PREDICTING FLY ASH ACCUMULATION IN A FOSSIL POWER PLANT FLUE GAS DUCT USING COMPUTATIONAL FLUID DYNAMICS

During the startup of a new fossil power plant, a high level of fly ash accumulation (higher than predicted) was encountered in the flue gas ducting upstream of a fluidized bed scrubber. The level of fly ash accumulation made it necessary to manually withdraw fly ash using a vacuum truck after short periods of operation, at less than 80% maximum continuous rating (MCR). This paper presents a simple method for rapid assessment of fly ash accumulation in flue gas ducts using steady state single phase Computational Fluid Dynamics (CFD) simulation of flue gas flow. The propensity for fly ash accumulation in a duct is predicted using calculated wall shear stresses from CFD coupled with estimates for the critical shear stresses required for mobilization of settled solids. Critical values for the mobilization stresses are determined from the Shields relations for incipient motion of particles in a packed bed with given fly ash particle size and density as inputs. Solids accumulation is possible where the wall shear stress magnitude is less than the critical shear stress for mobilization calculated from the Shields relations. Predictions of incipient fly ash accumulation based on the coupled CFD/Shields relations model correlate well with plant startup field observations. Fly ash accumulation was not observed in a related physical scale model test. A separate CFD/Shields relation analysis of the scale model physical tests show that the wall shear stresses in the scale model are several times larger than the critical value required for the mobilization of the fly ash simulant. This study demonstrates that a simple steady state, single phase CFD analysis of flue gas flow can be used to rapidly identify and address fly ash accumulation concerns in flue gas duct designs. This approach is much simpler and computationally inexpensive compared to a transient Eulerian multiphase simulation of particle laden flow involving handling the dense 597 phase in regions of ash accumulation. Further, this study shows that physical model tests will be accurate for predicting fly ash accumulation, only if, the scaling maintains the proper ratio of wall shear stress to critical remobilization stress.