Modeling CO2 recovery for optimal dynamic operations

The development of amine scrubbing processes for coal and natural gas-fired power plants is essential to reduce CO₂ emissions. The design of tailor-made dynamic models to predict CO₂ capture in amine scrubbing processes is fundamental for optimal control operations. This paper presents the use of SIMPCA, a subspace system identification technique used to develop a dynamic empirical model for an LQG controller with integral action. Such a controller is made to attain optimal operating conditions for a CO₂ capture pilot plant. Reference signals are used in conjunction with the controller integral action to bring few process outputs towards their set-points. The results illustrate the importance of reliable model prediction in order to provide desirable closed loop response and appropriate CO₂ emission reduction.