Evaluating the Effect of Computational Time Steps on Livestock Odor Dispersion Using a Computation Fluid Dynamics Model

A Computational Fluid Dynamics (CFD) model with Eulerian-Lagrangian approach was applied to predicted odour dispersion from a 3000-sow farrowing operation. Eulerian-Lagrangian approach solved turbulent air flow within Eulerian reference frame, and then calculated trajectories of discrete particles in continuous flow field within Lagrangian reference frame. Weather variables were taken every minute by the on-site weather station. The measured weather variables were synthesized into meteorological data with time intervals of 1, 10, 20, 30, and 60 min by Meteorological Processor for Regulatory Models (MPRM). Odour concentrations predicted by CFD model with different model computational time intervals were compared to evaluate the effects of time interval on odour dispersion. Predictions with 1 h time interval overestimated downwind odour concentration and travel distance compared with the predictions with sub-hour time intervals. The longer the time interval, the higher the odour concentration and the longer travel distance would be. Short time intervals (e.g. Δ t = 1 min) used in the prediction were more sensitive to wind direction shift and provided more accurate instantaneous odour plumes than longer time intervals. However application short time interval required more refined meteorological data. Keyword. Computational Fluid Dynamics, Odour, Swine, Air dispersion model, Time interval.