TABLE 2-11 Distribution of Pounds of Criteria-Pollutant-Forming Emissions per Megawatt-Hour by Coal-Fired Power Plants, 2005

 

Mean

Standard Deviation

5th Percentile

25th Percentile

50th Percentile

75th Percentile

95th Percentile

SO2

12

11

1.5

5.4

8.9

16

33

NOx

4.1

2.3

1.3

2.6

3.7

4.9

9.0

PM2.5

0.59

0.58

0.092

0.20

0.35

0.81

1.8

PM10

0.72

0.67

0.12

0.28

0.48

0.94

2.1

ABBREVIATIONS: SO2 = sulfur dioxide = NOx, oxides of nitrogen; PM = particulate matter.

concentration-response function for premature mortality instead of Pope et al. (2002), our damages would have been approximately three times as large as what is reported above.

How do our estimates of damages compare with the literature? Levy et al. (2009) estimated the criteria-air-pollutant damages associated with individual coal-fired power plants using a methodology similar to what is used here; however, their estimates of damages are much higher, ranging from $0.02 to $1.57 per kWh, with a median estimate of 14 cents per kWh (1999 USD).30 Converting the results of Levy et al. to 2007 USD, their median estimate is almost 6 times as high as our median estimate of 2.9 cents per kWh (Table 2-9).31 It is, however, possible to reconcile the two sets of estimates. Two notable differences are that Levy et al.’s estimates are based on emissions data for 1999 rather than 2005 and their estimates depend on a concentration-response function for premature mortality based on Schwartz et al. (2008) rather than Pope et al. (2002).32 Emissions of NOx from coal-fired power plants were approximately 50% higher in 1999 than in 2005; emissions of SO2 were approximately one-third higher. The concentration-response function in Schwartz et al. (2008) yields about three times more deaths associated with a microgram of PM2.5 than those estimated using Pope et al. (2002)—the concentration-response function used in APEEP. These differences lead to much higher estimates of mortality associated with PM2.5, and over 90% of the damages associated with air emissions in our study come from PM2.5 mortality. Levy et al. (2009) also performed uncertainty propagation involving asymmetric triangular distributions, which would contribute modest upward bias to the median

30

The mean value of a statistical life used in Levy et al. (2009) was identical to ours—$6 million USD. They reported monetary values in 1999 USD.

31

The figures in Levy et al. (2009) were unweighted by electricity production.

32

The concentration-response function for premature mortality in APEEP is the all-cause mortality function in Pope et al. (2002).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement