Carbon Dioxide Emission Factors for Coal
B.D. Hong and E. R. Slatick
January-April 1994, DOE/EIA-0121(94/Q1) (Washington, DC, August 1994), pp. 1-8.
Coal is an important source of energy in the United States, and the Nation's reliance on this fossil fuel for electricity generation is growing. The combustion of coal, however, adds a significant amount of carbon dioxide to the atmosphere per unit of heat energy, more than does the combustion of other fossil fuels.(1) Because of a growing concern over the possible consequences of global warming, which may be caused in part by increases in atmospheric carbon dioxide (a major greenhouse gas), and also because of the need for accurate estimates of carbon dioxide emissions, the Energy Information Administration (EIA) has developed factors for estimating the amount of carbon dioxide emitted as a result of U.S. coal consumption.
Carbon dioxide emission factors for U.S. coals have previously been available from several sources. However, those emission factors have shortcomings because they are based on analyses of only a few coal samples. Most are single factors applied to all coals, regardless of rank (i.e., whether anthracite, bituminous, subbituminous, or lignite) or geographic origin. Because single factors do not account for differences among coals, they fail to reflect the changing "mix" of coal in U.S. coal consumption that has occurred in the past and will occur in the future. Lacking standardization, the factors previously available also differ widely from each other.(2)
EIA's emission factors will improve the accuracy of estimates of carbon dioxide emissions, especially at State and regional levels, because they reflect the difference in the ratio of carbon to heat content by rank of coal and State of origin. EIA's emission factors are derived from the EIA Coal Analysis File, a large database of coal sample analyses. The emission factors vary significantly by coal rank, confirming a long-recognized finding, and also within each rank by State of origin. These findings were verified statistically.
Two types of carbon dioxide emission factors have been developed. First are basic emission factors covering the various coal ranks by State of origin. These basic emission factors are considered as "fixed" for the foreseeable future until better data become available. Second are emission factors for use in estimating carbon dioxide emissions from coal consumption by State, with consuming-sector detail. These emission factors are based on the mix of coal consumed and the basic emission factors by coal rank and State of origin. These emission factors are subject to change over time, reflecting changes in the mix of coal consumed.
EIA's emission factors will not only enable coal-generated carbon dioxide emissions to be estimated more accurately than before, but they will also provide consistency in estimates. Energy and environmental analysts will find EIA's emission factors useful for analyzing and monitoring carbon dioxide emissions from coal combustion, whether they are estimated by the State of origin of the coal, consuming State, or consuming sector.
Coal Combustion and Carbon Dioxide Emissions
The amount of heat emitted during coal combustion depends largely on the amounts of carbon, hydrogen, and oxygen present in the coal and, to a lesser extent, on the sulfur content. Hence, the ratio of carbon to heat content depends on these heat-producing components of coal, and these components vary by coal rank.
Carbon, by far the major component of coal, is the principal source of heat, generating about 14,500 British thermal units (Btu) per pound. The typical carbon content for coal (dry basis) ranges from more than 60 percent for lignite to more than 80 percent for anthracite. Although hydrogen generates about 62,000 Btu per pound, it accounts for only 5 percent or less of coal and not all of this is available for heat because part of the hydrogen combines with oxygen to form water vapor. The higher the oxygen content of coal, the lower its heating value.(3) This inverse relationship occurs because oxygen in the coal is bound to the carbon and has, therefore, already partially oxidized the carbon, decreasing its ability to generate heat. The amount of heat contributed by the combustion of sulfur in coal is relatively small, because the heating value of sulfur is only about 4,000 Btu per pound, and the sulfur content of coal generally averages 1 to 2 percent by weight.(4) Consequently, variations in the ratios of carbon to heat content of coal are due primarily to variations in the hydrogen content.
The carbon dioxide emission factors in this article are expressed in terms of the energy content of coal as pounds of carbon dioxide per million Btu. Carbon dioxide (CO2) forms during coal combustion when one atom of carbon (C) unites with two atoms of oxygen (O) from the air. Because the atomic weight of carbon is 12 and that of oxygen is 16, the atomic weight of carbon dioxide is 44. Based on that ratio, and assuming complete combustion, 1 pound of carbon combines with 2.667 pounds of oxygen to produce 3.667 pounds of carbon dioxide. For example, coal with a carbon content of 78 percent and a heating value of 14,000 Btu per pound emits about 204.3 pounds of carbon dioxide per million Btu when completely burned.(5) Complete combustion of 1 short ton (2,000 pounds) of this coal will generate about 5,720 pounds (2.86 short tons) of carbon dioxide.
Methodology and Statistical Checks
EIA's carbon dioxide emission factors were derived from data in the EIA Coal Analysis File, one of the most comprehensive data sources on U.S. coal quality by coalbed and coal-producing county. Most of the samples in the file were taken from coal shipments to U.S. Government facilities, from tipples and from mines. From the more than 60,000 coal samples in the File, 5,426 were identified as containing data on heat value and the ultimate analysis(6) needed for developing the relationship between carbon and heat content of the coal, that is, the carbon dioxide emission factors. Coal rank was assigned to each sample according to the standard classification method developed by the American Society for Testing and Materials. These data observations (samples) covered all of the major and most of the minor coal-producing States (Table FE1). Except for Arizona, North Dakota, and Texas, all of the major coal-producing States were considered to have a sufficiently large number of data observations to yield reliable emission factors.
The ratio of carbon to heat content was computed for each of the 5,426 selected coal samples by coal rank and State of origin under the assumption that all of the carbon in the coal is converted to carbon dioxide during combustion.(7) Variations in the ratios were observed across both coal rank and State of origin. Analysis was performed to determine whether these variations were statistically significant and to ensure that other factors pertaining to the samples (that is, the year the sample was collected and the degree of cleaning the sample received) were not significantly responsible for the observed variations.
Table FE1. Number of Observations by Coal Rank and State of Origin
Distributions of the data observations by year of collection and degree of cleaning were compiled (Table FE2). Because the dates of the samples range from 1900 through 1986, it was thought that changes in laboratory analysis techniques over the years might have influenced the resultant carbon-to-heat-content ratios. A regression analysis found that, with a R2 value of only 0.01 (Table FE3), the year the sample was collected was not a useful factor in explaining the variation in the ratio, although there were small changes in the ratio over time.(8) This finding indicated that samples from earlier time periods could be combined with more recent samples to derive carbon dioxide emission factors.
Table FE2. Distribution of Observations by Year and Degree of Cleaning
Of the total samples, 83 percent were raw coal, with the remainder either washed or partially washed. Cleaning should not materially affect the ratio of a coal's heat-to-carbon content because the process removes primarily non-combustible impurities. This was confirmed by an analysis of variance. There were differences in the carbon-to-heat-content ratios between washed or partially washed and raw coal, but with a R2 value of 0.06, the differences did little to explain the variation in the ratios. Therefore, no data correction was warranted to account for the small effect that coal cleaning had on emission factors.
Analysis of variance was used to test the statistical significance of differences in the carbon-to-heat-content ratios across coal rank and across State of origin within coal rank. The continuous response variable (the carbon dioxide emission factor) was related to classification variables of rank and State of origin. The carbon dioxide emission factor was assumed to be a linear function of the parameters associated with the coal rank and State of origin.(9)
The statistical analyses (Table FE3) indicated that: (1) there are statistically significant differences in carbon dioxide emission factors across both coal rank and State of origin; (2) coal rank and State of origin each explain approximately 80 percent of the variation in carbon dioxide emission factors; and (3) State of origin combined with coal rank is a slightly more powerful explanatory variable than either coal rank or State of origin alone.
Table FE3. Summary of Statistical Analyses Carbon Dioxide Emission Factors by Coal Rank and State of Origin
Carbon Dioxide Emission Factors by Coal Rank and State of Origin
The (arithmetic) average emission factors obtained from the individual samples (assuming complete combustion) (Table FE4)(10) confirm the long-recognized finding that anthracite emits the largest amount of carbon dioxide per million Btu, followed by lignite, subbituminous coal, and bituminous coal. The high carbon dioxide emission factor for anthracite reflects the coal's relatively small hydrogen content, which lowers its heating value.(11) In pounds of carbon dioxide per million Btu, U.S. average factors are 227.4 for anthracite, 216.3 for lignite, 211.9 for subbituminous coal, and 205.3 for bituminous coal.
Table FE4. Average Carbon Dioxide Emission Factors for Coal by Rank and State of Origin
In general, the carbon dioxide emission factors are lowest for coal produced in States east of the Mississippi River (Figure FE1), where the predominant coals are bituminous in rank and therefore have relatively low emission factors. By comparison, the coal deposits in the West are largely subbituminous coals, which have relatively high emission factors. In a broad sense, the geographic differences reflect the greater degree of coalification--the process that transformed plant material into coal under the influence of heat and pressure--in the coal-bearing areas in the East.
In the Appalachian Coal Basin, the emission factors for bituminous coal range from a low of 202.8 pounds of carbon dioxide per million Btu in Ohio to a high of 210.2 in Maryland.(12) Pennsylvania anthracite, which is produced in small amounts, has the highest emission factor among all coal ranks (227.4). For Illinois Basin coal, all bituminous in rank, the emission factors are relatively uniform, ranging from 203.2 in western Kentucky to 203.6 in Indiana.
Figure FE1: Average Carbon Dioxide Emission Factors for Coal by Rank and State of Origin
Pounds of Carbon Dioxide per Million Btu
West of the Mississippi River, the emission factors for bituminous coal range from more than 201 pounds of carbon dioxide per million Btu in Missouri, Iowa, and Nevada to more than 209 in Arizona, Arkansas, and Montana. About 16 percent of the 1992 coal output west of the Mississippi was bituminous coal, with production chiefly from Utah, Arizona, Colorado, and New Mexico.
Subbituminous coal is the predominant rank of coal produced west of the Mississippi River, accounting for 62 percent of the region's total coal output in 1992. Subbituminous coal in Wyoming's Powder River Basin, the principal source of this rank of coal, has an emission factor of 212.7 pounds of carbon dioxide per million Btu. This is the same as for subbituminous coal in Colorado, but slightly below that in Montana. The lowest emission factor for subbituminous coal is in Utah (207.1) and the highest is in Alaska (214.0).
The emission factor for lignite from the Gulf Coast Coal Region in Texas, Louisiana, and Arkansas is 213.5 pounds of carbon dioxide per million Btu. This is 1 to 3 percent lower than the emission factors for lignite in the Fort Union Coal Region in North Dakota, South Dakota, and Montana and for lignite in the Powder River Basin in Wyoming. The 1992 output of lignite accounted for 22 percent of coal production west of the Mississippi River, with two-thirds from Texas and most of the balance from North Dakota.
All of EIA's carbon dioxide emission factors for coal by rank and State of origin should be considered as "fixed" for the foreseeable future. This is because detailed coal analysis data are not widely available annually, and because the EIA emission factors, as developed from the EIA Coal Analysis File, are considered to effectively represent the relationship between the carbon and heat content of the various U.S. coals. However, the basic emission factors will be reviewed when sufficient additional coal analysis data are accumulated.
Carbon Dioxide Emission Factors by Coal-Consuming Sector and State
Coal use among the consuming sectors and States varies in quantity as well as in rank and State of origin. Therefore, emission factors by consuming sector in each State were derived by weighting the emission factors by coal rank and State of origin by the respective amounts received by sector.(13),(14) For comparison, emission factors for 1980 and 1992 are reported in this article (Table FE5). It should be noted that the amount of coal received in a certain year may not equal the amount consumed during that year because of stock additions or withdrawals. Furthermore, because data on the origin and destination of coal are available only for coal distribution, EIA's emission factors for coal consumption by sector assume that the mix of coal received during a certain year was the same as that consumed in that year.
The emission factors for coal consumption involving combustion are based on the assumption that all of the carbon in coal is converted to carbon dioxide during combustion. Actually, a very small percentage of the carbon in coal is not oxidized during combustion. The emission factors in Table FE5 can be adjusted to reflect incomplete combustion.(15)
In coke plants, coal is carbonized, not combusted, to make coke, which is used in the manufacture of pig iron by the iron and steel industry. Although most of the carbon in the coal carbonized remains in the coke, a small amount is retained in byproducts, some of which are consumed as energy sources and others as non-energy raw materials.(16) Examination of historical data for coke plant operations indicates that about 10 percent of the carbon in coking coal remains in non- energy byproducts.(17) However, no allowances have been made in the emission factors for coke plants (Table FE5) for carbon retained in non-energy byproducts, leaving any adjustments to the user's stipulations.
Table FE5. Average Carbon Dioxide Emission Factors for Coal-Consuming Sector and State, 1980 and 1992
The mix of rank and origin of coal consumed in the United States has changed substantially in the past two decades, reflecting shifts to Western low-sulfur subbituminous coal and lignite, predominantly for electricity generation. Further changes are expected in the coming years, especially due to the Clean Air Act Amendments of 1990, which will encourage switches from high-sulfur Eastern bituminous coal to low-sulfur Western subbituminous coal.
The shift in the mix of coal ranks consumed becomes apparent when production by coal rank in 1980 is compared with that in 1992, as most production was for domestic consumption.(18) In 1980, bituminous coal comprised 76 percent of the total, but by 1992 its share dropped to 65 percent. By contrast, the share for subbituminous coal rose from 18 percent in 1980 to 25 percent in 1992, while the share for lignite grew from 6 percent to 9 percent. Anthracite's share was about 1 percent in both years. Because lower rank coals have relatively high carbon dioxide emission factors, increased use of these coals caused the national average carbon dioxide emission factor to rise from 206.5 pounds per million Btu in 1980 to 207.6 pounds per million Btu in 1992.
The change in mix of coal ranks produced reflects the large sectorial and regional shifts in coal consumption that have occurred in the past two decades. The electric utility sector dominates coal consumption, and its share has grown substantially. Of total coal consumption in 1992, electric utilities accounted for 87 percent, up from 81 percent in 1980, due mostly to increases in utility coal consumption west of the Mississippi River.(19) The share held by low-rank coals in the electric utility sector increased substantially.(20) Subbituminous coal rose from 24 percent in 1980 to 31 percent in 1992, and lignite grew from 7 to 10 percent during the period. In contrast, bituminous coal fell from 69 percent in 1980 to 58 percent in 1992. The share held by anthracite (about 1 percent) did not change.
Coal used to produce coke is virtually all bituminous in rank; less than 1 percent is anthracite. Only a few States, mostly in Appalachia, supply coking coal. The coke industry, which has been declining, accounted for only 4 percent of total coal consumption in 1992, down from 9 percent in 1980.
All ranks of coal are used by the other industrial and the residential/commercial sectors.(21) The other industrial sector accounted for 8 percent of total coal consumption in 1992, slightly less than in 1980. However, the emission factor for this sector increased sizably during the period, due mainly to the rising use of low-rank coals in the West, and contributed to the increase in emission factors for the overall national average. The residential/commercial sector is a relatively minor component of coal consumption, with about 1 percent of the total in 1980 and 1992.
As with coal consumption by sector, the amount of carbon dioxide emitted from total coal combustion in a particular State--and hence the carbon dioxide emission factor for that State--depends on the mix of coal consumed by various consuming sectors in that State during a particular year. When the total energy in Btu from coal consumption by State is known (with no breakdown by coal-consuming sector), the State average emission factors can be used to estimate the total amount of carbon dioxide emissions by State.
Publication of Carbon Dioxide Emission Factors
EIA's carbon dioxide emission factors by consuming sector and State will be updated periodically to reflect changes in the mix of U.S. coal consumption. EIA plans to report these updates in the Quarterly Coal Report, the State Energy Data Report, and the annual issue of Emissions of Greenhouse Gases in the United States.
1Coal combustion emits almost twice as much carbon dioxide per unit of energy as does the combustion of natural gas, whereas the amount from crude oil combustion falls between coal and natural gas, according to Energy Information Administration, Emissions of Greenhouse Gases in the United States 1985-1990, DOE/EIA-0573 (Washington, DC, September 1993), p. 16.
2Examples of previously published emission factors include, in pounds of carbon dioxide per million Btu, single emission factors of 205.7 in "United States Emissions of Carbon Dioxide to the Earth's Atmosphere," Energy Systems Policy, Vol. 14, 1990, p. 323; 210.2 in Changing by Degrees, U.S. Congress, Office of Technology Assessment, February 1991, p. 333; 205.6 for bituminous coal in Greenhouse Gases, Abatement and Control, IEA Coal Research, June 1991, p. 24; and 183.4 in Limiting Net Greenhouse Gas Emissions in the United States (Executive Summary), U.S. Department of Energy, Office of Environmental Analysis, September 1991, p. 37. EIA's first reported emission factors by coal rank, published in Electric Power Annual 1990, DOE/EIA-0348(90) (Washington, DC, January 1992), p. 124, were as follows: anthracite, 209; bituminous coal, 209; subbituminous coal, 219; and lignite, 213.
4The relationships of the various heat-producing components of coal are given in Dulong's formula, which provides a method for calculating the heating value of solid fuels. Dulong's formula is as follows: Btu per pound = 14,544C + 62,028(H - O ÷ 8) + 4,050S. C is carbon, H is hydrogen, O is oxygen, and S is sulfur, all expressed in percent by weight. The coefficients represent the approximate heating values of the respective components in Btu per pound. The term O ÷ 8 for hydrogen is a correction applied to account for the portion of hydrogen combined with oxygen to form water. For a further discussion of Dulong's formula, see Babcock and Wilcox Co., Steam/Its Generation and Use, 40th edition, 1992, p. 9-9.
5Potential carbon dioxide emissions can be calculated by use of the following formula: percent carbon ÷ Btu per pound x 36,670 = pounds (lbs) of carbon dioxide per million (106) Btu. Multiply pounds of carbon dioxide per million Btu by 0.123706 to get million metric tons (MMT) of carbon per quadrillion (1015) Btu.
6Ultimate analysis refers to the determination of carbon, hydrogen, sulfur, nitrogen, oxygen, and ash. By comparison, proximate analysis determines fixed carbon, volatile matter, moisture, and ash. Fixed carbon is principally carbon, but it may contain appreciable amounts of sulfur, hydrogen, nitrogen, and oxygen. Volatile matter comprises hydrogen, carbon dioxide, carbon monoxide, and various compounds of carbon and hydrogen.
8For details, see "Analysis of the Relationship Between the Heat and Carbon Content of U.S. Coals," prepared for the Energy Information Administration, Office of Coal, Nuclear, Electric and Alternate Fuels, by Science Applications International Corp., September 1992.
9Because of the unbalanced nature of the data being analyzed (i.e., unequal numbers of observations for the different levels of the classification variables), the General Linear Models procedure in the Statistical Analysis System was used to perform the analyses.
10The EIA Coal Analysis File did not contain data for bituminous coal in Texas, subbituminous coal in Alaska and New Mexico, or lignite in Arkansas, California, and Louisiana. The emission factor for Alaska subbituminous coal was derived from information obtained from the sole producer of coal in Alaska. The others were assigned appropriate average factors for their coal ranks, as noted in Table FE4.
11For the coal analyzed in the EIA Coal Analysis File, the average hydrogen content was as follows, by weight (dry basis): anthracite, 2.5 percent; bituminous coal, 5.0 percent; subbituminous coal, 4.8 percent; and lignite, 4.4 percent.
12For information on States that produce coal, see Energy Information Administration, Coal Production 1992, DOE/EIA-0118(92) (Washington, DC, October 1993), and State Coal Profiles, DOE/EIA-0576 (Washington, DC, January 1994).
13The amount of coal distributed by State of origin and State of destination is reported on Form EIA-6, "Coal Distribution Report," for consuming sectors other than electric utilities, and on Federal Energy Regulatory Commission (FERC) Form 423, "Monthly Report of Cost and Quality of Fuels for Electric Plants," for utility coal by rank. The amount and energy content of coal consumption by State and sector are detailed in Energy Information Administration, State Energy Data Report, DOE/EIA-0214, published annually.
15Adjustments can be made by multiplying the factors by the estimated percentage of carbon converted to carbon dioxide. This has been estimated as 99 percent by G. Marland and A. Pippin, "United States Emissions of Carbon Dioxide to the Earth's Atmosphere by Economic Activity," Energy Systems and Policy, Vol. 14, (1990), p. 323. EIA's Emissions of Greenhouse Gases in the United States 1985-1990 (DOE/EIA-0573, September 1993) also assumed 99 percent combustion for carbon emission estimates.
16Byproducts include coke oven gas, benzene, creosote, and other hydrocarbons. See, for example, Energy Information Administration, Coke and Coal Chemicals in 1980, DOE/EIA-012(80) (Washington, DC, May 1981), for production and disposition of coal chemical materials.
17Another source, Greenhouse Gas Inventory Reference Manual--IPCC Draft Guideline for National Greenhouse Gas Inventories (IPCC/OECD Joint Programme, 1993), Volume 3, part 2, 1.29, states that on average 5.91 percent of coal going to coke plants ends up as light oil and crude tar, with 75 percent of the carbon in these products remaining unoxidized for long periods.
19Energy Information Administration, Quarterly Coal Report July-September 1993, DOE/EIA-0121(93/3Q) (Washington, DC, February 1994), p. 77; and Quarterly Coal Report October-December 1987, DOE/EIA-0121 (87/4Q) (Washington, DC, May 1988), p. 46.
20Energy Information Administration, Cost and Quality of Fuels for Electric Utility Plants 1992, DOE/EIA-019(92) (Washington, DC, August 1993), and Cost and Quality of Fuels for Electric Utility Plants 1980 Annual, DOE/EIA-0191(80) (Washington, DC, June 1981).
21Information on the rank
of coal distributed to the other industrial and residential/commercial
sectors from States producing more than one rank is not available. Therefore,
based on available EIA data, the following coal ranks were assigned to
distributions of nonutility coal from the following coal-producing States:
Arkansas, bituminous; Colorado, Montana, Washington, and Wyoming, subbituminous;