Vacuum swing adsorption (VSA) capture of CO 2 from flue gas streams is a promising technology for greenhouse gas mitigation. In this study we use a detailed, validated numerical model of the CO2VSA process to study the effect of a range of operating and design parameters on the system performance. The adsorbent used is 13X and a feed stream of 12% CO 2 and dry air is used to mimic flue gas. Feed pressures of 1.2 bar are used to minimize flue gas compression. A 9-step cycle with two equalisations and a 12-step cycle including product purge were both used to understand the impact of several cycle changes on performance. The ultimate vacuum level used is one of the most important parameters in dictating CO 2 purity, recovery and power consumption. For vacuum levels of 4 kPa and lower, CO 2 purities of >90% are achievable with a recovery of greater than 70%. Both purity and recovery drop quickly as the vacuum level is raised to 10 kPa. Total power consumption decreases as the vacuum pressure is raised, as expected, but the recovery decreases even quicker leading to a net increase in the specific power. The specific power appears to minimize at a vacuum pressure of approximately 4 kPa for the operating conditions used in our study. In addition to the ultimate vacuum level, vacuum time and feed time are found to impact the results for differing reasons. Longer evacuation times (to the same pressure level) imply lower flow rates and less pressure drop providing improved performance. Longer feed times led to partial breakthrough of the CO 2 front and reduced recovery but improved purity. The starting pressure of evacuation (which is not necessarily equal to the feed pressure) was also found to be important since the gas phase was enriched in CO 2 prior to removal by vacuum leading to improved CO 2 purity. A 12-step cycle including product purge was able to produce high purity CO 2 (>95%) with minimal impact on recovery. Finally, it was found that for 13X, the optimal feed temperature was around 67°C to maximize system purity. This is a consequence of the temperature dependence of the working selectivity and working capacity of 13X. In summary, our numerical model indicates that there is considerable scope for improvement and use of the VSA process for CO 2 capture from flue gas streams.

fast, solution-adaptive finite-volume technique for the simulation of a single adsorption step for a variety of boundary conditions was described. In this study, we apply this method to the simulation of nonisothermal PSA/VSA cycles of general complexity. Using successive substitution, stage-wise node refinement, and control algorithms, rapid cyclic steady state can be achieved even for problems with very long dynamics. We illustrate the advantages of node refinement as a useful tool for accelerating convergence to cyclic steady state. Simultaneous incorporation of control techniques into the successive substitution framework allows for rapid convergence of the PSA process to design specifications. We compare the results of our simulator to experimental data for a two-bed VSA process and find good agreement in pressures, flows, and temperatures.

This paper investigates the thermal profiles that arise in oxygen VSA, which is a prominent example of a PSA bulk gas separation process. Experimentally, it is demonstrated that the severe axial thermal profile or "cold spot" that frequently characterizes oxygen VSA can only arise if there are multilayered adsorption beds or if there are readily adsorbed trace components (such as water) that create a de facto multilayered bed. A qualitative explanation is offered to explain how this cold spot is formed. This paper also details a technique for predicting the penetration of a water-loaded zone into an oxygen VSA adsorption bed based on the method of characteristics. The results of this technique compare well with experimental and numerically simulated results. Finally, this paper demonstrates that a water-loaded zone and an inert zone of activated alumina result in very similar cyclic steady-state thermal profiles, even though the transient behaviors are markedly different.

Simulation of cyclic adsorption processes for gas separations relies on accurate models of the bed pressure drop during dynamic pressurization, depressurization, and breakthrough steps. This is especially true for rapid pressure swing adsorption (RPSA), where large gas flows can cause significant changes in the axial pressure gradient through the bed, influencing process performance. Under these conditions, there is some question as to the validity of conventional steady-state models used to represent the dynamic, rapidly changing pressures. In this study, the ability of the steady-state Ergun equation to represent pressure drop under dynamic adsorbing and nonadsorbing conditions was tested. The Ergun equation accurately reproduced dynamic depressurization and breakthrough pressure profiles using values of κ viscous and κ kinetic (the two Ergun parameters) determined experimentally from the steady-state flow of a variety of gases through a packed bed of LiLSX pellets. The full momentum equation was also tested, and errors of <0.1% were observed between the Ergun equation and the full momentum balance under dynamic conditions for a nonadsorbing packed bed. These observations and simulations suggest the Ergun equation can be reliably used to reproduce experimental profiles under dynamic adsorbing conditions where high gas flows result.

No abstract

No abstract

No abstract

A simple volumetric uptake apparatus was used to determine uptake data of N 2 on a sample of LiLSX zeolite for two different particle sizes, two temperatures, and a variety of different dosing pressure levels. Using a mass and energy conservation model for the dosing and sample volumes and the Dusty Gas Model + viscous flow for the mass transfer description at the pellet level, the Knudsen and viscous flow structural parameters were derived. Our analysis gave structural coefficients C K = 0.0827 ± 0.018 and C v = 0.0608 ± 0.026 which gave good agreement across all of the experimental runs conducted for both particle sizes and all pressure ranges. From these, tortuosity coefficients for Knudsen and viscous flow were derived and gave τ K = ε P,macro /C K = 3.7 ± 0.8 and τ v = ε P,macro /C v = 5.1 ± 2.2 respectively. These are in good agreement with reported values. The apparatus and procedure is not very sensitive to the viscous flow coefficient but is sensitive to the Knudsen coefficient. All other parameters of the model were measured or determined by calibration experiments. This study suggests that the apparatus may be useful for determination of some of the fundamental structural coefficients employed in the Dusty Gas Model.