Quantitative Evaluation of the Effect of Uncertainty in Property Models on the Simulated Performance of Solvent-Based CO2-Capture

Abstract

Process modelling has increasingly become the basis for the quantitative design and evaluation of chemical processes in general and CO2-capture processes in particular. How good are the available tools? Where are they weak and where are they strong? Which data are best to develop reliable models? How accurate should the data be? How large should the safety factors be? An effective way to answer these important questions is through uncertainty analysis, and this work attempts to do so. The approach taken here is first to ensure that the model used properly meets asymptotic limits (e.g., predicts equilibrium behavior for high kinetics or long contact times), and then to devise practical perturbation variables that provide quantitative engineering insight into process uncertainties. Our preliminary results indicate that uncertainty analysis can help evaluate the reliability and accuracy of the process models used for solvent-based CO2 capture.