|By Robynn Andracsek, P.E., Burns & McDonnell|
When it comes to the National Ambient Air Quality Standards (NAAQS), dispersion modeling is protective but not necessarily predictive of actual air quality. New shorter 1-hour averaging periods compound this issue and lead to restrictive and intricate permit conditions that have little to do with actual air quality.
When the 1-hour NO2 and SO2 NAAQS were set, it was only a matter of time before coal plants would have SO2 emission limits with averaging periods to match NAAQS modeling. In May 2013, the Homer City power plant in Pennsylvania was given a “precedent setting” SO2 limit of 6,360 pounds per hour for its three coal-fired boilers, including during periods of startup and shutdown. Also included was a prohibition on allowing more than one unit to start up at the same time. This last condition smacks of a dispersion modeling work-around. Likely, during the worst meteorological conditions the higher startup emissions from multiple boilers modeled as a violation of the 1-hour SO2 NAAQS. However, it is extremely unlikely that in the real world such a startup scenario would occur under those precise meteorological conditions. Nevertheless, this example encapsulates the disparity between modeling and reality.
At its heart, dispersion modeling is an attempt to mathematically represent a complex natural phenomenon. By knowing the initial emission characteristics, dispersion models statistically predict the behavior and movement of emission plumes and chemical air concentrations at selected downwind receptor locations. By design, dispersion models encompass all possible operating conditions at all possible meteorological conditions, yielding overly conservative results. If those results meet the NAAQS, then it can safely be said that the NAAQS are protected.
EPA recognized the overly restrictive nature of dispersion modeling in its draft SO2 NAAQS Designations Modeling Technical Assistance Document, which is “primarily for use by air agencies to assess likely areas of attainment and nonattainment with the 1-hour SO2 NAAQS.” Many state agencies will use dispersion modeling to help determine non-attainment areas, something that was previously only done based on actual monitored air quality. In its draft guidance, EPA is allowing not only the use of actual emission rates (instead of potential emission rates) but is allowing the use of actual stack heights, even when they exceed Good Engineering Practice (GEP) stack heights.
Sufficient information exists in the public domain for an intervener group to produce a rough dispersion model on any facility they wish to oppose on the basis of a perceived NAAQS violation. This opportunity might arise during the Title V operating permit renewal, which is required every five years and requires a public comment period. The effective halt of new coal-fired boiler permitting might free up environmental activists to pursue such efforts. This is where accurate modeling, using the best available data, could serve a facility owner to defend against any frivolous claims of a violation.
As the NAAQS have been revised downwards with shortened averaging periods, an individual facility must take on more operating restrictions to achieve a passing dispersion model. Take haul roads, for example. The least restrictive operating scenario to model is that any haul road could be used at any time. With the lower Pm2.5 NAAQS, however, this often models as a violation. The model can then be modified to exclude nighttime truck deliveries. In reality, this might be outside of normal operations and present little inconvenience to the facility. At night, ambient wind conditions are calmer, which provides less dispersion and higher ground-level emission concentrations. Excluding nighttime hours may be sufficient to resolve the exceedances. However, the facility now will need to accept a permit condition that truck deliveries may never be allowed at night, creating the need to forever monitor and keep records to show compliance.
A second example focuses on emergency diesel equipment operation. The new 1-hour NO2 NAAQS means that almost no fire pump or emergency generator models as in compliance, which is why EPA needed to write a specific policy carving out an exemption to modeling for these short-stack intermittent sources. Combustion equipment at rural municipal power plants might rarely operate but still may not meet conditions of EPA’s modeling exemption. These facilities are potentially subject to expensive control devices or mandatory shutdowns, shifting power production to larger utilities and reducing grid reliability. Fortunately, these small plants are not the target of environmentalist scrutiny. Yet.
Dispersion modeling is a valuable tool for protecting air quality when conducted by the right professionals, using the proper data and methods.
Power Engineerng Issue Archives
View Power Generation Articles on PennEnergy.com