|Fig. 1: Electricity Generation Sources for Three States. |
The measure of the fuel economy of a passenger vehicle is difficult to quantify for numerous reasons. Even for traditionally fueled vehicles there is wide variance in fuel economy for a single vehicle type based upon driving habits, fuel mixture (gasoline, corn ethanol, sugar cane ethanol, etc.) and even climate (use of air conditioning). This problem becomes vastly more complicated when partially or fully electrified vehicles are included. The EPA has arrived at a series of measures such as Miles Per Gallon Equivalent (MPGe), meant to simplify decisions for consumers by wrapping the complexities of this analysis in a simple numeric value. This value takes on an important significance, however, and many consumers are unable to critically evaluate their meanings. This is particularly important as many of the vehicles with more complicated fuel consumption patterns are sold on their eco-friendly merits.
In this report we begin by presenting some of the qualitative reasons for a more thorough investigation of the energy consumption and greenhouse gas emissions. We then examine some data concerning electrical power generation in a few regions of the United States and use this to demonstrate the more subtle conclusions on the impact of electric vehicles.
The gas mileage of a gasoline fueled vehicle, typically the miles-per-gallon (MPG) rating assigned by the United States Environmental Protection Agency (EPA) has long been held as a relevant proxy for the energy consumption and environmental impact of a passenger vehicle. This is despite the fact that it is a measure of the so-called Tank-to- Wheel (TTW) energy consumption of the vehicle. This TTW measure only considers the efficiency of the vehicle in converting the chemical energy held in the storage tank into motion and distance traveled. It entirely neglects the energy cost of extracting, refining, transporting and loading the fuel. This assumption may be justified in two ways.
First, the well-to-tank (WTT) energy consumption is typically quite similar for different vehicles, drivers and geographic locations, at least within the United States. No vehicle design can, in any straightforward manner, affect the amount of energy required to pump fuel. There might be some differences in transportation effort required based upon distance from refineries and shipping ports, but the differences are not large. For instance, a gasoline tanker truck may consume 100 gallons of fuel in a trip from a Texas refinery to a South Dakota gasoline station, but with 10,000 gallons of fuel on board, this is only a 1% difference from no fuel transportation at all. This is possible because the energy density of gasoline is so high that transportation energy costs are relatively modest compared to the energy released by the final consumption of the fuel. The majority of energy consumption and carbon emissions do in fact occur at the vehicle.
The onboard use of energy in a traditional gasoline fueled vehicle further justifies the use of MPG as a proxy for energy consumption of the vehicle. A very large fraction of the energy consumed onboard goes to activities directly in support of moving the vehicle (including both useful work and associated losses). Additional losses that are not accounted for include electrical loads (a typical alternator can demand as much as ~1kW/1.5hp), and air conditioning. These two loads account for a few percent of the energy consumption.
For these reasons, though some adjustments must be made (see section 2), the MPG rating of a gasoline vehicle is at least quite indicative of its true energy consumption, and thus its carbon footprint.
Some of the ways in which these assumptions break down are obvious, while others are more obscure. Beginning with the obvious, we can certainly not claim that the majority of carbon emissions occur at the vehicle, as electric vehicles make no direct emissions. A "tank"-to-wheel measure of energy consumption for an electric vehicle will be unfairly favorable, as it fails to account for any of the costs associated with generating the electric power and delivering it to the vehicle. The MPGe standard determined by the EPA has been adjusted to include charging losses, so it replicates the intent of MPG in allowing consumers to estimate the costs of the fuel required to drive a certain distance, but neglects the majority of the true energy and carbon emission costs of the electricity.
In the United States a large fraction of electricity is generated from fossil fuels, which have associated efficiencies and carbon emissions. To better understand the full energy consumption of an electric vehicle we must begin by considering exactly what fraction of the electricity is generated at different efficiencies. and the carbon released from the fuel. We must then consider transmission losses, as it is not immediately obvious that they are negligible.
Onboard an electric vehicle, the majority of energy consumed is still directly in support of moving the vehicle, but the other penalties are larger than for a gasoline vehicle. For instance, in a gasoline vehicle heating is provided exclusively by waste heat from the internal combustion engine. Since the electric motors may not provide enough waste heat, a significant electrical demand can be caused by electrical Joule heating, or driving a heat pump. Additionally, battery performance is strongly affected by temperature, so additional energy is devoted to heating and cooling of batteries. Finally, in a gasoline vehicle the fuel degrades extremely slowly, so an interval between vehicle usages does not impact the overall energy consumption. Current state-of-the-art batteries have a non-trivial self-discharge rate, which imposes an additional energy cost if the vehicle is left to sit unused.
Another additional consideration for electric vehicles is that, unlike gasoline vehicles, these parameters do vary significantly by location, even within the United States. The amount of climate control required in a location like Illinois, where temperatures reach both high and low extremes is much higher than in a location like Northern California, where temperatures are restricted to a small range. Similarly, the electrical generation schemes are almost entirely non- overlapping in renewable heavy Washington and traditional coal powered Indiana. For all these reasons, we must look well past the simplistic measure of MPGe provided with electric vehicles to understand the true energy costs and carbon emissions.
California is often considered to be a leader in progressive movements, and has a favorable climate for some types of renewable energy generation. Despite this, the majority of electricity generated in state is fueled by natural gas. See the figure above.
Of the 2 × 108 MWhr generated in 2012, 1.2 × 108 MWhr were generated by natural gas fueled plants, with an associated 1026 lb CO2/MWhr. Though natural gas is far from carbon neutral, it is considered to be among the least carbon emitting fossil fuel sources. In addition, a modern natural gas turbine electric power plant has relatively high thermal efficiency (see tables below). A quick calculation reveals that 1MWhr of electricity generated in California releases 655 lb of CO2. [1,2]
A megawatt of generation does not equate to a megawatt of supply, however, due to losses in transmission. Using an estimate for average transmission loss, we compensate the emissions.
Washington state has the combination of a large proportion of hydroelectric power, and strict emissions controls for its fossil fuel power plants.
Of the 1.2 × 108 MWhr generated in 2012, 8.9 × 107 MWhr was generated by emission-free hydroelectric sources. Less than 10% of the generation was provided by coal or natural gas. At a summary level, a megawatt hour of electricity generated in Washington releases only 132 lb. CO2 (before transmission losses). [1,2]
On the opposite end of the spectrum, Indiana is heavily reliant upon coal as a primary source of power.
Over 80% of the power is generated from coal. Though these coal plants are less polluting than those found in California when normalized for capacity, they still produce a large amount of carbon dioxide and other pollutants such as sulfur and nitrogen oxides. Each megawatt hour of electricity generated in Indiana releases 1,914 lb CO2 (before transmission losses). [1,2]
For the most rigorous comparison in this scheme we will compare the carbon emissions associated with providing power to a vehicle from the source of the fuel (well/mine/renewable source) to final consumption. This requires an estimate of the additional carbon emissions that come as a result of fuel acquisition, processing, transportation and delivery.
We begin with two gasoline fueled vehicles - the Toyota Prius (a gasoline-electric hybrid rated at 50 MPG combined) and the Honda Fit (a gasoline vehicle rated at 36 MPG combined). To calculate the approximate CO2 emissions resulting with this fuel consumption, we must calculate the carbon emissions directly from the combustion of gasoline (simple chemistry) and those resulting from the extraction of petroleum, refining, and transportation. This is demonstrated in the table below, and is assumed to be constant throughout the United States.
|Table 1: Carbon Emission Adjustments for Gasoline. [5,6]|
Using this value, and the energy demands of the vehicle, it is simple to compute the carbon emissions emitted for one typical mile of driving.
|Table 2: Carbon Emissions for Toyota Prius and Honda Fit. [7,8]|
While we have already examined the carbon emissions created directly by the generation of electricity, and considered the transmission losses, we must adjust these values to compensate for additional emissions accrued during the extraction, processing and transportation of the base fuels (in particular, natural gas and coal). Fortunately these processes have been studied extensively, making it easy to estimate the parameters within a reasonable range.  These parameters are not uniform across the country, depending upon the fuels used for electricity generation, so we calculate the adjusted CO2 emissions for each state in our case study. The results are shown in the table below.
|Table 3: Carbon Emission Adjustments for Electrical Power. [1,2,4,7]|
As in the above section, we consider the case for two prototypical vehicles - the Tesla Model S and BMW i3. The Tesla Model S is one of the least efficient mainstream, full electric vehicles available, consuming 38kWh/100mi. The BMW i3, on the other hand, is one of the most efficient, achieving only 27kWh/100mi. With some simple unit conversions we find the CO2 emissions for each car, in each state.
|Table 4: Carbon Emissions for Tesla Model S and BMW i3. [9, 10]|
|Fig. 2: Vehicle estimated carbon dioxide emissions for three US states. Values as calculated in Tables 1-4.|
In the above section we have seen that there is a large variance in the carbon emissions from a given electric vehicle in different parts of the United States, even neglecting other regional differences like use of climate control. This difference is so significant that it can negate the improved energy efficiency of the electric vehicle. Figure 2 shows the relative carbon emissions of the four vehicles and three states calculated in the above section.
The electric vehicles we examine are much more effective in converting the onboard "fuel" to motion than the gasoline vehicles, consuming less than half the energy per mile. Unfortunately, in some cases the electricity required to power the vehicle emits enough carbon dioxide to negate this advantage. Particularly in states which generate a large majority of their electricity from high-carbon sources like coal the net carbon emissions of an electric vehicle are higher than those a gasoline vehicle. In order to ensure a particular vehicle choice has a given environmental impact potential consumers must carefully consider extrinsic factors and the source of their electrical power.
© David Heinz. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
 "Electric Power Monthly with Data for September 2014," U.S. Energy Information Administration, November 2014.
 "State-Level Energy-Related Carbon Dioxide Emissions, 2000-2011," U.S. Energy Information Administration, August 2014.
 P. Jaramillo, W. M Griffin, and H. S. Matthews, "Comparative Life-Cycle Air Emissions of Coal, Domestic Natural Gas, LNG, and SNG for Electricity Generation," Environ. Sci. Technol. 41, 6290 (2007).
 "Electric Power Annual 2012," U.S. Energy Information Administration, December 2013.
 E. Hirst and R. Herendeen, R., "Total Energy Demand for Automobiles," SAE Technical Paper 730065, 1 Feb 73.
 R. Edwards, J. F. Larivé, and J.-C. Beziat, "Well-to-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Contex," European Commission Joint Research Center, EUR 24952 EN-2011, July 2011.
 J. Garrett, "Hybrid Superstar Shines Brighter," New York Times, 26 Mar 09.
 L. Ulrich, "Inside the Frog, a Prince Awaits," New York Times, 13 Jun 14.
 A. MacKenzie "2013 Motor Trend Car of the Year: Tesla Model S," Motor Trend, January 2013.
 J. Gall, "BMW i3 First Drive," Car and Driver. October 2013.