Chonicol Engineering Science, Vol. 45, No. 9, pp. 28952900,1990. Printed in Great Britain. AN IMPROVED MODEL FOR MASS TRANSFER THREE-PHASE FLUIDIZED BEDS otxs2509/90 53.00 + 0.00 0 1990 Pergamon Press plc IN ABDUL-FATTAH A. ASFOUR’ and ABDULGHANNL H. NHAESI Chemical Engineering Department, University of Windsor, Windsor, Ontario, Canada N9B 3P4 (First received 26 June 1989; acceptedfir publication in revised form 11 3anuary 1990) Abstract-A model for describing mass transfer in three-uhase fluid&d beds has been develoDed and tested using experimental data. The presence of two distinct mass transfer zones led to the idea o> interfacing a plug flow model (PFM) with an axial dispersion model (ADM) at the separation boundary between these zones. The model has been validated for a wide range of operating conditions and proved to perform better than previous models. INTRODUCTION Models describing mass transfer in three-phase fluid- ized beds are currently receiving considerable interest. Several models have been suggested in the literature and are being employed for the calculation of volu- metric mass transfer coefficients. Most studies of mass transfer in three-phase fluid- ized beds, e.g., 0stergaard and Suchozebriski (1971), Ostergaard and Fersbol (1972), Lee and Worthington (1974), Dhanuka and Stepanek (1980), Cherry et al. (1978), have adopted the axial dispersion model (ADM) or the plug flow model (PFM). Other invest- igators, e.g., Deckwer et al. (1974, 1983) and Alvarez- Cuenca (1979) and Alvarez-Cuenca et al. (1979), re- ported experimental tests for validating such models. Alvarez-Cuenca (1979) and Alvarez-Cuenca et al. (1984) utilized the system water-oxygen-glass beads in their three-phase fluidization studies and reported the existence of two well-differentiated mass transfer zones in three-phase fluidized beds. The first zone is near the distributor and is termed the “grid zone”. In this zone plug flow conditions prevail and rapid mass transfer (oxygenation) takes place. Much less mass transfer (oxygenation) takes place in the second zone, termed “the bulk zone”. Axial dispersion is more predominant in this zone. Alvarez-Cuenca (1979) suggested utilizing concentration contour diagrams (similar to those shown in Figs 1 and 2) to identify the different mass transfer zones. The existence of two mass transfer zones led Alvarez-Cuenca (1979) to develop a two-zone model (TZM), which was claimed to represent a significant improvement over the axial dispersion and plug flow models. The two-zone model results from interfacing two plug flow models at the boundary of separation between the two mass transfer zones. Deckwer et al. (1983) analyzed, from an experi- mental standpoint, the validity of the axial dispersion model and the two-zone model. Unfortunately, they deliberately ignored the experimental data obtained iAuthor to whom correspondence should be addressed. in the immediate vicinity of the distributor where a great deal of oxygenation takes place. The existence of two easily distinguishable mass transfer zones in three-phase fluidized beds has also been observed in our own work. Therefore, it is diffi- cult to believe that a single zone model, e.g., PFM or ADM, can describe mass transfer in such situations. This is because different conditions prevail in each zone. Therefore, the application of either the PFM or the ADM is tantamount to assuming that plug flow or axial dispersion conditions, respectively, prevail in the column which is contrary to reality. The main problem with the ADM, as we see it, is its constant axial dispersion coefficient. To overcome this problem we attempted to develop a “modified” ADM with a variable coefficient. However, the final form of the model obtained was very cumbersome and difFi- cult to use, even when it was assumed that the change in the axial dispersion coefficient was linear. Conse- quently, such an approach was discarded. Moreover, Alvarez-Cuenca’s development of the TZM by interfacing two plug flow models at the separation boundary between the two mass transfer zones does not, in reality, solve the problem since it assumes that plug flow conditions prevail in each zone. We show later, on the basis of Figs 1 and 2, that plug flow conditions only prevail in the grid zone. Some backmixing or dispersion conditions prevail in the bulk zone. Consequently, the TZM fails to repres- ent physical reality. Such a situation leads us to interface a PFM and ADM at the separation bound- ary. This results in a more realistic and physically meaningful model than the aforementioned models. Consequently, the objective of this study is to pro- vide the development of such a model and to validate the model using data generated in our laboratory as well as data reported by Alvarez-Cuenca (1979). It is recognized, however, that our proposed model contains more parameters than any of the afore- mentioned models. Whereas some may consider this as a disadvantage, such a disadvantage is outweighed by the fact that such a model describes mass transfer in three-phase fluidized beds more accurately and realistically than any other model. 2895