ERI home  New sectors home  Impacts home           Last updated 3/27/03

Greenhouse Gas Estimates for Selected Industry Sectors

The most detailed source of information on U.S. greenhouse gas (GHG) emissions is the Inventory of U.S. Greenhouse Gas Emissions and Sinks, issued by the EPA Office of Atmospheric Programs. This source provides detailed emissions data, broken down by industrial process, for about a dozen processes whose contributions to total GHG emissions are of particular significance.

In a few cases (cement manufacture, for example), the Inventory’s breakdown by process corresponds to breakdown by NAICS code sector. However, in general, it is not possible to derive sector-specific information directly from the process data provided in the Inventory.

A process-specific breakdown is very useful for gaining an understanding of GHG emissions from a material flow standpoint. However, there are circumstances under which a sector-specific breakdown may be desirable. For example, when designing programs for encouraging the deployment of resources to decrease overall GHG emissions, it may be most effective to consider groups of companies that are under similar economic constraints in order to evaluate the most cost-effective approaches. Companies in the same sector often must operate under similar market conditions, and can be conveniently considered as a unit. It would therefore be useful for such purposes to develop a sector-specific analysis, to help quantify the potential for improvement that might be expected to result from programs targeted to specific sectors.

The approach adopted for this analysis is based on the fact that, for many sectors, the total contribution of the sector processes to overall GHG emissions is dominated by the carbon dioxide generated from fossil fuel combustion. Information on energy consumption for specific fossil fuel types, broken down by sector, is available from the Energy Information Administration (EIA) of the U. S. Department of Energy. The carbon content of each fossil fuel type is known. Since virtually all of the carbon consumed when fuel is burned is released to the atmosphere in the form of carbon dioxide, it is possible to convert fuel usage figures into GHG emissions for each sector covered by the EIA energy usage information.

It should be noted that EIA also provides estimates of annual U. S. GHG emissions.  However, just as in the case of the EPA Inventory, the EIA estimates for GHG emissions do not provide data down to the level of individual NAICS codes.  It is still useful to compare the EIA with the EPA numbers.  Such comparisons provide consistency checks for conversion factors, and for the respective assumptions on which the separate data sources are based.  For example, the total greenhouse gas emissions in 2000 are given in the EIA document Emissions of Greenhouse Gases in the United States:  2000 Summary as 1,906.3 million metric tons carbon equivalent.  This may be compared with the total for 2000 listed in the EPA Inventory document, given as 7,001.2 Tg CO₂ equivalent.  One teragram (Tg) is the same as one million metric tons (10⁶ metric tons X 10⁶ grams per metric tons = 10¹² grams = 1 teragram.)  Using the factor 44/12 to convert from carbon (atomic weight 12) to carbon dioxide (molecular weight 44) changes the EIA total, expressed as Tg CO₂ equivalent, to 6,989.8, in reasonable agreement with the EPA number.  (EIA data were used in part to derive the Inventory estimates, so this degree of agreement is not surprising, but it is still a useful check on the conversion factors used here.)

Sectors covered by EIA data

The table below lists the sectors of interest for this analysis, together with the corresponding sectors and NAICS codes for which EIA fossil fuel consumption data are available.  In some cases, the mapping is not exact, but a reasonably close correspondence has been selected.

Table 1.  Mapping between sectors of interest and NAICS codes for sectors with EIA energy consumption data

Those sectors for which no EIA data are available are, for the most part, less significant contributors to GHG emissions, and will be omitted from consideration below.  An exception is the utility sector, which is treated separately.

Conversion to carbon dioxide equivalent units

Turning to the EIA fossil fuel consumption data, the first task is to convert the numbers (given both in physical units such as tons, barrels, and cubic feet, and in energy equivalents expressed in Btu) into units suitable for GHG emission comparisons.  The EPA units, Tg CO₂ equivalents, will be used in what follows.

Carbon content values for various fuel types, in terms of Tg carbon per quadrillion Btu, are provided in Annex A of the Inventory, in Tables A-14 and A-15, pp. 249-250.  These values are converted to Tg CO₂ per quadrillion Btu with the factor of 44/12, as in the previous section.

The result of applying these conversion factors to the sectors of interest covered by the EIA data is summarized in the table below.

Table 2.  Impacts of energy consumed by sectors of interest on greenhouse gas emissions, 1998 EIA data, converted to Tg CO₂ equivalent.  (See Table 3 below for additional detail on fuel types and conversion factors.)

*In a few cases, EIA has withheld data on specific fuel types to avoid revealing the identity of a particular facility.  The total consumption for the sector is, however, provided by EIA.  For the purpose of this analysis, the difference between the total and the columns for which fuel type is known has simply been apportioned equally among the unknown quantities.  (This is an arbitrary procedure, and could be improved by using known information about the specific sector to adjust the proportions, but since the results will not materially affect the conclusions, such a refinement has not been carried out here.)

The fuel types are spelled out in more detail below, along with the conversion factors used.  In some cases, the conversion factor represents an estimate, since the category does not refer to a well-defined material.  More details are available from the EIA documents, available in spreadsheet or in HTML formats.

Table 3.  Conversion factors for fuel types

The calculation of CO₂ emissions from fuel consumed by the electric utility sector was carried out using the same conversion factors, with primary data from the EIA Annual Energy Review, and specifically from Table 2.1f, Electric Power Sector Energy Consumption, 1949-2000.  The total CO₂ emission due to the electric utility sector for 2000 works out to 1926.8 Tg CO₂.

Non-Fuel Greenhouse Gas Emissions

Certain industries are responsible for significant GHG emissions from processes other than fuel combustion.  The EPA Inventory document includes a detailed presentation summarizing U.S. emissions, in Tg CO₂ equivalent, on a process-by-process basis.  In this section, the process information from the Inventory will be mapped to the corresponding sectors.

Table 4 includes the sectors of interest here, along with the processes listed in Table 3-1 of the Inventory that are assumed to be associated with them, and the GHG emissions for 2000, in Tg CO₂ equivalent.  Some rearrangement of the data as presented in the Inventory is necessary, since in some cases processes associated with the same sector of interest are included in the Inventory in separate categories, involving different greenhouse gases.  In addition, in some cases the emissions shown represent only the fractions of the amounts in Inventory, Table 3-1 that are attributable to the sectors of interest for this analysis.

Table 4.  Major industrial non-fuel GHG Emissions, 2000, in Tg CO₂ equivalent

^(a)Hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur hexafluoride (SF₆)

^(b)According to a statistical summary from the US Geological Survey (USGS, see index page for Commodity Statistics and Information, under nitrogen), about 89% of domestic ammonia consumption was for fertilizer use.  Total included in this table is 89% of total listed in Inventory, Table 3-1.

^(c)Lime manufacturers have been included in the stone, clay, & glass sector, rather than the chemical sector, since their NAICS code, 32741, falls within the 327 code for which fuel consumption data are available.

^(d)According to the Inventory, p. 92, 27% of soda ash produced in the U. S. was used in chemical production.  Total included in this table is 27% of total listed in Inventory, Table 3-1.  The same source indicates that the pulp and paper sector used about 2% of total soda ash production, which is below the limit of resolution of the table.

^(e)The NAICS code for petrochemical manufacturing, 32511, puts it within the 325 code (chemical manufacturing), rather than the 324 code (petroleum and coal products manufacturing).  The associated GHG emission is thus attributed to the chemical sector, for consistency with fuel consumption data.

^(f)According to an EPA analysis, nitric acid is used mostly for fertilizer and explosives production, with a lesser amount used for adipic acid production, and other industrial uses.  On page 4 of the EPA analysis, 80% of the nitric acid produced in the U.S. is used to produce ammonium nitrate, 9% for adipic acid, and 11% for aniline and toluene.  Assuming that most of the ammonium nitrate is used for fertilizer, most of the adipic acid for plastics, and most of the aniline and toluene for explosives and other industrial uses, the table attributes 80% of the nitric acid production total in  Inventory, Table 3-1 to the agricultural chemicals sector, 9% to the plastics sector, and 11% to the chemicals sector.

^(g)According to an EPA analysis, most adipic acid is used in the production of nylon and other plastics.  Since fuel consumption data are available for NAICS code 325211, plastics materials and resins, it has been broken out as a separate category.  Note that, although adipic acid is produced from nitric acid, which is associated with a GHG emission of its own, there is no double counting.  Catalytic oxidation of ammonia is used to produce nitric acid.  In the process, some of the nitrogen present in the ammonia is released as nitrous oxide (the GHG emission), while some is incorporated into the nitric acid product.  In adipic acid production, the remaining nitrogen is released, and that fraction which is released as nitrous oxide and is not recaptured is counted as a separate GHG emission.

The following processes are included as major sources in Inventory Table 3.1, but are not considered to be significantly attributable to the sectors of interest for this analysis:

• carbon dioxide consumption (1.4 Tg CO₂):  A fraction of this total is probably attributable to sectors of interest, particularly the chemical sector, but the amount is uncertain according to the Inventory, and would probably be below the limit of resolution of the table.
• silicon carbide production:  Amount below the limit of resolution of the table.
• magnesium production and processing

Summary of Results

We summarize the results in two tables below.

• Table 5 is lists the sectors covered by the analysis in order of their total GHG emissions.
• Table 6 gives a profile of GHG emissions according to fuel type for each of the sectors.   The entries in the table represent the percentage of the total GHG emissions for that sector due to each fuel type.  Breaking down the data in this way may be useful in evaluating potential benefits from fuel conservation and substitution strategies.

The information on emissions due to energy consumption in table 2 is given in some cases for three digit NAICS codes, and in other cases for four and six digit subcategories.  The impact information in table 2 follows the EIA data in including all emissions at each level.  Thus, the three digit lines include the contributions from the four through six digit lines that are their subcategories.  In Table 5, the data that appears in four and six digit subcategories has been subtracted from the three digit lines, to avoid double counting.  This situation occurs in Table 5 for NAICS codes 325 (chemicals) and 327 (nonmetallic mineral products).  (The non-fuel GHG data in Table 4 has already been separated according to sectors of interest, so no subcategory subtractions are necessary.)

For the case of electric utilities, no attempt has been made to subtract off the electric energy already accounted for by industrial use in other sectors of interest.  The number would not be meaningful without some measure of how much of the electricity consumed by a particular sector was generated by fossil fuels, and how much by non-GHG sources (hydroelectric power, for example).  This can vary considerably from sector to sector (and probably from day to day within a given sector).  Thus, such a refinement would not be meaningful.  For that reason, there is some unavoidable double counting between the first line of Table 5 and the remaining lines.

Table 5.  Sectors listed in order of quantity of GHG emissions, in units of Tg CO₂ equivalent.

*Excludes petrochemicals.

**Excludes plastics, rubber, and agricultural chemicals, includes petrochemicals.

***Excludes cement.

Table 6.  Profile of GHG emissions due to fuel consumption

Fuel types accounting for >10% of a given sector's consumption in boldface.

*Estimates of withheld data (see Table 2, above).  The data on the phosphatic fertilizer sector is particularly in need of refinement, given the relatively large percent involved.