No abstract

A two-bed pressure swing adsorption (PSA) process using activated carbon was studied to recover hydrogen from the effluent gas (H2/CO/CO2, 39.3:35.4:25.3 vol.%) of a melting incineration process. The adsorption dynamics of the activated carbon bed were investigated by breakthrough experiments. Since the product purity needs depend on demands of the incinerator site, various PSA operating conditions, such as purge to feed (P/F) ratio, adsorption (AD) step time, adsorption pressure, and feed flow rate, were investigated to recover H2 with a wide range of purities. The purity varied with P/F ratio, and the recovery varied with AD step time asymptotically because bed purification by H2 product purge approached a limitation. On the other hand, the variations in purity and recovery with P/F ratio, AD step time, and adsorption pressure were almost linear or only slightly curved. The variation in purity with feed flow rate was similar to the variation with P/F ratio while the recovery trend was more similar to the variation with the AD step time. Because the propagation velocity was significantly different for CO and CO2, the PSA performance was mainly affected by CO propagation, but the contribution of the CO2 heat of adsorption to the bed should be considered. The PSA process in this study produced hydrogen with a purity of 75.43–99.99% and a recovery of 90.99–49.29%. When a syngas (H2 and CO) is needed, the PSA process can result in high recovery and productivity because the major impurity in the product is CO.

The dynamic adsorption and desorption behaviors of lithium-modified mesoporous silica adsorbents (YSP-Li and MCF-Li) were investigated to remove sulfur compounds from natural gas. A mixture of methane and methyl mercaptan (291 μmol/mol) was used as feed gas. The adsorption capacities were determined from breakthrough experiments conducted at different temperatures and flow rates. The dynamic thermal desorption of both adsorbents was studied by applying stepwise temperature increases using nitrogen or methane. Both adsorbents could easily be regenerated at 100 °C with a purge gas. MCF-Li exhibited stronger adsorption affinity and higher adsorption capacity to methyl mercaptan than YSP-Li even though its surface area and the doping amount of Li were smaller. The desorption from MCF-Li was more affected by the applied purge gas (CH 4 or N 2 ) than YSP-Li.

It was aimed to derive rigorous momentum and energy balance equations where the change of kinetic energy in both spatial and temporal domains of a fixed-bed adsorption column was newly taken into account. While the effect of kinetic energy on adsorption column dynamics is negligible in most cases, it can become more and more influential with an adsorption column experiencing a huge pressure drop or with the gas velocity changing abruptly with time and along the column. The rigorous momentum and energy balance equations derived in this study have been validated with two limiting cases: (1) an inert gas flow through a packed column with a very high pressure drop and (2) blowdown of an adiabatic empty column. The new energy balance including the kinetic energy effect paves a way for simulating with an improved accuracy a Rapid Pressure Swing Adsorption process that inherently involves a very high pressure drop along the column and requires very high pressure change rates for column blowdown and pressurisation.

No abstract