(423c) Can the Bed Size Factor (BSF) for a Medical Oxygen Generator Employing a Pressure Swing Adsorption (PSA) Process Be Indefinitely Reduced? | AIChE

(423c) Can the Bed Size Factor (BSF) for a Medical Oxygen Generator Employing a Pressure Swing Adsorption (PSA) Process Be Indefinitely Reduced?

Authors 

Chai, S. W. - Presenter, Lehigh University
Kothare, M. - Presenter, Lehigh University
Sircar, S. - Presenter, Lehigh University


Reduction of the adsorber bed size for a medical oxygen generator by PSA producing ~ 90 mole % oxygen from ambient air is an ongoing research goal, and several U.S patents have been granted on the subject in recent years. A common approach is to design a rapid PSA system where the BSF (total pounds of adsorbent used in the system per ton of contained oxygen produced per day, lbs/TPDc O2) is reduced by lowering the total cycle times (tc, seconds) and using very small adsorbent particle sizes. The purpose of this article is to experimentally demonstrate that the BSF for a PSA process cannot be indefinitely reduced by lowering the process cycle time because of potential limitations introduced by several factors such as adsorptive mass transfer kinetics, gas-solid heat transfer resistance, column pressure drop, etc. Consequently, there may be a practical limitation on how small an adsorber bed can be for a given oxygen production rate. The limit will be governed by the cyclic process design, adsorber vessel design, sorptive properties of the adsorbent used, adsorbent particle size, product specifications, process operating conditions, etc. A very small adsorber (0.4 cm diameter x 10.8 cm long) containing ~ 1.0 gm of a commercial LiX zeolite, which is currently considered to be the best available sorbent for air separation by PSA, was used for generating the data to demonstrate the performance limitations. A Skarstorm?like PSA process consisting of four steps, viz., air pressurization, adsorption at high pressure to produce the O2 enriched product gas, depressurization to near ambient pressure, and ambient pressure back purge with a part of the product gas, was used in a simulated experimental protocol. Tests were carried out using dry and CO2 free air feed, different sorbent particle sizes, different adsorption pressures, and total cycle times ranging between 2 ? 12 seconds. Our preliminary results show that the above-described simple PSA cycle can be used to obtain a BSF of less than 50 by employing optimum adsorbent particle size and cycle time. The O2 recovery was comparable to similar PSA processes published in the literature. Comparable BSF was achieved in the past using a rapid vacuum swing adsorption (VSA) or a rapid pressure-vacuum swing adsorption (PVSA) cycle, and employing LiX zeolite. However, use of a vacuum pump may not be desirable for a compact, light weight, portable O2 generator design. This work indicates that a ?snap on? PSA O2 generator having those advantages can be designed (no moving machinery needed) for use in locations where piped compressed air is available, e.g. in civil and remote military hospitals, ambulances, air craft cabins, battlefield tanks, fish tankers, cruise ships, etc.