The significance of particle clustering in pulverized coal flames

Abstract

Evidence for the presence of preferential concentration or “clustering” of particles in lifted 2.5 MW pulverized fuel (PF) flames from an annular nozzle is presented. Images obtained using laser sheet Mie scattering show that clusters form in the preignition region and persist through the combustion zone. The images also show that increasing the momentum of air through a central precessing jet nozzle causes both the size of the clusters and the mean spreading rate of the particles to increase. These changes also cause a large reduction in flame ignition distance, an increase in peak heat flux to the furnace walls, and a translation of the heat flux profile toward the burner. Simplified models, based on energy balances of idealized clusters, are proposed to assess the significance of the clustering phenomenon. A sensitivity analysis is undertaken using the models, experimental measurements, and estimated data. The effects of clustering and the increased rates of entrainment of hot furnace gases are shown to be important in explaining the observed changes in these lifted flames. The sensitivity analysis revealed benefits from local segregation of both particle size and distributions. It showed that ignition is favored in clusters with diameters less than the characteristic nozzle diameter (here approximately 50 mm), located on the outside of the flow, and in the outer shell of clusters larger than the characteristic nozzle diameter. Ignition is favored in clusters containing fuel-rich concentrations of particles from the small end of the PF size distribution. Some experimental evidence of these conditions in the precessing jet flame is found, althoug further quantification and optimization is required. The results suggest that optimization of clustering effects may be used to provide NOx reduction by “self-staging” of the combustion within flames.