The effect of bubble growth by coalescence on predictions of reactant conversion in a fluidized bed catalytic reactor has been investigated for reactions that are described by either Langmuir-Hinshelwood or a simple first order kinetics. The two-phase model of Orcutt et nl. with continuous change of bubble size with height has been employed in all calculations. The results of these calculations have been compared with t-hose of models using a constant bubble diameter. These effective bubble diameters have been determined in three ways: by calculating an integral average bubble diameter, by calculating the bubble diameter which gives the same bed expansion and by calculating the bubble diameter which gives the same total number of mass transfer units.
It was found that, in the absence of significant interphase mass transfer resistance, predicted values of conversion, and/or multiplicity of steady states obtained from models using the effective bubble diameter determined from the bed expansion are in very good agreement with those obtained using models which incorporate bubble size variation with height. However, an accurate description of bubble size variation with height is required when the bubble growth is fast and the local mass transfer rate decreases rapidly with distance from a distributor. Under these conditions, the values of conversion obtained using the model with perfect mixing of the dense phase gas may be higher than the ones obtained from the model which assumes plug flow of the dense phase gas. This type of behaviour has not been observed when models with constant bubble size are used.
Introduction
It is generally agreed upon that the interphase rate of mass transfer between gas bubbles and the emulsion phase has great effect on a fluid bed reactor performance. The rate of mass transfer is related to bubble size, and for prediction of the reactor performance it is important to know how the bubble size varies with operating conditions and the reactor geometry (i.e. bed diameter and height). In the early models for fluidized bed reactors the bubble size was assumed to be constant and was treated as a model parameter. The effective bubble size was determined from experimental data by trying to match the observed values of reactant conversion. However, the models of this type are of limited value for use in reactor design and scaleup or in process optimization studies due to the lack of knowledge of the effective bubble size variation with operating conditions and/or the reactor size.
In recent years a number of empirical correlations for prediction of bubble size growth due to coalescence, as a function of superficial gas velocity, distributor design and the bed height, have been proposed (e.g.