Experimental investigation and modeling of the impact of random packings on mass transfer in fluidized beds

This study investigates the impact of two novel settings of bubbling fluidized beds equipped with random metal packings on the mass transfer. To do this, a comprehensive approach is applied that integrates experiments and modeling and explores the relevance of the different underlying mechanisms involved in interphase mass transfer in fluidized beds. Firstly, the mass transfer of water from moisturized silica gel particles to dry air is studied in both a packed-fluidized bed and a freely bubbling bed with no packing. The experimental set-up consists of a cylindrical bubbling fluidized-bed column with an inner diameter of 22 cm. Silica-gel particles with a mean particle diameter of 797 mu m are used as bed material. The total bed amount ranges from 4 to 8 kg, while the fluidization number (F) varies between 1.7 and 2.3. Two types of packing, RMSR (stainless steel thread saddle rings) and Hiflow (stainless steel pall rings) are examined and compared to the reference case of a bubbling bed with no packing. The height of the packed section is maintained at 60 cm. The results show that, at all operating conditions, the use of packings enhances the amount of desorbed water in the fluidized bed. The increase is up to 17%, as compared to the bed without packing. The effect is believed to be inhibition of bubble growth in the packed-fluidized bed. To study this further, a mass-transfer model is introduced to analyze the different mass- transfer steps (intra-particle, particle surface to emulsion gas, and emulsion gas to bubble gas) in packed- fluidized beds compared to beds with no packing. TGA experiments are applied to describe the intra-particle mass transfer through a desorption kinetic model. Model analysis shows that the main resistance for mass transfer occurs across the bubble-emulsion boundary. The calculated value of the mass transfer coefficient for this, K b , at reference conditions (6 kg of silica gel and F = 2.3) is 7.6e-5 s-1 with packings (in average) and 6.2e-5 s- 1 without packings, i.e. a 23% improvement in the governing mass-transfer coefficient.