Design and Implementation of an Optimal Control Framework for Post-combustion CO2 Capture Process

Abstract

The high penetration of renewable sources into fossil fuel-based thermal power plants demand for operational flexibility with Carbon Capture and Storage (CCS) technologies. Post-combustion CO2 Capture (PCC) processes using chemical absorption of CO2 in flue gas need to cope with flexible operation and the CO2 capture rate must be controlled under fluctuating flue gas conditions. To fill this gap, in this work, an optimal control framework is proposed and implemented on PCC with monoethanolamine (MEA)-based CO2 capture process simulation. A two-input-two-output control structure is selected from PCC that consist of flue gas flow rate and lean MEA flow rate as the input/manipulated variables while CO2 capture rate and reboiler duty are considered as the output/controlled variables. Open-loop simulations are performed in which simulated step tests are designed by individually moving the input variables as steps and collecting the resulting data for the output variables. The classical autoregressive model with exogenous inputs (ARX) method is used for deriving the data-driven simplified dynamic model that can be embedded inside the Biologically Inspired Optimal Control Strategy (BIO-CS) casted as Model Predictive Control (MPC) to compute control moves for simultaneous control of both output variables in the dynamic simulation. The results are compared to the standalone Proportional-Integral-Derivative (PID) controller existed in the simulation in terms of the time required to reach new steady state and output tracking error. The proposed approach improves the overall performance of the process resulting in faster and flexible setpoint tracking during ramp decrease in the flue gas flow rate case study and thus providing a promising alternative.