Report Contents

Report#:EIA/DOE-0573(98)

November 5, 1999 
(Next Release: November,  2000)

Overview

Motor Gasoline

Crude Oil

Liquefied Petroleum Gases

The first edition of Emissions of Greenhouse Gases in the United States, published in September 1993, applied emissions coefficients for petroleum and natural gas developed by Marland and Pippin⁽¹⁴⁶⁾ and adopted by the Intergovernmental Panel on Climate Change.⁽¹⁴⁷⁾ Those coefficients, developed for broad international use, covered only the six petroleum product categories in the International Energy Agency's taxonomy. The Energy Information Administration (EIA) collects data on more than 20 petroleum products, and U.S. petroleum products often differ in composition from those consumed abroad. In the first edition of this report, EIA estimated emissions coefficients for the remaining petroleum products based on their underlying chemical composition. EIA also used emissions coefficients for coal by rank (anthracite, bituminous, subbituminous, and lignite) and State of production, developed using 5,426 coal samples from the EIA coal analysis file.

In 1994, EIA developed specific and updated emissions coefficients for all petroleum products in their data collection system, based on their density, heat content, and carbon share. These variables were estimated on the basis of the underlying chemical composition of the fuels and, where available, ultimate analyses of product samples.⁽¹⁴⁸⁾ An emissions coefficient for natural gas was also generated, based on 6,743 gas samples in a Gas Research Institute database. The magnitude of potential variation in emissions coefficients for fossil fuels is constrained by the limits imposed by the chemical properties of the hydrocarbon compounds that define the fuels.⁽¹⁴⁹⁾ In all but a few cases, the revised emissions coefficients differed from those developed by Marland and Pippin by less than 5 percent.

The composition of marketed petroleum products varies over time because of changes in exploration, recovery, and refining technology, economic changes (e.g., changes in the price of oil), and regulatory changes (e.g., requirements for reformulated gasoline in the Clean Air Act Amendments of 1990). The composition of two petroleum products which make significant contributions to U.S. greenhouse gas emissions, motor gasoline and jet fuel, has changed significantly over the last decade. Thus, EIA provides annual updates of the emission coefficients for these fuels (Table B1). In addition, EIA publishes annual updates of emissions coefficients for crude oil and liquefied petroleum gases.

Table B1.  Carbon Emissions Coefficients at Full Combustion, 1990-1998

Crude oil consumption in the United States is a very small portion of carbon emissions, because nearly all crude is refined into finished petroleum products. However, crude oil refinery input can be used to develop a national mass balance estimate of carbon emissions that may be used as a benchmark for the more disaggregated estimation approach used by EIA. EIA has developed a regression equation reflecting the relationship between the density, sulfur content, and carbon content of crude oil. From these data a crude oil emissions coefficient can be calculated. This regression equation is applicable for any nation that collects these basic petroleum statistics and supports a mass balance approach to estimating emissions where detailed product data are not collected. The density of crude oil entering U.S. refineries has risen gradually over the last decade. Through 1996, this resulted in a greater carbon content in each barrel of crude oil. However, because sulfur content in crude oil rose in 1997 and 1998, some carbon was crowded out and the emission coefficients declined slightly.

In 1997, EIA began publishing separate emissions coefficients for LPG fuel use and LPG nonfuel use. LPG may be used as fuel or as a petrochemical feedstock. About three quarters of the carbon in petrochemical feedstocks will be sequestered. Further, if the mix of  paraffinic hydrocarbons used for petrochemical feedstock differs substantially from those used for fuel, using a single emissions coefficient for LPG will bias estimates of emissions. Thus, EIA now adjusts the emissions coefficient for LPG based on the mix of compounds used as fuel and feedstock.

Motor gasoline consumption accounts for about 20 percent of all U.S. greenhouse gas emissions. Thus, changes in composition can have important effects on national emission levels. In 1995, a requirement for reformulated gasoline in nonattainment areas implemented under the Clean Air Act Amendments changed the composition of gasoline consumed in the United States. During 1995, 25 percent of all gasoline consumed in the United States was reformulated, rising to 32 percent in 1997. Because the additives contained in reformulated gasoline have much lower carbon shares than typical gasoline, the national average emissions coefficient for motor gasoline has declined by about 0.25 percent over the past 3 years (Table B1). At the same time, the heat content of gasoline additives is lower than that of standard gasoline. Thus, an average gallon of gasoline sold today has a lower energy content than an average gallon sold 4 years ago. Holding car weight, horsepower, and miles traveled constant will require increased consumption of gasoline.

Jet fuel consumption is responsible for more than 4 percent of U.S. carbon dioxide emissions. Like motor gasoline, jet fuel consumed in the United States has undergone a dramatic change in composition over the past decade. Until 1993, two types of jet fuel were widely used in the United States. Kerosene-based jet fuel was generally used in the commercial airline industry and naphtha-based jet fuels were used primarily by the U.S. Department of Defense. The emissions coefficient for naphtha-based jet fuels was about 3 percent higher than that for kerosene-based jet fuel. In 1989, 13 percent of all jet fuel consumed was naphtha-based. By 1996, that figure had fallen to 0.3 percent, and in 1997 total naphtha-based jet fuel consumption was negligible. Thus, the emissions coefficient for jet fuel, weighted by consumption of each fuel type, fell steadily between 1988 and 1996 and has now stabilized at the level of kerosene-based jet fuel.⁽¹⁵⁰⁾ Notably, the emissions coefficient for jet fuel is now the same as the coefficient for motor gasoline.

As with all petroleum products, the emissions coefficient for motor gasoline is a function of its density and carbon content. This relationship is particularly clear in the case of motor gasoline because the share of impurities found in the fuel must be kept very low to maintain the operating condition of modern automobile engines and limit the environmental effects of vehicle use. Motor gasoline density varies between summer and winter grades and from low octane to high octane. This variation takes into account the differing performance requirements of gasoline associated with temperature changes. Partly as a result of the leaded gasoline phaseout, the density of gasoline increased slowly and steadily across all octane grades and in all seasons from 1987 through 1994.⁽¹⁵¹⁾ In order to maintain the "anti-knock" quality and octane ratings of motor gasoline in the absence of lead, the portion of aromatic hydrocarbons used in gasoline was increased. Aromatic hydrocarbons take the form of C_nH_2n-2, a lower ratio of hydrogen to carbon than other hydrocarbons typically found in gasoline. Because carbon is much heavier than hydrogen, this lower ratio results in increased fuel density and higher shares of carbon. As a result, the emissions coefficient for motor gasoline rose slowly from 19.39 million metric tons carbon per quadrillion Btu in 1988 to 19.45 million metric tons carbon per quadrillion Btu in 1994. Table B2 shows the increasing densities and emissions coefficients between 1990 and 1994.

Table B2.  Changes in Motor Gasoline Density, 1990-1998

Reformulated gasoline was consumed in large volumes (about 25 percent of overall gasoline consumption) for the first time during 1995. The density of reformulated gasoline is about one percent less than standard gasoline and the much lower carbon contents of the principal additives to reformulated gasoline (Table B3) reduce the overall share of carbon in reformulated fuel. Taking into account the 25 percent of fuel consumed with a lower emissions profile, the emissions coefficient for motor gasoline dropped from 19.45 million metric tons per quadrillion Btu in 1994 to 19.38 million metric tons per quadrillion Btu in 1995. In 1996, the share of motor gasoline that was reformulated rose to 30.8 percent. By the summer of 1996, most reformulated gasoline no longer contained any tertiary amyl methyl ether (TAME) or any ethyl tertiary butyl ether (ETBE). Nearly all additive used in reformulated gasoline was MTBE, the additive with the lowest emissions coefficient (Table B3). The increase use of reformulated gasoline continued through 1997 and 1998, driving the overall emissions coefficient to 19.33 million metric tons per quadrillion Btu. The downward trend in emission coefficients associated with market penetration by reformulated fuels is accompanied by a decrease in the energy content of fuels on a volumetric basis.

Table B3.  Characteristics of Major Reformulated Fuel Additives

To derive an overall emissions coefficient for gasoline, individual coefficients for standard motor gasoline consumed in the winter and summer months, respectively, were developed. These coefficients were based on the densities of product samples collected by the National Institute on Petroleum and Energy Research used in conjunction with a carbon share of 86.6 percent as estimated by Mark DeLuchi.⁽¹⁵²⁾ Emissions coefficients for reformulated fuels consumed during the summer and winter were calculated using the following procedure. First, the carbon share of each additive used in reformulated gasoline was calculated from its chemical formula and combined with the additive's density and energy content as provided by the California Air Resources Board to produce individual coefficients for each fuel additive. Next, the reformulated fuel was separated into its standard fuel components and its additive portions based on fuel samples examined by NIPER.⁽¹⁵³⁾ The additive portions were defined as the net increase in MTBE, ETBE, or TAME as compared with the additives in standard fuel, since small amounts of these compounds are present in standard gasoline. The emissions coefficients for standard gasoline and for each of the additives were then weighted by their proportion in reformulated fuel to arrive at a coefficient for reformulated fuel in each season.

After independent coefficients were developed for both standard and reformulated fuel, each season's coefficients were combined by weighting according to the ratio of standard vs. reformulated consumption. The combined summer and winter coefficients were then weighted based on seasonal consumption, with just over half occurring in summer, to derive an overall emissions coefficient for motor gasoline.

While crude oil composition is highly heterogeneous, the share of carbon in a fixed amount of crude oil (e.g., a gallon or barrel) varies somewhat systematically with such commonly available identifying characteristics as density and sulfur content. Because the economic value of a barrel of crude oil is largely a product of the oil's density and sulfur content these values are regularly recorded. Further, EIA maintains detailed data on the average density and sulfur content of crude oil entering U.S. refineries.⁽¹⁵⁴⁾ Thus, the annual emissions coefficient for crude oil is pegged to these two variables.

Ultimate analyses of 182 crude oil samples were used to derive a relationship between crude oil density, sulfur content, and the percentage of carbon in crude oil. The sulfur content and density of these samples was regressed against their carbon content. This regression analysis produced the following equation, which is used to estimate the carbon content of crude oil:

Percent Carbon = 76.99 + (10.19 * Specific Gravity)
                                  + (-0.76 * Sulfur Content) .

Annualized emissions coefficients are developed by inserting the average density and sulfur content for crude oil entering U.S. refineries for each year from 1987 through 1998. This provides the share of carbon in an average barrel of oil during each year. After the share of carbon is derived, it is used in conjunction with fuel density to estimate the total mass of carbon in a barrel of crude oil. An emissions coefficient per unit of energy is then calculated using EIA's standard energy content for crude oil of 5.8 million Btu per barrel.

The 1998 emissions coefficient for crude oil is 20.23 million metric tons carbon per quadrillion Btu, a slight decrease from the 1997 value. Despite an increase in density for crude oil entering U.S. refineries, carbon content has declined due to crowding out caused by growing sulfur contents (Table B4).

Table B4.  U.S. Crude Oil Characteristics, 1990-1998

EIA identifies four categories of paraffinic hydrocarbons as LPG: ethane, propane, isobutane, and n-butane. Because each of these hydrocarbons is a pure paraffinic compound, their carbon shares are easily derived by taking into account the atomic weight of carbon (12) and the atomic weight of hydrogen (1). Thus, for example, the carbon share of ethane, C₂H₆, which has an atomic weight of 30, is 80 percent. The densities of these compounds are also well known, allowing emissions coefficients to be calculated easily. EIA collects data on consumption of each compound and then reports them as LPG in the Petroleum Supply Annual.⁽¹⁵⁵⁾ By weighting each compound's individual emissions coefficient by its share of energy consumed, an overall emissions coefficient for LPG is derived.

More than 95 percent of all ethane and just under 85 percent of butane consumed goes to nonfuel uses. In contrast, nearly all LPG used as fuel is propane. Thus, the emissions coefficient for LPG used as fuel is 17.20 million metric tons carbon per quadrillion Btu, which is the emissions coefficient for propane (Table B5). On the other hand, the carbon emissions coefficient for LPG for nonfuel use is pulled down to 16.86 million metric tons carbon per quadrillion Btu by the large presence of the lighter ethane and its emissions factor of 16.25 million metric tons per quadrillion Btu. These coefficients have remained unchanged since 1996.

Table B5.  Emissions Coefficients for Liquefied Petroleum Gases, 1990-1998