377864 A Framework for Equation Based Optimization of Coal Oxycombustion Power Plants

Monday, November 17, 2014: 3:59 PM
M101 (Marriott Marquis Atlanta)
Alexander W. Dowling1, John P. Eason1, Jinliang Ma2, David C. Miller3 and Lorenz T. Biegler1, (1)Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA, (2)URS Corp., National Energy Technology Laboratory, Morgantown, WV, (3)U.S. Department Of Energy, National Energy Technology Laboratory, Pittsburgh, PA

In an oxy-fired power plant, coal is combusted in a N2 lean atmosphere, producing primarily CO2 and water. With a few additional processing steps, CO2 can be purified and compressed to supercritical conditions for utilization and/or sequestration. This approach may be more economical than post- and pre-combustion capture options for pulverized coal and IGCC power plants, respectively. When comparing these technologies optimized designs should be considered, lest a suboptimal design improperly disqualify a carbon capture approach.

From a systems perspective, optimizing an oxy-fired coal plant with carbon capture is very challenging, as the subsystems of the power plant are tightly integrated. For example, adjusting the flue gas recycle rate will impact the air separation unit (ASU), boiler, steam cycle and CO2 process unit (CPU) – essentially the entire power plant. Thus, it is inappropriate to optimize subsystems of the power plant individually. Instead, the entire process must be optimized at once to realize the tightest integration and best performance.

In this paper, we propose an equation-based framework for the optimization of coal oxycombustion power plants. By describing the entire process as equations instead of using commercial simulation software, we are able to use state-of-the-art nonlinear programming (optimization) algorithms capable of considering 100,000+ variables and constraints. The framework features several notable features:

- Embedded cubic equation of state thermodynamics with disappearing phases
- Simultaneous heat integration and process optimization
- Detailed distillation models with variable number of stages (important for ASU)
- Systematic initialization procedure
- Modular design for extension to other power systems and flowsheets

Using the framework, we designed cryogenic systems (ASU & CPU) for a coal oxycombustion power plant and conducted sensitivity analysis. For example, a Pareto optimality frontier was constructed by minimize the ASU specific separation energy for several O2 product purity specifications. These results provide other engineers vital information when selecting ASU operating conditions in oxy-fired power plants designs.

Progress in extending the framework to include a first-principles boiler model will also be discussed. We are developing a hybrid boiler model that provides 1-D resolution along the height of the radiant furnace for main flow and reaction related calculations, while radiative heat transfer calculations are resolved in 3-D. In contrast to CFD simulations, this model requires approximately 1 minute to evaluate, not days. Comparisons of this model with traditional CFD will be presented. We are also developing a trust-region based optimization framework to embed the hybrid model results via kriging interpolation into the full, equation-based power plant model. This is important, as it enables the use of efficient, large-scale nonlinear programming (optimization) algorithms.


Dowling, A. W., & Biegler, L. T. (2013). Optimization-based Process Synthesis for Sustainable Power Generation. Chemical Engineering Transactions, 35, 1–12.

Dowling, A. W., & Biegler, L. T. A Framework for Efficient Large Scale Equation-Oriented Flowsheet Optimization. Submitted to Computers & Chemical Engineering.

Extended Abstract: File Not Uploaded