(272c) PSA Performance for Varying Productivity Levels: Impact of Adsorption Step Time, Purge Flow and Pellet Size

PSA processes are generally associated with a high energy efficiency and a flexible operating range, making them an attractive separation technique. However, their design and optimization remain time consuming due to their complexity. Indeed, PSA processes can be manipulated by a large variety of parameters, including valve settings, cycle sequence, or even pellet size. This is often done to target performances combining a minimum purity requirement with high recovery and high productivity. Additionally, for demand-driven processes, the productivity rather than the feed rate is set, whilst targeting in-spec purity and good energy efficiency. Multi-objective optimization algorithms aide in the search for suitable PSA settings, yet these can obscure the impact of individual parameters as these are altered simultaneously. To rationalize the impact of the adsorption step time, purge flow rate and pellet size on these PSA process performance indicators, a biogas upgrading PSA process is studied.

The case study is focused on a simulated dual column PSA unit for CO₂/CH₄ (biogas proxy) separation, operating on a modified Skarstrom cycle with top-top equalization and delayed purge. The focus is on the parameters’ impact on purity and recovery whilst varying the total productivity. Accordingly, each parameter’s variation can be shown to define trajectories within a three-dimensional shape in the productivity-purity-recovery space. These shapes indicate the existence of optima, which can be rationalized by investigating elements as the mass transfer zone penetration in the beds, the extent of regeneration, mass transfer kinetics and pressure drop. Cycle features can play an important role in the observed trends, for example the top-equalization step employed here.