Membrane Reactors for Hydrogen Production Processes deals with technological and economic aspects of hydrogen selective membranes application in hydrogen production chemical processes. Membrane Reactors for Hydrogen Production Processes starts with an overview of membrane integration in the chemical reaction environment, formulating the thermodynamics and kinetics of membrane reactors and assessing the performance of different process architectures.
Then, the state of the art of hydrogen selective membranes, membrane manufacturing processes and the mathematical modeling of membrane reactors are discussed. A review of the most useful applications from an industrial point of view is given. These applications include:.
The final part is dedicated to the description of a pilot plant where the novel configuration was implemented at a semi-industrial scale. Plant engineers, researchers and postgraduate students will find Membrane Reactors for Hydrogen Production Processes a comprehensive guide to the state of the art of membrane reactor technology. Marcello De Falco is an academic researcher and an expert in reactor modelling. He has about 25 publications on reactor simulations and membrane technology assessment. He has 40 years academic research experience and about publications on chemical thermodynamics, kinetics and reactor modelling.
Sep Purity Methods — Wiley Interscience, New York US patent 3,, Desalination — J Memb Sci —50 J Memb Sci —27 Adv Eng Mat —15 Adv Catal — Reichelt K, Jiang X The preparation of thin films by physical vapor deposition methods. Thin Solid Films — Mattox DM Handbook of physical vapor deposition PVD processing: film formation, adhesion, surface preparation and contamination control.
RSC Publishing, London. Biswas DR Review: deposition processes for films and coatings. Mohler JB Electroplating and related processes. Chemical Publishing Co. Wise EM Palladium-recovery properties and uses.
- Membrane Reactors for Hydrogen Production Processes?
- Smart Graphics: 9th International Symposium, SG 2008, Rennes, France, August 27-29, 2008. Proceedings?
- Falco, Marcello De;
- CHAPTER 1 - Membrane Engineering for the Treatment of Gases (RSC Publishing).
- Hidden Realms, Lost Civilizations, and Beings from Other Worlds?
- Writing Compilers and Interpreters: A Software Engineering Approach, Third Edition;
Academic Press, New York Sturzenegger B, Puippe JC Electrodeposition of palladium—silver alloys from ammoniacal electrolytes. Reid HR Palladium—nickel electroplating. Effects of solution parameters on alloy properties. Platinum Met Rev —62 Loweheim FA Modern electroplating.
Wiley, New York, pp — and — Malygin AA The molecular layering nanotechnology: basis and application. J Ind Eng Chem —11 Catal Today —89 Huang Y, Dittmeyer R Preparation of thin palladium membranes on a porous support with rough surface. J Memb Sci — 52 A. Altinisik O, Dogan M, Dogu G Preparation and characterization of palladium-plated porous glass for hydrogen enrichment.
Catal Today — Liang W, Hughes R The effect of diffusion direction on the permeation rate of hydrogen in palladium composite membranes. Chem Eng J —86 Okada S, Mineshige A, Kikuchi T, Kobune M, Yazawa T Cermet-type hydrogen separation membrane obtained from fine particles of high temperature proton-conductive oxide and palladium.
J Memb Sci —18 Catal Today 93—— Sep Purif Technol — J Memb Sci —89 J Memb Sci —84 Chem Eng J —22 Cole MJ The generator of pure hydrogen for industrial applications.
Membrane Reactors for Hydrogen Production Processes | Ebook | Ellibs Ebookstore
Platinum Met Rev —13 Philpott J Hydrogen diffusion technology. Commercial applications of palladium membrane. Platinum Met Rev —16 Rev Chem Eng — Lin YM, Rei MH Process development for generating high purity hydrogen by using supported membrane reactor as steam reformer. Experimental and modeling. Appl Catal B — Appl Catal — Catal Today —81 Chem Eng J —39 Tong J, Matsumura Y Effect of catalytic activity on methane steam reforming in hydrogen-permeable membrane reactor. Appl Catal A — Application to the water gas shift reaction. Chem Lett — Catal Today 25 3—4 — J Memb Sci —23 Gas Sep Purif — Fusion Eng Des 49—— Criscuoli A, Basile A, Drioli E An analysis of the performance of membrane reactors for the water—gas shift reaction using gas feed mixtures.
Catal Today —64 Appl Thermal Eng — Haga F, Nakajima T, Yamashita K, Mishima S Effect of cristallite size on the catalysis of alumina-supported cobalt catalyst for steam reforming of ethanol. React Kinet Catal Lett — 54 A. J Catal — Energy Fuels — Fuel Cells —68 J Power Sources —17 Int J Hydrogen Energy. Iulianelli A, Basile A An experimental study on bio-ethanol steam reforming in a catalytic membrane reactor. Part I: temperature and sweep-gas flow configuration effects.
Energy Conver Manag — Lin YM, Rei MH Study on the hydrogen production from methanol steam reforming in supported palladium membrane reactor. Catal Today —84 Catal Today —72 Catal Today —22 Asia Pac J Chem Eng. The use of a Pd—Ag membrane reactor at middle reaction temperature. Int J Hydrogen Energy J Mol Catal A —48 J Anal Appl Pyrolysis — J Memb Sci —52 Iulianelli A, Longo T, Basile A CO-free hydrogen production by steam reforming of acetic acid carried out in a Pd—Ag membrane reactor: the effect of co-current and countercurrent mode. Katsir 3. H2 separation and production by membranes and membrane reactors at high temperatures and pressures was and still seems to be one of the most promising developing areas.
Although the major development effort, if dense Pd membranes are excluded due to their low permeation flux, thin palladium films supported on porous substrates have not yet reached a commercial stage due to some technical issues as long term permeance and selectivity stability, but also to the cost related to their manufacture. In this chapter, a short review of palladium membranes characteristics, hydrogen transport phenomena together with deposition techniques is provided.
Borruto and G. Lollobattista Processi Innovativi S. Katsir Acktar Ltd. Iaquaniello et al. On such a basis, a comparison of film and supports deposition technologies was illustrated to better understand how from a lab technique it is possible to move to an industrial manufacturing process and reach a volume high enough to sustain important economics of scale. The more interesting configurations for a hydrogen selective membrane are then: 1. Moreover, this materials show a relevant weldability and a low cost [1, 2]. A problem related with metallic materials used at high temperature is the intermetallic diffusion of the palladium in the metallic support [6—8].
In order to avoid this problem, it is necessary to realize a specific interdiffusion barrier  between the Pd-based selective layer and the metallic support. These materials can also be used as supports for palladium composite membranes, but their poor physical strength and, in particular, their incompatibility with conventional techniques for joining parts e.
Its unique rule is to avoid the Pd alloy interdiffusion in the steel support. One of the main objective to aim in realizing an interdiffusion barrier is to obtain a layer extremely adherent, dense, homogeneous, and with continuous thickness. TiN was considered the most promising type of barrier. Tecnimont-KT SpA, has focused on a traditional manufacturing path, realizing an innovative substrate for metallic membranes.
Two different configurations for the substrate have been considered: — metallic tubular support with titanium nitride barrier; — metallic planar support with alumina barrier. To realize the support, sintered austenitic stainless steel AISIL with a porosity [1 lm has been used. GKN technology consists in spraying the sintered stainless steel with a fine layer of nano-powder. The second sinterization step allows realizing a steel support with a surface porosity of 0. Figures 3. The GKN supports have been coated with TiN as intermetallic diffusion barrier with three different thicknesses 0.
The 2. Figure 3. It is highlighted the excellent adhesion of TiN on the low porosity support. Two optimum substrates for final membrane synthesis have been identified. The morphology of the surface on which the selective layer has to be deposited is a key issue for a proper integration. Pore size and porosity of the deposition surface influence the membrane thickness and consequently its permeability: increasing the deposition surface pore size, the thickness of the alloy required to obtain a dense and continuous layer increases reducing hydrogen flow.
Moreover, 3 Hydrogen Palladium Selective Membranes 63 increasing porosity, the effective available area for hydrogen permeation increases together with hydrogen flow . Another parameter that has to be considered is the support roughness: higher is the roughness of the deposition surface, more irregular is the deposition surface. The gas transport mechanism is the key to the high selectivity.
Hydrogen permeation through a metal membrane follows the multistep process illustrated in Fig. Hydrogen molecules from the feed gas are adsorbed on the membrane surface, where they dissociate into hydrogen atoms. Each individual hydrogen atom loses its electron to the metal lattice and diffuses through the lattice as an ion. Hydrogen atoms emerging at the permeate side of the membrane reassociate to form hydrogen molecules, then desorb, completing the permeation process.
Only hydrogen is transported through the membrane by this mechanism; all other gases are excluded . The palladium—hydrogen system is a two-component system Pd and H whose degrees of freedom are determined by the number of existing phases at working temperature and pressure. The palladium—hydrogen system Fig. In both the phases, hydrogen occupies, randomly, the interstitial octahedral sites of the f.
Pd lattice. The a phase is a low hydrogen concentration phase and it can be seen as a solution of atomic hydrogen in palladium. The a0 phase is a high hydrogen concentration phase and it can be seen as an expanded phase hydride. The a and a0 phases are separated by a region where both coexist.
Join Kobo & start eReading today
Dashed line gives a best fit 4th order polynomial to the limit of the coexistence region as determined by the authors cited. To draw the Pd—H phase diagram, we need to determine the absorption isotherms, indeed various authors, Wicke, Nernst, Frieske, Lasser, Blaurock [17—21], observed a hysteresis in the plateaus region, absent in monophasic regions, which translates into absorption process equilibrium pressure higher then desorption process equilibrium pressure with the same concentration.
In the a? This unit cell volume change can result in mechanical strains, physical distortions, and possibly failure of the palladium if cycled through the palladium hydride phase transition region. When the desorption step occurs, hydrogen is preferentially released from regions subjected to stress. So, during the desorption both the mechanical deformation that the disordered state would lack and the curve would be close if not identical to the equilibrium condition. Since the hydride is the only phase present at these 3 Hydrogen Palladium Selective Membranes 65 conditions, palladium will not be subjected to the stresses caused by the phase transition .
If the sorption and dissociation of hydrogen molecules is a rapid process, then the hydrogen atoms on the membrane surface are in equilibrium with the gas phase. Holleck et al. The lattice-diffusional mode of mass transfer for hydrogen results in the essentially infinite selectivity observed with dense palladium membranes. The hydrogen permeability of palladium increases with temperature because the endothermic activation energy for diffusion dominates the exothermic adsorption of hydrogen on palladium [25, 28].
Many disputes were created about the adsorption speed of hydrogen on the sample. Of course, as already Wagner noticed , the sample history preparation method, purity, crystalline state, cleaning of the membrane surface by oxidation or sulfur deposition from trace amounts of hydrogen sulfide dramatically affects the absorption mechanism.
Indeed, recent studies of Kay et al. This circumstance excludes that this step is the slow step of the absorption process. By thermal desorption measurements, the above-mentioned authors argued that the phenomenon speed is governed by atomic hydrogen diffusion from the surface towards the massive phase. At higher H pressure, the system no longer behaves ideally and the absorption initial speed increases with charge. This is interpreted as an effect due to lattice expansion which accompanied the absorption of hydrogen: indeed, the same effect of speed raising is achieved mechanically expanding the palladium lattice or introducing atoms of Ag Pd—Ag alloy.
The silver, in addition to increasing the solubility of hydrogen in palladium, contributes to mechanical stability and lower cost of the membrane. There are many processes today used for film deposition. However, often these are variants of two basic processes: physical process and chemical process. Only the most widely used processes for the production of composite membranes for hydrogen separation are mentioned below.
The thermal evaporation processes are classified as : a Vacuum deposition Resistive heating is most commonly used for the deposition of thin films. The source materials are evaporated by a resistively heated filament or boat, generally made of refractory metals such as W, Mo, or Ta, with or without ceramic coatings.
Crucibles of quartz, graphite, alumina, beryllia, boron-nitride, or zirconia are used with indirect heating. The refractory metals are evaporated by electron-beam deposition since simple resistive heating cannot evaporate high melting point materials. In laser deposition, a high-power pulsed laser is irradiated onto the target of source materials through a quartz window.
A quartz lens is used to increase the energy density of the laser power on the target source. Atoms that are ablated or evaporated from the surface are collected on nearby sample surfaces to form thin films . This phenomenon is called back-sputtering, or simply sputtering. When a thin foil is bombarded with energetic particles, some of the scattered atoms transmit through the foil. The phenomenon is called transmission sputtering . Several CVD processes are proposed to increase the efficiency of the chemical reaction at lower substrate temperatures . The ions in the plasma show slightly higher energy than the neutral gas molecules at room temperature.
Typically the temperature of the ions in plasma is around K. This is a relatively complex technology that involves a large number of steps . The chemical reactions are accomplished when hydrogen is released by a reducing agent, normally sodium hypophosphite, and is oxidized to produce a negative change on the surface of the substrate. This autocatalytic deposition method enables metal coating of non conductive textile material which can be used for precision work in conventional manufacturing. Unlike electroplating, the absence of electric field contributes to a uniform plating thickness.
Under properly controlled conditions all the above-mentioned methods produce good quality thin layers but electroless plating has the advantage of easy scale-up and the flexibility to coat the metal film on supports of different geometry. However, the main disadvantage is the difficulty to control the composition of the alloy. PVD sputtering has several advantages like: a b c d synthesis of ultrathin films with minimal impurity; easily controllable process parameters; flexibility for synthesizing alloys; the ability to generate nanostructured films.
The last two points are very important in membrane preparation for hydrogen separation because fabricating membrane alloys helps to overcome the problem of hydrogen embrittlement, while the nanostructured films may have unique sizedependent properties, e. Here, below a short review of some membranes providers. The technology is based on Palladium membranes which are capable of separating high purity hydrogen from a gas mixture. MRT produced membranes either as rolled foils or as deposited thin films 8— 15 lm.
In addition, patented bonding techniques have been developed to permanently attach membranes to support modules with a perfect, hydrogen-tight seal. For membranes thinner than 15 lm, MRT uses a proprietary coating technique. Prototype membranes as thin as 8 lm, tested by MRT, have been produced and show excellent performance and longevity.
Each module consists of two double sided, planar 30 cm 9 12 cm membrane panels welded in series. Each panel has a palladium Pd alloy active membrane area of 0. The modules are housed in a rectangular core which, along with the inlet distributor, promotes uniform reformate flow across the membrane modules. The core assembly Fig. The key is to simultaneously achieve cost effectiveness and high hydrogen selectivity, by making expensive palladium membranes thinner.
At laboratory scale, the JC has realized Pd—Ag membranes on ceramic substrates. The layered Pd—Ag membrane is finally heat-treated to obtain the Pd—Ag alloy membrane. The resulting membranes are tubular with an external diameter of about 1. In a second step, the film is removed from the wafer. These films may subsequently either be used self-supported or integrated with various supports of different pore size, geometry, and size Fig.
This allows, for example, the preparation of very thin approximately 2—3 lm high-flux membranes supported on macroporous substrates, which can operate at high pressures. The efficiency of these membranes has been investigated by SINTEF at elevated temperatures and pressures and reported in a remarkable number of publications on scientific journals. In pure H2, applying a H2 feed pressure of 26 bars, one of the highest H2 fluxes reported, ml cm-2 min-1 STP or 6. In water gas shift WGS conditions Operating the membrane for more than 1 year under various conditions WGS and H2?
N2 mixtures at 10 bars indicated no membrane failure. Changes in surface area are relatively small. This has been demonstrated by Bredesen and Klette  and shown in Sect. Moving ahead such a concept, the manufacturing process should consist of a batch process on three independent steps. The choice of a batch process is a logical one because it provides similar items on a repeat basis, usually in larger volume.
Batch procedure divides the manufacturing task into a series of appropriate operations, which together will make the product involved. It is then not so difficult to define the main steps in such a process: first of all, the Pd-alloy selective layer preparation, and secondly, the support preparation and finally the membrane module assembling and testing. One way to approach the membrane and support fabrication is to consider two distinct lines, each one delivering the distinctive product, which at the end of the process is tested for quality control before moving to the integration.
A roll-in process as proposed by Acktar, and detailed in Fig. Once the two single specific components have been prepared and tested, the membrane can be formed by laying the membrane on the support. Several membranes can be assembled together to build a module. Because of the metal structure of the support, the proposed composite membranes provide a solution to the problems that result from high welding temperatures or high mechanical compressing force caused by the joining of a composite membrane with other parts through Swagelok, welding, brazing, and gasket, etc.
A flat configuration of the membranes provides solutions to the problems associated with module assembling. For planar porous metal substrate, a solid frame, useful for realizing the module sealing, can be welded to its perimeter. Process choice concerns the features for hardware, the tangible ways in which the products are manufactured; but the task is more than this. The associated structures, controls, procedures, and other systems within manufacturing are equal necessary for successful, competitive manufacturing performance.
In our MMS, quality controls for instance related to the thin films but also to the support and the membrane overall, constitute an essential part in the manufacturing task. Creating a quality function in the organization to supervise such operational controls is an essential key issue to develop and manage. Fail to develop a proper infrastructure on the complex process of making membranes may result in the impossibility to reach the target costs. Graphically, the experience curve is characterized by a progressively declining gradient, which, when translated into logarithms, is linear.
The size of experience effect is measured by the proportion by which costs are reduced with subsequent doublings of aggregate production. Of course for the Pd-based or ceramic membrane such dates are limited to minimal surface less than 1 m2 , which can, however, be used as starting point of the curve. It is, however, a fact that costs decline systematically with increases in cumulative output.
Using such a data is possible to forecast the cost for m2 of membrane module versus the cumulative value of production, expressed in terms of m2. Table 3. The first and more important question to answer is when a 1,, m2 of membrane module cumulative production could be reached in order to have a unit cost around 1. In order to answer such a question, further considerations need to be developed, to relate surface to membrane module to the H2 production and to the introduction of such a new technology in the market.
With such cumulative production around year , the membrane cost per m2 could reach 76 Cumulative 2 production m Fig. The Fig. The approach used to determine the growth of the membranes market, together with the cumulative production does not, however, identify the real factors that determine its dynamics. As matter of fact, the experience curve combines four sources of costs reduction: learning, economics of scale, process innovation, and improved production design.
Economics of scale, conventionally associated with manufacturing operations, is probably the most important of these costs drivers and exists wherever as the scale of production increases unit costs fall. A plant capacity has then an economic sense if a minimum efficiency plant capacity is reached. This will imply that to reach the required reduction in the membrane cost, not only a few specialized technologies must emerge, but the production market will be concentrated in few highly specialized production plants. A precondition for such behavior is the emerging of one or two technologies which can sustain costs reduction based on economics of scale.
Physical vapor deposition technology to produce membranes on a roll-in process and a metallic support is indicated by the authors as one of the more promising technology for such industrial mass production. Preparation of membranes by electroless plating and characterization. J Membr Sci — 2. Ind Eng Chem Res — 3. Surf Coat Tech ——79 4. J Membr Sci —14 5. J Membr Sci — 6. Thin Solid Films —79 7. Catal Today 93—— 8. J Membr Sci — 9. Chin J Chern Eng 15 5 — Yan S, Maeda H, Kusakabe K, Morooka S Thin palladium membrane formed in support pores by metal-organic chemical vapor deposition method and application to hydrogen separation.
AIChE J 55 3 — Huang Y, Dittmeyer R Preparation and characterization of composite palladium membranes on sinter-metal supports with a ceramic barrier against intermetallic diffusion. Desalination —89 Baker RW Membrane technology and applications. Wiley, New York In: Manchester FD ed Phase diagrams of binary hydrogen alloy. Ber Bunsenges Phys Chem — Ber Bunsenges Phys Chem 77 1 —52 Phys Rev B — J Phys Chem Solids —37 J Less-Common Met — 78 G.
Plat Met Rev 27 4 — Ubbelohde AR Some properties of the metallic state I—metallic hydrogen and its alloys. Proc R Soc Lond A — J Membr Sci —97 Holleck GL Diffusion and solubility of hydrogen in palladium and palladium silver alloys. J Phys Chem 74 3 — Volkl J, Alefeld G Hydrogen diffusion in metals.
Academic Press, New York, pp — Zeitschrift Phys Chem A— Phys Rev B 34 2 — Surf Sci 1 — Jayaraman V, Lin YS Synthesis and hydrogen permeation properties of ultrathin palladium—silver alloy membranes. William Andrew Inc. McClanahan D, Laegreid N Production of thin films by controlled deposition of sputtered material. Topics in applied physics, vol 64, Springer Verlag, Berlin, p. Almeida E Surface treatments and coatings for metals. A general overview. Coatings: Application processes, environmental conditions during painting and drying, and new tendencies.
Ind Eng Chem Res —20 Principles, technology, and applications, 2nd edn. DOE Hydrogen separation—technical targets. Thanks to the strong development of both software and hardware, the computational loads of mathematical modeling have been increased in the years and nowadays the behavior of reactors and all chemical and physical phenomena occurring inside reaction environment can be evaluated in detail.
Practically, mathematical models are based on the conservation laws of mass, energy and momentum, which lead to mass, energy and momentum balances. The balances, together with transport and kinetics equations, form a set of equations ODE or PDE whose solution gives the component concentrations, temperature and pressure profiles inside the reactor. Mass and heat transport coefficients, reactants and products physical properties, catalyst efficiency factor and all parameters appearing in model equations have to be expressed.
In a rigorous model formulation, radial and axial mixing should be taken into account due to radial and gradients of compositions, pressure and temperature. As an example, if the elementary reaction 82 M. De Falco A! The concentration drop is due to the reaction but at the same time the gradient associated to this profile leads to a mass flux in the direction shown in the figure. This flux would reduce the gradient itself and its entity should be taken in consideration in reactor modeling.
Moreover, catalysts are typically porous and the concentration profile of reaction components inside the particles, shown in Fig. It is clear that a rigorous formulation of all phenomena occurring inside a reaction environment leads to a level of complexity difficult to be managed. The ability of reactor designers is mainly in introducing proper assumptions which reduce the formulation complexity but at the same time do not lead to unacceptable approximations. Therefore, various models types can be developed and applied in each specific case. In the following section, a classification of fixed bed reactor models is proposed and a brief description of each model type is reported.
In cylindrical symmetry, as for tubular reactors, the control volume is selected as shown in Fig. Axial mixing? Radial mixing Interfacial gradients? Intraparticle gradients? Radial mixing Accumulation term takes into account variations with time of reactor conditions.
Fixed bed reactors are systems composed by two or more physical phases since a fluid phase reacts over a solid catalyst. Mass and heat fluxes between the solid and fluid phases are expressed in terms of particle-to-fluid mass and heat transport coefficients. Fluid and solid phases are considered as a single pseudo-phase and the balances are imposed for only one phase. Heat and mass transport coefficients inside the bed are calculated by expressions which account for the simultaneous presence of two phases.
According to Froment—Bischoff  for each category, models can be classified in order of their growing complexity, as reported in Table 4. In the following, a survey of model typologies is reported starting from the simplest ones and removing assumptions gradually to formulate more and more complex models. Only steady-state conditions are taken into account. For a deeper models description, please refer to [1—3]. The pseudo-homogeneous assumption reduces the complexity of the model implementation and resolution, but of course this model category gives less accurate results with respect to heterogeneous one.
The simplest fixed bed reactor model formulation is the pseudo-homogeneous ideal model, by which: 1. Assuming the elementary reaction 4. The global heat transport coefficient U is calculated from the sum of different heat resistances in a series. Equations 4. Obviously, the ideal model is very simple and consequently solutions obtained are approximate. If a more accurate analysis of the phenomena inside the packed bed has to be performed, some of ideal model assumptions has to be released.
First of all, the axial gradients of concentrations and temperature lead to backmixing effects and diffusive contributions to mass and heat transport, which bring about axial gradients smoother than those obtained by an ideal model, have to be included.
- Membrane Reactors for Hydrogen Production Processes.
- Membrane Reactors for Hydrogen Production Processes | ytajacynemap.cf?
- A Review on Patents for Hydrogen Production Using Membrane Reactors | Bentham Science;
- Enumerability · Decidability Computability: An Introduction to the Theory of Recursive Functions.
Taken as reference the mass balance, the Eq. Pea value lies within the range 0. In this kind of models second-order differential terms appear in mass, energy, and momentum balances and two boundary conditions on the axial coordinate have to be imposed. The set of differential equations together with the boundary conditions represents a boundary value problem BVP , since the constraints are imposed at the inlet and at the outlet of the tubular reactor.
If radial variations are considered as well, the hypothesis of plug flow and consequently the one-dimensional nature of model equations falls through. Twodimensional models are usually required for energy balances in packed bed reactors, since the scarce conductivity of packed bed leads to strong temperature radial changes. Der and ker are the effective radial mass diffusivity  and thermal conductivity, respectively. The set to be solved is composed by partial differential equations PDEs , which can be handled by different approaches, as the Finite Elements Method or the Orthogonal Collocations.
If the system to be modeled verifies one of these conditions at least, the heterogeneous approach has to be followed since a pseudo-homogeneous model would lead to unacceptable errors. The simplest heterogeneous model considers only the interfacial gradients between solid and fluid phases, imposing mass and energy balances for all phases involved and introducing mass and heat transport from the gas bulk to catalyst surface and vice versa.
The assumption made is that the catalyst is not-porous and fluid components cannot diffuse through solid material. However, usually catalysts are porous materials to strongly increase the active reaction area per unit of reactor volume. Therefore, the intra-particle gradients, i. Fluid phase equations are the same as Eqs. The terms De and ke are the effective diffusivity and thermal conductivity in the solid particle: methods to calculate these coefficients are reported in the literature [1, 11]. Effectiveness factor is the ratio between the real rate of reaction observed in the presence of pore diffusion resistance and the rate of reaction that should be observed if the whole particle be at surface conditions.
Its value is within the range 0, 1 and it multiplies the rate of reaction at surface conditions to consider the reduction of reaction rate due to the diffusion resistance. Obviously, such a complex formulation is developed only if an extremely rigorous analysis is required. In fact, hydrogen permeation through selective membranes is a leading phenomenon for membrane reactors, and errors in assessing membrane behavior would lead to unreliable results. In this paragraph, the strategies to develop permeation models for Pd-based membranes and experimental protocol to evaluate parameters appearing in model equations are described.
As for the solution-diffusion mechanism, the hydrogen permeation through a palladium layer is a complex process consisting of adsorption and dissociation of hydrogen molecules, followed by diffusion of hydrogen atoms through the metal 90 M. De Falco Fig. Depending on the thickness of the membrane, step 5 diffusion of hydrogen atoms through the lattice results the controlling step in many cases. For thick membranes thickness of tens lm , the dissociative chemisorption of hydrogen on the membrane surface can be considered very fast compared with diffusion of atomic hydrogen in the membrane.
The exponent 0. From Eqs. Experimental values of n have been determined in the range 0.
An experimental apparatus for membrane permeability tests is shown in Fig. Thermocouples TI monitor the temperature. Usually, the task of experimental phase is threefold: 1. About point 1, the scope is to verify if exponent n in Eq. The following testing procedure has to be followed: 1. Then, the Arrhenius type temperature dependence of membrane permeability has to be verified as well, and at the same time the values of parameters P0 and Ea have to be assessed.
The experimental procedure is: 1. De Falco 2. The Eq. If the data arrange on a straight line, the Arrhenius dependence on the temperature is verified and the values of P0 and Ea can be derived from the slope and the intercept of the fitting straight line. Refer to Sect. At the end of the experimental procedure, a validated selective hydrogen membrane permeability is available for membrane reactors modeling. A detailed model applied to natural gas steam reforming is reported in Chap. An integrated membrane reactor IMR is usually composed by a tube, packed with catalyst pellets, inside which one or more membrane modules are assembled Fig.
The catalyst could be also packed in the inner tube, but generally it is preferred assembling the membrane in such a way that the higher pressure reaction pressure is imposed outside membrane wall. In order to study the reactor behavior, mass, energy, and momentum balances have to be imposed for both reaction zone, where catalyst is packed and reactions are promoted, and permeation zone, where an inert gas is sent to sweep the hydrogen permeated through the selective membranes. A potential alternative configuration is the application of a vacuum pump in the inner zone to reduce membrane downstream pressure, supporting hydrogen permeation without feeding a sweep gas.
In the following sections both one- and two-dimensional models are described. De Falco 4. As aforementioned, mass, energy, and momentum balances have to be imposed both in reaction and in permeation zone. Reaction rate equation and all physical properties as heats of reactions, thermal transport coefficient, density of packed bed and fluid mixture, fluid mixture viscosity, reactor and catalyst particles diameters, void fraction, etc.
JH2 is expressed according to Eq.
Membrane Reactors for Energy Applications and Basic Chemical Production
The term us is included in the derivative since in IMR the gas velocity can change both for gas volume variation due to reactions and for hydrogen flow outgoing from the reaction environment through the selective membrane. Obviously, hydrogen mass balance in reaction zone must include the hydrogen flux leaving the reaction environment thanks to selective membrane integration. In Eq. Likewise to reaction zone energy balance, the enthalpy flow leaving the reaction zone with the permeated hydrogen flux is neglected.
Momentum Balance permeation zone : Usually, pressure drops in permeation zone can be neglected. However, typical expressions are available for fluids flowing in a tube, both in laminar and turbulent conditions, and could be applied. The ODE set composed by Eqs. De Falco 1. ODE set is solved; 3. Therefore, the radial mixing effect should be always considered in IMR modeling, and 2D model are certainly more reliable and realistic. A 2D pseudo-homogeneous model for IMR is presented here.
Usually, the radial diffusive transport term is taken into account for mass and energy balances, while for momentum balance it could be neglected since the presence of catalyst pellets produces a mixing effect leading to a uniform gas velocity radial profile. Moreover, the radial diffusive terms are always much greater than the convective radial terms, which can be neglected.
For the reaction zone, mass and energy balances are Eqs. By the 2D mathematical model, axial and radial profiles both of components concentrations and of reactor operating temperature can be calculated. Therefore, an alternative configuration is analyzed in the present section. Hydrogen selective membrane is assembled in separation modules located downstream to reaction units. This configuration, called staged membrane reactor SMR and shown in Fig. Of course, it is possible to replicate the SMR until the desired reactant conversion is achieved.
SMR configuration is more deeply modeled and assessed in Chap. Here, equations and boundary conditions of a model useful to describe plant behavior are presented. The reactor model equations are those reported in Sect. The improvement of hardware and software allows reactor designers to work with more and more detailed and reliable algorithms, by which it is possible to understand behavior and performance even before fabricating the reactor itself.
In the present chapter, membrane reactors modeling strategies have been presented, and various models of tubular fixed bed reactors, Pd-based membrane separators, IMR and SMR are reported and explained. Different complexity levels can be applied, from ideal to much complex structured models accounting for radial and axial mixing, intra-particle components diffusion through catalyst particles, inter-facial gradients between solid and fluid phases.
For membrane reactors, many crucial phenomena have to be included, as membrane permeability mechanism and hydrogen flux, reactions kinetics, heat and mass transport inside the reactor and from the external to the reactor. Therefore, a proper simulation certainly requires a deep study and a careful evaluation and definition of the system. The model developers have to work in a strict connection with test drivers, since reliable model parameters and coefficients definition is crucial: designers can address reactors experimentation clarifying which information from test-benches are required.
At the same time, a proper model development allows the number of experimental tests to be reduced drastically to those ones required for a complete reactor validation. By this chapter, the author does not pretend to run out a so wide topic as reactor modeling and simulation but he wants only to give to readers the basis for a further and deeper study. Wiley, New York 2. Fogler HS Elements of chemical reactor engineering, 3rd edn.
De Falco 3. In: Simulation of membrane reactors. Nova Science Publishers Inc. Chem Eng Sci — 5. Chem Eng Sci — 6. Li C, Finlayson B Heat transfer in packed beds—a reevaluation. Chem Eng Sci — 7. Yagi S, Wakao S Heat and mass transfer from wall to fluid in packed beds. AIChE J 5 1 —85 8. Yagi S, Kunii D Studies on effective thermal conductivities in packed beds.
AIChE J 3 3 — 9. Int J Heat Mass Transfer 27 10 — Cat Rev Sci Eng 22 3 — Satterfield CN Mass transfer in heterogeneous catalysis. Press, Cambridge Lewis FA The palladium hydrogen system. Academic Press, London 94 The deteriorating quality of crude oils, the more stringent petroleum product specifications, and environmental problems are leading to larger needs of hydrogen in hydroprocessing.
Moreover, there is an increasing interest about hydrogen as energy carrier and as feedstock for fuel cells, which further stimulates hydrogen industry. Hydrogen can be obtained from different sources as fossil fuels natural gas reforming, and coal gasification , renewable fuels biomass , algae, and vegetables or water electrolysis and thermo-chemical cycles. Many different energy sources can be used in most of these processes: heat from fossil fuel or nuclear reactors, electricity from several sources as solar energy. Figure 5. Nowadays, NG steam reforming appears to be the only process able to produce large amounts of hydrogen at a competitive cost and to promote hydrogen technologies in the next years.
However, some uncertainty comes from the volatility of natural gas price spot prices can double or triple in a short period of time that affects the total cost of hydrogen production for a half. However, usually heavier hydrocarbon ethane, propane, etc. The second reaction is known as water—gas shift reaction WGS , an exothermic reaction favored at low temperature and not affected by the operating pressure.
Globally the steam reforming process should be supported by high temperature in the steam reformer the task is the conversion of the methane, the WGS is supported by one or two steps water—gas shift reactor which follow the steam reformer, as described in the next paragraph and by low pressure. However, usually the reactions are conducted at 25—40 bar in industrial plants to reduce the total volume of the devices and favor the heat transfer Fig.
In industry, the ratio is usually within the range 2. The effect of the temperature, pressure, and steam to carbon ratio on the thermodynamic equilibrium for the reactions 5. Two alternatives are available: 1. In tubular fired reforming, process heat duty is supplied by burning a share of NG feedstock in burners placed at top, down, or sideways of a furnace. In autothermal reformers, a part of NG feedstock is directly burned in the first section of the reactor by adding an oxygen or enriched air stream and heat duty is supplied without any external source.
The autothermal process, described in Chap. Natural gas is mixed with appropriate amounts of steam and recycled hydrogen before entering the reformer reactor. The recycle of an amount of H2 produced is necessary to keep the catalyst in the early part of the reformer tubes in the reduced active state. Reactions 5. Then, the H2 is separated from the off-gases CH4, CO, CO2 in a pressure swing adsorption PSA section, which involves the adsorption of impurities onto a fixed bed of adsorbents at high pressure.
The impurities are subsequently desorbed at low pressure into an off-gas stream. This operation allows an extremely pure hydrogen to be reached: H2 purities over Off-gases, mixed with additional fuel, are sent to the burners to supply a part of the process heat duty. If CO2 sequestration is required for environmental reasons, a Methyl-DiEthanol-Amine MDEA unit has to be inserted to absorb the carbon dioxide and to separate it from the other off-gas mixture components. Considering that in a traditional industrial plant hundreds of parallel tubes are installed, the total heat duty is some tens of MWth.
Moreover, the high NG fuel consumption is the main reason of the strong dependence of the final H2 cost on the NG market price Fig. The high operating temperature forces to use expensive high alloy steels whereas the large thermal gradient in axial direction, together with the pressure 25—30 bar in the tube, produce hard stresses on materials. Reducing the pellets size should improve the effectiveness factor but particle dimension cannot be too small, M. De Falco since an excessive packing of the catalyst reduces the bed void fraction and increases pressure drops, which are usually important 2—3 barg.
As it is widely reported in literature, the random distribution of the pellets in the packed bed is characterized by a high void fraction in the zone near the tube wall, moreover which is a rigid boundary inhibiting lateral gas mixing. Consequently, in the zone near the hot tube wall the molecular conduction is the only heat transfer mechanism, causing a strong temperature drop immediately in the first layer of the gas-catalyst phase.
Then, the heat flux encounters the high heat transport resistances of the gas mixture and of the catalytic pellets. Catalyst particles placed in the central zone of industrial large reformers usually do not work well since the temperature is too low for promoting the endothermic reactions. The very high temperature reached in the traditional steam reforming reactor changes the direction of the exothermic water—gas shift. Therefore, in the traditional steam reforming industrial plants, the reformer is followed by a water—gas shift reactor to reduce the carbon monoxide contents in the outlet stream.
The selective membrane integration allows a reaction product to be removed and equilibrium conditions are never reached.
As a result, a specific feedstock conversion can be obtained at a lower operating temperature with the main benefits of a better heat integration, as the use of gas exhausts from a gas turbine as suggested by  or solar heated molten salts [4, 5]. By membrane integration, the design criteria of steam reforming plant have to be completely re-thought, carefully facing the membrane integration drawbacks, mainly coupling catalyst and membrane operating conditions and meeting a compromise optimization to promote both kinetics and permeability, without damaging the membrane which always requests a stringent thermal threshold.
Two plant configurations can be used: 1. MR is then modeled and simulated to assess the effects of operating conditions on reactor behavior, while for a description of a real RMM installation refer to Chap. The simplest configuration is composed by two concentric tubes, where catalyst pellets are packed in the annular zone while the inner tube is the membrane itself, as shown in Fig. Through the inner tube a sweeping gas water steam is sent, co-currently or counter-currently, to drag the hydrogen permeated.
Of course, the membrane integration can also be made by assembling many smaller tubes, thus increasing the specific membrane surface per unit volume of reactor and consequently the overall hydrogen flow permeating. Two different zones can be recognized: the reaction zone, which is the annular section where catalyst is packed, and the permeation zone, where the sweeping gas is sent.
Moreover, the MR configuration avoids packing catalyst in the central zone of the tubular reactor, thus improving the average catalyst effectiveness. Furthermore, the lower operating thermal level required in MR leads to new procedures of process heat duty supply, avoiding the furnace and allowing hot fluids from other processes or solar thermal fluids [4, 5] to be used. Membranes that can be used in MRs are: 1. In modeling and simulating MR, Pd-based membranes are selected in this chapter to exploit their very high selectivity and because their development is at a pre-commercial status.
For typical Pd-based selective membrane performance, Fig. De Falco refer Table 2. In Sect. The feed is sent to a convective steam reformer where it is partially converted into hydrogen; then hydrogen is recovered through a Pd alloy membrane separation module, while the retentate is sent to the next step or recycled to the first module. It is possible to replicate the RMM until the desired natural gas conversion is achieved. Reforming and separation module temperatures can be optimized independently, both increasing methane conversion for each reaction step and membrane stability and durability.
On the other hand, the main drawbacks in respect to MR configuration can be summarized as follows: — the lower compactness of the plant: RMM configuration is composed by reactors, separation modules and heat exchangers between them, while in MR reaction and separation are performed in a single compact device. This leads to a greater cost. Performance of such an innovative architecture is evaluated in a real pre-industrial application. Chemical—physical properties of reactions and components involved are taken from literature [17, 18] as functions of temperature, pressure, and mixture composition.
The kinetic equations for the reactions scheme composed by 5. The model is implemented in MatLab environment. The validation is made on the basis of industrial data sets  of a traditional steam reforming plant. Then, the dependence of membrane permeability on the temperature, given by Shu  and valid for a 20 lm Pd—Ag membrane, has been assumed and hydrogen down and upstream partial pressures see Eq.
The configuration depicted in Fig. The following simulations are performed fixing the MR geometry and the permeation zone operating conditions and varying: 1. The simulations are performed for a single MR reactor. The benefits of integrating a selective membrane inside the reaction environment are self-evident: in the range of explored temperatures, the improvement of Fig. On the other hand, increasing wall temperature leads to an increase of maximum membrane temperature as well.
For this reason, the wall temperature, and consequently the heat flux supplied, has to be limited to avoid the overtaking of Pd-based membrane temperature threshold, fixed at K with the present technology. Therefore, the wall temperature has to be maintained at values lower than K. The residence time is calculated as the ratio between the catalyst packed volume and the inlet volume flow rate of reactants. The reason is that at high GHSV the residence time is too low compared to the characteristic time of permeation kinetics, and the hydrogen recovery percentage, i.
Pure hydrogen produced has not a monotonic trend, since GHSV has a double effect: 1. Globally, if the GHSV value corresponding to the maximum pure hydrogen recovery is taken as optimal, the benefits of MR compared to TR are not enough to justify membrane integration methane conversion improvement of Therefore, the optimal value can be taken equal to 6,—6, h On the other hand, steam-to-carbon ratio has a double effect on permeation behavior: 1.
Globally, at low pressure the negative effect is prevailing, while at higher pressure the two effects are balanced Fig. Also the pressure has a double effect on MR behavior: 1. Altogether, it is a worth assessment that high pressure enhances the advantage of MR over TR, since the selective membrane works better. A pressure of 10 bars and a steam-to-carbon ratio of 4 seem to be a good compromise for reaction and permeation performance.