EPA (see table 5-3) offers the following breakdown for the year 2000 (on an estimated total of 5858 millions of tons of CO2 emissions from the United States not including land use changes) : EPA has also just released a new and more detailed inventory of all greenhouse gas emissions, updated through 2005.
| Heating/Direct | Electric/Indirect | Total | |
| Residential | 6.4% | 12.8% | 19.1% |
| Commercial | 4.1% | 12.2% | 16.3% |
| Industrial | 18.5% | 11.5% | 29.9% |
| Transportation | 32.2% | 0.2% | 32.3% |
| Non-energy uses | 2.3% | 0.0% | 2.3% |
| Total | 63.3% | 36.7% | 100.0% |
CAIT offers a slightly different breakdown and slightly different total for the United States in 2000 (5,774 millions of tons of CO2 not including land use changes):
| Electricity & Heat — power plants, auto producers, refineries | 46.5% |
| Manufacturing & Construction | 11.4% |
| Transportation | 29.8% |
| Fuel combustion in Residential, Commercial, Agricultural and other | 11.1% |
| Fugitive Emissions | 0.4% |
| Industrial Processes | 0.8% |
| Total | 100.0% |
These two breakdowns are roughly consistent: CAIT includes some industrial uses within the “electricity and heat category” — see CAIT definitions at page 15. If one combines the “electricity and heat” and the “manufacturing and construction” categories in the CAIT and compares that total (57.9%) to the electric/indirect total plus the industrial direct amount from the EPA (55.2%), one is probably comparing slightly different estimates for roughly the same emissions sources. There is no apparent reason why the transportation sectors should differ, but the numbers are close — 32.3% vs. 29.8%. Finally, the heating/direct components from the EPA total 10.5%, close to the 11.1% for other fuel combustion from the CAIT.
The federal EIA’s Monthly Energy Review is another good source. According to MER Table 1.3, total consumption of energy in the United States was 99.7 quadrillion BTUs in 2006. According to MER Table 2.1, 6.4 were generated within residences, and another 14.7 were consumed as electric power delivered to residences. This happens to check reasonably well with the EPA proportions above.
However, shares of BTUs used do not translate directly to shares of carbon emissions because different fuels have different carbon/BTU ratios — natural gas yielding about 25% less carbon than petroleum fuels and coal about 25% more. Applying Carbon/Energy coefficients for different fuels from EPA Greenhouse Gas Emissions Inventory, Table A-204 to BTU’s consumed in the energy sector as in MER Table 2.6 yields an estimate of carbon production for electricity production which is 15% below the CAIT number, but this is probably due to the sector definition differences. Compare a similar conversion usingcarbon coefficients for power consumption (see table 2) and MER Table 7.6, which comes in 8% below the CAIT number.) For the transportation sector, the difference from the CAIT totals is slightly in the other direction — see MER Table 2.5 and convert using the EPA coefficients. On a straight BTU basis, of the total US energy consumption per the MER, 28.4 QBTU’s (28.5% of the total) were consumed in the transportation sector, according to MER Table 2.1 Of that, in turn, almost all, 27.7 QBTUs, is from petroleum product consumption, per MER Table 2.5.
For additional detailed data on carbon emissions and climate change, see the Carbon Dioxide Information Analysis Center of the Oak Ridge National Laboratory.
Note that electric power statistics include an allocated share of the approximately 80% of energy input lost in thermal production and transmission of electricity — see MER notes to section 2, note 11.) The International Energy Agency charges for most of its data, but does make a summary available online.
For an elegant mapping of world emissions for all sectors and including other greenhouse gases, see the Stern Report, Chapter 7, Figure B at page 174.