Modeling and Optimization of a Novel Chilled Ammonia Absorption Process and Amine-Appended Metal-Organic Frameworks for CO2 Capture

Abstract

Post-combustion capture is one of the leading technologies for CO2 abatement from anthropogenic sources which have contributed significantly to the rise of atmospheric greenhouse gases. Specifically, solvent-based capture post-combustion processes are the industry standard but can suffer drawbacks such as high energy penalties and corrosion. In this work, two possible improvements are investigated which have been recently proposed in the literature. The first is aqueous ammonia as a capture solvent which has been shown to have several advantages including, but not limited to, a lower regeneration energy. The second is a novel solid sorbent, an amine-appended metal-organic framework (MOF). The MOF exhibits several promising attributes, namely, a step-shaped adsorption isotherm which leads to lower working capacities and lower regeneration energies when compared to traditional solid sorbents. The overall goal of this work is to develop rigorous mathematical models which can be used for process design and economic evaluation of these technologies. First, an integrated mass transfer model is developed for the chilled ammonia process (CAP). This model is developed using a simultaneous regression approach that has been recently proposed in the literature with parameter estimation performed using data from a pilot plant source and wetted-wall column. The optimally estimated parameters are shown to have a lower prediction error to validation data than parameters found in literature. The integrated mass transfer model is then used to develop a model for a novel chilled ammonia process. The process includes a NH3 abatement system which utilizes a reverse osmosis membrane to aid in separation and reduce the energy penalty. Simulation of the process shows that the membrane can significantly reduce the energy requirement of the reboiler, condenser, and cooler in the abatement section. Uncertainty of the estimated mass transfer parameters is quantified using a fully Bayesian approach which is demonstrated to show a significant reduction in the prediction uncertainty of key process indicators. Second, isotherm and kinetic models are developed for amine-appended MOFs, dmpn-Mg2(dobpdc) and Mg2(dobpdc)(3-4-3). The step-shaped adsorption isotherms exhibited by these MOFs present a modeling challenge since many of the traditional isotherm models are unable to capture step transitions. Three isotherm models are examined in this work, a weighted dual-site Langmuir model found in literature, a dual-site Sips model developed in this work, and an extended weighted Langmuir model also developed in this work. Parameter estimation is performed using available isotherm data and it is shown that the models are able