NUMERICAL SIMULATION OF PARTICULATE MATTER EMISSIONS FROM MECHANICALLY VENTILATED SWINE BARNS

Abstract

Numerical modeling using computational fluid dynamics (CFD) was applied to predict the transport of particulatematter (PM) in a mechanically ventilated airspace. A three-dimensional model was used to predict the PM concentrationsand air velocities at selected locations within the ventilated space using two turbulence models (standard and realizable k-.models) and particle tracking methods (time-averaged and stochastic tracking). The PM concentrations predicted using therealizable k-. model and stochastic particle tracking showed better agreement with experimental data than the standard k-.model and time-averaged particle tracking. The PM concentrations and air velocities at selected points within the exhaustduct were also predicted. The calculated normalized mean square error (NMSE) values ranged from 0.012 to 0.021, indicatingclose agreement between the corresponding predicted and measured values. The modeling parameters were applied todetermine the PM emission rates from a typical swine barn ventilated by a bank of fans. The PM emission rates were calculatedbased on PM concentrations determined: (a) inside the exhaust duct, (b) at regions near the exhaust, and (c) along the alley.Emission rates based on PM concentrations determined at locations within the barn other than the exhaust could differ (byas much as 40%) from PM emission rates determined at the exhaust. Numerical simulation showed that one possible approachto the measurement of PM emission rates from mechanically ventilated swine barns equipped with multiple fans would beto determine the PM emission rate at one exhaust fan and estimate the emission rates in the other fans based on the airflowrate ratios between the fans.