Carbon dioxide separation from natural gas: evaluation of adsorbents and influence of process variables
AffiliationChemical and Biomedical Engineering
Document TypePhD thesis
Access StatusOpen Access
© 2018 Dr. David Danaci
The use of natural gas in place of coal offers a method to effectively reduce greenhouse gas emissions. Adsorption processes, specifically, pressure swing adsorption (PSA), offers an energy efficient method to perform bulk removal of CO2 from high pressure sour natural gas. A selection of zeolitic imidazolate frameworks (ZIFs) were chosen and evaluated using PSA process simulation. ZIFs -8, -14 and -71 were synthesised and gas adsorption isotherms measured. Using these isotherms, process simulation over a range of feed conditions from 15 %mol to 35 %mol CO2 at 100 bar(a) and 303 K, process performance metrics including CO2 and CH4 purity and recovery were observed. Considering CH4 is the saleable product, it was decided that its purity target of 98 %mol could not be compromised, and was thus set as a requirement. Using a 9 step, 3 bed PSA cycle, ZIF-14 was not able to meet this CH4 purity requirement; however, ZIF-8 and ZIF-71 were able to meet it while achieving CO2 purities of 24 – 45 %mol with good recovery (89 – 96 %mol), and CH4 recoveries of 48 – 34 %mol. A thorough investigation of the adsorption properties and thermodynamics of the ZIF adsorbents was also carried out. It was known that ZIF-8 demonstrated a structural transition upon adsorption; based on this work, it is quite likely that such transitions are also taking place in ZIF-14 and ZIF-71, however, definitive experimental evidence of this is required such as in-situ X-ray diffraction. Some artefacts seen in the adsorption isotherms of ZIF-71 were found to correlate with the adsorbate liquid phase surface tension. It was also found that these ZIF adsorbents deteriorated over time, contrary to existing claims in the literature. During the thermodynamic analysis, it was found that temperature invariant properties such as the differential enthalpy of adsorption were affected by temperature, and this was attributed to the structural transition. It was also found that the adsorbed phase was not ‘liquid like’ at low loadings, but was at higher loadings. A published method (osmotic potential) was used in evaluating the thermodynamics of the structural transition, and it was found that this method was inconsistent and inconvenient when multiple isotherms are used. An alternative method based on van ‘t Hoff plots was proposed and better results were observed, however, further application of this method is required to confirm its general applicability. In order to form a more general view on the topic, the literature was reviewed for high pressure adsorption isotherms of CO2 and CH4. These adsorbents, in addition to the ZIF adsorbents synthesised in this work, and a family of zeolite-Y adsorbents for which adsorption isotherms were measured for, were evaluated for process performance using a simple PSA model. It was found that ZIF-71 was most often the best performing adsorbent over the range of conditions investigated. The adsorbents were evaluated over a range of process conditions such that the output could be used as an adsorbent screening method. Feed temperature, pressure, and desorption pressure were all varied for feed compositions of 10 %mol and 30 %mol CO2. This finding implied that ZIF-71 should be investigated further for high pressure CO2/CH4 separations, with the belief that a more advanced PSA cycle could be developed or used to give better performance than was found earlier in this work. It was originally thought that the bed void fraction would be a significant limitation for high pressure separations. It was found, however, that the void fraction of the adsorbent bed did not have as great an influence as was imagined, although minor gains in CO2 purity and CH4 recovery could be found. Finally, a range of model adsorbents/isotherms were made in order to uncover the key adsorbent properties that result in good process performance. It was found that adsorbents with either a moderate loading, low heat of adsorption and low selectivity or low loading, moderate heat of adsorption and high selectivity yielded the best results. The unexpected outcome of this finding was that these characteristics align very well with ZIF-71. A range of future work was also recommended. ZIF-71 should be investigated further in different PSA cycles, and high pressure binary adsorption data should also be measured to investigate the true effects of multicomponent adsorption as they are currently unknown. A series of experiments were also suggested to confirm findings regarding the adsorption properties of the ZIF adsorbents, including in-situ X-ray diffraction for the suspected structural transitions of ZIF-14 and ZIF-71, and isotherm measurements using a wider variety of adsorbates regarding the suggested surface tension related phenomenon in ZIF-71. A selection of adsorbent development tasks were also recommended including, post-synthesis modification of the ZIF adsorbents in an effort to tune adsorption properties, further investigation into an adsorbent called ZSM-25 for which adsorption properties are not well known at the time of writing, and a hierarchical LTA-FAU adsorbent was also suggested. Regarding the separation process, it was speculated that a hybrid PSA-membrane based separation process may offer enhanced process performance, with the PSA system helping to overcome issues such as membrane plasticisation and the membrane system increasing the attainable CO2 purity and CH4 recovery.
Keywordscarbon capture; adsorption; porous materials; adsorption thermodynamics; natural gas; pressure swing adsorption; flexible adsorbent
- Click on "Export Reference in RIS Format" and choose "open with... Endnote".
- Click on "Export Reference in RIS Format". Login to Refworks, go to References => Import References