Pressure swing adsorption (PSA) is a promising technology for carbon capture and sequestration (CCS). However, while there has been much interest in PSA process development, the choice of adsorbent for the separation is just as important as the process configuration. Therefore, it is important to develop PSA processes in conjunction with development of the adsorbent. One class of adsorbents that received significant interest in the past few decades are metal-organic frameworks (MOFs). MOFs are crystalline, porous materials synthesized through self-assembly of metal nodes and organic linkers. Due to the great variety of organic linker and metal node combinations, thousands of potential MOFs can be synthesized and specifically tailored for any applications including CCS. This work focuses on the simultaneous development of MOFs and PSA processes to better understand and improve the process to reduce the cost of CO2 capture.', '\tFirst, we investigated the impact of water on the performance of different materials in a two-stage Skarstrom cycle. In this investigation, we found that the dehydration method that resulted in the lowest CO2 cost was to feed the wet flue gas directly into the column. This resulted in the first section of the column acting as a desiccant, capturing the water, while the remainder of the bed separated the CO2 from the N2. From this work we also found that the inclusion of water can shift the ranking of the ideal material, as we saw that zeolite 13X was the best performing material under dry conditions, but zeolite 5A performed equally well under wet conditions. Next, we developed a new general evaluation metric from CO2 capture cost data for 190 MOFs using the four step Fractionated-Vacuum Pressure Swing Adsorption (FVPSA) cycle. This metric has a higher Spearman correlation coefficient with the cost data than several other metrics previously proposed, making it useful for future work in quickly evaluating a MOF’s potential for CCS applications. In developing the metric, we discovered that the most important feature of this metric is the working capacity of N2, followed by the working capacity of CO2. We also evaluated 16 MOFs that are reported in the literature to be promising for CCS using a modified Skarstrom cycle to rank them based on economic performance metrics, finding UTSA-16 as the best performer. Finally, we showed that to reduce the computational time for simulating PSA cycles, artificial neural networks (ANNs) are a promising surrogate model that are able to simulate PSA steps that may be used in a given cycle.

Last modified

• 10/14/2019

Creator

• Karson Troy Leperi

DOI

• https://doi.org/10.21985/n2-p3nf-yw87

Subject

• Chemical and Biological Engineering

Keyword

• Chemical engineering

Date created

• 2018-01-01

Resource type

• Dissertation

Rights statement

• In Copyright