Hydrogen separation by multi-bed pressure swing adsorption of synthesis gas

The performance of multi-bed pressure swing adsorption (PSA) process for producing high purity hydrogen from synthesis gas was studied experimentally and theoretically using layered beds of activated carbon and zeolite 5A. Nonisothermal and nonadiabatic models, considering linear driving force model and Dual-site Langmuir adsorption isotherm model, were used. The effects of the following PSA variables on separation process were investigated: linear velocity of feed, adsorption time and purge gas quantity. As a result, we recovered a high purity H₂ product (99.999%) with a recovery of 66% from synthesis gas when the pressure was cycled between 1 and 8 atm at ambient temperature.     

Backfill Cycle of a Layered Bed H2 PSA Process

The backfill cycle of two-bed PSA process using activated carbon beds, zeolite 5A beds, and layered beds was studied experimentally and theoretically to recover high purity H₂ from coke oven gas. In a layered bed PSA, a comparison was made between two PSA processes with/without a backfill step before the feed pressurization step. Since the backfill step made the adsorption bed rich in H₂ and this led to a rather steep concentration wave front at the feed pressurization step, incorporating a backfill step resulted in an increase in product purity with a decrease in recovery. Each step of the single-adsorbent and layered bed PSA processes with a backfill step was simulated with a dynamic model incorporating mass, energy, and momentum balances. The model agreed well with the experimental results in predicting the product H₂ purity and recovery, thus giving a basic understanding of the bed dynamics of a backfill cycle. While the concentration and temperature profiles of a layered bed in each step showed characteristic behavior of each adsorbent in each layer, the product purity of a layered bed was not between the limits of two single-adsorbent bed processes. The concentration profiles predicted by simulation showed that CO and N₂ played an important role in obtaining high H₂ purity.     

Design of a H2 PSA for cogeneration of ultrapure hydrogen and power at an advanced integrated gasification combined cycle with pre-combustion capture

A novel hydrogen pressure swing adsorption system has been studied that is applied to an advanced integrated gasification combined cycle plant for cogenerating power and ultrapure hydrogen (99.99+ mol%) with CO₂ capture. In designing the H₂ PSA, it is essential to increase the recovery of ultrapure hydrogen product to its maximum since the power consumption for compressing the H₂ PSA tail gas up to the gas turbine operating pressure should be minimised to save the total auxiliary power consumption of the advanced IGCC plant. In this study, it is sought to increase the H₂ recovery by increasing the complexity of the PSA step configuration that enables a PSA cycle to have a lower feed flow to one column for adsorption and more pressure equalisation steps. As a result the H₂ recovery reaches a maximum around 93 % with a Polybed H₂ PSA system having twelve columns and the step configuration contains simultaneous adsorption at three columns and four-stage pressure equalisation.     

Heuristics for synthesis and design of pressure-swing adsorption processes

Simulation based synthesis and design of adsorptive enrichment of CO from tail gas having 51?% CO are presented. The adsorption breakthrough curve simulation using this feed gas composition, helped to provide a starting guess of the adsorption step duration in a pressure-swing adsorption cycle for meeting the purity and recovery targets. Use of smaller bed dimensions facilitated the simulation of many cycles. These simulations helped to decide the operating pressure range, operating temperature, constituent steps of the cycle, their sequence, direction of pressurization of the bed, number of beds in the cycle and the composition of the streams to be used for pressurization and/or rinse and/or purge steps. Only an optimally designed pressure-vacuum-swing adsorption cycle achieves the stiff separation targets of getting an extract having 80?% pure CO at 80?% recovery in a single-stage with an adsorbent that uses physical adsorption and offers CO/CH₄ sorption selectivity of just 2.44. Additional simulations are done wherein the bed sizes and velocities are increased to predict the performance of a large-scale unit. These require deciding only the durations of the steps that are finalized from the small-scale unit simulations. These durations were kept fixed and the bed dimensions were varied till the separation targets are obtained for the particular feed rate. The scale-up criteria was matching residence times in the bed. A strategy for treating a feed gas having only 30?% CO is also discussed and a novel concept of cascaded PSA is evaluated using simulations. Some heuristics are evolved from the studies.     

Heavy reflux PSA cycles for CO2 recovery from flue gas: Part I. Performance evaluation

This study evaluated nine stripping PSA cycle configurations, all with a heavy reflux (HR) step, some with a light reflux (LR) step, and some with a recovery (REC) or feed plus recycle (F+R) step, for concentrating CO₂ from stack and flue gas at high temperature (575 K) using a K-promoted HTlc. Under the process conditions studied, the addition of the LR step always resulted in a better process performance; and in all cases, the addition of a REC or F+R step surprisingly did not affect the process performance except at low feed throughputs, where either cycle step resulted in a similar diminished performance. The best cycle based on overall performance was a 5-bed 5-step stripping PSA cycle with LR and HR from countercurrent depressurization (CnD) (98.7% CO₂ purity, 98.7% CO₂ recovery and 5.8 L STP/hr/kg feed throughput). The next best cycle was a 5-bed 5-step stripping PSA cycle with LR and HR from LR purge (96.5% CO₂ purity, 71.1% CO₂ recovery and 57.6 L STP/hr/kg feed throughput). These improved performances were caused mainly by the use of a very small purge to feed ratio (γ=0.02) for the former cycle and a larger one (γ=0.50) for the latter cycle. The former cycle was good for producing CO₂ at high purities and recoveries but at lower feed throughputs, and the latter cycle was useful for obtaining CO₂ at high purities and feed throughputs but at lower recoveries. The best performance of a 4-bed 4-step stripping PSA cycle with HR from CnD was disappointing because of low CO₂ recoveries (99.2% CO₂ purity, 15.2% CO₂ recovery and 72.0 L STP/hr/kg feed throughput). This last result revealed that the recoveries of this cycle would always be much lower than the corresponding cycles with a LR step, no matter the process conditions, and that the LR step was very important to the performance of these HR cycles for this application and process conditions studied.     

Separation of a binary gas mixture by pressure swing adsorption: Comparison of different PSA cycles

Gas separation of a binary gas mixture by various pressure swing adsorption (PSA) cycles was studied by a numerical simulation in order to provide a guidance in selecting PSA cycles. PSA cycles considered in this study are 3, 4-step cycles for production of only one component and a cycle with pressure equalization for production of a light component. 4 and 5-step cycles for simultaneous production of both components of a binary gas mixture are also considered. Separation of a CH₄/CO₂ gas mixture with zeolite 5A was chosen as a case study. Performances of cycles were examined and compared in view of purity, recovery and productivity. Their relative advantages were discussed. Inclusion of a purging step to a 3-step cycle for production of only one component improves a cycle performance. Further performance improvement of a cycle for production of a light component can be achieved by employing pressure equalization. Sircar's 4-step cycle with a recycle of effluent shows the best performance in view of purity and recovery among cycles for simultaneous production of both components.Nomenclature B Langmuir adsorption constant, bar^–1 - C concentration of sorbate in gas phase, mol/m³ - D defined by Eq. (7) - n amount of sorbate in solid phase, mol/kg - n _s monolayer amount adsorbed, mol/kg - P pressure, bar - R gas constant, J/mol K - T temperature, K - t time, s - U effluent gas velocity, m/s - z height of one cell, m - bulk density of a bed, kg/m³ - bed void fraction - A CH₄ - B CO₂ - H high pressure feed step - P purge step - R heavy-component rinse step - i cell number (i=1 toN)     

Optimizing the separation of gaseous enantiomers by simulated moving bed and pressure swing adsorption

The resolution of racemic gas mixtures by simulated moving bed (SMB) and pressure swing adsorption (PSA) is investigated by dynamic simulation and optimization. Enantiomer separation of inhalation anesthetics is important because there is evidence that the purified enantiomers may have different pharmacological properties than the racemate. The model parameters reported in an experimental investigation performed elsewhere are used to study the feasibility of this separation using SMB and PSA configurations. Both processes were modeled in gPROMS^® as systems of differential algebraic equations. Operating conditions are optimized such that the feed throughput and product recovery for each process were maximized subject to equal constraints on the pressures and superficial gas velocities. SMB was found to be capable of resolving racemic feed mixtures with purity and recovery exceeding 99%. On the other hand, PSA was also able to provide a single purified enantiomer with low recovery of about 30% which may limit its application to enantiomer separation. Nevertheless, PSA consumes less desorbent, and achieves higher throughput at the sacrifice of lower recovery.     

Parametric Study of a Pressure Swing Adsorption Process

The performance of a pressure swing adsorption (PSA) process for production of high purity hydrogen from a binary methane-hydrogen mixture is simulated using a detailed, adiabatic PSA model. An activated carbon is used for selective adsorption of methane over hydrogen. The effects of various independent process variables (feed gas pressure and composition, purge gas pressure and quantity, configuration of process steps) on the key dependent process variables (hydrogen recovery at high purity, hydrogen production capacity) are evaluated. It is demonstrated that many different combinations of PSA process steps, their operating conditions, and the feed gas conditions can be chosen to produce an identical product gas with different hydrogen recovery and productivity.     

Carbon monoxide isotope enrichment and separation by pressure swing adsorption

Simulations of three different 3-bed 3-step pressure swing adsorption (PSA) cycles were carried out to study the enrichment and recovery of ¹⁴CO from an isotopic mixture of ¹⁴CO, ¹³CO and ¹²CO using NaX zeolite. Each PSA cycle included feed pressurization/feed (FP/P), heavy reflux (HR) and countercurrent depressurization (CnD) steps; they differed only in the way the CnD step was carried out: PSA Cycle I was carried out under total reflux (i.e., with no ¹⁴CO heavy product production); PSA Cycle II was carried out with discontinuous ¹⁴CO heavy product production; and PSA Cycle III was carried out with continuous ¹⁴CO heavy product production. The effects of the CnD step valve coefficient (c _v ), heavy reflux ratio (R _R ), and cycle time (t _cyc ) on the PSA process performance were determined in terms of the ¹⁴CO enrichment, ¹⁴CO recovery and CO feed throughput. The results showed that there was essentially no limit to the extent of the ¹⁴CO enrichment, despite the inherently low ¹⁴CO/¹²CO (1.05) and ¹⁴CO/¹³CO (1.12) separation factors for these isotopes on NaX zeolite. Under total reflux an optimum c _v was found for the CnD step and ¹⁴CO enrichments as high as 152 were obtained. Using the optimum c _v under finite reflux, a ¹⁴CO enrichment approaching 20 and a ¹⁴CO recovery approaching 100 % were easily achieved with discontinuous (PSA Cycle II) or continuous (PSA Cycle III) ¹⁴CO heavy product production. There was essentially no difference in the performance of PSA Cycles II and III, a counterintuitive result. The ¹⁴CO enrichment and the ¹⁴CO recovery both increased with decreasing CO feed throughputs and higher R _R , which were always very close to unity.     

Pressure-Swing Adsorption Using Layered Adsorbent Beds with Different Adsorption Properties: I—Results of Process Simulation

A simulation study was conducted on layered-bed pressure-swing adsorption, PSA, processes with adsorbents that differ in their adsorption properties. As an example, an oxygen, O₂, vacuum-swing adsorption, VSA, process was analyzed to investigate relationships between process performance and adsorption properties of the adsorbents used. For two adsorbents with identical nitrogen-to-oxygen, N₂/O₂, selectivity but different N₂ and O₂ capacities, placing the high-capacity adsorbent at the product end and the low-capacity adsorbent at the feed end of the adsorption bed gives a better performance than the case of reversing layering of these adsorbents. However, for two adsorbents with different values of N₂/O₂ selectivity but identical N₂ capacity, changing the bed-layer configuration does not show a significant difference in O₂-VSA performance. The advantages of layering a high-capacity adsorbent on product end of the bed are demonstrated by an examination of the N₂-loading difference in a VSA cycle. The modeling study also reveals an effect of cycle features (e.g., equalization step) on the effectiveness of using layered-bed configurations in VSA/PSA processes. It suggests that layering appropriately two adsorbents with different adsorption properties could result in better VSA/PSA-process performance than using a single-layer bed with either of the two adsorbents.     

Sensitivity of PSA process performance to input variables

Mathematical models for pressure swing adsorption (PSA) processes essentially require the simultaneous solutions of mass, heat and momentum balance equations for each step of the process using appropriate boundary conditions for the steps. The key model input variables needed for estimating the separation performance of the process are the multicomponent adsorption equilibria, kinetics and heats of adsorption for the system of interest. A very detailed model of an adiabatic Skarstrom PSA cycle for production of high purity methane from a ethylene-methane bulk mixture is developed to study the sensitivity of the process performance to the input variables. The adsorption equilibria are described by the heterogeneous Toth model which accounts for variations of isosteric heats of adsorption of the components with adsorbate loading. A linear driving force model is used to describe the kinetics. The study shows that small errors in the heats of adsorption of the components can severely alter the overall performance of the process (methane recovery and productivity). The adsorptive mass transfer coefficients of the components also must be known fairly accurately in order to obtain precise separation performance.     

A Robust Metal-Organic Framework with Scalable Synthesis and Optimal Adsorption and Desorption for Energy-Efficient Ethylene Purification

Adsorptive separation is an energy-efficient alternative, but its advancement has been hindered by the challenge of industrially potential adsorbents development. Herein, a novel ultra-microporous metal-organic framework ZU-901 is designed that satisfies the basic criteria raised by ethylene/ethane (C₂H₄/C₂H₆) pressure swing adsorption (PSA). ZU-901 exhibits an “S” shaped C₂H₄ curve with high sorbent selection parameter (65) and could be mildly regenerated. Through green aqueous-phase synthesis, ZU-901 is easily scalable with 99 % yield, and it is stable in water, acid, basic solutions and cycling breakthrough experiments. Polymer-grade C₂H₄ (99.51 %) could be obtained via a simulating two-bed PSA process, and the corresponding energy consumption is only 1/10 of that of simulating cryogenic distillation. Our work has demonstrated the great potential of pore engineering in designing porous materials with desired adsorption and desorption behavior to implement an efficient PSA process.     

Ethanol-Water Separation in the PSA Process

An investigation was carried out on ethanol-water separation employing a PSA adsorption cycle with zeolite 3A as the adsorbent. The cycle was operated under the following operating variables: feed flow (2, 4, 6 and 8 L/h), adsorption temperature (200°C) and adsorption pressures (2, 4 and 6 bar). All experimental runs were performed under vacuum of 0.2 bar in the desorption step. The effect of these variables on the enrichment and recovery percentage of the product, on the productivity and on the total cycle time was studied, using the adsorption pressure as a parameter. The results showed that these variables significantly affect the responses of interest. We also studied the influence of such variables as: adsorption pressures, desorption pressures, flow rates and adsorption temperatures (T), on the enrichment, recovery, productivity and on the total cycle time. The data were obtained from operational cycles in a Kahle's system. The experiments were organized by a three level factorial design. The results obtained were compared with from the fitted empirical equations as well as with the corresponding surface responses and all variables showed to be influential. Last, by an optimal combination of the variables was obtained by means of the ridge analysis method and the optimized multi-response method.     

Pressure-Swing Adsorption Using Layered Adsorbent Beds with Different Adsorption Properties: II—Experimental Investigation

An experimental study was conducted on a layered-bed pressure-vacuum-swing adsorption, PVSA, process with adsorbents that differ in their adsorption properties. An oxygen, O₂, PVSA process was employed as an example for investigating how the process performance is affected by bed-layering configuration under different operating conditions for specific purge, product purity, and cycle feature. For two adsorbents with similar nitrogen-to-oxygen, N₂/O₂, selectivity but different N₂ and O₂ capacities, placing the high-capacity adsorbent at the product end and the low-capacity adsorbent at the feed end of the adsorption bed results in a better performance than in the case of reversing the layer positions of those adsorbents. The benefit of placing the adsorbent with higher capacity at the product end becomes more significant at high O₂ product-purity levels. The experimental data obtained in this investigation agree well with simulation results reported earlier.     

Capture of CO2 from flue gas streams with zeolite 13X by vacuum-pressure swing adsorption

Vacuum swing adsorption (VSA) capture of CO₂ 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% CO₂ 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 CO₂ purity, recovery and power consumption. For vacuum levels of 4 kPa and lower, CO₂ 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 CO₂ 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 CO₂ prior to removal by vacuum leading to improved CO₂ purity. A 12-step cycle including product purge was able to produce high purity CO₂ (>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 CO₂ capture from flue gas streams.     

Piston-driven ultra rapid pressure swing adsorption

The piston-driven ultra rapid pressure swing adsorption (URPSA) equipment was developed and oxygen enrichment from air was examined as an example. The adsorbent bed is directly connected to the cylinder where a piston moves at high frequency. Thus pressurization and depressurization in the bed are driven by mechanical piston motion, which can achieve far more rapid cycles compared with the conventional pressure swing operation using valves. The cycle time is usually on the order of seconds or sub seconds. Oxygen enrichment from air up to about 60% or higher of oxygen concentration was achieved by small-scale equipment using zeolite 5A with a oxygen production capacity of 100 Nm³-product gas/m³-zeolite/hr, which is about ten times larger than those of commercialized PSAs for this purpose.A simplified numerical model describing the mass transfer taking place in URPSA was developed. The model could simulate fairly well the air separation characteristics in terms of oxygen concentration, oxygen production capacity and oxygen yield. The proposed model helps in the understanding of the basic nature of URPSA and possible applications. This novel PSA is promising as a compact yet high-capacity PSA to be utilized in a wide variety of applications.     

Effect of Operation Symmetry on Pressure Swing Adsorption Process

In a multi-bed pressure swing adsorption (PSA) process, cycle steps with gas flow transferring from one bed to another such as equalization, purge, etc. are generally practiced to enhance the product recovery. However, if the flows for the connected beds in these steps are not balanced, the PSA process may not operate in a symmetrical manner. In the modeling of the PSA process, most of the simulations consider only one bed and assume that the rest of the beds would behave in a same way. In order to assess the impact of bed symmetry on the PSA performance, a new PSA model capable of studying bed symmetry in a two-bed system is developed. Experimental results from this paper show that uneven equalization flow can result in a lower product purity and a peculiar purity curve at different equalization levels. This phenomenon can be successfully predicted by this model. Simulation results also show that in large-scale PSA units, asymmetrical operation can cause drastically different temperature profiles in different adsorbers and hence a much lower performance. This paper demonstrates the importance of maintaining operation symmetry in PSA processes.     

Hydrogen PSA process product purity control method and controller

It is well known in the industry that a primary means for controlling the pressure swing adsorption (PSA) process product gas purity is the adjustment of PSA feed time or adsorption time. If the product impurity is too high, the feed time is shortened and if the impurity level is below the target the feed time is increased. Conventionally, the plant operator monitors the product purity and manually adjusts the feed time. Several control methodologies such as classical feedback and feedforward systems were suggested to automate this task with limited success. A novel control methodology based on the measurement of impurity fronts within the adsorber bed was developed by the Praxair Adsorption R&D team. The response of the concentration measurements inside the adsorber vessel to the process upsets and changes in feed time is more rapid than in the product stream. Consequently, closed loop control performance can be made much more effective and the operating impurity set points for product gas can be more aggressive resulting in longer PSA feed times, higher bed utilization and thus higher hydrogen recovery. The control methodology will be discussed in greater detail along with the advantages it has to offer such as improved process performance, disturbance rejection capability and improved process robustness. The control methodology will be illustrated using the hydrogen PSA process as an example.     

Adsorption and diffusion of nitrogen, methane and carbon dioxide in aluminophosphate molecular sieve AlPO4-11

The knowledge about the adsorption and diffusion properties (specially about diffusion) of aluminophosphate molecular sieves is very scarce in the literature. These materials offer interesting properties as adsorbents as they have a polar framework and do not contain charge-balancing cations. In this work, the adsorption isotherms of nitrogen, methane and carbon dioxide over an AlPO₄-11 sample synthesized in our laboratories have been measured with a volumetric method at 25, 35, 50 and 65 °C over a pressure range up to 110 kPa. The adsorption capacities of each gas are determined by the strength of interaction with the pore surface (carbon dioxide > methane > nitrogen). The equilibrium selectivity to carbon dioxide is quite high with respect to other adsorbents without cations due to the polarity of the aluminophosphate framework. The adsorption Henry’s law constants and diffusion time constants of nitrogen, methane and carbon dioxide in the synthesized AlPO₄-11 material have been measured from pulse experiments. A pressure swing adsorption (PSA) process for recovering methane from a carbon dioxide/methane mixture (resembling biogas) has been designed using a dynamic model where the measured adsorption equilibrium and kinetic information has been incorporated. The simulation results show that the proposed process could be simpler than other PSA processes for biogas upgrading based on cation-containing molecular sieves such as 13X zeolite, as it can treat the biogas at atmospheric pressure, and it requires a lower pressure ratio, to produce high purity methane with high recovery.     

Removal performance and mechanism of Fe3O4/graphene oxide as an efficient and recyclable adsorbent toward aqueous Hg(II)

Decontamination of aqueous heavy metal is a challenging task of environmental remediation. Herein, we demonstrated an adsorptive method for efficient removal of aqueous Hg(II) using a magnetic nanocomposite Fe₃O₄/graphene oxide (Fe₃O₄/GO). Adsorption of Hg(II) onto Fe₃O₄/GO equilibrated in 4 min, with the adsorption percent and quantity of 91.17% and 547.01 mg g^?1, respectively. Fe₃O₄/GO can be easily recovered from solution via magnetic separation for reuse, and retaining 73.5% of its original capacity after five consecutive cycles. The Temkin model and PSO model were most suitable for describing adsorption in equilibrium and non-equilibrium state, respectively. Both GO and Fe₃O₄ adsorbed Hg(II) via donating electrons in oxygen atoms toward Hg(II). Moreover, GO made a major contribution, while Fe₃O₄ made a minor one to adsorption. The facile preparation, high adsorption efficiency, easy recovery, and reusability may enable Fe₃O₄/GO to be a promising adsorbent for aqueous Hg(II).