Erden, Lutfi, Ebner, Armin D., and Ritter, James A.
Energy & Fuels; March 2018, Vol. 32 Issue: 3 p3488-3498, 11p
Simulations were carried out to study the purification of CH4from pretreated landfill gas containing 88 vol % CH4and 12 vol % N2using BPL activated carbon and three different four-bed four-step pressure vacuum swing adsorption (PVSA) cycles. All three PVSA cycle schedules included feed (F), heavy reflux (HR), countercurrent depressurization (CnD), and light product pressurization (LPP) steps. The light-end heavy-reflux plus recycle (LEHR–Rec) cycle had a HR step fed to the light end of a bed by a partial reflux of the product from the CnD step and a full recycle of the product from the HR step blended back with the feed. The heavy-end HR plus recycle (HEHR–Rec) cycle was the same as the LEHR–Rec cycle except the HR step was fed to the heavy end of a bed. The heavy-end HR (HEHR) cycle was the same as the HEHR–Rec cycle, except that it did not have Rec, so the product from the HR step was taken as light product. For all three PVSA cycles, increases in either the feed throughput or the HR reflux ratio caused the CH4recovery to decrease or the CH4purity to increase, and concomitantly, the feed throughput did not have any effect on the vacuum pump/compressor energy penalty, while increasing the HR reflux ratio caused the energy penalty to increase. The energy penalty was essentially the same for all three PVSA cycles. Recycle-to-feed from the HR step was also more important than whether the HR step was carried out cocurrently or countercurrently, but the cocurrent approach was generally better. Overall, pipeline-quality CH4with a purity greater than 98 vol % could be produced with both the HEHR–Rec and LEHR–Rec at feed throughputs as high as 500 L(STP) h–1kg–1, with the HEHR–Rec generally exhibiting the better performance and the HEHR cycle exhibiting the worst performance. The best performance exhibited by the HEHR–Rec had a CH4purity of 99.4 vol %, a CH4recovery of 99.2%, a feed throughput of 500 L(STP) h–1kg–1, and an energy penalty of 27.0 kJ mol–1CH4produced.
Abdollahi Govar, Anahita, Ebner, Armin D., and Ritter, James A.
Energy & Fuels; December 2016, Vol. 30 Issue: 12 p10653-10660, 8p
The effect of H2O vapor on the adsorption and desorption of CO2and vice versa, both on a commercially viable solid amine sorbent (SAS), were studied using a thermogravimetric analyzer at constant temperature. The SAS was prepared by physically immobilizing PEI (Mn423) in a porous silica (CARiACT G10). Studies were carried out at two concentrations of CO2(2 vol % in N2and 100 vol %), one concentration of H2O vapor (2 vol %), and four temperatures (40, 60, 80, and 100 °C). The results revealed three separate and identifiable mechanisms or processes taking place on the SAS, i.e., the individual adsorption of CO2or H2O and the coadsorption of H2O and CO2. The individual adsorption of either CO2or H2O was noncompetitive or completely independent of each other thermodynamically for both concentrations of CO2and all four temperatures. In addition, the individual adsorption of CO2was fully reversible, while the individual adsorption of H2O was only partially reversible. The coadsorption of CO2and H2O was exothermic, with either irreversible or very slow desorption kinetics. Coadsorption was observed appreciably only at 60 and 40 °C at both CO2concentrations and only very slightly at 80 °C and 2 vol % CO2. It was not observed at 100 °C at either CO2concentration or at 80 °C at 100 vol % CO2. The influence of CO2or H2O on the kinetics of any of these processes was revealed only at the two lower temperatures, with H2O having a positive influence on the desorption kinetics of CO2at 40 °C and with CO2, if adsorbed previously, having a negative influence on the adsorption kinetics of both H2O and the coadsorption of both CO2and H2O at both 60 and 40 °C. When these effects existed, they were not that significant, at least not at 2 vol % H2O vapor. Overall, these results thus indicated that the effect of H2O vapor in the feed of a CO2stream does not have to be accounted for mechanistically when modeling a cycle adsorption process based on this commercially viable SAS.