Robust multi-objective optimization of methanol steam reforming for boosting hydrogen production

Abstract

Methanol steam reforming (MSR) has been considered as a promising method for producing pure hydrogen in recent decades. A comprehensive two-dimensional steady-state mathematical model was developed to analyze the MSR reactor. To improving high purity hydrogen production, a triple-objective optimization of the MSR reactor is performed. Non-dominated Sorting Genetic Algorithm-II (NSGA-II) is employed as a robust optimization approach to maximize the three objectives, termed as, methanol conversion, CO selectivity, and H2 selectivity. The Pareto optimal frontier has also been provided and the ultimate solution of the Pareto front has been found by the three decision-making methods (TOPSIS, LINMAP, and Shannon's Entropy). Among the three distinct decision-making approaches, LINMAP presents better results according to the deviation index parameter. It has been shown that a perfect agreement is available between the plant and simulation data. Operating under the optimum values based on the LINMAP method confirms an almost 47.04% enhancement of H2 mass fraction compared to the conventional industrial MSR reactor. The predicted results advocate that the key superiority of the optimized-industrial reactor is the remarkable higher production rate of hydrogen compared to the conventional MSR reactor which makes optimized-industrial reactor both feasible and beneficial.