On March 24, 2011, the U.S. EPA released long-awaited guidance on modeling approaches to demonstrate compliance or address nonattainment with the 1-hour primary National Ambient Air Quality Standard (NAAQS) for sulfur dioxide (SO2). The new 1-hour standard was set at the level of 75 parts per billion (ppb) calculated as the three-year average of the 99th percentile (typically the 4th high value) of the annual distribution of daily maximum 1-hour average concentrations. The modeling guidance described the basics of how modeling should be conducting using AERMOD. The remainder of the guidance focused on attainment designations and strategies.

In the publication of the final rule, the U.S. EPA states that attainment with the 1-hour SO2 NAAQS is demonstrated through a combination of compliant monitoring data and/or dispersion modeling results. Since the existing SO2 monitoring network is so sparse, the U.S. EPA expects that the identification of NAAQS violations and compliance with the 1-hour SO2 NAAQS, in those areas without currently operating SO2 monitors, will primarily be done through refined, source-oriented air quality dispersion modeling. This modeling would initially focus on larger sources of SO2 (100 tons per year of emissions or more), but could be expanded to include any SO2 source (e.g. smaller but poorly dispersed stacks) that could reasonably be expected to contribute to a violation of the NAAQS. Given the relatively short time period between promulgation of the NAAQS and initial area designations (due June 3, 2011), states have not been able to complete detailed compliance analyses. As such, it is anticipated that states will propose that most areas be designated as “unclassifiable”. Going forward, states may even elect to delay modeling until such a time when the U.S. EPA makes such demonstrations mandatory in nature.Those areas with existing monitor violations (e.g. New Hanover County, NC), may be forced to designate certain areas as nonattainment. The states are working quickly to determine which sources are potential contributors to any violations through modeling. Those modeling results will likely be used by the agencies to support their proposed area designations, should they choose to define areas other than the traditional approach of using county or other local jurisdictional boundaries.