Dynamic process modeling and simulation of the CO2 Technology Center Mongstad MEA pilot plant
MetadataVis full innførsel
In recent years, CO2 capture and storage has risen as a key part of the portfolio of solutions proposed to reduce emissions of greenhouse gases from fossil power plants and heavy industry. Various CO2 capture technologies exist, however post-combustion capture by absorption into amines is the most mature technology to this date. While CO2 capture is already realized in numerous projects of smaller scale, full scale deployment remains limited. As power plants are required to operate in a flexible manner it follows that steady-state operation of a post-combustion capture plant is unlikely. To realize the commercialization of large scale CO2 capture it is essential to gain a broad understanding of the transient behaviour of the capture plant. Dynamic process simulation is a time and cost-efficient tool to aid in this process. In this work a dynamic process model of the Technology Center Mongstad amine plant was validated with transient data from a test campaign. Simulations were performed with the process simulation software UniSim Design. The model was further used to study the dynamic response of the plant to step changes in flue gas flow rate, rich solvent flow rate and reboiler heat duty. Solvent flow rates were found to adjust fast, however a change in the solvent flow rate showed to act as the largest disturbance to the plant as a whole. Calculated dead times and settling times suggest that the main inertia lies in the absorber sump as well as the cross heat exchanger and piping. Dead times and settling times were up to 16 minutes and 4.5 hours respectively. The same step responses were also performed for a slightly different configuration including a solvent buffer tank. The buffer tank caused increased transport delay and for changes in reboiler duty it was found an extensive increase in settling time of 2-3 hours for increasing buffer tank holdup. Finally, a supervisory control layer was implemented and the performance of three different control structures was compared. The control objective was to maintain the CO2 capture rate for varying flue gas flow rate. Pairing the CO2 capture rate with the reboiler heat duty showed to give fast disturbance rejection although inefficient operation at increased loads. Pairing the CO2 capture rate with the solvent flow rate led to slower disturbance rejection, but more stable reboiler conditions.