Equation Oriented Coal Oxycombustion Flowsheet Optimization

Flowsheet optimization and synthesis improves process design by systematically exploring nearly uncountable process configurations. This allows for the selection of the best designs, ultimately reducing costs, increasing efficiency and improving operability. Applied to tomorrow’s power systems, large scale optimization will help reduce the capital and operational costs of carbon capture, utilization and sequestration. Furthermore optimization frameworks allow for an automated, systematic screening of promising technologies, ranging from single systems, such as membranes for CO2 separation, to entire new flowsheets, such as coal oxycombustion. This provides insight to the potential value of new technologies throughout the development cycle. Unfortunately reliable commercial packages for full flowsheet optimization are not readily available. Most optimization algorithms require derivative information to handle a large number of variables, but exact derivatives are not provided in many popular flowsheet simulation packages. To leverage state-of-the-art optimization algorithms not available in commercial simulators, we are developing a rich, equation based framework for flowsheet optimization. We are using this framework to explore complex design trade-offs in the oxycombustion flowsheet. In this paper we present our work modeling various sections of the oxycombustion process. These models include cubic equations of state (non-ideal thermodynamics), making them comparable to commercial flowsheeting tools. By exploiting the “open” nature of the models, accurate derivatives are obtained through automatic differentiation. This enables the use of modern optimization algorithms not available in most commercial simulators. In the paper we focus on optimization results for the air separation unit (ASU). Using the framework we minimize the specific energy for oxygen production from while simultaneously considering heat integration of the cryogenic double-column system. We also discuss planned extensions of the framework to consider heat integration in the entire flowsheet. For example many CO2 processing units proposed in literature feature a cryonics flash sequence or distillation column. Using the optimization framework we intend to systematically explore heat integration options between the ASU and CPU, along with other sections of the flowsheet. This offers a new, systematic way to consider integration in the coal oxycombustion flowhseet.