Fuzzy Technology Application in a Bauxite Digestion Unit

Abstract

Bauxite slurry and heated caustic liquor are fed into digestion reactors for alumina extraction. The main parameter of the reactor output, the A/C (alumina/caustic) ratio is adjusted manually through the bauxite flow, based on laboratory analyses that are made to calibrate the system every two hours. Since many variables interfere in the A/C ratio, such as bauxite and liquor composition and scaling process, the measurements have low reliability, making this control not effective. This case focuses on Fuzzy technology application in a bauxite digester in order to control the bauxite slurry.

The software does not require phenomenological knowledge to control the A/C ratio and uses empirical rules to estimate it. By anticipating production variances, the process variability was reduced by 50%, which saved 3.2% of the steam used for caustic liquor heating.