(751a) Single Bed Apparatus Designed for Studying Rapid Pressure Swing Adsorption Concepts
Pressure swing adsorption (PSA) is a cyclic gas separation technique that is just over 50 years old. It has many successful commercial applications including gas dehydration, air separation, hydrogen purification, and solvent vapor recovery. More applications are also on the horizon, including CO2 capture from flue gas
Traditional, commercial PSA systems utilize cycle times on the order of minutes to tens of minutes, with individual step times no shorter than 0.5 min (30 s) or so. However, since the mid 80’s there has been a great deal of interest in trying to reduce the cycle time to make the columns smaller and thus increase the feed throughput, which is the amount of gas processed per unit time per unit mass of adsorbent. These so called rapid PSA processes have met with limited success, however, until only recently.
Within the past 10 years, SeQual Technologies in San Diego and QuestAir in Vancouver (among less notable others) developed the rapid PSA concept into viable, commercial processes using rotary valve and rotary bed technology, respectively. They both have commercial processes that operate with cycle times on the order of a few seconds, with individual step times of less than 1 s. SeQual has been focusing on medical oxygen technology, while QuestAir has been focusing on hydrogen technology.
However, a paucity of work has been done on other rapid PSA applications; and overall, the understanding of the rapid PSA concept is quite limited with little fundamental information available. This leaves considerable room not only for the development of new theories and mechanisms underlying the rapid PSA concept, but also for the development of new applications for rapid PSA processes. The goal of this project has been to shed some light on both of these notions.
One of the shortcomings that have apparently limited the development of rapid PSA processes is a fundamental understanding of the controlling mass transfer mechanism that governs the performance. It has become clear to Ritter and co-workers over the past few years that the kind of information needed cannot be obtained by conventional measurement techniques, such as analyzing a column breakthrough curve. However, it can be obtained by building a prototypic rapid PSA process and then calibrating a rapid PSA process simulator to the performance data by varying the mass transfer coefficient. A 1-bed, rapid PSA system has been constructed for this purpose. So far, this rapid PSA system has been characterized by carrying out dynamic cycling experiments with single gases such as He, CO2 and N2, both with an empty bed and with the bed filled with beaded 13X zeolite over a wide range of cycle times and process conditions. Pressurization and depressurization step times as short as 0.5 s are possible in this system under both positive and negative gauge pressures and at different temperatures. The separation of a model binary gas mixture will be studied next. This presentation will discuss the latest results obtained with this unique 1-bed rapid PSA apparatus.