In both these methods, the impurities are removed from the air stream by adsorbing them onto the surface of a packed adsorbent bed. The concentration of each component removed will be highest at the upstream end of the bed and tail off over a mass transfer zone. If the process is conducted indefinitely, then the mass transfer zone moves progressively downstream until it breaks through at the outlet of the bed. This would result in impurities entering the distillation process and it is therefore necessary to regenerate the adsorbent bed before this occurs.
In the PSA system, regeneration is achieved by first stopping the air flow, depressurizing the adsorbent and, usually, by passing regeneration gas through the bed counter-current to the feed direction. The pressure of the regenerating gas is generally at a lower pressure than that of the air and free of impurities.
During the feed step, the adsorption process generates heat which causes a thermal pulse to progress downstream through the bed. During the regeneration process, this same amount of heat must be supplied to desorb the impurities which have been adsorbed. In PSA, the aim is to commence regeneration before the heat pulse mentioned above has reached the downstream end of the bed. The direction of the heat pulse is reversed by the regeneration process and the heat which derived from the adsorption stage is then used for desorbing the impurities during regeneration. This avoids having to add heat during the regeneration step.
In the alternative procedure of TSA, the cycle time is extended and the heat pulse mentioned above is allowed to exit the adsorbent bed during the feed period. To achieve regeneration it is therefore necessary to supply heat to desorb the impurities. To this end the regenerating gas is heated for a period to produce a heat pulse which moves through the bed counter-current to the feed direction. This flow of heated regenerating gas is usually followed by a flow of cool regenerating gas which continues the displacement of the heat pulse through the bed toward the upstream end.
Each procedure has its own characteristic advantages and disadvantages. TSA is energy intensive because of the need to supply heat to the regenerating gas. Typically, there will be more than one unwanted gas component which is removed in the process and generally one or more of these components will adsorb strongly and others much more weakly. The temperature used for regenerating in TSA needs to be sufficient for the desorption of the more strongly adsorbed component. The temperatures required for the regenerating gas are typically sufficiently high, e.g. 150°C to 200°C, as to place demands on the system engineering which increases costs.
Whilst the PSA system avoids many of these disadvantages by avoiding the need for coping with high temperatures, the short cycle time which characterizes PSA brings its own consequences. In each cycle of operation the adsorbent is subjected to a feed period during which adsorption takes place followed by depressurization, regeneration and repressurization. During depressurization, the feed gas in the bed is vented off and lost and this known as the "switch loss". The short cycle time in the PSA system gives rise to high switch losses. Also, because the cycle is short it is necessary that the repressurization be conducted quickly. This rapid repressurization causes transient variations in the feed and product flows which can adversely affect the plant operation, particularly the operation of processes downstream from the adsorption system.
In order to address some of the problems related with TSA and PSA, a new technology has been developed for air prepurification known as TPSA (temperature-pressure swing adsorption). This is a hybrid regeneration cycle, which combines many of the benefits of PSA and TSA whilst minimizing the associated costs. The TPSA process works in a similar manner to TSA, by supplying a heat pulse to drive off impurities from the bed. However, with TPSA, the heat pulse energy is less than the energy required to desorb all the impurities from the bed. The remaining impurities are removed by continuing to supply cool regeneration gas after the heat pulse has been extinguished. This second step works because the stream of dry regeneration gas contains heat energy even if it is not heated above its supply temperature. It is found that by suitable adjustment of the conditions it is possible to achieve repeated cycles of adsorption and regeneration with only a fraction of the heat of desorption being supplied by heating the regenerating gas.
A theoretical adsorption model has been developed of the TPSA process to calculate the required regeneration conditions needed for a cyclic steady state to be achieved. This has been tested against a wealth of pilot plant and field data and shown to accurately predict the performance of these systems. Using this model, optimized TPSA designs can be obtained quickly and easily based on any given set of operating conditions.
Compared with a TSA system, it has been shown that heater power savings in the order of 20% - 60% can be achieved by moving to a TPSA cycle. It is also possible with TPSA to supply the reduced heat pulse energy by using a lower regeneration temperature (<130°C). This can reduce the cost of the vessel and accompanying pipe-work compared with TSA. The major advantage of TPSA over PSA, is that this thermal energy saving can be realized without requiring a significant decrease in cycle time. Therefore, considerable power reductions can be achieved without having to pay the cost of increased switch losses and dealing with operating problems due to rapid pressure transients.