Abstract

Post combustion CO2 capturing holds an important position in the area of carbon capture and se-questration (CCS). Research in this area range from experimental work to modeling work. Dynamic models are interesting since these describe the plant operation during variations, up-stream or down-stream, and due to their use-fulness in control design. To take full advantage of state space models in control design, it is necessary to have on-line knowledge of all states, also states that are not measured directly. Techniques for state estimation, such as Kalman filter based methods, thus form key technology for advancing control solutions. But state estimation is also of interest in its own right for making available on-line knowledge of states. In this study, a dynamic model of an amine based CO2 capture plant is used as a basis for a state estimator. A high order version of the model is used to represent the “real” plant. A reduced order model of the plant is then used for state estimation, and the Ensemble Kalman filter is used.