Hybrid separation processes are of commercial interest for many applications. Airseparation is a prime target because cryogenic distillation, adsorption, and membranes are usedcommercially, and increasing demand for oxygen for alternative energy applications willincrease construction of new plants. Hybrid adsorption-cryogenic distillation processes havebeen developed and potentially have lower operating and capital costs than their conventionalcounterparts.1In this paper, a novel membrane-cryogenic distillation air separation process for oxygenproduction is developed. This process uses a membrane gas permeator to increase the oxygenconcentration of the feed to 23.5 % or less before the main air compressor of the cryogenicdistillation plant (Figure 1). The reason for a 23.5% limit on oxygen is that above thisconcentration more expensive materials of construction are required. Although 23.5 % is a lowconcentration, it represents a more than 11 % reduction in gas flow rate. This reduction in flowrate results in reduced power requirements for compression and reduced sizes and costs of thedownstream equipment. For a basis of 1.0 m3 air/s, the power requirement to compress air from1.0 atm. to a typical distillation column operating pressure of 6.0 atm. is 322.7 kW. For the sameamount of oxygen product the power requirement to compress air that has been enriched to 23.5% oxygen from 1.0 to 6.0 atm. is 287.0 KW. The difference, 35.7 kW, is the maximumpower that can be used in the membrane permeator to have the membrane-cryogenic distillationsystem not require more energy. The hybrid approach can be applied either to new designs, orfor retrofitting and debottlenecking existing plants.Achieving the 23.5% concentration with very low energy and reasonable membrane areasis surprisingly difficult with currently available membranes. To reduce the permeateconcentration to 23.5% oxygen either a very high cut or a bypass stream is required. Very highcuts are not economical because the membrane area becomes too large. The use of a bypassstream and power recovery from the retentate proved to be the best configuration (Figure 2).Abbreviated results for this configuration are given in Table 1. Note that all of the membranesystems have an optimum cut which minimizes the power. Power is also reduced by operating atas low a feed pressure as possible. On the other hand, membrane area is reduced by operating ata low cut with a higher feed pressure.High flux, low selectivity membranes (e.g., silicone PDMS membrane in Table 1)resulted in low membrane areas, but the power requirements were greater than the savings fromreduced gas flow rates. Highly selective, low flux membranes (e.g., polystyrene membranes)had low power but huge areas (not shown in Table 1). The TMHFPSF composite membrane(Table 1) had reasonable power (net power < 35.7 kW), but based on capital cost estimate didnot appear to be economical. A combination of high flux and high selectivity appears to benecessary to be economical for this application. Carbon sieve membranes (Table 1) have acombination of high flux and high selectivity that appears to be viable if they can be made withthin active layers and sell for a reasonable cost.
展开▼
