Prediction of Stack Plume Downwash Using Computational Fluid Dynamics

Accurate prediction of the fluid dynamic and thermodynamic characteristics of saturated buoyant plumes at power plant chimneys is important in developing reliable methods for controlling stack plume downwash. In particular, the accurate prediction and abatement of stack plume downwash is critical in northern climates where, under downwash conditions, the interaction of the saturated, warm plume with the cold outer chimney surface can lead to hazardous ice formation and buildup near the top of the chimney. When a stack is in downwash mode the plume leaving the stack turns downward and flows down along the leeward side of the shell. This is a direct consequence of the wind dynamic pressure acting on the plume and the low pressure in the wake of the shell. In downwash model it is not uncommon to see the plume travel down the shell one third to one half the chimney height and extend radially away from the shell a distance of twenty to thirty feet. This complex interaction of a turbulent thermally buoyant jet entering a cross wind has been studied extensively in the past both experimentally and theoretically with emphasis on the introduction of the jet through an orifice in an infinitely long flat plate. In the case of stack plume downwash the drag of the cylindrical stack in cross flow interacts with the plume under certain “worst-case” ambient wind conditions for the geographic location of the plant and draws the swirling plume into the wake region behind the stack. Once in this region, the distance the plume will travel down the leeward side of the chimney is a function of the ambient wind velocity and the plume velocity. Prediction of this complex, turbulent, three dimensional swirling flow including mixing of different temperature gases and the development of remedial devices to control, in particular, the problem of plume downwash has traditionally required an extensive and expensive wind tunnel model study. Results of these wind tunnel tests include empirical correlations and charts which have been used in the industry for decades. Advances in the capabilities of Computational Fluid Dynamics (CFD) have allowed engineers the ability to reliably study this flow phenomena in greater detail than attainable in a typical wind tunnel model study. In this paper Computational Fluid Dynamics (CFD) is used to predict downwash as a function of flue gas discharge velocity, wind velocity and temperature and the geometry of the stack near the discharge elevation. Further, two devices for minimizing plume downwash in a prototype stack installation are discussed and evaluated by the authors using CFD. Model validation simulations against experimental data and theoretical predictions of buoyant jets in cross flow are also presented and discussed.