(732h) Optimization of the Amines-CO2 Capture Process By a Nonequilibrium Rate-Based Modeling Approach to Determine Operating Policies

Amine based chemical absorption has stood as the most mature and studied technology among Carbon Dioxide capture processes, featuring major advantages such as large separation efficiencies and simplicity to attach itself to existent industrial facilities. Nevertheless, its operation requires vast amounts of energy, especially at the solvent regeneration unit, making the process expensive and poorly sustainable, thus stopping its practical implementation. To tackle this problem we propose a deterministic nonlinear parameter optimization of a typical CO2-amines process, aiming to reduce the overall energy consumption while meeting CO2 capture targets as well. For such purpose, we deploy a rigorous first principles nonequilibrium rate-based and chemically reacting steady-state model of the aforementioned system.

The mathematical model (consisting of approximately 32 800 equations and variables) was implemented in the GAMS optimization environment. Critical output parameters were validated against experimental data from a pilot plant, which consists of two columns interconnected (an absorber and a desorber). Moreover, relevant profiles were successfully compared against results from simulations in Aspen Plus. Once validated, the model was used to perform several sensitivity analyses on interesting input parameters in order to study the response of the system’s separation degree and energy demand. Afterwards, a series of deterministic nonlinear programming (NLP) problems were proposed, where three different objective functions were studied in order to obtain optimal operational policies which enable both CO2 sequestration and low energy expenditure.

Up to our best knowledge, optimization studies with this degree of modeling complexity (i.e. rate-based models along with NLP) have not been reported in the open literature. Results reflect the importance of optimizing both columns simultaneously instead of modeling them separately, since a complete model like the one at hand takes into account complex interactions between variables all across the process, which may be counterintuitive. By applying the optimal operating policies suggested by the model, an energy reduction of up to 25% was achieved when compared against conventional industrial data from plants operating under heuristic policies. Furthermore, this work emphasizes the significance of performing optimization studies in advance of practical implementations of emerging technologies. As can be seen, the problem is challenging and timely due to global warming concerns.