Evaluation of the WIND System atmospheric models: An analytic approach

An analytic approach was used in this study to test the logic, coding, and the theoretical limits of the WIND System atmospheric models for the Savannah River Plant. In this method, dose or concentration estimates predicted by the models were compared to the analytic solutions to evaluate their performance. The results from AREA EVACUATION and PLTFF/PLUME were very nearly identical to the analytic solutions they are based on and the evaluation procedure demonstrated that these models were able to reproduce the theoretical characteristics of a puff or a plume. The dose or concentration predicted by PLTFF/PLUME was always within 1% of the analytic solution. Differences between the dose predicted by 2DPUF and its analytic solution were substantially greater than those associated with PUFF/PLUME, but were usually smaller than 6%. This behavior was expected because PUFF/PLUME solves a form of the analytic solution for a single puff, and 2DPUF performs an integration over a period of time for several puffs to obtain the dose. Relatively large differences between the dose predicted by 2DPUF and its analytic solution were found to occur close to the source under stable atmospheric conditions. WIND System users should be aware of these situations in which the assumptions of the System atmospheric models may be violated so that dose predictions can be interpreted correctly. The WIND System atmospheric models are similar to many other dispersion codes used by the EPA, NRC, and DOE. If the quality of the source term and meteorological data is high, relatively accurate and timely forecasts for emergency response situations can be made by the WIND System atmospheric models