An Investigation of the Electric Vehicle's Impact on CO2 Emissions

ohmexcited@comcast.net

Readers, thank you very much for your feedback, as you are continuing to help me make these calculations more accurate. Your continued feedback is most welcome by e-mailing ohmexcited@comcast.net.

1.0 Introduction and Objective

1.1 Why Do We care?

Following the documentaries acclaimed by environmentalists, "Who Killed the Electric Car" and Al Gore's "An Inconvenient Truth", the announcement of the Tesla Roadster pure electric sports car, and more recently General Motors' Chevrolet Volt serial hybrid concept, said to go 40 miles on battery power alone, the electric car has received a lot of attention lately.

The driving forces behind the excitement are primarily the potential mitigating impact on global warming and also a reduction of the importation of oil from foreign countries such as Saudi Arabia, Iran, and Venezuela -- oil trading that sometimes involves turbulent relationships and that adds significantly to the US trade deficit. In an October, 2001 PBS interview former Bush Sr.'s Secretary of State, James Baker III, stated that he, "worked for four administrations under three presidents. And in every one of those, our policy was that we would go to war to protect the energy reserves in the Persian Gulf. That is a major and very significant national security interest that we have."

National security may be reason enough for some of you considering the purchase of an electric vehicle. However, this paper will focus solely on the global warming impact, examining the documented facts and claims of various auto makers and performing an independent analysis of the CO2 emissions under various different scenarios.

1.2 What's My Dog in the Fight?

Most of the analyses I've reviewed (an am able to comprehend!) have been performed by auto makers, environment enthusiasts, politicians with agendas, and newspaper reporters with little or no technical insight. Other organizations, such as the non-profit  utility-funded Electric Power Research Institute, have published studies on electric vehicles. However, they are typically extremely long, or do not include the methods of their calculations in an easy to follow way. Therefore, I wanted to do my own study, and see how that compares with other claims out there and bandied about.

I am not affiliated with auto companies, the energy industry, or environmentalist groups. Although I have been interested in electric vehicle technology lately, I have no vested interest in seeing any one side "win". I do, however, believe in global warming, and am interested in solutions that would impact it. As for my technical background, I have an M.S. in Engineering from UC Berkeley.

1.3 What Am I Trying to Find Out?

Tesla Motors has documented well-to-wheel analyses of their own Roadster, comparing net energy input and CO2 emissions, compared to that of some internal combustion engines. However, the comparisons in the Tesla whitepaper tend to be theoretical, enthusiastic, and have primarily assumed the source of electricity to be natural gas, and using the most advanced combined cycle reactor available from General Electric. In fact, dirtier burning coal represents approximately 50% of the national grid. Detailed information with which to make additional comparisons does exist for other electric vehicles, including General Motors' discontinued EV1, and Toyota's discontinued RAV4 EV, and with the announcement of the Chevrolet Volt concept vehicle, we do have some ballpark information to work with on that car as well.

I want to answer the following basic question: How does an electric vehicle compare to other vehicles in terms of CO2 emissions under various "real world" scenarios? For example, what is the impact of an electric vehicle assuming the ultimate source of power is (1) the average of the US national grid, (2) purely coal, (3) purely natural gas, or (4) the grid in various locations throughout the United States?

2.0 Efficiency of Electric Vehicles

The electric vehicle energy efficiency, E, is taken in this paper to mean the amount of energy demanded from your A/C electrical outlet to drive a certain distance (units are energy per distance). A low E number is good.

2.1 2007-1/2 Tesla Roadster

• 2 door, 2 seat sports car.
• 3-phase A/C induction motor rated for 250 hp.
• Liquid-cooled lithium-ion battery pack.
• 250 mile driving cycle range.
• MSRP: $92,950

The Tesla Motors whitepaper states that their Roadster prototype uses "110 watt-hours (0.40 mega-joules) of electricity from the battery to drive a kilometer, or 2.53 km/MJ." It also states that the charging efficiency of the lithium-ion batteries is 86%. The efficiency from A/C outlet to the wheels, E_TESLA, is therefore:

2.2 1999 General Motors EV1

• 2 door, 2 seat subcompact.
• 3-phase A/C induction motor rated for 137 hp.
• NiMH battery pack.
• 140.3 mile driving cycle range.
• MSRP: not for sale

Idaho National Laboratory performed vehicle testing of many different electric vehicles. Their results for the 1999 GM EV1 with nickel-metal-hydride (NiMH) batteries documented that 25.14 kWh of energy was used to drive 140.3 miles of the driving cycle range, or 179 watt-hours per mile (111 watt-hours per kilometer). For a 1999 vehicle, quite impressive, in fact, when compared to Tesla Roadster's 110 watt-hours per kilometer. However, it took significantly more power at the A/C outlet to charge the NiMH battery pack, compared to the lithium-ion cells available today. The same study showed that the charging efficiency was 373 watt-hours from A/C outlet per mile. Put another way, if your power company charged you 10 cents a kilowatt-hour for power demand as recorded on your home's meter, it would cost you 3.73 cents a mile to operate the car.

2.3 2010-2012 Chevrolet Volt concept car

• Class C sized 4 door, 4-5 seat sedan
• A/C induction motor rated for 160 hp.
• 16 kWh (minimum) advanced lithium-ion battery system from A123Systems.
• 40 mile driving cycle range pure electric mode; 640 mile driving cycle range using onboard internal combustion engine range extender at 50 mpg sustained.
• MSRP: "Priced nicely under $30,000" -- Bob Lutz, GM.

Although the car does not exist yet, many different news outlets and blogs reported the intended specifications of the vehicle.  Autobloggreen.com states some specs, that we assume came straight from GM. The advanced lithium-ion battery system from A123Systems/Cobasys is intended to be 16 kWh minimum and has a 40 mile range in pure electric mode, where all power comes solely from energy stored in the battery starting from its maximum charge state.  WIRED magazine's John Garner said in his online blog that according to GM's Tony Posawatz, the batteries are "managed to not be charged more than 80 percent or less than 30 percent so that they will last the lifetime of the vehicle." Other reports state that the target lifetime is approximately 10 years, and 4,000 cycles.

Given that information, we can estimate that half of the battery's 16 kWh stored energy, or 8 kWh, drives the vehicle 40 miles using the driving cycle range in pure EV mode. I don't know the charging efficiency of the A123Systems nano-based lithium-ion batteries, but I will assume it is similar to conventional lithium-ion batteries -- the 86% number provided by Tesla Motors (if anyone knows a more exact figure, please let me know).

This calculated efficiency is not as good as the Tesla Roadster. There is probably conservatism in GM's claims. But for now, we'll use this number as a best estimate for comparison purposes. If GM decides to provide more precise information as development of the vehicle continues, these calculations can be revisited.

2.4 1998 Toyota RAV4 EV

• 4 door, 4-5 seat small SUV
• 94.0 mile driving cycle range.
• Permanent magnet D/C brushless motor rated at 67 hp.
• NiMH battery pack.
• MSRP: Not for sale new. Available on eBay for approximately $50,000!

RAV4 vehicles are one of the few EV's mandated by California in the 90's to still be on the roads. Owners have fetched high numbers for the used vehicles on eBay. Like with the EV1, Idaho National Laboratory performed detailed testing.  23.01 kWh of energy from the NiMH batteries were used to drive 94.0 miles of the driving cycle range. However, the A/C wall outlet to wheels efficiency was reported to be 432 Wh-AC per mile.

Note that it takes more than twice as much energy from the wall outlet to travel a mile as the Tesla Roadster.

There is also a 2003 model of the RAV4 EV available. If you do the calculations, you arrive at an efficiency of 302 W·hr/mi. An apparent significant improvement in charging/battery system technology of the vehicle since 5 years prior. Note that puts its results somewhere between the Volt and the EV1.

Now that we've determined the operating efficiencies of the vehicles, we will figure out what the CO2 emissions could possibly be using various different power sources, and compare them to conventional gasoline internal combustion engines and hybrid vehicles currently on the roads.

3.0 Calculation of CO2 Emissions

The Tesla whitepaper used a theoretical, scientific approach in calculating carbon emissions from various power sources. This paper will use data from a new software tool developed Argonne National Labs called GREET and the United States Environmental Protection Agency's eGRID database. I believe these account for more accurate estimations of the efficiency of existing power plants throughout the United States. 

3.1 CO2 Emissions from Internal Combustion Engine Vehicles

Per the Code of Federal Regulations at 40 CFR 600.113-78 (consistent with the Intergovernmental Panel on Climate Change (IPCC) guidelines and used by the EPA to calculate fuel economy of vehicles), a gallon of gasoline is assumed to produce 8.8 kg or 19.4 pounds of CO2. This takes into account the carbon content of gasoline, and an oxidation factor that accounts for a portion of the carbon remaining un-oxidized. However, this value does not take into account emissions during the process of extraction, refinement, and transportation to retail outlets. According to a 2006 article in the New York Times, it states that a "gallon of gasoline produced from conventional oil emits 11 kilograms carbon, from the day it is pumped out of the ground to the day it is burned by a car." That turns out to be 24.25 pounds per gallon of CO2. So only about 80% of the overall CO2 output comes from the tailpipe.

The data in Table 1 below calculates vehicle CO2 (vCO2) in pounds per mile as: 24.25 lb/gal divided by the combined mpg in mi/gal. I've included a number of popular models from Toyota that range from a small SUV to their most efficient Prius hybrid. Note that some people hotly contest that the Prius actually gets 55 mpg. So, I've also included a line for 47 mpg. And for a sanity check, I also included a 2007 Hummer H3 4-wheel drive.

Table 1. Internal Combustion Engine Vehicle CO2 Emissions

Excel Spreadsheet: CO2calc_ICE.xls

3.2 CO2 Emissions from Electric Vehicles

The EPA website includes a number of resources for estimating CO2 emissions from various energy sources and various parts of the country where you would draw power from a local grid.  One of the most useful is the Emissions & Generating Resource Integrated Database (eGRID). It includes data from 24 Federal databases: EPA, EIA, FERC, and claims to be the most comprehensive database of almost all electric generators in the US. However, the data appears to be taken from the 1999-2000 time period.

I looked into a way to quantify the power plant's full fuel cycle efficiency, or the total emissions produced from the well/mine to A/C electrical outlet. Argonne National Laboratory has produced software to compute just that called GREET.

Unfortunately it does not contain data for different parts of the United States. Only California, the average national grid, and specific fuel sources like coal and natural gas.

However, with a hybrid approach we can attempt to make use of both sources of information for a wider comparison.

3.2.1 Method of Calculation

The GREET software provides full fuel cycle emissions data in terms of grams per mmBTU. 1000 grams per mmBTU equals 7.522 pounds per megawatt hour. In this paper "pCO2" is a variable representing power plant CO2 emissions from well to A/C outlet in pounds per megawatt-hour.

eGRID provides emissions data in terms of pounds of CO2 per megawatt-hour of energy produced for various regions of the country.  Using the Power Profiler, it can also be seen that eGRID assumes line losses (the amount of energy lost during transmission and distribution of electricity, including unaccounted for uses) of 9% nationally. That is, if you use 91 MWh of energy from your outlet, the power plants had to provide 100 MWh (on average) to meet the demand. pCO2 for eGRID is therefore taken as the documented value of lb/MWh divided by 0.91.

The following are FULL fuel cycle emissions rates -- mine/well to A/C outlet -- according to GREET. Also shown are PARTIAL fuel cycle emissions rates -- power plant to A/C outlet -- according to eGRID. Multiple factors explain the discrepancies. The eGRID data appears to be from the 1999-2000 time period, while GREET is more recent. The California numbers below are very similar, but the eGRID value from an earlier time frame did not account for extraction and transportation.

We will use GREET values when we have them (national grid and California), and eGRID when we do not (other parts of the country). Note that eGRID values do not take into account emissions from extraction, processing, and transportation, however, they are assumed to be relatively small compared to the outpuet.

The vehicle CO2 (vCO2) emitted to the atmosphere in terms of pounds per mile is then calculated as:

vCO2 (lb/mi) = pCO2 (lb/Wh) x E (Wh/mi) = pounds of CO2 emitted per mile driven in the electric vehicle.

3.2.2 Power Plant Emission Rates

As described above, GREET calculates the 2007 CO2 emissions rate of the average national grid as 1604 pounds per megawatt hour (213,280 grams per mmBTU) to the outlet, and 883 lb/MWh (117,421 g/mmBTU) for California.  Using eGRID, data is also available for region-specific CO2 emissions at the plant level.

In addition, we can look at emission rates of specific types of energy supplies. GREET computes that the average CO2 emissions rate in the US from natural gas-fired generation are 1238 lbs/MWh average and 949 lbs/MWh for combined cycle generators.  For coal the average value is 2606 lbs/MWh. An advanced coal fueled Integrated Gasification Combined Cycle (IGCC) plant, has a mine-to-outlet emissions rate of 1954 lbs/MWh.  IGCC plants are "sequestration-ready" -- designed to be able to capture and store the CO2 -- but none have been implemented to-date.  The GREET emissions data DO take into account extraction, treatment, and transport.

3.2.3 Results of Calculations

The following table summarizes CO2 emissions for various electric vehicles using power derived in various conditions. For comparison purposes, the internal combustion vehicles are shown above in Table 1. Note that Tesla Motors performed their own calculation of emissions assuming the most efficient natural gas power plant available from General Electric. They concluded with a different methodology that 0.261 grams of CO2 would be emitted per kilometer driven, or 0.164 pounds per mile. My calculation using GREET numbers of combined cycle gas-fired plants arrived at 0.195 pounds of CO2 per mile.

Table 2. Vehicle Ultimate CO2 Emissions

The numbers above for the Chevrolet Volt assume operation in pure EV mode. According to GM 78% of Americans drive 40 miles a day or less, which is the range of their battery-only mode. Past 40 miles if you don't plug-in again, the battery is periodically recharged with an internal combustion engine at an optimized RPM, resulting in sustained fuel efficiency of 50 MPG, or approximately the same as a Prius Hybrid (note that these numbers from GM are preliminary and subject to change). 

4.0 Conclusions

The results for electric vehicles are generally very positive in terms of reducing CO2 in the US, but a mixed bag.

4.1 The good news for electric vehicles:

• The Tesla Roadster is an extremely efficient vehicle. Of the electric fuel sources examined, a Tesla Roadster charged in Washington State produces the fewest CO2 emissions of any vehicle by far. Thanks to all the rain, the majority of power in the Pacific Northwest is hydroelectric. The Tesla Roadster using a mix of power from the national grid average is also the most efficient vehicle (0.330 lb CO2 per mile compared to 0.441 lb per mile for the Prius).
• Electric cars charged up in California also produce unbeatable results when compared to any internal combustion vehicle on the road. According to the eGRID statistics I used, the California fuel mix is 45% natural gas, 17% hydro, 15% nuclear, 12% coal, and 9% non-hydro renewables.
• Assuming the average mix of power plants across the United States, the modern electric vehicles using lithium-ion batteries like the Tesla Roadster cause fewer CO2 emissions than any other car on the road. The Chevy Volt (estimate) was calculated to be slightly better than a Prius.
• Electric vehicles have some other intrinsic values: Much better torque at lower RPM, resulting in effortless and superior acceleration even with superior efficiency. Convenient avoidance of gas stations. Less maintenance. Reduction in noise pollution, smog, and other noxious fumes around areas where humans tend to live. Improvement in national security with transformation of oil being a strategic commodity to a mere commodity. Reduced trade deficit.

4.2 The bad news for electric vehicles

• The most efficient electric car, the Tesla Roadster, is hard to beat when it comes to emissions (and in drag races!). But charged in Texas (area 16 of the eGRID Subregion map) it is not environmentally more competitive in a very obvious way than a Prius.
• An electric vehicle charged purely from an average US coal plant is generally less polluting than an average or even above average in efficiency internal combustion vehicle. A Tesla Roadster powered by an average coal plant is cleaner than a Toyota Corolla. However, it is not better than a Toyota Prius Hybrid in this contest.  An advanced, state of the art coal plant makes electric vehicles environmentally competitive even with conventional hybrids, but no by much. In countries like Australia, where around 80% of the electric power is derived from coal, it might not make sense getting an electric vehicle instead of a Prius, if you're purely concerned about global warming.
• Electric vehicles tend to cost more than conventional vehicles because of battery packs and lower sales volume. Buying an electric vehicle is not the only way to reduce CO2 with a given dollar to spend.

4.3 Final Thoughts

4.3.1 No Free Lunch With Existing Infrastructure

A recent study by Pacific Northwest National Laboratory on behalf of the Department of Energy concluded that "off-peak" electricity production, or idle capacity, could power 84% of cars and light trucks (84% of 198 million vehicles) if they were plug-in hybrid electrics. Although it is true some off-peak power is "wasted", it should not be assumed that it has no cost, either economical or environmental. For example, during California heat waves, the grid is taxed above 50,000 megawatts to near capacity. Currently demand is around 30,000 megawatts. Although the grid is able to add extra power on demand, it's highly unlikely that the fossil fuel portion of that 20,000 megawatt gap is "wasted" and burning away for nothing. Another way of putting it is that, yes, the grid has idle capacity that could charge electric cars without the need to build more plants. However it also has the capacity to produce more CO2 without building more plants!

While the numbers I've calculated above are likely conservative, and that stabilizing power production with plug-in vehicles could help increase transmission efficiency of the national grid, there is a cost, both in dollars and to the environment, of plugging in an EV.

4.3.2 But an affordable lunch is achievable

The good news is that in most "real world" conditions in the US, electric vehicles could reduce CO2 emissions significantly. However, coal remains a very dirty polluter. EV's with modern batteries compete against internal combustion engines even assuming coal as the fuel source. But coal can reduce the electric vehicle's environmental return on investment significantly. It's the single biggest emitter of CO2 in the US. If the government is ever to get a handle on American CO2 emissions, in a world of EV's or not, something serious needs to be done about coal plants. Either by carbon sequestration aka "clean coal" or phasing out with replacement of nuclear, wind, solar, or other renewables. In fact, if all American cars and light trucks were eliminated from US roads, it would only reduce American CO2 output by 20-25%.

Even large power generation companies such as Duke Energy are receptive to national cap and trade legislation. If American transport infrastructure were electrified, in combination with a cap and trade scheme in the electric industry, this would have a dovetailing effect that could potentially make a serious dent in CO2 emissions.

Regards,
OhmExcited
ohmexcited@comcast.net

Index