Constrained optimization of cellulose acetate membrane using two‐level factorial design

This work presents a method, the SUMT algorithm, to optimize the performance of the cellulose acetate (CA) membrane by maximizing the permeate flux subject to an equality constraint of salt rejection and a pair of inequality constraints for each variable. Three variables, i.e., (1) the concentration of formamide in the casting solution, (2) the time of evaporation of the membrane prior to gelling, and (3) the annealing temperature of the membrane, were selected for this optimization study. Experiments based on the two‐level factorial design were conducted for the determination of regression equations for both premeate flux and salt rejection. At a level of 97.00% rejection of a 0.5% sodium chloride solution at 600 psig, the maximum permeate flux of the CA membrane was predicted to be 14.52 gallons per day per sq. ft (gfd). This was later confirmed by experiments. The effects of the casting variables on membrane performance concluded in this study were consistent with those reported by other researchers.