hence we also report damages per kWh of electricity produced.27 Table 2-9 and Figures 2-7 and 2-8 show damages per kWh for all four pollutants. Mean damages per kWh (2007 USD) from four criteria-pollutant-forming emissions are 4.4 cents per kWh if all plants are weighted equally and 3.2 cents per kWh if plants are weighted by the electricity they generate. The lower figure reflects the fact that larger plants are often less damaging per kWh.28 What is equally important as mean damages is the distribution of damages across plants. As Table 2-9 indicates the 95th percentile of the distribution—damages of 12 cents per kWh—is more than an order of magnitude larger than the 5th percentile. The distribution of damages per kWh (Figure 2-7) is very skewed: There are many coal-fired power plants with low damages per kWh as well as a small number of plants with high damages. Using generation-weighted figures, the damages per kWh from the least damaging 5% of plants were very small: 94% lower than the average coal-fired plant and almost as low as the average damage per kWh at natural gas power plants (0.16 cents). Figure 2-8 maps damages per kWh for each power plant. As in the case of aggregate damages, the plants with lowest damages per kWh are in the West. Plants with the largest damages per kWh are concentrated in the Northeast and the Midwest.

What explains variation in damages per kWh across plants? Damages per kWh associated with a criteria air pollutant (for example, SO2) are the product of emissions per kWh and the damage per ton of pollutant emitted. For the 406 plants examined, variation in damages per kWh is primarily due to variation in pollution intensity (emissions per kWh) across plants, rather than variation in damages per ton of pollutant, which varies with plant location. In the case of SO2, emissions per kWh reflect the sulfur content of the coal burned, adoption of control technologies (for example, scrubbers), as well as the vintage of the plant. Pounds of SO2 emitted per MWh (see Tables 2-10 and 2-11) vary greatly across plants, and this variation explains approximately 83% of the variation in damages attributed to SO2 emissions per kWh. As Table 2-11 indicates, pounds of SO2 and NOx emitted per MWh vary significantly with plant vintage, reflecting the fact that newer plants are subject to more stringent pollution controls. Variation in damages per ton of SO2 emitted (see Table 2-8) accounts for only


It is, however, the case that less than half of the variation in damages is explained by variation in the amount of electricity generated. A regression of damages on net generation yields an R2 = 0.32; the R2 is 0.48 when the logarithms of the variables are used.


The correlation coefficient between damages per kWh and net generation is = −0.26, significant at <0.01 level of significance.

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