Computational study of an alumina flash calciner.

The mineral processing sector of industry is very fuel-intensive. For instance, aluminium making is one of the nine processes which together account for about 50% of the industrial energy used in the United States (Warshawsky,1976). In aluminium production, the first stage is alumina manufacture. In alumina processing, i e. bauxite refining, the final step is calcination. Calcination was defined and the evolution in the development of today's alumina calcination plants was reviewed. The whole processes (unit configuration and operation, equipment capabilities, energy consumption and efficiency, product quality, etc..,) were described. Operational difficulties and successive improvements were also analysed. In these units, the major component of the equipment is the furnace (rotary kiln or stationary furnace) wherein the water of constitution of precalcined alumina crystals is expelled at high temperature (900 to 1200°C). Overall aluminium trihydrate dehydroxylation is expressed by the following theoretical chemical reaction: Al2O3, 3H2O → AI2O3 + 3H2O(v) Special attention was given to five commercial stationary calciner furnaces which were detained and compared. However, the scope of this work was focused on the study of the Mark III Alcoa flash furnace : a cylindrical vertical vessel composed of three sections -the bottom and top section of smaller dimensions than the central main body- in which fuel oil or gas are introduced and burnt directly into the suspension of alumina particles. Computational Fluid Dynamics simulations---the best technique available to obtain a description of the flow of solid in the furnace since adequately reliable physical modelling experiment does not exist for two phases situations---were carried out to ascertain the media behaviour in the chamber. A description of the numerical tool, together with the various models used in the computations were given. Numerical models of the above existing furnace were built: Isothermal single (gaseous flow) and two phase flows, combustion reaction. Swirling flow field was tested in a partial model of equivalent dimension. Heat and mass transfer between the two phases have not been implemented. Predictive results were validated against physical tests and plant operating data. The models allowed the cause of the current operational inefficiencies to be addressed. Design changes were suggested and actually introduced in the new generations of calciner furnaces which combine an increased combustion efficiency and a better processing of particles