Dynamic simulation, optimization and control of multi-stage, counter- current washing systems in alumina-producing plants

Abstract EN

Entrapped caustic liquor in the red mud residue, a by-product of the Bayer ?rocess used to produce alumina from Bauxite ores, is recovered in a series of counter-current, multi-stage washer trains using the process diluting liquor.

Co-ordinated control of the operation of the washer trains is difficult owing to both the long-time constants of elements of the plant, which cause delays between corrective actions and observable effects, and the imprecisely-known interaction of disturbances in composition and rate of the counter-current flows.

A dynamic simulation model has been developed for a multi-stage washer train using SLAM II simulation package in order to gain understanding of the dynamic behavior of the washing system.

Optimization of the simulation model has identified sets of operating conditions under which the caustic soda gain from the washing operation would be maximized.

Based on optimization results, a ?C-based feedforward feedback control system has been designed and implemented to control liquor level and mud level in a washer tank, polymer flow rate to a washer tank and the ‘optimal’ washer in the train to return the fine seed filtrate.

About 20% increase in caustic soda gain from the counter-current washing has been achieved as a result of controlling the operation.