Pressure Swing Adsorption (PSA) is a well known technology for gas separation and purification. It has been widely used for H2 production from the effluent stream of a shift converter, which predominately comprises H2 and CO2, with other components in negligible amounts. Most of the commercial PSA cycles have been developed to recover H2 at an extremely high purity, and do not focus on enriching the strongly adsorbed CO2. Thus, a major limitation exists with the use of these conventional PSA cycles for high purity CO2 capture. Furthermore, multi-bed PSA systems are typically operated in a cyclic manner with each bed repeatedly undergoing a sequence of steps. Their complex dynamic behavior, together with the numerical difficulties of the model governed by partial differential and algebraic equations (PDAE), makes the evaluation and assessment of different operating steps and cycle configurations very difficult and time consuming. Therefore, a systematic methodology is essential to develop, evaluate and optimize PSA cycles to recover both H2 and CO2 at high purity.
In this work, we present a systematic optimization-based formulation for the synthesis and design of novel PSA cycles for pre-combustion CO2 capture. Here, we propose a superstructure-based approach to simultaneously determine optimal cycle configurations and design parameters for PSA units. The superstructure approach can be effectively used to assess the usefulness of PSA processes for CO2 capture. The superstructure consists of two beds, one of which acts as an adsorbing bed and the other as a desorbing bed. The interconnections between the two beds are governed by time-dependent control variables, such as fractions of the light and the heavy product recycle. The superstructure is rich enough to predict a number of different PSA operating steps (e.g., pressurization/depressurization, adsorption/desorption, pressure equalization, light and heavy product recycle), which are accomplished by varying these control variables. An optimal sequence of operating steps is achieved by solving an optimal control problem for the superstructure. A partial discretization approach (or sequential approach) is used in this work to solve the optimal control problem as a dynamic optimization problem. The sequential approach decouples the PDAEs of the PSA system from the optimization problem. As a result, the partially discretized PDAEs are integrated outside the optimization problem using sophisticated dynamic simulators which are able to capture the steep adsorption fronts with high accuracy. After this, we solve the optimization problem of relatively smaller size using a reduced space sequential quadratic programming (rSQP) algorithm. A case study for maximizing CO2 recovery while ensuring a specified purity level for H2 and CO2 is presented.