Modelling and multi-objective optimization for simulation of hydrogen production using a photosynthetic consortium

Abstract

This study was aimed at finding the optimal conditions for hydrogen production based on statistical experiments and using a simulation approach. A Plackett–Burman design and steepest ascent were used to screen the key factors to obtain the best hydrogen concentration. According to the regression analysis, cysteine, acetate, and aeration had the best effect. The optimal conditions, using the method of steepest ascent, were aeration (0.125 L/min), acetate (200 mg/L), cysteine (498 mg/L). Once this was determined, an experiment with more than two factors was considered. The combinations: acetate + cysteine without aeration and cysteine without aeration increased hydrogen concentration. These last two criteria were used to validate the dynamic model based on unstructured kinetics. Biomass, nitrogen, acetate, and hydrogen concentrations were monitored. The proposed model was used to perform the multi-objective optimization for various desired combinations. The simultaneous optimization for a minimum ratio of cysteine-acetate improved the concentration of hydrogen to 20 mg/L. Biomass optimized the concentration of hydrogen to 11.5 mg/L. The simultaneous optimization of reaction time (RT) and cysteine improved hydrogen concentration to 28.19 mg/L. The experimental hydrogen production was 11.4 mg/L at 24 h under discontinuous operation. Finally, the proposed model and the optimization methodology calculated a higher hydrogen concentration than the experimental data.