(657a) Performance-Based Sorbent Selection for Helium Purification

Helium is a colorless, odorless, and non-toxic gas. One of the main challenges in purifying Helium from depleted gas wells is very low concentration (0.5-10 %) of Helium. However, the expected targets for a promising helium purification by adsorption-driven gas separation process are 90-95% purity with 90% recovery of helium [1]. Several adsorbents with various nitrogen adsorption capacities (between 1 bar and 20 bar) and selectivity are available in the literature. Most of the published studies in helium purification chose adsorbents solely based on high N₂ uptake. There are few studies that have reported co-optimizing PSA cycle design along with the adsorbent properties [2,3]. Traditionally this has been done using experiments, but it can be solved by optimization. A multi-objective optimization showing trade-offs among objectives such as purity and recovery for different cycles for various adsorbents is central to designing an efficient helium purification process. As the helium present in the feed gas is very low, it requires a minimum of two stages of PSA for producing helium (> 95%) with high recoveries (> 90%). Hence, we have done a systematic analysis of several different PSA cycles with different adsorbents to purify feed containing dilute helium, 1-3% He (stage-1) to concentrated helium (stage-2), 10% - 25% He in nitrogen. Genetic algorithm-based optimization was performed to obtain purity and recovery of Pareto fronts. We have performed a parametric study by varying the number of pressure equalization cycles required to achieve the target separation. The optimum adsorbent cycle configurations that satisfy constraints were ranked based on the performance that can be achieved for the dilute and concentrated helium in the feed. A 5-bed high-pressure experimental test system was designed in such a way that it could perform multiple pressure equalization cycles and monitor axial bed temperature, bed pressure, and product and exhaust composition and flow rates. The results from the process optimization study were validated using the test system.

[1] Rufford, T. E.; Chan, K. I.; Huang, S. H.; May, E. F. Adsorp. Sci. Technol 2014, 32, 49–72.

[2] Haghpanah, R.; Nilam, R.; Rajendran, A.; Farooq, S.; Karimi, I. A. AIChE J 2013, 59, 4735–4748.

[3] Jahromi, P. E.; Fatemi, S.; Vatani, A.; Ritter, J. A.; Ebner, A. D. Sep Purif Technol 2018, 193, 91–102.