Development of a fluctuating plume model for odour dispersion around buildings

This paper presents a fluctuating plume model, which incorporates the PRIME algorithm in order to include the effect of plume elevation and downwash due to buoyancy and the presence of obstacles. The Gaussian fluctuating plume model has the ability to predict both mean concentrations and concentration fluctuations. Therefore, it is useful for modelling dispersion in cases where concentration fluctuations are important for environmental impact assessment, such as for odorous compounds. The model is validated using two different experimental datasets, one involving dispersion around a complex building in a wind tunnel and the other involving dispersion around an isolated cube in the field. The results suggest that the model in general predicts adequately mean concentrations and concentration fluctuation statistics, such as concentration peaks and intermittency, downwind of the near-wake. However, mainly due to the formulation of the PRIME algorithm, it underestimates concentrations in the near-wake recirculation region of the obstacle and although it can adequately predict the maximum intermittency value, it does not predict accurately its location. In general, the model appears to over-predict dispersion if compared to the wind tunnel data. This can be partly attributed to the larger scales of turbulence not reproduced in the wind tunnel as also suggested from the comparison of the model results with field data. Finally, due to the assumptions incorporated in PRIME, the model cannot capture the effect of the complex shape of a building on near-field dispersion.