Developing a physically consistent model for gibbsite leaching kinetics

Leaching aluminium hydroxide from bauxite ores is the first step of the Bayer process which is widely used in the aluminium industry. This paper aims to develop a better understanding of the gibbsite leaching kinetics which can allow for subsequent process optimisation. Comparison of many conventional kinetic models developed for gibbsite leaching with experimental data showed poor agreement. Several new kinetic models have been developed and validated experimentally. The model equations were numerically integrated applying the fourth-order Runge–Kutta technique. Unknown model parameters were estimated by comparing numerical solutions with available experimental data using nonlinear regression analysis with the Levenberg–Marquardt algorithm. The new model for the gibbsite leaching controlled by diffusion through the surface ash/product layer did not compare well with the data. Taking into account the actual irregular surface morphology of the dissolving solids, the fractal geometry of this irregular surface was considered in the development of newer models. With the inclusion of the fractal dimension, the new models agreed better with the experimental data. Finally, the residual aluminium concentration remained in the solid phase after leaching was given consideration in the development of the new model which showed perfect agreement with the data. A physically consistent model for gibbsite leaching was thus obtained by considering these last two factors: fractal geometry of the shrinking gibbsite particles and residual aluminium concentration.