An MINLP (mixed integer nonlinear programming) model for optimal design of multipurpose batch processes with multiple production routes is developed. It allows the existence of intermediate storage

parallel units in and out of phase

as well as discrete equipment sizes. By combining simulated annealing (SA) with linear programming (LP)

SA/LP algorithm is proposed for solving the MINLP model. SA/LP algorithm takes advantage of both SA in the good global optimization and LP in the ability to deal with continuous variables as well as constraints equations. The illustrating examples show that SA/LP method gives better global convergence and is far more efficient in computing time compared with some published methods.