(524c) A Plantwide Dynamic Model of Post-Combustion Amine Scrubbing With Aqueous Piperazine

Authors: 
Walters, M. S., The University of Texas at Austin
Edgar, T. F., McKetta Department of Chemical Engineering, The University of Texas at Austin
Rochelle, G. T., The University of Texas at Austin



Dynamic process models are important for designing control strategies and optimizing off-design operating conditions. An equation-based dynamic model has been developed for a post-combustion amine scrubbing process including CO2 absorption, CO2 desorption, and heat exchange. Aqueous piperazine was used as the amine solvent for CO2 absorption in this model. Piperazine is a second generation solvent that has superior energy performance compared to the first generation solvent monoethanolamine, which was used in previously reported first principle dynamic modeling works. Unit level dynamic models for alternative process configurations, which have been shown to reduce capital and operating costs compared to the simple absorber/stripper base case, were developed here. For the absorber, an intercooling option was included to increase mass transfer by decreasing the temperature bulge that occurs as a result of the exothermic absorption of CO2. An advanced flash stripper configuration with CO2-rich solvent bypass options for tighter energy integration was used in place of a simple stripper. The increased energy recycle from advanced configurations creates more complicated dynamics and added complexity. Steady state representations were used for other unit operations, including heat exchangers, pumps, and control valves.

The individual unit models were combined into one flowsheet. This plantwide dynamic model was implemented in gPROMS©, an advanced process modeling platform. The model was validated using pilot plant data from the Separations Research Program at the University of Texas at Austin. In order to validate the model with pilot plant data, reasonable systematic adjustments to CO2-loading measurements and heat loss estimates were required.  After the adjustments were made, validation with temperature, pressure, and CO2-loading measurements observed at the pilot plant had adequate agreement to the predicted model outputs. Both steady state runs and dynamic behavior between runs were used for validation purposes. The multiple time scale behavior of this process was demonstrated by showing that the ratio of the amount of solvent recycled to the amount of CO2 and solvent makeup continuously added to the system greatly exceeds unity. A process control hierarchy is proposed based on the observed fast time scale associated with the large recycle flows and the slow time scale associated with total energy and CO2 holdups in the system.