Modeled response of ozone to electricity generation emissions in the northeastern United States using three sensitivity techniques

ABSTRACT

Electrical generation units (EGUs) are important sources of nitrogen oxides (NO_x) that contribute to ozone air pollution. A dynamic management system can anticipate high ozone and dispatch EGU generation on a daily basis to attempt to avoid violations, temporarily scaling back or shutting down EGUs that most influence the high ozone while compensating for that generation elsewhere. Here we investigate the contributions of NO_x from individual EGUs to high daily ozone, with the goal of informing the design of a dynamic management system. In particular, we illustrate the use of three sensitivity techniques in air quality models—brute force, decoupled direct method (DDM), and higher-order DDM—to quantify the sensitivity of high ozone to NO_x emissions from 80 individual EGUs. We model two episodes with high ozone in the region around Pittsburgh, PA, on August 4 and 13, 2005, showing that the contribution of 80 EGUs to 8-hr daily maximum ozone ranges from 1 to >5 ppb at particular locations. At these locations and on the two high ozone days, shutting down power plants roughly 1.5 days before the 8-hr ozone violation causes greater ozone reductions than 1 full day before; however, the benefits of shutting down roughly 2 days before the high ozone are modest compared with 1.5 days. Using DDM, we find that six EGUs are responsible for >65% of the total EGU ozone contribution at locations of interest; in some locations, a single EGU is responsible for most of the contribution. Considering ozone sensitivities for all 80 EGUs, DDM performs well compared with a brute-force simulation with a small normalized mean bias (–0.20), while this bias is reduced when using the higher-order DDM (–0.10).

Implications: Dynamic management of electrical generation has the potential to meet daily ozone air quality standards at low cost. We show that dynamic management can be effective at reducing ozone, as EGU contributions are important and as the number of EGUs that contribute to high ozone in a given location is small (<6). For two high ozone days and seven geographic regions, EGUs would best be shut down or their production scaled back roughly 1.5 days before the forecasted exceedance. Including online sensitivity techniques in an air quality forecasting model can provide timely and useful information on which EGUs would be most beneficial to shut down or scale back temporarily.

Acknowledgments

The authors thank Kirk Baker (EPA) for his help in obtaining the CAMx simulations and with configuring the direct decoupled method within CAMx.

Additional information

Funding

This publication was made possible by EPA STAR grant number 8351901. Its contents are solely the responsibility of the grantee and do not necessarily represent the official views of the EPA. Further, the EPA does not endorse the purchase of any commercial products or services mentioned in the publication.

Notes on contributors

Evan Couzo

Evan Couzo is an assistant professor in the Department of Education at the University of North Carolina at Asheville.

James McCann

James McCann contributed to this work while completing an MSEE in the Department of Environmental Sciences & Engineering at the University of North Carolina at Chapel Hill, and is presently an associate with Ramboll Environ in Arlington, VA

William Vizuete

William Vizuete is an associate professor in the Department of Environmental Sciences & Engineering at the University of North Carolina at Chapel Hill.

Seth Blumsack

Seth Blumsack is an associate professor in the Department of Energy and Mineral Engineering at Penn State University.

J. Jason West

J. Jason West is an associate professor in the Department of Environmental Sciences & Engineering at the University of North Carolina at Chapel Hill.