Models of a post-combustion absorption unit for simulation, optimization and non-linear model predictive control schemes

In English An increasing demand on load flexibility in power
supply networks is the motivation to look at flexible,
and possibly optimal control systems for power
plants with carbon capture units. Minimizing the
energy demand for carbon dioxide removal under
these circumstances reduces the cost disadvantage of
carbon capture compared to conventional production.
In this work a dynamic model in Modelica of a
chemical absorption process run with an aqueous
monoethanolamine (MEA) is developed, and used
for solving optimal control problems. Starting from a
rather detailed dynamic model of the process, model
reduction is performed based on physical insight.
The reduced model computes distinctly faster, shows
similar transient behavior and reflects trends for optimal steady-state operations reported in the literature. The detailed model has been developed in Dymola, and the reduced model is used in JModelica.org, a platform supporting non-linear dynamic
optimization. First results are shown on the dynamic
optimization of the desorption column, the main
cause of energy usage in the process.