Gas separation by means of synthetic polymeric membranes is a well-known technology for several industrial applications such as CO2 capture from industrial emissions, hydrogen recovery and purification, or nitrogen removal from natural gas.

Due to the limitations of the solution-diffusion separation mechanism in polymeric membranes, several separation stages are often necessary when high levels of recovery and purity are required. It is therefore essential to find the optimal design in multi-stage membrane separation processes, which involves the determination of various factors such as the number of membrane stages, the membrane surface at each stage, upstream or downstream pressure of each stage, stream connections between stages, etc.

Artelys Knitro embedded into a global optimization algorithm is employed to solve the nonlinear programming problem coming from such a complex design problem allowing the determination of optimal configurations at minimum separation costs.