Source characterization refinements for routine modeling applications

Abstract

Abstract Steady-state dispersion models recommended by various environmental agencies worldwide have generally been evaluated with traditional stack release databases, including tracer studies. The sources associated with these field data are generally those with isolated stacks or release points under relatively ideal conditions. Many modeling applications, however, involve sources that act to modify the local dispersion environment as well as the conditions associated with plume buoyancy and final plume rise. The source characterizations affecting plume rise that are introduced and discussed in this paper include: 1) sources with large fugitive heat releases that result in a local urbanized effect, 2) stacks on or near individual buildings with large fugitive heat releases that tend to result in buoyant “liftoff” effects counteracting aerodynamic downwash effects, 3) stacks with considerable moisture content, which leads to additional heat of condensation during plume rise – an effect that is not considered by most dispersion models, and 4) stacks in a line that result in at least partial plume merging and buoyancy enhancement under certain conditions. One or more of these effects are appropriate for a given modeling application. We present examples of specific applications for one or more of these procedures in the paper. This paper describes methods to introduce the four source characterization approaches to more accurately simulate plume rise to a variety of dispersion models. The authors have focused upon applying these methods to the AERMOD modeling system, which is the United States Environmental Protection Agency's preferred model in addition to being used internationally, but the techniques are applicable to dispersion models worldwide. While the methods could be installed directly into specific models such as AERMOD, the advantage of implementing them outside the model is to allow them to be applicable to numerous models immediately and also to allow them to remain applicable when the dispersion models themselves are updated. Available evaluation experiences with these techniques, which are discussed in the paper, indicate improved model performance in a variety of application settings.