APPENDIX: METHODOLOGICAL APPENDIX

CO2 Emissions Data

The Clean Air Act Amendments of 1990 require electric utilities to install on power plant stacks Continuous Emission Monitoring Systems, which measure each stack's emissions of sulfur dioxide, nitrogen oxides, and carbon dioxide. The requirement is being phased in, and in 1995 virtually all utility-owned units of over 25 megawatts reported their emissions to EPA.

We based our analysis on the data we received from EPA. But because this year many units were reporting for the first time, we also subjected the data to our own quality check. From the Energy Information Administration's Form 759, we obtained generation data for each power plant which reported to EPA. We used average efficiency and carbon dioxide emission factors to obtain a very rough estimate of the carbon dioxide emitted from the plants.^[28] We then compared these rough estimates to the measurements reported to EPA. We flagged as suspicious the five percent of plants which emitted over one million tons according to either EIA or EPA data, and for which the EIA estimates and EPA reported values differed by more than a factor of two. For these plants, we turned to their FERC Form 1 submissions to determine exactly how many British Thermal Units (BTUs) were burned of each type of fuel, and then used average carbon emission factors to estimate the CO2 emitted.^[29] The EIA and FERC estimates agreed for the most part, and because both forms have been used for many years and utilities have independent financial reasons to monitor their generation and BTU consumption, we used the FERC-based estimates in these cases of discrepancies.

Concerns also have been raised about possible calculation problems with some Continuous Emission Monitors, particularly when there is turbulent flow in the stack. It has been suggested that reported emissions for some units could be biased high by 10-30 percent. EPA and the Electric Power Research Institute are investigating this issue and have reached no conclusions at this time. If a problem is confirmed, calibration methods may be revised, but we do not expect any resulting adjustments to affect the overall rankings significantly.

Ownership Data

For ownership data, we relied on the Energy Information Administration's latest Inventory of Power Plants in the United States. We also used EIA's Inventory of Power Plants to identify jointly owned units and apportion the emissions from these units among their owners. (For example, a utility owning 25 percent of a given unit would be assigned 25 percent of the carbon emissions from that unit.) The most recent ownership data available are for 1994, but we have no reason to believe that the results were affected significantly by subsequent transactions.

Revenues Data

We relied on utilities' Form 1 submissions to the Federal Energy Regulatory Commission for their 1995 electric operating revenues. For the Tennessee Valley Authority, Bonneville Power Administration, New York Power Authority, and the major consumer-owned utilities, which do not submit FERC Form 1, we used their annual reports.

Future Data Sources

The rankings presented here are based on preliminary data as reported to EPA from the continuous emission monitoring system established under the Clean Air Act amendments of 1990. In addition to CO2, continuous monitoring is also conducted for NOx and SO2 for virtually all electric generators with a capacity over 25 megawatts. Although this monitoring system was given a test run for most of the largest power plants in 1994 (Phase I Affected Units under Title IV of the Clean Air Act), 1995 was the first year for which virtually all power plants over 25 megawatts were required to report their emissions. This start up process has resulted in inevitable delays in acquiring quality-assured data as generators established their monitoring and reporting procedures and EPA established its quality assurance, data reduction and distribution systems. Later this Fall, EPA plans to release its final 1995 emissions inventory for all three monitored pollutants and NRDC plans to expand the coverage of its environmental liability rankings to include the financial liabilities associated with NOx and SO2 emissions as well as CO2. NRDC also plans to update the rankings in future years and we expect to be able to reduce the time lag between the end of the reporting period and the availability of the data needed to compute the rankings.

Notes

28. For each plant, the EIA database reported the generation by fuel type. For this rough screening calculation, we assumed that each unit had an efficiency of 10,000 British Thermal Units (BTU) burned to generate 1 kilowatt-hour of electricity. We then used average emission factors of 0.104 tons of CO2 emitted per million BTU for coal, 0.087 tons per Million BTU for oil, and 0.058 tons per million BTU for gas.

29. Again, we used average emission factors of 0.104 tons of CO2 emitted per million BTU for coal, .087 tons for oil, and 0.058 tons for gas.