Optimization of simulated moving bed and varicol processes for glucose-fructose separation

Abstract

A comprehensive optimization study was carried out to evaluate the performance of a simulated moving bed (SMB) system and its modification, Varicol process, for an industrially important separation problem, the isolation of fructose (which is of interest in the food industry as sweetener) from a mixture of glucose and fructose solution. Contrary to the operation of SMB, where the inlet and outlet ports are switched synchronously at a predetermined time interval, the Varicol system is based on a non-synchronous shift of these ports. An existing model that can predict published pilot-scale experimental results of the system is used in the optimization study. The optimization problem is complex owing to the complex interplay of relatively large number of decision variables that include continuous variables like flow rates in different sections and column length, and discrete variables like column number and configuration. Several two-objective functions optimization was performed both for existing system and for systems at the design stage. A typical example is simultaneous maximization of productivity and purity of the desired fructose product or simultaneous maximization of productivity of both the products. An adaptation of the state-of-the-art artificial intelligence-based non-traditional optimization method known as non-dominated sorting genetic algorithm (NSGA) is used in obtaining Pareto optimal solution. It was observed that, after performing rigorous optimization, significant improvement in terms of performance is possible for both SMB and Varicol process.