One of the common methods for hydrogen production is the natural gas reforming process, where the resulting product is a mixture of hydrogen as the desired output and one or more impurity gases. A suitable method for purifying this gas on an industrial scale is the Pressure Swing Adsorption (PSA) process. PSA is one of the most cost-effective methods for hydrogen purification at an industrial scale; however, due to high operational costs and laboratory constraints, the simulation of this process is crucial. Various methods have been proposed by researchers to simulate this process.
In this research, a comprehensive model for simulating PSA units was examined. The equations formed for the PSA process were solved using numerical methods. This was carried out using the finite difference method for discretizing the equations and coding in MATLAB. The developed software is capable of simulating dynamic adsorption processes as well as continuous cyclic adsorption processes with any number and type of stages. First, the simulation results of dynamic adsorption processes for a gas mixture were compared with experimental data from a published study. Subsequently, the effects of flow rate, adsorbent type, and component composition were examined, with the results aligned with experimental data. Finally, a five-bed, 12-step PSA cycle process for hydrogen purification from an olefin unit was simulated. The feed gas mixture entering the PSA unit for hydrogen purification consisted of 94.84% hydrogen, 0.01% nitrogen, 4.84% methane, 10 ppm carbon dioxide, and 0.29% carbon monoxide. The process operates at a pressure of 32.30 bar and a temperature of 35°C with a flow rate of 37186 〖Nm〗^3/h. The simulation results of the PSA unit showed that hydrogen with a purity of 99.50% and a recovery rate of 85.68% was obtained from the activated carbon adsorbent bed, while a purity of 95.07% and a recovery rate of 87.38% was achieved from the zeolite adsorbent bed.