Search

The simulations are compared of six widely-used dense gas dispersion models of downwind chlorine gas concentrations following three railcar accidents. The six models are ALOHA, HGSYSTEM, SLAB, SCIPUFF, PHAST and TRACE. The three railcar accidents, where as much as 60 t of chlorine were released, are Festus, MO (release from a ruptured 1 in. line while offloading), and Macdona, TX, and Graniteville, SC (release from a large hole due to an accident). Input data were obtained from public sources. Source emissions rates were refined based on source modeling with PHAST and TRACE and derivations using fundamental thermodynamic equations. When using the same source emissions rates, the models΄ simulations of 10-minute averaged cloud centerline concentration, at downwind distances ranging from 0.1 to 10 km, agree with each other within plus and minus a factor of two most of the time. For a very large release (Graniteville), the 10-minute averaged 2000 ppm, 400 ppm, and 20 ppm contours are predicted to extend downwind about 1.3, 3.1, and 14 km, respectively, from the source. There is also agreement among the models simulations of the plume widths and heights to the 2000, 400, and 20 ppm contours. A major conclusion of the study is that estimation of the source or release term is important for reliable results. [ABSTRACT FROM AUTHOR]

Results are presented of an evaluation of the Hazard Prediction and Assessment Capability (HPAC) suite of models in an urban environment using data from the Joint Urban 2003 (JU2003) Field Experiment in Oklahoma City (OKC). JU2003 included 29 separate SF6 tracer continuous releases (of 30-minute duration) on ten days from a point source near ground level in or immediately upwind of the built-up downtown area. The ten Intensive Operating Periods (IOPs) consisted of six daytime periods and four nighttime periods. Tracer was sampled at over 100 locations at distances ranging from 0.1 to 4 km from the source. The current study tests two alternate urban configurations of HPAC and four optional meteorological inputs. The two HPAC configurations were the Urban Dispersion Model (UDM) and the Urban Canopy (UC) options. The four meteorological data options were basic default National Weather Service (NWS) data, a single averaged wind, a single upwind anemometer and radiosonde, and detailed three-dimensional winds from a meteorological model, MM5. The evaluations of the maximum 30-minute averaged concentrations on six downwind distance arcs are summarized in this paper. In most cases, the MM5 meteorological inputs yielded the best HPAC results. Also, in general, the UC urban option produced higher concentrations, by about a factor of two, than the UDM urban option. The UDM urban option performed better during the night IOPs and the UC urban option performed better during the day IOPs. There is an obvious day-night difference in the model biases, with most options overpredicting during the night and most options underpredicting during the day, suggesting that they are overstating the relatively-small observed day-night difference in near-ground urban stability and in tracer concentrations. [ABSTRACT FROM AUTHOR]

A comprehensive review is given of uncertainty and sensitivity methods as they are applied to atmospheric transport and dispersion models, including an example of a recent Monte Carlo (MC) uncertainty study involving simulations of air toxics by the dispersion models AERMOD and ISC. The four components of uncertainty are described and set of definitions given for a variety of the more widely used uncertainty and sensitivity options. The special needs of atmospheric models are distinguished from the procedures followed for general environmental or health risk models. Examples of the use of ensembles to estimate the uncertainty of linked meteorological and dispersion models are given. [ABSTRACT FROM AUTHOR]