Standing wave annealing technique: For the design and optimization of nonlinear simulated moving bed systems with significant mass-transfer effects

This paper introduces a flexible and computationally efficient technique for the optimization of nonlinear simulated moving bed (8MB) systems with significant mass-transfer effects. The efficiency results from a combination of standing wave design equations (SWD), with a stochastic optimization algorithm, simulated annealing. Standing wave annealing technique (SWAT) extends the applicability of the SWD to the simultaneous optimization of a large number of variables that include material parameters. Several interrelated issues regarding the design of an SMB system are addressed through an example, the resolution of racemic mixtures of FTC-esters. Models containing 16, 18, and 19 decision variables are considered in terms of two alternative objectives: maximum productivity or minimum purification cost. S WAT's computational efficiency (each optimization takes minutes rather than hours or days) helps identify the important role that maximum operating pressure plays in determining the economical design of an SMB system. Among the key findings are the following: (1) The costs of building a high-pressure SMB system that maximizes productivity can be nearly 60% greater than those of a cost-minimizing medium-pressure system. (2) For optimal separation, the maximal binding capacity should be as high as possible, and the adsorption affinity for the low-affinity solute should be as low as possible. The affinity for the high-affinity solute should be moderate in order to maximize productivity while keeping the solvent cost low. (3) Optimizing the material parameters has the potential of increasing productivity by 7-10-fold, reducing the solvent cost by 60%, and reducing the average purification cost by 60-70%.