(123d) An Equation-Oriented Framework for Optimal Design of Amine-Based Carbon Capture Processes

Separating carbon dioxide from gas streams emitted by point sources such as power plants is important for reducing greenhouse gas emissions during the energy transition. Cost is a critical factor for the adoption and widespread deployment of such technologies. Reducing cost calls for careful optimization of the process design, with a view toward tailoring the capture plant for each application. In turn, this requires that the models of the process units be accurate in both predicting the phenomena occurring in the capture process and in estimating the capital cost of the equipment as a function of scale and process conditions.

Motivated by the above, this presentation describes a framework for design optimization of amine-based capture processes, as the most mature and accepted capture technology. Specifically, an equation-oriented set of models is described and used to formulate a design optimization problem aimed at minimizing capital and/or operating cost. The models are implemented in gPROMS Process^® and are available open source to interested parties^[1,2].

Figure 1 shows the model structure and main material streams. For the absorber and stripper, CO₂ absorption and desorption are represented using a rate-based mass transfer model. The liquid-film mass transfer coefficient () is correlated from data generated by a model that involves detailed speciation in the solvent phase. Experimental measurements of the PZ-H₂O-CO₂ equilibrium are used to regress a thermodynamically consistent set of equations for the necessary thermodynamic quantities.

Publicly available front-end engineering design (FEED) studies for retrofitting commercial natural gas combined-cycle (NGCC) power plants with aqueous amine carbon capture were used to develop realistic capital cost estimates^[3].

A design optimization case study for a natural gas-fired cogeneration unit (combined heat and power) is presented. Several cases are considered. As a baseline, a design optimized for typical NGCC conditions is used directly, with design variables like the intercooling configuration, rich solvent bypass ratios, and lean/rich solvent loadings held unchanged. This reflects a typical industry practice of using a standardized plant for each application. Subsequently, the design is optimized for the specific flue gas conditions of the plant. It is shown that the optimized design achieves substantial savings compared to the standardized plant.