Statistical Modeling of the Industrial Sodium Aluminate Solutions Decomposition Process

Abstract

This article presents the results of the statistical modeling of industrial sodium aluminate solution decomposition as part of the Bayer alumina production process. The aim of this study was to define the correlation dependence of degree of the aluminate solution decomposition on the following parameters of technological processes: concentration of the Na2O (caustic), caustic ratio and crystallization ratio, starting temperature, final temperature, average diameter of crystallization seed, and duration of decomposition process. Multiple linear regression analysis (MLRA) and artificial neural networks (ANNs) were used as the tools for the mathematical analysis of the indicated problem. On the one hand, the attempt of process modeling, using MLRA, resulted in a linear model whose correlation coefficient was equal to R2 = 0.731. On the other hand, ANNs enabled, to some extent, better process modeling, with a correlation coefficient equal to R2 = 0.895. Both models obtained using MLRA and ANNs can be used for the efficient prediction of the degree of sodium aluminate solution decomposition, as the function of the input parameters, under industrial conditions of the Bayer alumina production process.