Slide Background
Slide Background

FEED WELL DESIGN SUPPORTED BY CFD FOR FLUID-BED BAUXITE CLASSIFIERS

Boudeville, D. Simard, G. Grillot, T. Labrum, D. Bienvenu, C.

The bauxite production of the Andoom mine, located in the Weipa region, has significantly increased in recent years, without major equipment upgrade. One critical piece of equipment is the bauxite fluidised-bed classifiers which currently undergo periodic rat holing phenomena at the underflow, which can cause instability and lead to production losses.

A CFD-assisted (Computational fluid dynamics) hydrodynamic study of the classifier was conducted to understand the bauxite slurry behaviour and identify the causes for this observed instability. The turbulence model k-e was used in this study to accurately solve the high-speed gradients present in this equipment, which lead to the observation that the issue was caused by a misalignment of the feed and a bypass of “coarse” bauxite towards the overflow.

The original feed well design, made obsolete by the increased feed rate, was evaluated in order improve the classifier stability. For this, a two-phase Euler-Euler’s study simulating the transport of the different classes of bauxite particles in water was modelled and leads to an improved feed system design to eliminate the coarse bauxite bypass to the overflow and maximizes bauxite particles diffusion in the vessel.

Finally, a two-phase Volume of Fluid (VoF) study, simulating the interaction between the slurry and the air in the classifier was conducted by CFD, to make sure that this new design did not lead to flow dumping or splashing reducing the classifier performance and causing HSE risks.