assigning an emission value for electricity use in carbon-footprinting exercises or GHG emissions inventory efforts. Rather, eGRID subregion-level total output emission rates are recommended for estimating emissions associated with electricity use (scope 2 emissions).”

  • Martin Boyd

    This is great news! I wonder though if you have overlooked an important aspect of EVs, regenerative braking. Take for example the performance report for this past month (February 2018) for my 2015 Nissan Leaf, a BEV. It used 156.11 kWh, was driven 474.65 miles. You might think then that this car got 3 miles per kWh (102.4 mpge). That would be wrong though because the regenerative braking system produced 43.57 kWh (28% of the total used). Taking the difference from what was used and what was generated, we see that my EV needed only 112.54 kWh. We then can take that number and divide into the miles driven to see that my EV averaged 4.2 miles per kWh for the month. That’s equivalent to getting 142 mpg using the EPA conversion formula.

    • neroden

      I believe the EPA numbers include regenerative braking. Unlike Nissan, Tesla reports the *net* energy used, not the gross energy used.

      • Correct, the EPA efficiency is calculated using the energy use from the wall over a set of driving conditions. Regenerative braking is included, though only at the default “drive” setting. Modes like the Chevy Bolt’s ‘low’ mode with extra regenerative braking is not considered for the EPA efficiency on the window sticker and at

  • MarkfromLexington

    Your first link will not open in a Chrome browser on a Mac. Says that browser is unsupported.
    When I use a Safari browser – I get a message saying I need to download and install some software before I can proceed.
    Not happening… Can you provide a friendlier link?

  • Arthur Burnside

    Nobody except global warming alarmists give a hoot about greenhouse gases. Peddle your junk science somewhere else.

    • Ash45

      Citation needed for your statement.