Economic and flexible operation of solvent-based carbon capture system with solvent storage based on absorption-desorption decentralized predictive control

The deployment of lean and rich solvent storage (SS) tanks in a solvent-based post-combustion CO2 capture (PCC) system can effectively decouple the CO2 absorption and desorption processes, thereby enhancing the process flexibility and operating performance. This paper investigates the dynamic characteristics of the PCC-SS system and proposes a novel decentralized predictive control technique for optimal operation. The absorption side applies a conventional model predictive controller, which adjusts the lean solvent flowrate according to changing flue gas conditions and CO2 reduction targets; while the desorption side employs an economic model predictive controller, which regulates the reboiler heat duty and rich solvent flowrate to enhance the economic performance of solvent regeneration during the dynamic load transitions. Simulation results on a 500, 000 ton/year PCC-SS system demonstrate the effectiveness of the proposed controller. The absorption side adapts to variations in flue gas and quickly tracks the desired CO2 capture level. The desorption side achieves significant improvement in operating economics. Under free operating mode, the proposed EMPC reduces the operating cost by 28.86 % and 1.58 % respectively, compared with conventional MPC and MPC with steady-state optimization. The performance improvement remains significant even in cases of given CO2 production demand and given reboiler heat duty.