Multivariable Self-tuning Control of a Binary Distillation Column

Abstract

Control of distillation columns remains a key concern of industry because of the high energy consumption associated with distillation separations. The reported results, from a continuing program of experimental studies directed to improved distillation column control, are concerned with the evaluation of multivariable self-tuning control algorithms for the simultaneous control of both top and bottom product compositions. Evaluation of the performance of multivariable self-tuning control algorithms has been performed by experimental testing using a 8-tray, 30.5 cm diameter methanol-water pilot scale column. The column is controlled using a DLC LSI 11/03 microcomputer with continuous capacitance analysis of top product composition with bottom composition analyzed on a 3 min cycle, by gas chromatograph. Column control is studied for 25% step increases and decreases in feed flow rate. The performance of the multivariable algorithms is compared with that achieved using conventional digital PI/PH) control algorithms. The results have demonstrated that use of self-tuning control will provide better control performance for feed flow rate disturbances than can be achieved using well-tuned conventional multiloop PI/PTD control.