Review of models used for determining consequences of UF{sub 6} release: Model evaluation report. Volume 2

摘要

Three uranium hexafluoride-(UF{sub 6}-) specific models--HGSYSTEM/UF{sub 6}, Science Application International Corporation, and RTM-96; three dense-gas models--DEGADIS, SLAB, and the Chlorine Institute methodology; and one toxic chemical model--AFTOX--are evaluated on their capabilities to simulate the chemical reactions, thermodynamics, and atmospheric dispersion of UF{sub 6} released from accidents at nuclear fuel-cycle facilities, to support Integrated Safety Analysis, Emergency Response Planning, and Post-Accident Analysis. These models are also evaluated for user-friendliness and for quality assurance and quality control features, to ensure the validity and credibility of the results. Model performance evaluations are conducted for the three UF{sub 6}-specific models, using field data on releases of UF{sub 6} and other heavy gases. Predictions from the HGSYSTEM/UF{sub 6} and SAIC models are within an order of magnitude of the field data, but the SAIC model overpredicts beyond an order of magnitude for a few UF{sub 6}-specific data points. The RTM-96 model provides overpredictions within a factor of 3 for all data points beyond 400 m from the source. For one data set, however, the RTM-96 model severely underpredicts the observations within 200 m of the source. Outputs of the models are most sensitive to the meteorological parameters at large distances from the source and to certain source-specific and meteorological parameters at distances close to the source. Specific recommendations are being made to improve the applicability and usefulness of the three models and to choose a specific model to support the intended analyses. Guidance is also provided on the choice of input parameters for initial dilution, building wake effects, and distance to completion of UF{sub 6} reaction with water.