In general, this report includes all known significant anthropogenic sources of greenhouse gases . However, a number of potential emissions sources have been excluded due to lack of data, or because they have been defined as naturally occurring.
As the product of atmospheric uptake by organic matter, the carbon present in biofuels is, arguably, a component of the terrestrial carbon cycle . Thus, no net additions to the global carbon budget result from their combustion. Carbon emissions from biofuels consumption are, therefore, excluded from the emissions estimates in this report. However, had they been included, they would have raised the estimate by about 70 million metric tons of carbon.
While emissions from the combustion of municipal solid waste grew almost threefold between 1985 and 1990 and emissions from the consumption of alcohol fuel almost doubled, those increases were more than offset by a large drop in emissions from wood combustion, the largest portion of biofuel consumption (Table D1).
Estimates of emissions from the combustion of municipal waste were calculated using an emissions coefficient of 24.7 million metric tons of carbon per quadrillion Btu. (209) Quantities of municipal solid waste burned are found in the EIA report, Estimates of U.S. Biomass Energy Consumption 1992.
Emissions from the consumption of alcohol fuels were calculated using an emissions coefficient of 19.67 million metric tons of carbon per quadrillion Btu, based on the carbon content by weight of the ethanol molecule. (210) Ethanol is 52.2 percent carbon by weight, and it was assumed that 99 percent of its carbon is oxidized when ethanol is consumed.
Emissions from the use of wood fuels were calculated using an emissions coefficient of 27.43 million metric tons of carbon per quadrillion Btu. (211) Data for alcohol fuel and wood fuel consumption were unavailable for the years 1985, 1986, 1988, and 1991. Consumption was interpolated from data for the surrounding years.
The volume of crop residues and agricultural waste burned in the United States is very large. One estimate, drawn from the scientific literature, put the amount at about 300 million metric tons in the mid-1970s. (212) The authors suggested that about 10 percent of the material was left unburned, and that the carbon content of agricultural waste was about 45 percent by weight, implying annual carbon emissions from this source of about 120 million metric tons, equal to 8 to 9 percent of emissions from fossil fuels . However, as in the case of biofuels, the carbon emitted from this source would, in principle, be reabsorbed by the growth of crops during the next crop cycle and, hence, would not constitute a net source of emissions.
Methane and nitrous oxide emissions from burning agricultural wastes would be considered a controllable anthropogenic source of emissions. The latest draft Intergovernmental Panel on Climate Change (IPCC) guidelines suggest procedures for estimating carbon dioxide, methane, and nitrous oxide emissions from agricultural burning. The major uncertainty in this effort is estimating the extent to which stubble is actually burned in the United States, as open burning is prohibited by air quality regulations in much of the United States.
Using the default coefficients provided in the IPCC draft report, it is possible to produce estimates of the amount of methane and nitrous oxide that might be emitted from agricultural burning. A calculation of this nature suggests that methane emissions would be on the order of 50,000 metric tons annually, while nitrous oxide emissions would be on the order of 5,000 metric tons annually.
An EPA report indicates that coal mine operators burn coal waste (mined material with an energy content too low to qualify as a commercial product) to dispose of it. (213) However, available information dates from the early 1970s, and the EPA report indicated that "the number of coal refuse piles had decreased to negligible by 1975." There are no data, and we have found no recent evidence to indicate that this activity continues to take place on any significant scale.
Forest fires are a large source of both carbon dioxide and methane emissions. However, they are both a temporary source of carbon dioxide and usually a biogenic one.
Fermentation is a chemical process in which complex organic compounds, such as sugar, are broken down into simpler substances, namely, alcohol and carbon dioxide. The carbon emitted during this process does not represent a net source of emissions, however, since the agricultural materials used during fermentation are renewable.
A recent EPA draft report has suggested that methane emissions from human waste treatment plants might produce emissions of 200,000 metric tons annually. The EIA has not yet been able to investigate methods of estimating emissions from this source, but they will probably be included in next year's report.
Table D2 reports estimated oil consumption by U.S. military forces. Energy consumption by the U.S. military is something of an anomaly in energy statistics. Domestic military energy consumption is included in domestic energy consumption statistics. Overseas U.S. military energy consumption of nonpetroleum fuels and some petroleum sources is generally included in the national energy statistics of the countries (such as Britain, Germany, Japan, and Korea) where the U.S. military operates. However, a substantial portion of U.S. military oil consumption is treated as an export in national energy statistics, but without any corresponding record of imports or consumption, since the fuel is loaded on tankers (recorded as an export) and then transferred directly to U.S. warships and military facilities without ever reappearing in the energy statistics of any country.
The amount of oil involved is problematic, because the Defense Department does not keep its books for the purpose of clarifying ambiguities in international energy statistics. Total oil consumption by U.S. military forces in 1990 amounted to about 500,000 barrels per day, according to records of fuel sales by the Defense Department's Defense Fuel Supply Center. The Defense Department does not maintain central records of the location of fuel sales, but reports that 65 to 75 percent of this amount (approximately 350,000 barrels per day) was acquired domestically, and hence was presumably counted in U.S. domestic oil consumption. The accuracy of these calculations may be impaired by the fact that estimates of the percentage acquired domestically is based on bulk fuel purchases, while consumption estimates are based on sales figures. Motor gasoline is presumably acquired almost entirely for motor vehicles, and is largely counted in national energy statistics, whether domestic or foreign. However, most of the oil consumption by the U.S. military consists of jet fuel and "middle distillates" used both to power ships and diesel vehicles.
A reasonable "order of magnitude" estimate for U.S. military oil consumption not reported elsewhere might be 25 percent of total military oil consumption of jet fuel, middle distillates, and residual oil. Table D2 illustrates carbon emissions associated with this estimate: approximately 5 million metric tons of carbon per year. The range of uncertainty associated with this estimate is on the order of 20 percent, i.e., between 20 and 30 percent of total military oil consumption.
U.S. military fuel consumption has declined rapidly since 1991. It is likely that the division between domestic and foreign consumption has also shifted, with a larger portion of total consumption accounted for domestically.
Natural gas , as found in nature, usually consists predominantly of methane (CH4) but is actually a mixture of several different hydrocarbon and nonhydrocarbon gases in varying proportions. A typical blend of raw gas might be 85 percent methane, 14 percent other hydrocarbons (particularly propane, butane, and ethane), and 1 percent carbon dioxide. The precise proportions of the blend are unique to each of the tens of thousands of natural gas reservoirs in the United States. Natural gas as shipped by pipeline, however, must be a standard product of fairly precise quality and must consist largely of methane.
If the quality of raw natural gas varies appreciably from pipeline quality, it must be treated in a gas processing plant. In the plant, valuable heavier hydrocarbons are stripped out for separate sale as "natural gas liquids ." Hydrogen sulfide gas, a common contaminant, is always removed, and the sulfur is frequently recovered for separate sale. Excess carbon dioxide (if any) is also removed and recovered for industrial use or vented to the atmosphere. In natural gas statistics, these removals account for most of the difference between raw or "wet" gas production and marketed "dry" gas production.
In 1990 some 289 billion cubic feet of nonhydrocarbon gases were removed from natural gas production in 25 States. (214) Eight gas-producing States (including Louisiana, a major producer) do not report on nonhydrocarbon removals. While no State reports on the breakdown among nonhydrocarbon gases, Texas reports on commercial carbon dioxide recovery from gas plants. In the early 1980s, commercial recovery was less than 10 percent of total nonhydrocarbon gas removals. Since 1988, the value has ranged from 75 percent to 95 percent of removals and has totaled 1.3 million to 2.1 million metric tons of carbon per year. (215)
It is possible to make crude estimates of carbon dioxide removals from rich gas for other States based on the Texas data. Table D3 illustrates an attempt to do this. In Texas, nonhydrocarbon gas extraction accounts for about 2 percent of gross natural gas production. In Table D3, nonhydrocarbon gas extraction is assumed to be 2 percent of gross production for the eight nonreporting States. This produces a 33-State estimate of nonhydrocarbon gas extraction. In Texas, commercial production of carbon dioxide from gas plants is as much as 95 percent of nonhydrocarbon gas extraction from 1988 through 1992. If it is assumed that other States have a similar ratio (90 percent), then national carbon dioxide extraction would be on the order of 400 billion cubic feet per year. From this amount, commercial carbon dioxide recovery from Texas is deducted (to avoid double counting with industrial carbon dioxide production). Commercial carbon dioxide recovery from gas plants in other States (if any) is double counting, although no data in this area have yet been found.
This very speculative estimation procedure produces net carbon dioxide emissions of 6 million metric tons in 1985, declining to 4 million metric tons by 1991. The main source of the decline is increasing commercial use of carbon dioxide in Texas. Some of this gas was vented to the atmosphere, some was used for industrial purposes, resulting in its being vented to the atmosphere, and an indeterminate (but probably large) volume was captured for reinjection into oil reservoirs.
This estimate has not been included in the main report because of its excessive uncertainty. The entire estimate is really based on a single ratio: the ratio of commercial carbon dioxide recovery to nonhydrocarbon gases removed in Texas. There is no real basis for assuming that the nonhydrocarbon gas share of nonreporting States is equal to the share of those that do report, nor is there any real basis for assuming that the carbon dioxide share of nonhydrocarbon gases outside of Texas equals that within Texas. Finally, there is no basis for determining how much carbon dioxide is recovered for commercial use or in enhanced oil recovery outside of Texas. There are, however, some emissions from this source, and this estimate provides a first approximation of what their magnitude might be.
In the past 15 years, carbon dioxide gas has become a very popular substitute for natural gas in pushing crude oil out of its underground reservoirs. Typically, an oil company will acquire carbon dioxide from a dedicated carbon dioxide pipeline and inject the carbon dioxide into the top or periphery of an oil reservoir. Under the increased pressure, more oil is squeezed out of the ground and into producing wells. In the first instance, there are only incidental emissions of carbon dioxide, because all of the carbon dioxide goes back into the ground.
However, not all carbon dioxide can be expected to stay underground. As time passes, and more and more oil is produced from a particular reservoir, eventually the carbon dioxide gas will start to leak into the producing well, which will then begin to produce a mixture of oil, carbon dioxide, and natural gas. As long as the energy content is sufficiently high, the gaseous portion of the mix will probably be shipped to a gas plant. Low-Btu gas, however, is prone to be vented or flared. If the amount of oil produced drops too low, the well will be sealed and abandoned.
Conceptually, large volumes of carbon dioxide may be vented or flared by this process. However, the EIA has no basis for estimating the amounts. The total quantity of carbon dioxide used for enhanced oil recovery is not known with precision, but it is probably on the order of 12 million metric tons annually. (216) Carbon dioxide emissions from this source can only be a fraction of the total.
EPA reports that limestone is also calcined as part of the lead smelting process. (217) No data on the amount of limestone involved have yet been found, but the volume is likely to be very small, since less than 400,000 tons of lead is smelted annually.
From 1985 to 1990, Unocal Corp. operated a shale oil plant at Parachute Creek, Colorado. The plant produced some 4.6 million barrels of synthetic crude oil during this period. In order to extract synthetic crude oil from shale, the shale is normally crushed and heated in a giant retort. The heat is provided by burning the residual hydrocarbons in the waste shale. Since shale is not a commercial fuel, consumption is not normally recorded in EIA energy statistics; hence, emissions from this source are not included in the inventory. Based on some reasonable assumptions about the amount of carbon in shale oil and typical process efficiencies of other retorting process, the volume of carbon emitted was probably on the order of 30,000 metric tons in 1990. (218) As additional information is developed, this small source will be included in a future report.
Commercial geothermal power plants typically generate electricity by passing geothermally pressurized steam through a turbine. After passing through the turbine, the steam condenses into water. At the most prolific geothermal site in the United States (the Geysers, near Guerneyville, California) the geothermal steam contains a substantial amount of dissolved carbon dioxide. This carbon dioxide is emitted to the atmosphere. A "back-of-the-envelope" computation suggests that if all geothermal power in the United States were generated at the Geysers, the amount of carbon emitted would be on the order of 70,000 metric tons. (219) With better documentation, this small source will be included in a future report.
