The thermal effect on the flow and dispersion of pollutants emitted from a rooftop stack is investigated by means of CFD (computational fluid dynamics) models with wind tunnel experimental validations. The leeward wall and its nearby ground are heated simultaneously to mimic solar radiation. Seventeen Ri (Richardson number) cases with four inflow wind speeds (1, 3, 6, and 9 m/s) and five temperature differences (0, 60, 120, 180, and 240 K) between the heated surface and ambient air are considered to represent the interaction between thermal buoyancy force and inertia force. The results reveal that (1) the steady RANS (Reynolds Averaged Navier-Stokes) computations with Boussinesq approximation can generally reproduce the effect of thermal buoyancy on the wake flow and pollutant distribution in wind tunnel experiments; (2) the wake vortex flow is less affected by the thermal buoyancy force at small Ri (e.g., Ri ≤ 0.26) while an upward flow rather than a clockwise vortex structure is developed in the near wake at Ri ≥ 0.58; (3) it is inappropriate to place fresh air intakes on the leeward wall of the emitting building, but natural ventilation through windows on the leeward wall can be implemented at higher Ri (e.g., Ri = 2.33); (4) at the pedestrian respiration height downstream of the building, the distance between the location of maximum pollutant concentration and the leeward wall increases linearly with Ri while the maximum dimensionless concentration decreases exponentially with increasing Ri; (5) the air temperature is rapidly reduced away from the heated wall/ground and a heat accumulation zone is formed at the ground corner next to the leeward wall. This study can be helpful for determining the strategy for natural ventilation through windows and for evaluating the impacts of rooftop stack exhaust on air quality downstream of emitting buildings.
Keywords: CFD; Richardson number; Rooftop stack emission; Thermal effect; Wind tunnel experiment.