Model predictive control for amine-based CO2 capture process with advanced flash stripper

Abstract

Advanced flash stripper (AFS) has been suggested as a process alternative for conventional absorption based post-combustion carbon capture process, which can significantly reduce the solvent regeneration energy. However, such reduction is enabled through energy integration achieved by two additional heat exchangers, which can pose serious challenges to the control and operation of carbon capture processes. For complex energy-integrated processes, simple decentralized control scheme, where multiple proportional- integral-derivative (PID) controllers are employed, typically shows limited control performances. Thus, this study aims to propose an effective control structure for the carbon capture process with AFS, which can regulate the process under various dynamic scenarios involving significant changes in operational variables such as flue gas flowrate and carbon capture rate. Specifically, a dynamic model for the post-combustion carbon capture process is built in gPROMS, where 30 wt% Monoethanolamine (MEA) solvent is used in the absorber. Then, step responses of the controlled outputs with respect to the manipulated and disturbance inputs are analyzed to characterize the dynamic behavior of such process. A model predictive control (MPC) strategy is proposed on the basis of the understanding from the analysis. Finally, the closed-loop performances of the proposed control strategy and decentralized PID controllers are compared to demonstrate the effectiveness of the MPC strategy. The MPC strategy demonstrates that it can track the set-point change at least 20 min faster than the PID strategies, and stabilize the stripper section about 200 min faster.