Cluster Characteristics of Polydisperse Group B Particles In a CFB Riser

Cluster characteristics of Polydisperse Group B particles in a CFB riser

Jia Wei Chew^a, Drew M. Parker^a, Roy Hays^b, John G. Findlay^b, Ted M. Knowlton^b, S.B. Reddy Karri^b, Ray A. Cocco^b, Christine M. Hrenya^a,*

^aDepartment of Chemical and Biological Engineering, University of Colorado at Boulder, Boulder, CO 80309

^bParticulate Solid Research Incorporated, Chicago, IL 60632

The important role of clustering in high-velocity, gas-solid flow has been well documented in experiments and simulations alike, though previous efforts have focused on monodisperse systems. In this work, experiments with a focus on understanding the impact of polydispersity on clustering characteristics (appearance probability, duration and frequency) of Geldart Group B particles in a circulating fluidized bed (CFB) riser have been carried out. In a moderately dense CFB riser, three mixtures were investigated: (i) a density-difference binary mixture, with particles of different material density (ρ_s) but same particle sizes (d_ave), (ii) a size-difference binary mixture, with particles of different d_ave but same material density ρ_s, and (iii) a continuous particle size distribution (PSD). In addition, in a dilute CFB riser, varying widths of continuous PSDs and varying compositions of binary mixtures were investigated. A fiber optic probe was used for detection of solids, and cluster identification was accomplished via wavelet decomposition. Collectively, results show that: (i) local riser position has a significant influence on all three cluster characteristics in all cases, while effects of operating condition and material type (i.e., material property or effect of polydispersity) are secondary, (ii) among the three cluster characteristics, appearance probability appears to be largely a function of riser position only, and is insensitive to changes in operating condition or material type, and (iii) cluster duration and frequency are influenced by operating condition and material type, though only at some axial positions. The results presented provide a comprehensive picture of factors affecting cluster trends.

^*Corresponding author: hrenya@colorado.edu