CFD modelling of dispersion in neutrally and stably-stratified atmospheric boundary layers: results for Prairie Grass and Thorney Island

It is a known problem that CFD models using the standard k - epsilon turbulence model do not maintain the correct atmospheric boundary layer (ABL) profiles along a flat, unobstructed domain. The present work examines the impact of these errors in the ABL profiles on dispersion model predictions for three field-scale experiments from the Prairie Grass and Thorney Island datasets. The modified ABL profiles produced by the CFD model in the Prairie Grass experiments result in differences in the predicted concentrations of up to a factor of two, as compared to a reference model. For the Thorney Island experiment, the results for the standard k - epsilon turbulence model are sensitive to the ground surface roughness and problems are identified in relation to the grid resolution near the ground. Industrial risk assessments involving atmospheric dispersion of toxic or flammable substances using CFD models should take into account these limitations of the k - epsilon turbulence model.