Optimal design of pressure swing adsorption units for hydrogen recovery under uncertainty

Abstract The paper proposes an approach to the optimal design of pressure swing adsorption (PSA) units for hydrogen recovery under uncertainty, which provides a reasonable margin of the potential resource of the PSA hydrogen unit and compensates for the negative impact of a random change in uncertain parameters within specified limits. A heuristic iterative algorithm is proposed to solve the design problem with a profit criterion, which is guaranteed to provide the technological requirements for the PSA unit, regardless of the values that take uncertain parameters from the specified intervals of their possible change. An experimental verification of the approach with the root-mean-square error of 19.43 % has been carried out. Optimization problems of searching for a combination of mode and design parameters under uncertainty for a range of 4-bed 4-step VPSA units with a capacity of 100–2000 L/min STP have been solved taking into account the requirements for hydrogen purity of 99.99+ %, gas inlet velocity of 0.2 m/s, and bed pressure drop (no more than 1 atm). It has been established that taking into account uncertainties leads to an increase in energy costs by 8–10 %, a decrease in profit by 10–15 %, and a decrease in hydrogen recovery by 4–5 %, which is a payment for the uninterrupted operation of the PSA unit. The effect of uncertain parameters (percentage composition of the gas mixture; gas temperature; the diameter of adsorbent particles) on the key indicators of the PSA process (recovery, profit, hydrogen purity, unit capacity) has been established and trends in adsorption duration, adsorption and desorption pressure, P/F ratio, valve capacity, bed length, adsorber diameter for design of hydrogen PSA unit, which are necessary for subsequent design and scaling of units.