Gas centrifugation with wall condensation

Abstract

Current large throughput gas cleaning processes are not able to deal with highly contaminated gas fields. Amine treatment is typically used for removing these contaminants — CO2 and H2S — from gas, but is not economic or practicable for highly contaminated gas, that is, where the concentrations of CO2 or H2S are above ca. 10 and 5%, respectively.1,2 Another major disadvantage is the low pressure at which waste gas is produced — compression is then required for waste disposal processes, such as reinjection. An alternative candidate process which has recently been evaluated is centrifugal gas separation.3 A gas centrifuge is basically a cylinder, filled with a gas mixture, and rotating at high speed. Due to the large centrifugal forces, the gas is pushed to the wall, resulting in a pressure gradient. Gases with different molecular weights have different partial pressure profiles and their mole fraction profiles vary with radius. In a methane/carbon dioxide mixture, the concentration of the CO2 increases near the wall by diffusion along the concentration gradient. We have shown experimentally that a purely gaseous process is too slow for practical application. Increasing the pressure is not that helpful for component separation in the gas phase, because to a first approximation the product of diffusion constant and pressure pD is constant that is, at high pressures the diffusion constant decreases.4 However, at the higher pressures generated in a centrifuge, there is a second much more dominant mechanism which will also cause separation — namely condensation due to the radial compression.5 This effect has some similar physical properties to what happens in the so-called evaporative centrifuge which has been previously analyzed for isotope separation, and which we have recently shown to be quite different from natural gas separation behavior in a centrifuge.6 In this study we examine whether condensation speeds up the separation process for a contaminated natural gas scenario. We identify two mechanisms for this condensation. A model is constructed which simulates the effect of condensation by centrifugal enrichment. The results of this model are compared to results of simulations of a gas/gas centrifuge. It is investigated in how far the concept of wall condensation leads to a significant increase of the separator performance in comparison with that of pure gas/gas separation by centrifugation.