Using Backup Generators for Meeting Peak Electricity Demand: A Sensitivity Analysis on Emission Controls, Location, and Health Endpoints

摘要

Generators installed for backup power during blackouts could help satisfy peak electricity demand; however, many are diesel generators with nonnegligible air emissions that may damage air quality and human health. The full (private and social) cost of using diesel generators with and without emission control retrofits for fine par-ticulate matter (PM_(2.5)) and nitrogen oxides (NO_x) were compared with a new natural gas turbine peaking plant. Lower private costs were found for the backup generators because the capital costs are mostly ascribed to reliability. To estimate the social costs from air quality, the changes in ambient concentrations of ozone (O_3) and PM_(2.5) were modeled using the Particulate Matter Comprehensive Air Quality Model with extensions (PMCAM_x) chemical transport model. These air quality changes were translated to their equivalent human health effects using concentration-response functions and then into dollars using estimates of "willingness-to-pay" to avoid ill health. As a case study, 1000 MW of backup generation operating for 12 hr/day for 6 days in each of four eastern U.S. cities (Atlanta, Chicago, Dallas, and New York) was modeled. In all cities, modeled PM_(2.5) concentrations increased (up to 5 μg/m~3) due mainly to primary emissions. Smaller increases and decreases were observed for secondary PM_(2.5) with more variation between cities. Increases in NO_x emissions resulted in significant nitrate formation (up torn1 μg/m~3) in Atlanta and Chicago. The NO_x emissions also caused O_3 decreases in the urban centers and increases in the surrounding areas. For PM_(2.5), a social cost of approximately $2/kWh was calculated for uncontrolled diesel generators in highly populated cities but was under 10 ￠/kWh with PM_(2.5) and NO_x controls. On a full cost basis, it was found that properly controlled diesel generators are cost-effective for meeting peak electricity demand. The authors recommend NO_x and PM_(2.5) controls.