Simulation and advanced process control of an extractive distillation column and a vinyl acetate monomer plant

Dynamic simulation was used in an investigation of constrained model predictive control (CMPC) of two different chemical processes. The first process is an extractive distillation column that separates an acetone-methanol mixture using water as the solvent. For comparison purposes, results with traditional PID-type control have been obtained. The ability of CMPC to predict future process outputs allowed it to provide interaction compensation. This resulted in tighter control of the column with respect to PID-type control. CMPC was also able to demonstrate its optimization capability by maximizing column throughput while maintaining the other controlled variables within their respective bounds.; The second process is a vinyl acetate monomer plant. Vinyl acetate is produced by the gas phase reaction of ethylene, acetic acid, and oxygen. Plant parameters for the simulation were obtained from an existing plant in Mahad, India. The control objective is to regulate the production of vinyl acetate monomer, while maintaining other plant operating conditions within bounds. The plant is a five output - four input system. The results of the investigation demonstrate that CMPC can not only regulate the production of vinyl acetate monomer, but also maximize production while maintaining all other controlled variables within bounds.