AGGLOMERATION OF GIBBSITE CRYSTALS: INFLUENCE OF THE PROCESS PARAMETERS AND MODELISATION

摘要

An important stage of the Bayer process, leads to the formation of gibbsite particles (Al(OH)3) mostly grown by agglomeration (aggregation and crystal growth). In the first part of the present study, using a crystallizer working at constant supersaturation with time, we determined the agglomeration kinetics of gibbsite particles, as a function of different process parameters. The effect of the crystallization temperature, supersaturation, seed mass, stirring rate and seed size was investigated. The use of an experimentation plan first revealed that the temperature and supersaturation were the parameters controlling the agglomeration kinetic. The effect of other process parameters could be revealed, in a new set of experiments conducted at constant temperature and supersaturation. It notably appeared that there was an optimum seed mass, for which the agglomeration degree of gibbsite particles was maximum. In the second part of the work, a population balance (PB) model was developed to describe experimental data, and pointing out the effect of the seed concentration on the agglomeration kinetic of gibbsite, in the absence of nucleation. The PB equation is discretized and solved with the method of classes. The whole agglomeration process of gibbsite particles undergoing turbulent agglomeration is modelled by an agglomeration and a fragmentation kernel. The hypothesis of a size independent agglomeration kernel combined with a fragmentation kernel increasing with the mass of the particle is then tested. Results are compared and discussed with previous one drawn from the literature.