Distributor design and its effect on a fluidised bed reactor.

Abstract

The design of multi-orifice gas distributors and the effect of the distributor on the physical and chemical behaviour of a large two-dimensional fluidised bed reactor were studied. A rational procedure was proposed for the design of multi-orifice distributors. A new model was proposed to determine the necessary minimum distributor pressure drop. Model predictions were in good agreement with experimental data. Experimental work was carried out to examine the effect of five different multi-orifice distributors on the catalytic decomposition of ozone. For values of the dimensionless rate constant, NR greater than 1.0, the number of orifices, N, had a significant effect on conversion. Reactor conversion increased as N increased. The distributor with 25 orifices represented the limiting value of N, in terms of the influence of the distributor on reactor performance. Jet penetration and mass transfer were studied in a small two-dimensional fluidised bed. The results showed that jet penetration in two-dimensional beds was greater than that in three-dimensional beds. The measured rates of mass transfer were considerably less than those predicted by the empirical equation of Behie et al(17) , but were of similar magnitude to the rates of mass transfer associated with the bubble phase close to the grid. A model, developed to describe the jetting region, was found to underestimate both gas- solid contacting and mass transfer, particularly in the upper portion of the jet which fluctuated rapidly in shape and height. The jetting region was satisfactorily described using the equations for the bubbling region of the bed. A new model was proposed to explain the discrepancy between the observed visible gas flow and that predicted by the two-phase theories. This discrepancy was accounted for by a combination of an increased cloud size and gas short-circuiting through bubble chains. Two reactor models in literature were modified and the predictions compared with experimental data. The modified Kunii and Levenspiel model indicated that an increased cloud size was predominant, while the modified Orcutt and Davidson model suggested that gas 'short- circuiting' was more probable. It was not the aim of the present study to discriminate between reactor models and therefore no definite conclusion was drawn in favour of one model or the other. Evidence from literature suggests that the former model provides a more realistic account of gas-solid contacting and mass-transfer between the phases.

Date of Award	1981
Original language	English
Awarding Institution	University of Bath