CFD investigation of gas-solid flow dynamics in monolithic micro
Transcription
CFD investigation of gas-solid flow dynamics in monolithic micro
YININGWANG CFD INVESTIGATION OF GAS-SOLID FLOW DYNAMICS IN MONOLITHIC MICRO- CIRCULATING FLUIDIZED BED REACTORS Mémoire présenté à la Faculté des études supérieures de l'Université Laval dans le cadre du programme de maîtrise en génie chimique pour l'obtention du grade de maître ès sciences (M. Sc.) DÉPARTEMENT DE GÉNIE CHIMIQUE FACULTÉ DES SCIENCES ET DE GÉNIE UNIVERSITÉ LA V AL QUÉBEC 2008 © Yining WANG, 2008 RÉSUMÉ La biomasse est une des sources importantes d'énergie primaire et renouvelable. Le développement d'un procédé basé sur la conversion de celle-ci en énergie tout en demeurant respectueux de l'environnement, fait l'objet de recherches intenses aussi bien dans les mondes académique qu'industriel. La gazéification pour produire un gaz de biosynthèse est considérée comme une des options les plus prometteuses via la valorisation des sources de résidus de biomasse. La thermodynamique et la cinétique intrinsèque imposent que les réactions de gazéification de la biomasse doivent être effectuées à des températures élevées, exigeant la fourniture et la récupération de chaleur de manière efficace. Le concept de gazéification allotherme (par opposition à son pendant autotherme) offre une solution attrayante pour la mise en œuvre à haute température du couplage de réactions fortement endothermique avec des réactions exothermiques. Toutefois, la mise en œuvre pratique du concept sous haute température n'est pas aisée. Dans ce travail, un nouveau concept pour la gazéification de résidus de la biomasse est proposé impliquant l'hybridation de réactions à hautes températures de la gazéification et de la combustion dans un réacteur monolithique structuré. Clairement, le design et l'optimisation de ce nouveau procédé hybride requiert la compréhension précise, non seulement des phénomènes physico-chimiques de la conversion thermochimique de la biomasse, mais aussi du comportement hydrodynamique, complexe, des deux phases mises en œuvre dans un microréacteur monolithique à lit fluidisé. À cet égard, la caractéristique hydrodynamique de la distribution des écoulements des phases gaz-solide au sein du réacteur revêt une importance cruciale pour la prédiction du comportement des processus de gazéification/combustion et pour l'examen de stratégies d'opération du procédé. En particulier, en raison de la nature complexe de l'interaction entre le gaz et les particules solides ainsi que la phase stationnaire représentée par le microréacteur monolithique, un des défis dans le design et l'opération de ces réacteurs est la prévention de la maldistribution des phases. Dans ce travail, la mécanique des fluides numériques (MFN) est mise à profit comme outil de simulation permettant d'explorer les distributions des écoulements gaz-solide dans un réacteur monolithique. L'ensemble des sections structurée 111 (le monolithe) et les parties terminales non-structurées (lits fixes aléatoires permettant l'alimentation et l'évacuation de la suspension gaz-solide) est globalement considéré dans la simulation afin de capturer les tendances lourdes des mécanismes contribuant à la dynamique gaz-solide. Les résultats des simulations ont démontré la capacité de la MFN à capturer la caractéristique de non-uniformité de l'écoulement dans ce type de géométrie. iv ABSTRACT Biomass is one of the important pnmary and renewable energy sources. The development of a biomass-based but energy-efficient and environment-friendly system is seen to be very seductive. Gasification to produce biosyngas is regarded as one of the most promising options for utilizing biomass sources. Therrnodynamics and intrinsic kinetics dictate that endotherrnic biomass gasification reactions have to be carried out at high temperatures, which demands efficient heat supply and recovery policy. The concept of allothermal gasification offers an attractive solution for implementing high-temperature reactions by coupling strongly endothermic reactions with exotherrnic reactions. However, implementation of the concept under high-temperature conditions .in practice is not straightforward. In this work, an innovative process concept for biomass gasification is proposed, which involves the hybridization of high-temperature gasificationlcombustion reactions in a monolithic structured reactor. Evidently, the design and optimization of this novel hybrid process requires accurate understanding of not only the physicochemical phenomena of biomass thermochemical c0I?-version but also the two-phase hydrodynamics behaviour in the monolithic micro-fIuidized reactor which are highly complex in nature. In this regard, the fIow distribution characteristic of the gas-solid two-phase hydrodynamics in monolithic structured reactor is significantly important for prediction of gasification/combustion performance and examination of strategies for process operation. Especially, due to the complex nature of the interaction between gas and particulate phases and the stationary monolith backbone, one of the challenges in the design and operation of the monolith reactors is the prevention of fIow maldistribution. In this work, computational fIuid dynamics (CFD) is used as a tool to investigate the gas-solids two-phase fIow distribution in a monolithic structured reactor. The assemblage of monolithic structured packings with through-fIow gas-particulate fIows is globally considered in the simulation to capture the dominant possible mechanisms contributing to the final overall gas and granular dynamics. The simulation results demonstrated the ability of our CFD simulation to capture the nonuniform fIow characteristics in monolithic structured packings. v FOREWORD There are four chapters in this thesis. Among them, chapter 3 is composed of a research article which was submitted to the scientific journal lndustrial & Engineering Chemistry Research at the time of this thesis deposit for evaluation (August 2008). This research article is entitled: Yi-Ning Wang, Faïçal Larachi, Shantanu Roy. Simulating the Dynamics of Gas-Solid Flows in a Multichannel Micro-Circulating Fluidized Bed, lndustrial & Engineering Chemistry Research, 2008 (Accepted). From its integrity viewpoint, this chapter consists of the research article. Nonetheless, the figures and tables were displaced from the end of the research article to where the y are mentioned in the text. The size of the figures and tables as weIl as the size of the characters were also adjusted to fit the requirement of the thesis writing. The research article was prepared on my own and revised by my research supervisor, Prof. Faïçal Larachi and my research co-supervisor, Prof. Shantanu Roy, who were included in this article as co-authors. VI ACKOWLEDGEMENTS First of aU, 1 would like to express my sincere gratitude and appteciation to my research supervisor, Prof. Faïçal Larachi, for granting me the opportunity and resources to study the Master pro gram and offering his invaluable thoughtful insights and unique source of knowledge throughout this research project. 1 would like to sincerely thank my research co-supervisor, Prof. Shantanu Royat the Department of Chemical Engineering in Indian Institute of Technology (lIT) for his helpful discussions and suggestions as weIl as his consistent support. His experience and professional attitude inspired me throughout this work. 1 would like to express my appreciation to the technical and administrative staff at the Department of Chemical Engineering in Laval University for their continuous assistance and cooperation all along the Master program. 1 would like to thank and convey my gratitude to the graduate students, postdoctorates and colleagues in our research group (Soumaine, Cedric, Simon, Florin, Bora, Mugurel, Mohsen, David, Olivier, Insaf, Samira, Lyes, Aziz, Pouya and Elahe) with whom 1 had the pleasure to share great moments during the past years. 1 am also taking the occasion to specially thank Mf. Mohsen for his friendly help in the course of this project. FinaUy, 1 am deeply grateful to my family members for their eternal and implicit support during my study. Most important of aU, 1 would like to thank my wife, Ying SUN, for her endless love and encouragement. My special thanks go to my lovely daughter, YaXuan WANG, who always makes my coming home in the evening a joyful event with her smiling face and loving hug. VIl LIST OF TABLES Chapter 2 Table 2.1 List of heterogeneous and homogeneous reactions involved in biomass gasification ................................................................................................................... 12 Table 2.2 Summary of important investigations of the gasification of biomass in fluidized beds .............................................................................................................................. 13 Table 2.3 Summary of recent important attempts at reactor modeling of biomass gasification ................................................................ .................................................. 16 Table 2.4 Candidates of PCM for high temperature application (Maruoka et al.,2002) ..... 24 Table 2.5 Comparisons of numerical schemes for modeling phase change phenomena .... 26 Table 2.6 Recent attempts at CFD modeling of circulating fluidized bed reactor performances ............... .................................................................................... :........... 32 Chapter 3 Table 3.1 Basic simulation conditions used in this work .................................................... 58 Table 3.2 Effect of particle size and radial porosity distribution of nonstructured packings on the flow characteristics in monolith .................................................... .. ... .............. 69 VIn LIST OF FIGURES . Chapter 2 Figure 2.1 Paths for the conversion of raw materials to final products (via syngas production step) ....................................................................................................... .. .... 5 Figure 2.2 Three routes to syngas .......................................................................................... 7 Figure 2.3 Proposed novel process concept .......................................................................... 9 Figure 2.4 Van Krevelen diagram for various solid fuels (Prins et al.,2007) ...................... Il Figure 2.5 General reaction mechanism for the gasification of a biomass fuel (Higman and van der Burgt,2003) ..................................................................................................... 12 Figure 2.6 Schematic representation of the monolith reactor (Tomasic, 2007) .................. 18 Figure 2.7 Vertical distribution of solid in different contacting regimes (Kunii & Levenspiel, 1997) ....................................."............................................................. "....... 29 Chapter 3 Figure 3.1 Proposed process concept .................................................................................. 49 Figure 3.2 Radial variation of bed porosity in packed-bed sections ................................... 52 Figure 3.3 Two-dimensional computation al geometry with the assemblage of three-section structured/non-structured packings (yellow line, 2D symmetric plane) ..................... 57 Figure 3.4 Solids biomass flux of suspended phase in different packing sections .............. 60 Figure 3.5 Gas mass fluxes mirroring Figure 4 simulations ............................................... 60 Figure 3.6 Channel dependence of gas-phase velocity, solid velocity, and solid holdup (z= 0.4m) ...................... ~ ..................................................................................................... 61 Figure 3.7 The gas-phase velocity in single-phase flow simulation ................................... 63 Figure 3.8 Comparison of gas-phase velocities under single-phase/two-phase simulation conditions (z=0.4m) with and without the nonstructured packings ............................ 64 Figure 3.9 Comparison of monolith-section flow distribution characteristics (z=0.4m) with and without the nonstructured packing in the downstream section ............................. 65 IX Figure 3.10 Details of the channel locations and centerline-based pressure sampling in the three-section monolith system ..................................................................................... 67 Figure 3.11 Effect of particle size and porosity radial distribution on the solid mass flux distribution in the composite monolith system ............................................................ 70 x TABLE OF CONTENTS RÉSUMÉ .............................................................................................................................. iii ABSTRACT ................ .......................................................................................................... v FOREWORD ........................................................................................................................ vi ACKOWLEDGEMENTS ................................................................................................... vii LIST OF TABLES ............................................................................................................. viii LIST OF FIGURES .............................................................................................................. ix TABLE OF CONTENTS ..................................................................................................... xi Chapter 1 General Introduction ............................................................................................ 1 1.1 Research Background & Problem Statement .............................................................. 1 1.2 Research Objectives and Scope of the Thesis ............................................................. 2 Chapter 2 Literature Review ................................................................................................ 4 2.1 Introduction ................................................................................................................. 4 2.2 Hybridization of Gasification/Combustion Processes: A Novel Process Concept ..... 6 2.3 Physicochemical Processes in Biomass Gasification and Modelling ........................ Il 2.3.1 Physicochemical processes in biomass gasification ......................................... Il 2.3.2 Modeling of biomass gasification process ....................................................... 14 2.4 Monolithic Structured Reactor and Modelling Methodology ................................... 17 2.4.1 Monolithic structured reactors .......................................................................... 17 2.4.2 Modeling of monolithic structured reactors ..................................................... 19 \ 2.5 High-Temperature Phase-Change Material and Modelling Approaches .................. 23 2.5.} High-temperature phase-change material ......................................................... 23 2.5.2 Modeling of solidification and melting processes in phase-change-material .. 24 2.6 Gas-Solid Fluidization in Micro-Fluidized Bed Reactors and Modeling Methodology ......................................................................................................................................... 27 2.6.1 Monolithic micro-fluidized bed reactors and gas-solid fluidization ................ 27 Xl 2.6.2 Modeling of circulating fluidized bed reactors ................................................ 29 2.7 Summary and Conclu ding Remarks .......................................................................... 32 References .......................... ............................................................................................. 34 Chapter 3 Simulating the Dynamics of Gas-Solid Flows in .a Multichannel Micro- Circulating Fluidized Bed ................................................................................................ 46 Abstract ................................................................. '........................................................... 46 3.1 Introduction ............................................................................................................... 47 3.2 Hybridization of Gasification/Combustion Processes in Monolithic Structured Reactors ............. ,............................................................................................................. 48 3.3 Representation of Nonuniform Porosity Distribution for Packed-bed Sections ....... 50 3.4 Eulerian-Eulerian Multifluid Model for Gas-Solid Flow in Monolithic Structured Reactor ............................................................................................................................. 52 3.4.1 Continuity and momentum conservation equations ......................................... 52 3.4.1.1 Mass conservation equations of gas and particulate phases .. .................... 52 3.4.1.2 Momentum conservation equation of gas and particulate phases .............. 53 3 .4.2 Kinetic theory of granular flow equations ........................................................ 53 3.4.3 Closure relationships for interphase interactions ............................................. 54 3.4.4 Definition of maldistrioution quantities ........................................................... 56 3.5 Computational Geometry, Boundary Conditions and Numerical Solution ............... 56 3.6 Results and Discussion .............................................................................................. 58 3.6.1 Modeling of two-phase flow behavior in monolith structured packings .......... 59 3.6.2 Comparison of gas-solid two-phase flow with single-phase flow .................... 62 3.6.3 Effect of downstream-section packing mode on flow distribution in monolith64 3.6.4 Effect of particle size of nonstructured packings on flow characteristics in monolith ..................................................................................................................... 66 3.7 Conclusions ................................................................................................................ 71 Acknowledgement ....... ~ ................................................................................................... 72 Nomenclature .................................................................................................................. 72 Literature Cited ................................................................................................................ 75 XlI Chapter 4 Conclusions and Recommendations .................................................................. 79 4.1 General conclusions ........................................................ ..... ...................................... 79 4.2 Recommendations for future investigations .............................................................. 81 X1l1 Chapter 1 General Introduction 1.1 Research Background & Problem Statement Biomass is one of the important primary and renewable energy sources. With the depletion of fossil fuel sources as weIl as the evolving global warming issues, the need for utilization of biomass for energy is seen to be imperative, particularly because it is believed that energy obtained from biomass has a carbon-neutral cycle. This situation calls for the development of a biomass-based but energy efficient and environment friendly system with better environmental acceptability and economic viability. Gasification to produce biosyngas is regarded as one of the most promising options for utilizing biomass. However, due to the thermodynamic and kinetic limitations, endothermic biomass gasification reactions have to be carried out at high temperatures, which demands an efficient heat supply and heat recovery. The concept of allothermal gasification offers an attractive solution for implementing high-temperature reactions by coupling strongly endothermic reaction with exothermic reactions. However, implementing the concept in practice is not straightforward. > Steam gasification of solid carbonaceous fuels is highly endothermic, which demands the input of additional heat source to drive the reactor system. This is a challenge because the input of energy reduces the maximum efficiency of the process. A further challenge is the provision of the additional heat without compromising the quality of the products. Methods to meet this energy shortfall involve: (i) the combustion of a fraction of the biomass fuel or unconverted biomass residue to generate heat; (ii) the use of a fraction of the combustible product gases to generate energy. In convention al gasifiers, the energy required for heating the reactants and for the heat of reaction is supplied by burning a significant portion of the feedstock, either directly by internaI combustion or indirectly by external combustion. InternaI combustion, as applied in autothermal reactors, results in the contamination of the gaseous products, while external combustion, as applied in allothermal reactors, results in lower thermal efficiency because of the irreversibilities associated with indirect heat transfer. As far as biomass gasification conversion process is concemed, there are a number of potential problems which could be encountered in the energy management of the process: (i) If biomass is reacted with both air and steam in one reactor, then nitrogen is present in the product stream and is costly to remove; (ii) If trying to avoid this proble~ by using oxygen instead of air, then a source of pure oxygen would be needed, which is also a costly proposition; (iii) It is possible to circumvent the separation issues by running the oxygenless gasification and the combustion reactions in different locations (spatial segregation), in which transferring heat from one location to the other would be accompanied with heat losses; (iv) AIso, in aIl of these schemes, if the product gas is rapidly cooled, th en tar forms, which is also afflicting process stability and efficiency. To avoid this, the product gases must be kept hot for a while to let the tars crack into lower molecular weight compounds. tn view of the aforementioned problems, an innovative process concept which involves the hybridization of biomass gasification/combustion reactions in a monolithic . structured reactor is proposed in this work to address sorne of the above potential "showstoppers". In this novel process, the monolith is operated periodically between an endothermic gasification step and an exothermic combustion step. High-temperature phase-change-materials are used to intensify the process heat management. The heat released during combustion is stored using a high-temperature phase-change material (like LiF-CaF2), which is expected to discharge heat during the endothermic gasification step. The biomass is supplied to the monolith by fine granulation and subsequent pneumatic conveying, essentially creating monolithic micro-circulating fluidized beds. Hence, the process intensification is achieved both by temporal segregation of gasification and combustion as weIl as the use of a monolithic micro-fluidized bed reactor. 1.2 Research Objectives and Scope of the Thesis To effectively design and optimize this novel process, knowledge from different important fields (including biomass gasification, mon?lith reactor engineering, high2 temperature phase change material, and gas-solids fluidization) is required. Among them, modeling and understanding of gas-solid (biomass particles) flow hydrodynamics in monolithic structured reactor is very important, in view of the complexity of two-phase flow in structured packings. In this regard, the flow distribution characteristic of the gassolid two-phase hydrodynamics in monolithic structured reactor is significantly essential for prediction of gasification/combustion performance and examination of strategies for process operation. Especially, due to the complex nature of the interaction between gas and particulate phases and the stationary monolith backbone, one of the challenges in the design and operation of the monolith reactors is the prevention of flow maldistribution. In this thesis, following the proposaI of this novel process concept as weIl as the review of the relevant literature, the research focus is oriented to the CFD investigation of gas-solids (biomass particles) two-phase flow dynamics in monolithic multichannel micro-circulating fluidized bed. The computational fluid dynamics approach is used as a tool to investigate the gas-solids two-phase flow distribution in a monolithic structured reactor. A 2-D Euler-Euler multiphase model with the kinetic theory of granular flow has been solved for the detailed monolithic packing geometry. The assemblage of structured monolithic section with non-structured packed-bed sections is globally considered in the simulation, allowing comprehensive capture of various possible mechanisms contributing to the final overall aero/granular dynamics. 3 Chapter 2 Literature Review 2.1 Introduction Sustainable development requires sustainable energy resources. It is now widely acknowledged that combustion of fossil fuels contributes to the buildup of CO 2 in the atmosphere, which in turn contributes to the greenhouse effect, gradually warming the planet. Biomass is considered to be one of the most promising alternatives to replace fossi! fuels (Negro et al. , 2008). As a diverse energy carrier with a multitude of potential sources, biomass is the most important fuel worldwide following coal, oil and natural gas. Furthermore, it is considered to be a carbon-neutral and renewable energy source, offering substantial advantages for environmental protection and much shorter C02circuits compared to fossil fuels. Therefore, biomass has a considerable potential for future energy supply and to dramatically improve our environment, economy and energy security. In view of its remarkable contribution to the reduction of CO2 emission, the development of innovative utilization technologies of biomass has become increasingly important (Kobayashi et al., 2008; Florin and Harris, 2008). Biofuels are expected to become increasingly important in the future to reduce CO2 emissions, improve local emissions, and obtain security of supply. Much research and developemnt efforts worldwide focus on ways to produce so-called second generation biofuels, that are characterised by excellent environmental performance as weIl as high biomass feedstock flexibility. Making syngas (composed primarily of carbon monoxide and hydrogen) from biomass is a crucial step in the production of most second gener'ation biofuels (van der Drift and Boerrigter, 2006). The convention al way to convert biomass for energy production is direct combustion: Biomass can be combusted in grate firing systems, in fluidised bed combustion chambers, or even in pulverised co-combustion systems. However, the direct combustion of biomass raises certain issues su ch as high temperature chlorine corrosion, low-melting temperature of biomass ash (especially of straw) and the agglomeration in fluidised bed combustion chambers (Karallas et al., 2008). The gasification of biomass is generally considered to be one of the most 4 promising technologies to convert biomass into useful products. The gasification process can convert the carbonaceous materials into synthesis gas, and typical raw materials used in gasification include biomass, coal, petroleum-based materials (crude oil, petroleum coke, and other refinery residuals) and municipal solid waste (MSW). Energy source Conversion technology. Syngas production .. :a .. ~ .. :t . . . . .. . . . ~ ••• :'t :f( $ ~ la ... .~ a". 1l4 * .* .t ~ . ....... .. " ..... ~ ... .. .. ;,. ... .... 'il '"t ;~ ~ ~ ;J :+ ~ 'K: .t; :0 ....... " Products . " . .. . . . . . . . . . Of .. ., .. " .x ,." :~ .x. .x .;.; ". ~ .~ .x '* .. .;. .:.1 ~ ".A.. .......... .... .. ... .. ........ :0 .......... "<$ ":0; » ~ .li! 6: • :~ Je '* ~ . : :V • ..;r ~ Figure 2. 1 Paths for th e conversion of raw materials to final produ cts (via syngas producti o n step) . The syngas from biomass can be further upgraded into methanol, dimethylether, Fischer-Tropsch liquid fuels or other chemical products, as shown in Figure 2.1. The advantage of gasification is that using the syngas is more efficient than direct combustion of the original fuel and more of the energy contained in the fuel is extracted. Due to the thermodynamic and kinetic limitations, endothermic reactions like gasification of solid carbonaceous materials have to be carried out at high temperatures, which asks for an efficient heat suppl y and heat recovery. Multifunctional reactor concept offers an attractive solution for implementing high-temperature reactions by coupling 5 strong endothermic reaction with exothermic reaction, which has been a subject of vital research and development (Agar, 1999; Kolios et al., 2000; Kolios et al., 2002; Ramaswamy et al. ,2006). Gasification technologies are divided into autothenhal and allothermal ones. In the autothermal gasification, the partial combustion of biomass provides the required heat for the gasification. In the allothermal gasification process, the necessary heat is usually provided from an external source (Karallas et al.,2008). The key challenge of the allothermal gasification is the need to transfer the heat-of-reaction for the endothermic gasification reactions from an external heat source into the gasifier. In this chapter, we will first propose an innovative biomass gasification process concept in which the allothermal coupling of gasification/combustion processes with high-temperature phase change material will be implemented in a monolithic .structured reactor and intensified by periodic operation mode. Then, the literature work on the relevant aspects is reviewed, which includes biomass gasification, monolithic reactor, high-temperature phase change material, and fluidized bed reactor. Finally, a concluding remark is made. 2.2 Hybridization of GasificationlCombustion Processes: A Novel Process Concept Steam gasification of solid carbonaceous fuels is highly endothermic, which demands the input of additional heat source to drive the reactor system. This is a challenge because the input of energy reduces the maximum efficiency of the process. A further challenge is the provision of the addition al heat without compromising the quality of the products (Frolin et al., 2008). Methods to meet this energy shortfall involye: (i) the combustion of a fraction of the biomass fuel or unconverted biomass residue to generate heat; (ii) the use of a fraction of the combustible product gases to generate energy (Lutz et al., 2003). In conventional gasifiers, the energy required for heating the reactants and for the heat of reaction is supplied by burning a significant portion of the feedstock, either directly by internaI combustion or indirectly by external combustion. InternaI combustion, as applied in autothermal reactors, results in the contamination of the gaseous products, while external combustion, as applied in allothermal reactors, results in lower thermal efficiency because of the irreversibilities associated with indirect heat transfer. 6 For practical implementation of the gasification in converting solid carbonaceous materials (like biomass), there are a number of potential problems which could be encountered in view of the energy management and product control (Levenspiel, 2005): (i) If biomass is reacted with both air and steam in one reactor, then nitrogen is present in the product stream and is costly to remove; (ii) If trying to avoid this problem by using oxygen instead of air, then a source of pure oxygen would be needed, again costly; (iii) It is possible to avoid the nitrogen separation problem by running the two reactions in different locations, but then transferring heat from one location to the other will be a problem; (iv) AIso, in aIl of these schemes, if the product gas is rapidly cooled, then tar forms, and this is also costly to remove. To avoid this, the product gases must be kept hot for a while to let the tars crack into lower molecular weight compounds. In a recent paper, Levenspiel (2005) has suggested coal as a replacement for petroleum; and three distinct coal-to-syngas routes are identified for producing syngas steanl steam CO +H! +waste waste waste +coaJ +coal tife ~ gasifier COfIlbusto.f Ro ute 1 (sfmuftane ou.sJy coupfed) CO+Hz a ir a ir+steam +coal +coal Route 3 Route 2 {chrono fogi caUy s e·gœg:a ted) (spatia Hy.segŒga ted) ----------------~ -v? ~---------------- Fi gure 2.2 Three routes to syngas 7 from the solid carbonaceous fuel (coal), which are schematically illustrated in Figure 2.2. ln route l, both gasification and combustion reactions are gasification and combustion reactions are simultaneously coupled in one reactor, and then separate the wanted from the unwanted products. The co st of the units needed to separate the waste gases, especially nitrogen, from syngas is high. In route II, the gasification and combustion reactions are spatially segregated using two different reactors which require the transfer heat from one to the other. In route III, the gasification and combustion reactions taking place in one single reactor are chronologically segregated and the whole process operation is of cyclic nature. In a combustion step, only air (not oxygen) is used. Hence, there is no need for nitrogen removal either before or after the gasification step. In addition, there is no . need for an oxygen separation plant. Furthermore, because fresh syngas has to pass through the hot bed, tar formed at the heat front will hopefully be destroyed. However, up to now the giant corporations aIl takes the same route l, leaving the other alternative routes (route II and route III) untouched. In his paper, the author (Levenspiel, .2005) highlighted the necessity and importance of exploring these two alternatives. However, there is no mention of biomass materials in this paper. As a very promising and competitive option, the importance of syngas production from biomass through gasification has been widely recognized by scientific community (Wang et al. , 2008; Panigrapi et al., 2003). Therefore, it is of great importance to initialize the research efforts to address these aforementioned alternative routes which are equally interesting and important in the framework of biomass utilization and thermochemical conversion. ln the present work, a novel process concept is proposed for syngas production through biomass gasification, which involves the allothermal coupling of the biomass gasification and combustion processes in monolith structured reactors. The principle of the allothermal process concept is schematically illustrated in Figure 2.4. In this novel process, the monolithic micro-circulating fluidized bed will be used as the reactor unit for gasification-combustion of biomass. The exothermic combustion step and endothermic gasification step will be undertaken in one single monolithic reactor. The process intensification by periodic operation mode is used to chronologically segregate the gasification/combustion step. The wall of the monolithic reactor is constru·c ted by 8 Figure 2.3 Proposed novel process concept intemally encapsulating high-temperature phase-change-material, for example, LiF-CaF2 (Pletka et al., 2001 a,b) which serves for heat storage and heat release in cyclic operation. The di lute mixture of biomass (solid) and air/steam (fluid) flows in the monolithic reactor in pneumatic conveying fluidization mode. The gasification conversion of biomass is undertaken, using steam as the gasifying agent, the resulting product gas is rich in H 2 . The use of steam, instead of air or CO2 , leads to higher H2 yields due to the additional H 2 produced from the decomposition of H 2 0. In addition, compared with partial oxidation using substoichiometric air, the product gas has a higher heating value and the dilution with N 2 is avoided (Franco et al., 2003~ Frolin et al., 2008). The proposaI of this new process is supported by the recent advances in: (1) the development of micro-fluidized bed concept for biomass conversion (Potic et al., 2005)~ (2) the development of high- temperature PCM and its application in biomass gasification (Pletka et al., 9 200Ia , b)~ and (3) the pioneering experimental investigation in flow hydrodynamics of gas-solid twophase mixture in monolith (Ding et al., 2005, 2006). B Y checking our proposed process concept with the aforementioned three routes highlighted by Levenspiel (2005), it can be regarded that our present effort is an attempt to address one of the two important alternatives (i.e., route III). In this proposed process, the coal-to-syngas route is adapted for the biomass-to-syngas route. And the implementation of route III for syngas production is conducted through the chronologically-segregated hybridization of gasification/combustion processes in one single monolithic structured reactor. The main feature of the novel process is that the gasification and combustion of biomass are chronologically isolated from each other, and so are their gas streams. In this way, the product gas from gasification step is not diluted by the flue gas from combustion step. Furthermore, since there is no concern about dilution of the product gas by the flue gas, ait can be used as an oxidizing agent for the biomass combustion, instead of costly pure oxygen. U nderstanding and modeling complex flow hydrodynamics and thermochemical conversion behavior is very essential for effectively design and operate the suggested novel gasification-combustion coupling process. The knowledge from the different important fields (including biomass gasification, monolithic reactor, high-temperature phase change material, and fluidized bed reactor) is required for this purpose. Furthermore, the mathematical descriptions of these important aspects should be integrated together to realize a comprehensive capture of the fluidization hydrodynamics and reaction behavior. To this end, the literature review will be conducted in the following sections with a view to gaining a systematical understanding of the states-ofthe-arts in these relevant aspects. 10 2.3 Physicochemical Processes in Biomass Gasification and Modelling 2.3.1 Physicochemical processes in biomass gasification Gasification is a thermochemical conversion of solid carbonaceous materials by means of free or bound oxygen at elevated temperatures. This technology has been primarily used for coal gasification, but more recently it has been used for biomass and 1.3 1.6 1,4 • \Vood .. U gnin - • Cellulose Anthracite o 02 0,4 Figure 2.4 Van Krevelen diagram for various solid fuels (Prins et al.,2007) cellulose-rich wastes which have different C-H-O compositions from coal (see Figure 2.4). Several chemical aspects of the gasification of solid carbonaceous materials are summarized in the literature (Schlosberg, 1985; Vorres, 1999; Furimsky, 1999). Biomass gasification generally refers to the thermochemical conversion of solid biomass fuels using a gasifying agent (e.g. steam, substoichiometric air, or CO 2 ) to a mixture of combustible product gases (including H2, CH4 , CO and CO 2) along with heavy hydrocarbons with low dew points known as tar (Frolin et aL, . 2008). In the gasification processes, the fuel conversion takes place by various mechanisms, that is, drying, primary pyrolysis, secondary tar cracking, gasification, and combustion. During drying, fuel moisture evaporates followed by pyrolysis, which is the thermal Il decomposition of the solid fuel that forms gases, tar, and solid char residues (Figure 2.5). In addition to pyrolysis, thermal cracking of tar occurs. Gasification comprises a complex volatiles ~ il. ~c.rac~ng & fct onmng L pyrolysis B iom ass fùe} tar product gas: 1I2• CH 4 • CO. CO2·, Ci H4' C 1H 6 · char Figure 2.5 General reaction mechani sm for the gasification of a biomass fuel (Higman and van der Burgt,2003) set of heterogeneous reactions between CO 2 , H 2 0, and the solid char. Table 2.1 gives the possible heterogeneous and homogeneous reaction involved in the gasification of biomass (Radmanesh et al.,2006; Wurzenberger et al., 2002; Di Blasi, 2004). And sorne important experimental investigations of biomass gasification using fluidized bed are summarized in Table 2.2. Table 2.] List of heterogeneous and homogeneous reactions involved in biomass gasification No. ChemicaJ reaction Heterogeneous reactions Rl R2 R3 Homongeneous reactions R4 RS R6 R7 R8 Tar cracking reaction R9 Tar combustion 12 Table 2.2 Summary of important investigations of the gasification of biomass in fluidized beds Investigator Solid fuel Gasifying agent Pressure(kPa) Temperature(OC) Walawender et al. (1985) CelJulose H 20 101.325 592-787 Boateng et al . (1992) Rice hull s H20 111.457 700-800 Corella et al. (1991) Chips, straw H2 0 111.457 650-850 Herguido et al. (1992) Pine sawdust 90% H20 101.325 650-780 Tum et al. (1998) Sawdust Air and H 20 101.325 750-950 Gil et al. (1999) Pine wood chips Air, Air+H 20 , H20 101.325 750-830 Franco et a1. (2003) Pine and eucalyptus H 20 101.325 735-900 Cao et al. (2006) Wood sawdust Air 101.325 730-890 800-815 Radmanesh et al. (2006) Beech wood particles Air, H20 101.325 Ross et al. (2007) Eucalyptus and wood pellet Air + H 20 106.300 800-840 Lim et al. (2008) Wood chip Air 101.325 718-733 Campoy et al. (2008) Wood pellets Air, Air+ H 20 101.325 730-815 Gasification differs from combustion in several ways. When oxygen is limited in the gasification reaction, combustible products like hydrogen, carbon monoxide, and methane are produced. As a result, the carbonaceous feedstock is converted to a low- or medium-value synfuel gas, which is rich in carbon monoxide, methane, and hydrogen. If sufficient oxygen is provided, as it is in combustion, the products fully oxidize to water vapor and carbon dioxide. While combustion is useful for providing immediate heat, the produced gases have liule chemical energy remaining. As a complex thermochemical conversion process, the biomass gasification process is quite similar to that of coal gasification, yielding in both cases a mixture of gases \vith the same principal components (Zuberbuhler et al., 2005). However, the distribution of the resulting gases is different for biomass and coal, and the reaction conditions for biomass are milder than for coal gasification, due to the higher reactivity of biomass (Klass, 1998). As in the case of coal gasification, biomass gasification under elevated pressure conditions favors the production of methane and carbon dioxide, whereas increasing the temperature tends to increase the concentration of hydrogen and carbon monoxide. Steam is. often used as the gasification agent for syngas production. Blended with oxygen or air, it promotes the formation of H 2 and CO. Undesirable by-products and emissions encountered in the product gas, such as tar, are the main complications for its 13 use in the downstrearn synthesis or electricity production (Klass, 1998; Zuberbuhler et al. , 2005). Tar derived frorn biornass gasification or pyrolysis is condensible compound and causes sorne troubles in downstream equipment such as blocking and fouling of fuel lines, filters , engines and turbines. It was reported that tar content in the syngas from an air-blown circulating fluidized bed (CFB) biomass gasifier was about 10 g/m 3 . For other types of gasifier, tar content varied from about 0.5 to 100 g/m 3 (Asadullah et al. , 2003; Lopamudra et al. , 2003; Paasen, 2004). However, most applications of product gases require a Iow tar content of the order 0.05 g/m 3 or Iess. Hence, tar disposaI becomes one of the most necessary and urgent problems during biomass gasification (Lopamudraet al. , 2003 , Han and Kim, 2008). 2.3.2 Modeling of biomass gasification process Research on biornass pyrolysis, gasification and combustion processes has attracted growing attention during recent years due to the increasing use of renewable biomass energy. As important aspects, the optimization of thermal efficiency and the reduction of furnace emissions iequire accurate understanding of the physical and chemical effects in the reactor which are highly complex in nature. Mathematical modeling which allows quantitative representation of various phenomena is a powerful tool for process design, prediction of gasification performances, understanding of evolution of pollutants, analysis of process transients, and examination of strategies for effectivecontrol (Di Blasi, 2004). Although a lot of studies on the modeling of coal gasification can be found in the literature, modeling biornass gasification has not been amply addressed and only a very limited number of numerical models have been proposed for biomass gasification. Recently, several significant modeling efforts have been devoted to the simulation of biomass gasification in which much more comprehensive descriptions of related complex chemical and physical processes have been taken intQ account (see Table 2.3). Wurzenberger et al. (2002) developed a comprehensive transient model for packedbed biomass gasifier, which consists of a combined transient single particle and fuel-bed 14 model. Drying was modeled using an equilibrium approach, and primary pyrolysis was described by independent parallel-reactions. Secondary tar cracking, homogeneous gas reactions, and heterogeneous char reactions were modeled using kinetic data from literature. First, simulations of single particles, decoupled from the packed bed model, were performed. These simulations were compared with experimental investigations and showed the validity of the chosen overall approach for drying, pyrolysis, gasification, and combustion. Second, operation conditions of a moving bed combustor were chosen and the combined packed-bed and single-particle model were used to predict the overall behavior of this system. Di Blasi (2004) formulated a . one-dimensional, unsteady mathematical model to simulate countercurrent fixed-bed wood gasifiers, which coupled heat and mass transport with wood drying and devolatilization, char gasification, and combustion of both char and gas-phase species. The main processes modeled included: (1) moisture evaporation/condensation; (2) finite-rate kinetics of wood devolatilization and tar degradation; (3) heterogeneous gasification (steam, carbon dioxide, and hydrogen) and combustion of char; (4) combustion of volatile species; (5) finite-rate gas-phase watergas shift; (6) extraparticle mass transfer resistances, through the introduction of apparent rates for the heterogeneous reactions according to the unreacted core model; (7) heat and mass transfer across the bed resulting from macroscopic (convection) and molecular (diffusion and conduction) exchanges; (8) absence of thermal equilibrium (different solid and gas temperatures); (9) solid- and gas-phase heat transfer with the reactor walls; (10) radiative heat transfer through the porous bed; and (11) variable solid and gas flow rates. The model ~as used to simulate the structure of the reaction fronts and the gasification behavior of a laboratory-scale plant as the reactor throughput and. the air-to-wood (or char) weight ratio were varied. Predictions showed the existence of four main regions along the gasifier axis. In the first, gasification and combustion overlapped, the second was essentially the inert heating of a descending bed of char particles, and the last two were associated with wood devolatilization and drying, respectively. This structure of the reaction fronts was qualitatively similar to that reported for coal gasification. 15 Yang et al. (2006) developed a CFD-based model for simulating substoichiometric conversions of municipal solid wastes and as weIl as biomass fuel in packed-bed and moving-bed gasifier. The governing equations for mass, momentum and heat transfer for both solid and gaseous phases in a moving bed in a solid-waste incineration fumace were described and relevant sub-models were presented. Radiation heat transfer in the bed was simulated by a two-flux model. Mathematical simulation showed that countercurrent, substoichiometric conversion of both municipal solid wastes and biomass in movinggrate' systems was possible without loss in throughput or conversion efficiency. Char conversion rate was significantly lower than the devolatilization rate and the char conversion process occupies 1/2 of the total bed length, whereas fuel devolatilization occuped only around 1/3 of the bed length. The averaged devolatilization rate of biomass was twice as high as that for municipal solid wastes as a result of less moisture and ash contents. Biomass fuel also required a shorter distance to be ignited. Radmanesh et al. (2006) recently developed a one-dimensional transient model for biomass gasification in a bubbling fluidized bed reactor. The model took into account the pyrolysis and various heterogeneous and homogeneous reaction kinetics as weIl as the hydrodynamics of the bed and freeboard. A two-phase model w'as used to de scribe the gas phase in the bed, whereas a countercurrent back-mixing model was applied for the char mixing in the bed. It was shown that pyrolysis is an important step in the overall gasification model that can determine the distribution of products and thus the heating value of product fuel gas. The model also showed good agreement with experiments on steam gasification of wood, wheieby concentrations of H2, and CO 2 rise and that of CO drops. Table 2.3 Summary of recent important attempts at reactor modeling of biomass gasificati on Investi gators Fuel characteri sti cs Reactor type Remarks - - - - - - - - - --- - - - - - - - - - - - - - - - - - - - - -' - - - -, Wurzenberger et al. (2002) wood Moving bed 1D + tran sient model + detaiJed single-particJe model Di Blasi (2004) wood Counter-current fixed-bed ID + transient model + shrinking core particJ e model Yang et al. (2006) Bi omass and solid wastes Fixed-bed and rnoving bed 20 + transient model (CFD mode!) + two-f1u x radi ati on Radmanesh et al. (2006) beech wood particJ es Bubbling f1uidi zed bed 16 10+ transient model 2.4 Monolithic Structured Reactor and Modelling Methodology 2.4.1 Monolithic structured reactors Structured reactors/supports are increasingly considered for use ln multi-phase processes, because of the potential imp~ovements they offer with respect to the decoupling of heat and mass transfer phenomena, operation under reduced pressure drop conditions and at high gas/solid flow rates, and a greater resistance to attrition. One might also expect that the uniformity of channel structure may give improved homogeneity in performance compared to that of a fixed bed of traditional catalyst packing material, which is· inherently associated with significant flow heterogeneity. Monoliths, which contain catalysts with certain structures or arrangements, belong to the new family of the so-called structured catalysts and/or reactors (the border between 'catalyst' and 'reactor' vanishes in these reaction systems) (Tomasic, 2007). Usually monolith reactors refer to those containing catalysts with parallel straight channels inside the catalyst block (see Figure 2.6). The straight channels normally have circular, square or triangular cross-sections. A monolith structure is sometimes referred to as a (a) monolithic reactor /channel wal1 / washcoat with catalyst 17 SOLID PHASE .~~~!!~~~ Hète~rQgEmeôus GAS PHASE CenterHneof the m·ooQlith Qh~nne~ (b) transport/reaction phenomena in a monolith ch annel. Figure 2.6 Schematic representation of the monolith reactor (Tomasic, 2007). honeycomb structure, although in the technical context monolith has a much broader meaning, generally referred to as the large uniform block of a single building material. Monolith catalysts or monolith reactors have sorne cornmon features in most of their applications. These features or characteristics include (Chen et al, 2008): (i) low pressure drop especially under high fluid throughputs; (ii) elimination of external mass transfer and internaI diffusion limitations; (iii) low axial dispersion and backmixing, and therefore high product selectivity; (iv) larger external surface; (v) uniform distribution of ftow (gas phase); (vi) elimination of fouling and plugging, and thus extended catalyst lifetirne; (vii) easy scale-up. In view of their salient characteristics, monolithic catalysts and/or reactors appear to be one of the most significant and promising developments in the field of heterogeneous catalysis and chemical engineering of recent years. The use of monoliths in solid-catalyzed gas phase (single phase) chemical reactions is weIl established. In the last years, monoliths as multiphase reactors have receiveêl more and more attention (Roy et al, 2004a; Roy et al, 2004b; Irandoust & Andersson, 1988a). For example, monoliths can be used both for co-current and counter-CUITent operation in gasliquid reaction systems. They can combine the advantages of the slurry and trickle-bed reactor and eliminate the disadvantages such as discontinuous operation, stirring energy input, and catalyst attrition or ineffective catalyst use, liquid maldistribution, and local hotspots that may develop and cause runaways (Roy et al, 2004b, Charpentier, 2007). 18 However, the majority of the multiphase applications of monolith reactors have been mainly limited to the gas-liquid (or gas-liquid-solid) flow contact based reaction systems. The application of monoliths to gas-solid two-phase flow and reactions is not weIl advanced. Up to now, literature research on hydrodynamic studies for gas-solid twophase flow in monolith (particularly of the gas-solid flow distribution) is very scarce. Only very recently, the pioneering experimental work on gas-solid two-phase mixtures through multichannel monolithic geometry has been reported for the first time by a research group of University of Leeds (Ding et al., 2005, 2006). Ding et al. (2005) carried out the study on the macroscopic behavior of a gas-solid two-phase mixture flowing through monolith channels. The work showed that for pure gas 'flows, the laminar-to-turbulent transition in monolith channels occurred at a Reynolds number of about 620, much lower than the conventional transition criterion of 2200 for large pipes. For gas-solid two-phase flows, the pressure drop was shown to be significantly lower than that thr~ugh packed particle beds with even a lower specifie surface area. It was also shown that the measured pressure drop was considerably lower than the semi-theories developed for pneumatic conveying. In a subsequent paper, Ding et al. (2006) employed the non-intrusive positron emission particle tracking (PEPT) technique to investigate three-dimensional solids motion and microscopie behavior of suspended particles. Processing of the PEPT data gave solids velocity and occupancy in the monolith. channel. The results showed a nonuniform radial distribution of both the solids velocity and concentration. The highest solids concentration took place at a position approximately 0.7 times the column radius. 2.4.2 Modeling of monolithic structured reactors Mathematical modeling of monolithic catalysis has been an area attracting significant interest. The performance of the monolith reactor is a complex function . of design parameters (channel geometry, length and diameter of the channel, channel wall thickness), operating conditions (temperature, velocity) and the properties of both the catalyst (active species loading, washcoat loading, etc.) and the re.action mixture 19 (Tomasic, 2007). ) In addition, complexities arise from continuously changing inlet conditions that require a transient description of the monolith reactor. Therefore, modeling and simulation of monolith reactors can help to understand the complexity of interactions between various physical and· chemical processes that occur within the channels and in the channel walls (Tomasic, 2007; Chen et al.,2008). Till now, a great number of mathema"tÏcal models have been proposed to conduct various modeling and simulation for monolith reactors. However, the majority of modeling and simulation r~search has been focused on gas phase monolith reactors or catalytic converters (Chen et al., 2008). In literature, there is also several modeling research addressing on multiphase monolith reactors, which are mostly limited to gasliquid flow/reaction systems (Irandoust & Andersson, 1988b; Edvinsson & Cybulski, 1994; Stankiewicz et al., 2001, Roy et al., 2004a; Bauer et al., 2005). For modeling work on gas-solid two-phase mixture in monolith, there is no single report available in the open literature. Compared to the two-phase flow, the single~phase flow corresponds to a two- phase flow with zero solids holdup. From this standpoint, the knowledge from the modeling of single-phase flow behavior in gas-phase mORolith reactors could be to sorne extent helpful in understanding the two-phasè flow in monolith blocks. Therefore, in the following of this section, the modeling research on single-phase flow in monolith will be addressed. The models of monolithic reactors have been developed at di fferent levels of complexity. These models can be classified as one-, two-, or three-dimensional models, or classified as washcoat level , single-channel model, or multichannel model (Chen et al., 2008). The choice of complexity of the model is a tradeoff between specific modeling objectives and computational resource limitations. As the indi vidual channels within a monolith are separatecÎ from each other in terms of mass transfer, modeling of a single channel can often provide a wealth of information pertaining to the chemical behavior of the catalyst. In particular, it helps to identify and understand the rate limiting processes and the interplay between transport and heterogeneous surface reactions (Mazumder & Sengupta, 2002). Up to now, singlechannel modeling is the most extensively applied to describe the behaviors of a monolith 20 reactor, and much work has been done with the single-channel model (Deutschmann et al., 1999,2000; Tischer et al., 2001; Zerkle et al., 2000; Raja et al., 2000; Hayes et al. , 1996; Wilber & Boehman, 1999; Boehman &Dibble, 2000; Canu & Vecchi, 2002; Kumar & Mazumder, 2008). At this scale of modeling, it is assumed that every channel in the monolith reactor behaves exactly the same and can represent the entire reactor. However, under certain circumstances, modeling-a single monolith channel might be inadequate. Such circumstances include non-uniform inlet gas distribution, blocked or deactivated channels, -etc. (James et al., 2003; Chen et al. , 2008). In this case, all of the channels which interact with each other, because of the strong coupling of the individual channels through heat transfer and the inherent nonuniformities in flow distril)ution within monolith reactors. To address the differences in flow and temperature in different channels, multi-channel model has _to be chosen by accounting for a number of representative channels, or ev en the whole monolith block (Chen et al. , 2008; James et al., 2003). Although full-scale model provides more details and gives highest accuracy, it demands expensive computing facilities (Mazumder, 2007). As alternative modeling methodology, the equivalent continuum approach (Zygourakis and Aris, 1982; Chen et al., 1988; Zygourakis,1989) appears to be one of the most attractive solutions for simplifying the modeling of monolithic reactors. In recent years, computational fluid dynamics (CFD) has been introduced to model monolith reactors and has shown to be of significant importance in design and optimization of monolith reactors. However, much work has been done with the singlechannel CFD model. As opposed to the single-phase modeling abundantly available in literature, the studies on CFD modeling of whole monolith reactor are very limited in the published literature (Holmgren et al., 1997; Jeong and Kim, 1997,1998,2000; Shuai et al. , 2000; Chakravarthy et al., 2003; Liu et al, 2007; Mazumder & Sengupta, 2002; Mazumder, 2007). Chakravarthy et al. (2003) used a multi-channel model to study the impact of flow non-uniformity during cold-start transient operations of a catalytic converter. It was seen that inlet zone recirculation can lead to significant non-uniformity of the flow in the monolith , and this non-uniformity can lead to significant differences in ignition 21 characteristics among the channels. These ignition differences were especially pronounced at lower exhaust temperatures, where the axial location of ignition can vary from one channel to another. Liu et al. (2007) modeled a reverse flow catalytic converter used for a lean bum natural gas engine using a 3D model to study methane ignition. A dual zone approach was used for the heterogeneous model, where double ceIIs, or nodes are used to distinguish between fluid and solid temperatures. It is demonstrated that methane ignition can be achieved at a lower inlet gas temperature under conditions of reverse flow, compared to uni-direction al flow. The selection of flow mode must be selected depending on the inlet condition. Mazumder (2007) discussed and demonstrated two approaches that make simulation of full-scale catalytic converters with complex chemistry feasible. The proposed two different approaches were subgrid scale modeling and in situ adaptive tabulation. The first approach was one where only the larger sc ales were resolved by a grid, while the physics at the smallest scale (channel scale) were modeled using subgrid scale models whose development entailed detailed flux balances at the imaginary fluid-solid interfaces within each computational celI. The second approach made use of the in situ adaptive tabulation algorithm, after significant reformulation of the underlying mathematics, to accelerate computation of the surface reaction boundary conditions. Preliminary results shown for a catalytic combustion application indicated that both methods had the potential of improving computational efficiency by several orders of magnitude. It is important to note that for gas-solid two-phase flow/reactions through monolith 1 structured reactors, the modeling research work has not appeared in the open literature so far. However, the modeling methodologies and ideas which have been applied in singlephase monolith reactors could be borrowed to a .great extent and imparted into the modeling of gas-solid two-phase flow/reactions in monolithic reactors. 22 2.5 High-Temperature Phase-Change Material and Modelling Approaches 2.5.1 High-temp~rature phase-change material Thermal Energy Storage (TES) has received increasing attention over the last years. A widely used class of en erg y storage media is the so-called Phase Change Materials (PCMs). These media, characterized by a high value of latent heat per unit mass, seem to offer the better performance in thermal energy storage, due to their capability of absorbing/releasing high rates of energy as weIl as its relatively constant storage temperature. The phase changes of material are caused by the heat transfer to and from both of the phases on either side of the interface. This yields melting if the net heat is added to the solid part of the interface and solidification when the net heat is subtracted. / The observed addition al heat, which is involved in the conversion of one phase to another, is the latent heat; and the entire heat transport problem is usually referred to as the Stefan problem. Latent heat thermal storage using PCMs have been used in many applications as, for instance, in thermal control systems to reduce the temperature oscillations, or in space application for power production using closed Brayton cycle. In addition, thermal energy storage using PCMs is seen to be one of the effective ways for solar energy utilization (Hall et al., 1997), due to the following advantages: (i) the PCMs have high latent heat storage capacity (ii) the PCMs melt and solidify at a nearly constant temperature (3) a small volume is required for a latent heat storage system, thereby the heat los ses from the system maintains in a reasonable level during the charging and discharging of heat. Moreover, the application of PCMs for recovering high-temperature waste heat have attracted much interest in recent yearS (Maruoka & Akiyama,2006). Recently, the development of high-temperature phase change material has become a very interesting topic (Maruoka et al.,2002; Maruoka & Akiyama,2006). Table 2.4 gives the various properties of the PCM for the high temperature applications (Maruoka et al.,2002). 23 Table 2.4 Candidates of PCM for hjgh temperature application (Maruoka et a1.,2002) Malarial Comp. [m ol%) T",,,, [t<] D.H [kJ / mol M [g/mol] D.H [KJ / kg ] Density [kg! m") J bH (kJ/ m J Pri ce~ [~/kil [ kJ / lJ] k [ W/ m· k) 377 - Ag - 1235 11.3 108.0 104.6 10500 I.099E+06 22.600 0.005 NaF - 1269 - MgF2- NaF 64- 36 1273 42 .0 55.0 796.0 794.0 2780 3017 40,000 15,000 0.020 0.053 KF-MgF2 31-69 1281 6LO 71 0.0 2943 2.21 3E+06 2.395E+06 2.089E-t06 20,000 0.036 - Au Sm - 1337 12.7 197 .0 64.5 19300 0.000 272 1345 8.6 150.0 57.5 7700 1.244E+06 4.430E-tOS 10,000.000 - 2,500.000 0.000 - - 62.0 770,0 2390 L840E+06 - - 60A 922.0 3187 2.938E+06 12,000 0.0 77 - 55.0 62.3 265.5 942.0 7420 3150 1.970E -I-06 2.967E+06 170 11 ,200 1.561 0.084 157 .0 28.0 59.0 64.0 1414.3 29 1.5 7870 5.038E+05 3.309E+Q6 2.565E+06 2.500.000 6.000 6900 0.000 0.236 0.042 No2O - 1405 MgFz- MgO 91. 5- 8.5 1502 15 17 1536 Mn - M gF ~ - Gd - Si Co - 1535 1685 1767 -_. 14.6 10.1 39.6 17.2 2340 8800 - 8 - - 148 99 With its high storage density and small temperature variation from storage to retrieval, latent-heat thermal storage using high-temperature PCMs has been applied in gasification of biomass (Pletka et al., 200Ia,b; Cummer & Brown, 2005). In the process, heat released during combustion is stored as latent heat in phase change material sealed in tubes immersed in the reactor. This heat is released during the pyrolysis stage of the cycle .. The phase change material may be an inorganic salt or metal alloy. The reactor employs a fluidized bed to obtain uniform and rapid distribution of heat from the phase change material to the pyrolyzing fuel. It was demonstrated through the experimental results that the indirectly heated gasification of biomass is feasible to produce medium enthalpy producer gas (Pletka et al., 200Ia,b). 2.5.2 Modeling of solidification and melting processes in phase-change-material In a latent heat storage system, energy is stored during melting and recovered during solidification of the PCMs. Prediction of such altemating melting-solidification heat transfer processes is the key to optimal design of the energy storage system. However, theoretical analysis of problems involving melting or solidification is not an easy task. In fact, during the solid-liquid phase change, the interface between the two phases moves through the medium and its position is priori not known. The fact introduces a nonlinearity into the mathematical model which is very difficult to deal with, especially in 24 two- or three-dimensional problems. Moreover, many other factors such as variation of material properties and/or boundary conditions variable with arbitrary laws, increase the complexity of the problem (Pinelli & Piva, 2003). In the literature significant efforts have been devoted toward the development of mathematical models and numerical algorithms to study the transport phenomena occurring during the solidification/melting processes. Mathematically, the problem of solid-liquid change belongs to the class of the so-called 'moving boundary problems', due to the existence of moving phase-change boundary. Such problems are nonlinear and analytical solutions of the phase change problem are difficult to obtain except for a mere handful of physical situations with simple geometries and boundary conditions. Therefore, in most cases, the numerical methods have been resorted for the solution of phase-change problems. Basically, two different approaches have been used for numerical simulation of the phase change processes: (i) front-tracking formulation, and (ii) fixed-domain formulation. In the front-tracking approach, the position of the solid-liquid interface needs to be continuously tracked. The variable grid method (variable space grid and variable time step) provides the way to track the phase front explicitly. This approach works efficiently for pure substances. However, serious complications are encountered for solidification /melting problems involving multi-component systems, due mainly to topologically complicated diffusion interfaces characterizing the phase-transition morphology. In addition, this approach is poorly suited to multi-dimensional problems, due to the difficulties with algorithms of implementation and the penalty in computational cost. As an alternative, the fixed-domain formulation emerged as a more convenient strategy (VoIler & Swaminathan, 1990). With this approach, the need for explicit tracking of the solidificationlmelting fronts is eliminated and the entire computational dOll)ain is modeled with a single set of volume-averaged continuum conservation equations. One popular method akin to this approach is the enthalpy method, in which enthalpy is treated as ind~pendent variable. A fixed-grid is applied to the physical space and latent heat is accounted fro by using suitable source ' terms in the energy equation. 25 Table 2.5 Comparisons of numerical schemes for modeling phase change phenomena Investigators Grid Method Time-stepping Primary variable Murray & Landis (1959) Front track Finite difference Two-step Temperature Morgan et al. (1979) Fixed grid Finite element Two-step Apparent h Explicit Basic H Lemmon ' (1979) Front track Finite difference Rubisky & CravahJo (1981) Front track Finite element Explicit Fictitious h Voiler & Cross (1981) Front track Finite difference Explicit Basic H Rolph & Bathe (1982) Fixed grid Finite element ,Implicit Fictitious h Basic H Voiler & Cross (1983) Front track Finite difference ExplicitJImplicit Roose & Storrer (1984) Fixed grid Finite element Explicit Fictitious h Pham (1986) Fixed grid Finite element Two-step Basic H Crivelli & Idel sohn (1986) Fixed grid Finite element Implicit Temperature Dalhuijsen & Segal (1986) Fixed grid Finite eJement Two-step Apparent h Weaver & Viskanta (1986) Front track Finite difference Implicit Temperature Askar (1987) Front track Finite eJement C-N Temperature Dhatt et al. (1989) Fixed grid Finite element Explicit Basic H Comini et al. (1990) Fixed grid Finite eJement Two-step Apparent h Kim & Kaviany (1990) Front track Finite difference Explicit Basic H Vo]Jer (1990) Fixed grid Finite difference Implicit Apparent h Apparent h Temperature Tamma & Namburu (1990) Fixed grid Finite element Implicit Celentano et al. (1994) Fixed grid Finite element Implicit Esen & Ayhan (1996) Fixed grid Finite volume Implicit Apparent h Gong & Mujumdar (1997) Fixed grid Finite element three time-IeveJ scheme TeJTlperature Ha]] et al. (1997) Fixed grid Finite volume Explicit Apparent h Costa et al. (1998) Fixed grid Finite volume Implicit Apparent h Cui et al. (2003) , Fixed grid Finite volume Explicit Temperature Xing et al. (2004) Fixed grid Finite volume Explicit Apparent h Finite volume Explicit Temperature Elgafy-et al. (2004) Fixed grid Sharma et al. (2005) Fixed grid Finite difference Implicit Apparent h Xu et al (2005) Fixed grid Finite difference Implicit Temperature Trp (2005) Fixed grid Finite volume Implicit Temperature Halaw et al. (2005) Fixed grid Finite difference lmplicit Apparent h Frusteri et al. (2006) Fixed grid Finite difference C-N Temperature Fang & Chen (2007) Fixed grid Finite difference Implicit Apparent h Chen et al. (2008) Fixed grid Finite difference Implicit Apparent h This method is particularly suitable for alloys and plastics for which the change of phase occurs over a finite temperature range. Another method akin to the fixed-domain approach is the use of coordinate transformation. In the coordinate transformation method, the moving boundary is immobilized by using suitable transformation, which maps the physical plane onto the transformed plane. By scaling space and time, it permits 26 simplification of the solution which can be realized in the fixed domain. This method is particularl y useful for phase change at a fixed temperature, such as that for pure metals. In Table 2.5, the comparison of various numerical schemes used for modeling phasechange phenomena are detailed. Among them, sorne of the modeling works have been attempted to address the phase-change phenomena for high-temperature PCMs (Hall et al., 1997; Gong & Mujumdar, 1997; Cui et al., 2003; Xing et al., 2004; Eigafy et al. ,2004). 2.6 Gas-Solid Fluidization in Micro-Fluidized Bed Reactors and Modeling Methodology 2.6.1 Monolithic micro-fluidized bed reactors and gas-solid fluidization Very recently, miniaturization of fluidized beds is receiving increasing interest, due to that a small-size bed has good operability and availability for sorne particularly required characteristics. Such microfluidic-based microsystems represent the potential to 'shrink' convention al bench chemical systems to sm aIl size systems with major advantages in terms of performance, integration and portability. The concept of micro fluidized beds (MFBs) was first put forward by Potic et al. (2005) to refer to the beds with inner diameters of a few millimeters. The numbering-up concept is often regarded as a technique suited for increasing the throughput of a microreactor system. In this concept, the throughput is increased by parallelizing many identical microreactors or microchannels. Numbering-up is sometimes regarded as one of the advantages of microreactor technology. When the optimal microreactor design and its operating conditions are found in laboratory experiments, a commercial scale production plant can be designed in this concept more quickly than in the convention al scaling-up approach, which requires repetitive performance testing and process modification at several different throughput levels. In this context, the flow of gas-solid two-phase mixtures through monolith can be regarded as the consequence from the numbering-up (or scaleout) of a single micro-fluidized reactor. At the same time, the flow behavior in monolith could be fundamentally similar to that in microchannels, which has emerged as an important area of research in the past two decades due to their potential applications in 27 micropower generation, microelectro-mechanical systems (MEMS), biomedical use, biotechnology, and computer chips (Ding et ~l, 2005). Therefore, understanding of the gas-solid two-phase fIuidization hydrodynamics behavior in a single channel (of monolith) is of great interest In fact, gas-solid fIuidized beds are highly complex in fIuidization hydrodynamics (Kunii and Levenspiel, 1997), as in Figure 2.7. However, due to their favorable massand heat transfer characteristics and their continuous particle handling ability, they are extensively applied in a variety of industries. In view of chemical reaction processes, they are particularly suitable for highly exothennic and temperature-sensitive reactions , since the particle motion gives them a unique ability to rapidly transport heat and maintain a uniform temperature. In chemical reactors, not only the degree of particle mixing, but also the degree of gas mixing is of considerable importance. For biomass gasification application, the advantage of fIuidized bed reactors are (van der Drift et al., 2001; Yin et al., 2002; Wang et al., 2008): (i) short residence time; (ii) high productivity; (iii) uniform temperature distribution in gasifiers; (iv) low char or/and tar contents; (v) high cold gas energy efficiency; (vi) reduced ash-related problems; and (vii) the possibility of in-bed use of a catalyst for tar cracking. Fluidized bed gasification performs better than fixed bed gasification to reduce ash-related problems since the bed temperature of fluidized bed gasification can be kept uniformly below the ash slagging temperature. The low gasification temperature can also reduce the volatilization of ash elements such as sodium and potassium into 'the syngas, thus improving the quality of syngas (Wang et al., 2008). 28 Fast fluidization .. ---~_ ......-. low SOfld throughflow félle • high Solid throughflow rats Turbulent fluîdization Bubblîng bed o OA 0.2 0 .6 Volume traction solids: f Figure 2.7 Vertical distribution of so1id in different contacting regimes (Kunii & Levenspiel , 1997) 2.6.2 Modeling of circulating fluidized bed reactors With the advent of high-performance computers and the advances in numerical techniques and algorithm, computational fluid dynamics (CPD) analysis of multiphase systems has evolved to become a strong tool and approach for understanding the hydrodynamics and transfer mechanism as weIl as designing and developing equipment units. In recent years, the application of CFD for modeling and simulating gas-solid fluidized bed systems has been intensively explored to gain insight into the detailed local flow patterns and structures (Ding and Gidaspow, 1996~ Mathiesen et al., 2000~ Ibsen et al., 2001~ 1990~ Samuelsberg and Hjertager, Agrawal et al., 2001~ Zhang and Van der Heyden, 2001 ; Benyahia et al., 2007). GeneraIly, there are two different approaches dominating in the ~eld of numerical simulations of fluidized bed systems. Commonly, the y are referred to as the EulerianEulerian approach and the Eulerian-Lagrangian approach (Ibsen et al., 2004). The fundamental difference between them is how the particles are treated. In the Lagrangian 29 approach, the particles are treated individually and the motion of particles is obtained directly by solving Newton's second law for each particle. The reference frame thus moves with the particle as the individual tracked particles move through the domain. When applied to granular systems, such models are referred to as discrete element methods (DEM) or a particle-tracking approach. In contrast, in the Eulerian-Eulerian approach, the particle phase is modeled as interpenetrating continua, and its conservation equations have a form similar to those of the other phases. For gas-particle flows , the Eulerian-Eulerian model is often referred to as the two-fluid model (TFM). When solving the TFM, a set of models, either physical or empirical, is required in order to close the system of equations, including the interfacial terms and solid stress. One important closure is the particulate phase stress (namely, viscosity and normal stresses). Basically, two approaches exist today for treating the particulate phase stress. The first approach uses a constant particle viscosity (CPV) and an exponential power law for the particleparticle interaction force (Rietema, 1973; Gidaspow and Ettehadieh, 1983; Syamlal and Obrien, 1988; Bouillard et al., 1989). The second approach uses the kinetic theory of granular flow (KTGF), which is derived in analogy with the kinetic theory of gases (Lun et al., 1984; Ding and Gidaspow, 1990; Gidaspow et al, 2001). In the TFM models, the conservation equations for each of the two phases are derived to obtain a set of equations that have similar mathematical structure for both phases, which makes the mathematical manipulation of the system relatively easier and minimize the computation co st. From the point of view of computation, the TFM approach is much more feasible for practical applications to complex multiphase floes. The Eulerian-Lagrangian approach IS computationally intensive or even impossible for systems with a large number of particles. Thus, the Eulerian-Eulerian approach is convenient in simulating systems su ch as fluidized beds. A coupling between prediction of flow patterns and chemical reactions in riser flows is of great interest. In the past decade, significant progress has ' been made on CFD modeling of gas-solid circulating fluidized bed (CFB) systems. However, most of the works have been focused on modeling and simulating hydrodynamic behavior and flow patterns. So far, the attempts at coupling flow hydrodynamics with reaction kinetics by using CFD approach for the simulation of CFB reactors (riser) are still very limited in the 30 literature. Table 2.6 gives a summary of the recent attempts on CFD modeling of CFB reactors. Gao et al. (1999) developed a 3D two-fluid CFD flow-reaction model to predict flow and chemical reactions taking place in a FCC riser. This model combines a modified twophase turbulent model with realistic 13-lump reaction kinetics. The various key engineering aspects of the two-phase reacting flow in a catalytic riser reactor (including catalyst concentration distribution, the velocity distribution of both phases, interphase slip velocity, the temperature distribution of both phases, and the yield distribution over the entire reactor) can be predicted using this model. The predicted results showed that the gas-particulate turbulent reacting flow in the FCC riser reactors was very complicated due to feed efflux. The flow fields, particle concentration, temperature distribution, and yield distributions showed significant inhomogeneities in the axial, radial, and circumferential directions. Therdthianwong et al."(2003) developed a two-dimensional model for describing the performance of the ozone decomposition reaction in CFB system. The effect of solid viscosity on flow structure was explored by using two different models of sol id viscosity (namely, the constant solid viscosity coefficient model and the kinetic theory model). It showed that the solid viscosity calculated from different models had a significant effects on gas-solid flow pattern. The solid volume fraction profile calculated from the kinetic theory model with restitutive coefficient of 0.9999 matched the experimental value better than the constant solid viscosity coefficient model. Benyahia et al. (2003) used a transient isothermal gas/solid flow model to simulate the cracking reaction in an industrial FCC riser by using a 3-lumps reaction model. The hydrodynamic predictions, based on kinetic theory for granular flow, were compared to similar predictions found in literature. The cracking reactions" of heavy oil showed an increase in the gas axial velocity along the height of the riser, which had a significant impact on the gas/solid flow hydrodynamics. Das et al. (2004a,b) developed a three-dimensional simulation of a dilute phase riser reactor using a novel density based solution algorithm and following the Eulerian- 31 Eulerian approach. The kinetic theory of granular flow was applied. The gas phase turbulence was accounted for via a k-& model. The simulations showed a core-annulus flow pattern emerges on a time-averaged basis. Industrial data of the simultaneous adsorption of S02 and NO x in ariser were weIl simulated with a 3D reactor model. Comparison of simulations with a 1D and a 3D model showed that the use of 1D model was limited to riser configurations and conditions for which the effects induced by the outlet configuration were only small. For more restrictive outlet configurations, a 3D simulation was required. Hansen et al. (2004) modeled ozone conversion in a circulating fluidized bed (CFB) using a three dimensional multi-fluid CFD code. The gas phase was modeled using a LES model and the turbulent motion of the particulate phase was modeled by use of the kinetic theory of granular flow. The ozone conversion was modeled as a one-step catalytic reaction. The predicted ozone concentrations in the riser of the CFB were in good agreement with the experimental results. The radial variation in ozone concentration in 3D representation was better captured than in the 2D case. Table 2.6 Recent attempts at CFD modeling of circulating fluidized bed reactor performances Investigators Reactor type reactions Remarks Gao et al. (1999) FCC riser Catalytic cracking of crude oil Two fluid + ] 3-lump kinetics Therdthianwong et al. (2003) CFB riser 0 3 decomposition Two f1uid + simple kinetics Benyahiya et al. (2003) FCC riser Catalytic cracking of crude oil Two f1uid + 3-lump kinetics Das et al. (2004a,2004b) Dilute riser SOx-NOx adsorption Two fluid + adsorption kinetics Hansen et al. (2004) CFB riser 0 3 decomposition Two fluid + simple kinetics FCC= fluid catalytic cracking . 2.7 Summary and Conclu ding Remarks In this chapter, a novel process concept is proposed for coupling biomass gasification and combustion processes in monolithic structured reactors. Following the proposaI of this process concept, the literature review is performed to establish a global and systematical understanding of the states-of-the-arts in the relevant aspects (including biomass gasification, monolithic reactor, high-temperature phase change material, and fl uidized bed reactor). 32 Evidently, to effectively design and optimize this proposed process, an in-depth understanding of the coupling between themochemical reactions and fluid mechanics in monolithic structured reactor is very crucial. The treatment of generalized · local information demands the help of computational fluid dynamics (CFD) which can be used for simulating flow phenomena, understanding the impact of complex flow geometries on mixing and reaction phenomena, and obtaining information on the detailed quantitative flow pattern in multiphase flows. From product design and control point of view, those models which can be used to achieve a comprehensive description of the complex chemistry and transport phenomena occurring in biomass gasification are of primary importance and preferred. The use of monolith structured reactors allows decoupling of the chemistry, transport phenomena, and hydrodynamics, and the like to tailor the reactor independently to satisfy optimal operation conditions . .However, modeling of multiphase monolithic reactors is not an easy task and the choice of complexity of the models allows us to tailor the appropriate models for our purpose. The integration of biomass gasification/combustion processes with monolithic structured reactors will increase greatly the complexity of the whole system. This requires an integrated approach for modeling of coupled momentum-, heat-, and mass-transfer phenomena and complex kinetic processes which happen on different scales. Meanwhile, this calls for a further research into modeling strategies, methodologies, and tools to organize the levels of complexity and integrate the know ledge from the different fields of relevance. Testing the efficacy of the proposed process concept through modeling and experimentation is part of our ongoing project. Recognizing that understandinggas-solid flow distribution characteristics in monolith is significantly important for the development of the proposed process concept, the research focus of this Master thesis work is oriented to the CFD investigation of gas-solids (biomass particles) two-phase flow dynamics in monolithic multichannel micro-circulating fluidized bed (as detailed in Chapter 3). 33 References Agar DW. Multifunctional reactors: Old preconceptions and new dimensions, Chemical Engineering Science, 1999, 54 (10): 1299-1305 . Agrawal K, Loezos PN, Syamlal M, Sundaresan S. The role of meso-scale structures in . rapid gas-solid flows, Journal of Fluid Mechanics , 2001, 445: 151-185. Asadullah M, Miyazawa T, Kunimori K. Catalyst development for the gasification of biomass in the dual-bed gasifier, Appl Catal A: Gen, 2003,255: 169-180. Askar HG. The front tracking scheme for the one-dimensional freezing problem, Int. J. Numer. Methods Eng., 1987,24: 859-869. Bauer T, Guettel R, Roy S, Schubert M, AI-Dahhan M , Lange R. ModeBing and simulation of the monolithic reactor for gas-liquid-solid reactions, Chemical Engineering Research & Design, 2005, 83(A7): 811-819. Benyahia S, Ortiz AG, Paredes lIP. Numerical analysis of a reacting gas/solid flow in the riser section of an industrial fluid catalytic cracking unit, International Journal of Chemical Reactor Engineering, 2003, 1, A41. Benyahia S, Syamlal M, O'Brien Tl. Study of the ability of multiphase continuum models to predict core-annulus flow, AIChE~Journal, 2007, 53 (10): 2549-2568. Boateng AA, Walawender WP, Fan LT, Chee CS. Fluidized-Bed Steam Gasification of Rice Hull, Bioresource Technology, 1992, 40 (3): 235-239. Boehman AL, Dibble RW. Experimental and numerical investigation on the influence of tempora~ fuel/air unmixedness on NOx emissions of lean premixed catalytically stabilized and non-catalytic combustion, Catalysis Today, 2000, 59 (1-2):131-140. Bouillard lX, Lyczkowski RW, Folga S, Gidaspow D, Berry GF. Hydrodynamics of erosion of heat-exchanger tubes in fluidized-bed combustors, Canadian Journal of Chemical Engineering, 1989,67(2): 218-229. Campoy M, G6mez-Barea A, Villanueva AL, OBero P. Air-steam gasification of biomass in a fluidized bed under simulated autothermal and adiabatic conditions, Industrial & EngineeringChenûstry Research, 2008, DOl: 10.1021/ie800220t (in press). Canu P, Vecchi S. CFD Simulation of Reactive Flows: Catalytic Combustion ln a Monolith, AIChE Journal, 2002,48(12): 2921-2935. 34 Cao Y, Wang Y, Riley JT, Pan WP, A novel biomass air gasification process for producing tar-free higher heating· value fuel gas, Fuel Processing Technology, 2006 , 87:343-353. Celentano D, Onate E, OlIer S. A temperature-based formulation for finite element analysis of generalized phase-change problems, lnt. J. Numer. Methods Eng. , 1990, 37: 3441-3465. Chakravarthy VK, Conklin JC, Daw CS, D'Azevedo EF. Multi-dimensional simulations of cold-start transients in a catalytic converter under steady inflow conditions, Applied Catalysis A: General, 2003,241 (1-2): 289-306. Charpentier JC. Modern Chemical Engineering in the Framework of Globalization, Sustainability, and Technical Innovation, Ind. Eng. Chem. Res. ,2007, 46, 3465-3485. Chen CR, Sharma A, Tyagi SK, Buddhi D. Numerical heat transfer studies of PCMs used in a box-type solar cooker, Renewable Energy, 2008, 33(5): 1121-1129. Chen DKS, Oh SH, Bissett EJ, Van Ostrom DL. A three-dimensional model for the analysis of transient thermal and conversion characteristics of monolithic catalytic converters, in: Proceedings of the International Congress & Exposition, SAE Paper No. 880282, Detroit, MI, 1988. Chen JW, Yang H, Wang N, Ring Z, Dabros T. Mathematical modeling of monolith catalysts and reactors for gas phase reactions, Applied Catalysis A: General, 2008, 345:1-11 Comini G, Guidice SD, Saro O. A conservative algorithm for multidimensional conduction phase change, lnt. J. Numer. Methods Eng., 1990, 30: 697-709. Corella J, Aznar MP, Delgado J, Aldea E. Steam gasification of cellulosic wastes in a fluidized-bed with downstream vessels, Indus trial & Engineering Chemistry Research, 1991,30 (10): 2252-2262. Costa M, Buddhi D, Oliva A. Numerical simulation of a latent heat thermal energy storage system with enhanced heat conduction, Energy Conversion and Management, 1998, 39 (3-4): 319-330. Crivelli LA, Idelsohn SR. A temperature-based finite element solution for phase-change problems, lnt. J. Numer. Methods Eng., 1986,23: 99-119. 35 Cui HT, Hou XB, Yuan XG. Energy analysis of space solar dynamic heat rec~ivers , Solar Energy, 2003, 74:303-308. Cummer K, Brown RC. Indirectly heated biomass gasification using a latent-heat ballast part 3: refinement of the heat transfer model, Biomass & Bioenergy, 2005 , 28 (3): 321-330. Dalhuijsen Al, Segal A. Comparison of finite element techniques for solidification problems, Int. J. Numer. Methods Eng., 1986,23: 1807-1829. Das AK, De Wilde l , HeynderickxGl, Marin GB, Vierendeels l, Dick E. CFD simulation of dilute phase gas-solid riser reactors: Part l - a new solution method and flow model validation, Chemical Engineering Science, 2004a, 59 (1): 167-186. Das AK, De Wilde l, Heynderickx Gl, Marin GB, Vierendeels l, Dick E. CFD simulation of dilute phase gas-solid riser reactors: part II - simultaneous adsorption of S02-NOx from flue gases, Chemical Engineering Science, 2004b, 59 (1): 187-200. Deutschmann 0, Maier LI, Riedel U, Stroemman AH, Dibble RW. Hydrogen assisted catalytic combustion of methane on platinum, Catalysis Today, 200, 59 (1-2):141150. Deutschmann 0, Behrendt F, Warnatz l. FormaI treatment of catalytic combustion and . catalytic conversion ofmethane, Catalysis Today, 1998,46 (2-3): 155-163. Dhatt G, Song R, Cheikh AN. 1989. Direct enthalpy method for solidification calculation, in Gruber, R. et al. (eds.), Prods. of the Fifth Int. Symp. on Numerical Methods in Engineering, Boston, 487-494. Di Blasi C. Modeling wood gasification in a countercurrent fixed-bed reactor, AIChE Journal, 2004, 50 (9): 2306-2319. Ding l, Gidaspow D. A bubbling fluidization model using kinetic-theory of granular flow, AIChE Journal, 1990, 36 (4): 523-538. Ding YL, Wang ZL, Ghadiri M, Wen DS. Vertical upward flow of gas-solid two-phase mixtures through monolith channels, Powder Technology , 2005, 153 (1): 51-58. Ding YL, Wang ZL, Wen DS , Ghadiri M, Fan XF, Parker D. Solids behavior i~ a dilute gas-solid two-phase mixture flowing through monolith channels, Chemical Engineering Science , 2006, 61(5): 1561-1570. 36 Edvinsson RK, Cybulski A. A comparative analysis of the trickle-bed and the monolithic reactor for three-phase hydrogenations, Chemical Engineering Science, 1994, 49(24 )c: 5653-5666. Elgafy A, Mesalhy 0, Lafdi K. Numerical and experimental investigations of melting and solidification processes of high melting point peM in a cylindrical enclosure, 'Jol:lrnal of Beat Transfer- Transactions of the ASME, 2004, 126(5): 869-875. Esen M, Ayhan T. Development of a model compatible with solar assisted cylindrical energy storage tank and variation of stored energy with time for different phase change materials, Energy Conversion and Management, 1996,37(12): 1775-1785. Fang M, Chen GM. Effects of different multiple PCMs on the performance of a latent thermal energy storage system, Applied Thermal Engineering, 2007, 27(5-6): 9941000. Florin NH, Harris AT. Enhanced hydrogen production from biomass with in situ carbon dioxide capture using calcium oxide sorbents, Chemical Engineering Science, 2008, 63:287-3.16. Franco C, Pinto F, Gulyurtlu I, Cabrita I. The study of reactions influencing the biomass steam gasification process, Fuel, 2003, 82 (7): 835-842. Frusteri F, Leonardi V, Maggio G. Numerical approach to de scribe the phase change of an inorganic peM containing carbon fibres, Applied Thermal Engineering, 2006, 26: 1883 ~ 1892. Furimsky E. Gasification in petroleum refinery of 21 st century. Oil Gas Sci. Technol. . 1999,54,597-618. Gao JS, Xu CM, Lin SX, Yang GH, Guo YC. Advanced model for turbulent gas-solid flow and reaction in FCC riser reactors, AIChE Journal, 1999,45 (5): 1095-1113. Gidaspow D, Ettehadieh B. Fluidization in two-dimensional beds with a jet .2. hydrodynamic modeling, Industrial & Engineering Chemistry Fundamentals, 1983, 22 (2): 193-201. Gidaspow D, Jung JW, Singh RK. Hydrodynamics of fluidization usirtg kinetic theory: an emerging paradigm 2002 Flour-Daniel lecture, Powder Technology, 2001,148 (2-3): 123-141. 37 Gil J, Corella J, Aznar MP, Caballero MA. Biomass gasification in atmospheric and bubbling fluidized bed: effect of the type of gasifying agent on the product distribution, Biomass Bioenergy, 1999, 17: 389. Halaw E, Bruno F, Saman W. Numerical analysis of a PCM thennal storage system with varying wall temperature, Energy Conversion and Management, 2005, 46: 25922604. Hall CA, Glakpe EK, Cannon JN. Modeling Cyclic Phase Change and Energy Storage in Solar Heat Receivers, Proceedings of 32 nd AlAA Thermophysics Conference, Atlanta, USA, 1997. Han J, Kim H. The reduction and control technology of tar during biomass gasification/pyrolysis: an overview, Renewable and Sustainable Energy Reviews, 2008, 12:397-416. Hansen KG, Solberg T, Hjertager BH. A three-dimensional simulation of gas/particle flow and ozone decomposition in the .riser of a circulating fluidized bed, Chemical Engineering Science, 2004,59 (22-23): 5217-5224. Hayes RE, Kolaczkowski ST, Thomas WJ, Titiloye J. Transient experiments and modeling of the catalytic combustion of methane in a monolith reactor, lndustrial & Engineering Chemistry Research, 1996, 35(2): 406-414. Herguido J, Corella J, Gonzalezsaiz J. Steam Gasification of Lignocellulosic Residues in A Fluidized-Bed at A Small Pilot Scale - Effect of The Type of F,eedstock, lndustrial & Engineering Chemistry Research, 1992,31 (5): 1274-1282. Higman C, van der Burgt M. Gasification, Gulf Professional Publishing, United States of America, 2003. Holmgren A, Gronstedt T, Andersson B. Improved flow distribution in automotive monolith converters, React. Kinet. Catai. Leu., 1997, 60 (2): 363-371. Ibsen CH, Solberg T, Hjertager BH. Evaluation of a three-dimensional numerical model of a scaled circulating fluidized bed, lndustrial & Engineering Chemistry Research, 2001,40 (23): 5081-5086. Ibsen CH, Helland E, Hjertager BH, Solberg T, Tadrist L, Occelli R. Comparison of multifluid and discrete particle modelling in numerical predictions of gas particle flow in circulating fluidised beds, Powder Technology, 2004, 149 (1): 29-41. 38 Irandoust S, Andersson B. Monolithic catalysts for nonautomobile applications, Catai. Rev. -Sei. Eng., 1988a, 30: 341-392. Irandoust S, Andersson B. Mass transfer and liquid-phase reactions in a segmented twophase flow monolithic catalyst reactor, Chemical Engineering Science, 1988b, 43(8): 1983-1988. James A, Brindley J, Mclntosh AC. Multi-channel monolith reactors as dynamical systems, Combustion and Flame, 2003, 134(3): 193-205. Jeong S, Kim T, CFD investigation of the three-dimensional unsteady flow ln the catalytic converter, in: Proceedings of the International Congress & Exposition, SAE Paper No. 971025, Detroit, MI, 1997. J eong S, Kim W, A numerical approach to investigate transient thermal and conversion characteristics of automotive catalytic converters, in: Proceedings of the International Congress & Exposition, SAE Paper No. 980881, Detroit, MI, 1998. Jeong S, Kim W, Numerical analysis of light-off performance and thermo-fluid characteristics in a three-way monolithic catalytic converter, in: Proceedings of the Seoul 2000 FISITA' World Automotive Congress, 2000. Karellas S, Karla J, Kakarasb E. An innovative biomass gasification process and its coupling with microturbine and fuel cell systems, Energy, 2008, 33:284-291 Kim CJ, Kaviany M. A numerical method for phase change problems, Int. J. Heat Mass Transfer, 1990,33: 2721-2734. Klass DL, in Biomass for renewable energy, fuels and chemicals, Academic press, San Diego, London, 1998. Kobayashi N, Guilin P, Kobayashi J, Hatano S, Itaya Y, Mori S. A new pulverized biomass utilization technology, Powder Technology, 2008, 180: 272-283. Kolios G, Frauhammer J, Eigenberger G. Autothermal fixed-bed reactor concepts, Chemical Engineering Science, 2000, 55 (24): 5945-5967. Kolios G, Frauhammer J, Eigenberger G. Efficient reactor concepts for coupling of endothermic and exothermic reactions, Chemical Engineering Science, 2002, 57 (9): 1505-1510. 39 Kumar A, Mazumder S. Assessment of various diffusion models for the prediction of heterogeneous combustion in monolith tubes, Computers & Chemical Engineering, 2008,32 (7): 1482-1493. Kunii D, Levenspiel O. Circulating fluidized-bed reactors, Chemical Engineering Science, 1997,52 (15): 2471-2482. Lemmon EC. Phase change techniques for finite element codes. in Lewis, R.W. and Morgan, K. (eds.), Numerical Methods in Thermal Problems, Swansea, 1979, pp149158. Levenspiel O. What will come after petroleum? Industrial & Engineering Chemistry Research, 2005, 44 (14): 5073-5078. Lim MT, Alimuddin Z. Bubbling fluidized bed biomass gasification-performance, process findings and energy analysis, Renewable Energy, 2008, 33: 2339-2343. Liu B, Hayes RE, Yi Y, Mmbaga J, Checkel MD, Zheng M. Three dimensional modeling of methane ignition in a reverse flow catalytic converter, Computers & Chemical Engineering, 2007, 31(4): 292-306. Lopamudra D, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar elimination in biomass gasification processes. Biomass Bioenerg , 2003,24: 125-140. Lun CKK, Savage SB, Jeffrey DJ, Chepumiy N. Kinetic theories for granular flow inelastic particles in couette-flow and slightly inelastic particles in a general flow field, Journal of Fluid Mechanics, 1984,140: 223-256. Lutz AE, Bradshaw RW, Keller JO, Witmer DE. Thermodynamic analysis of hydrogen production by steam reforming, International Journal of Hydrogen Energy, 2003, 28, 159-167. Mamoka N, Akiyama T. Exergy recovery from steelmaking off-gas by latent heat storage for methanol production, Energy, 2006, 31 (10-11): 1632-1642. Mamoka N, Sato K, Yagi J, Akiyama T. Development of PCM for recovering high temperature waste heat and utilization for producing hydrogen by reforming reaction ofmethane, ISIJ International, 2002, 42 (2): 215-219. Mathiesen V, Solberg T, Arastoopour H, Hjertager BH. Experimental and computational study of multiphase gas/particle flow in a CFB riser, AIChE Journal, 1999, 45 (12): 2503-2518. 40 Mazumder S, Sengupta D. Sub-grid scale modeling of heterogeneous chemical reactions and transport in full-scale catalytic converters, Combustion and Flame, 2002, 131 (12): 85-97. Mazumder S. Modeling full-scale monolithic catalytic converters: Challenges and possible solutions, Journal of Hf!at Transfer- Transactions of the ASME, 2007, 129 (4): 526-535. Morgan K, Lewis, RW, Zienkiewicz, OC . An improved algorithm for heat conduction problems with phase change, Int. J. Numer. Methods Eng., 1978,12:1191-1195 Murray WD, Landis F. Numerical and machine solutions of transient heat conduction problem involving meltingor freezing, J. Heat Transfer, 1959, 81: 106-112. Negro SO, Suurs R AA, Hekkert MP. The bump road of biomass gasification in the Net.herlands: Explaining the rise and faB of an emerging innovation system, Technological Forecsting & Socil Change, 2008,75: 57-77. Paasen SVB. Tar formation in fluidized-bed gasification-impact of gasifier operating c.onditions. The 2nd world conference and technology exhibition on biomass for energy, industry and climate protection, 2004. pp. 130-138. Panigrahi S,Dalai AK, Chaudhari ST, Bakhshi NN. Synthesis gas production from steam gasification of biomass-derived oil, Energy Fuel, 2003, 17(3): 637-642. Pham QT. The use of lumped capacitance in the finite-element solution of heat conduction problems with phase change, Int. J. Heat Mass Tran., 1986,29: 285-291. Pinelli M, Piva S. Solid/liquid phase change in presence of natural convection: A thermal energy storage case study, Journal of Energy Resources Technology-Transactions of The ASME, 2003,125 (3): 190-198. Pletka R, Brown RC, Smeenk J. Indirectly heated biomass gasification using a latent heat ballast - 1: experimental evaluations, Biomass & Bioenergy, 2001a, 20 (4): 297-305. Pletka R, Brown RC, Smeenk J. Indirectly heated biomass gasification using a latent heat ballast. Part 2: modeling, Biomass & Bioenergy, 2001 b, 20 (4): 307-315. Potic B, Kersten SRA, Ye M, van der Hoef MA, Kuipers JAM, van Swaaij WPM. Fluidization with hot compressed water in micro-reactors, Chemical Engineering Science, 2005, 60 (22): 5982-5990. 41 Prins MJ, Ptasinski KJ, Janssen FJJG. From coal to biomass gasification: comparison of thermodynamic efficiency, Energy, 2007,32(7): 1248-1259. Radmanesh R, Chaouki J, Guy C. Biomass gasification in a bubbling fluidized bed reactor: Experiments and modeling, AIChE Journal, 2006, 52 (12): 4258-4272. Raja LL, Kee RJ, Deutschmann 0, Warnatz J, Schmidt LD. A critical evaluation of Navier-Stokes, boundary-Iayer, and plug-flow models of the flow and chemistry in a catalytic-combustion monolith, Catalysis Today, 2090, 59 (1-2):47-60. Ramaswamy RC, Ramachandran PA, Dudukovic MP. Recuperative coupling of exothermic and endothermic reactions, Chemical Engineering Science, 2006,61 (2): 459-472. Rietema K. Effect of Interparticle Forces On Expansion of a Homogeneous GasFluidized Bed, Chemical Engineering Science, 1973,28 (7): 1493-1496. Rolph WD, Bathe ~J. An efficient algorithm for analysis of nonlinear heat transfer with phase changes, Int. J. Numer. Methods Eng., 1982, 18: 119-134. Roose J, Storrer O. Modelization of phase changes by fictitous heat flow, Int. J. Numer. Methods Eng., 1984, 20: 217-225. Ross D, Noda R, Horio M, Kosminski A, Ashman P, Mullinger P. Axial gas profiles in a bubbling fluidised bed biomass gasifier, Fuel, 2007, 86 (10-11): 1417-1429. Roy S, Heibel AK, Liu W, Boger T. Design of monolithic catalysts for multiphase reactions, Chemical Engineering Science, 2004a, 59: 957- 966. Roy S, Bauer T, AI-Dahhan M, Lehner P, Turek T. Monoliths as multiphase reactors: a review, AIChE Journal, 2004b, 50 (11): 2918-2938. Rubinsky B, Cravahl0 EG. A finite element method for the solution of one-dimensional phase change problems, Int. J. Beat Mass Tran., 1981,24: 1987-1989. Samuelsberg A, Hjertager BH. Computation al modeling of gas/particle flow in ariser, AIChE Journal, 1996,42 (6): 1536-1546. Schlosberg RH. Chemistry of coal conversion; Pl~num: New York, 1985. Sharma A, Won LD, Buddhi D, Park JU. Numerical heat transfer studies of the fatty acids for different heat exchanger materials on the performance of a latent heat storage system, Renewable Energy, 2005, 30: 2179-2187. 42 Shuai S, Wang J, Zhuang R, Numerical simulation and optimum design of automotive catalytic converters, in: Proceedings of the Seoul 2000 FISITA World Automotive Congress, 2000. Stankiewicz A. Process intensification in in-line monolithic reactor, Chemical Engineering Science, 2001, 56:359-364. Syamlal M, Obrien TJ. Simulation of granular layer inversion in liquid fluidized-beds, International Journal of Multiphase Flow, 1988, 14(4): 473-481. Tomasic V. Application of the monoliths in DeNOx catalysis, Catalysis Today, 2007', 119: 106-113. Tamma K, Namburu R. Recent advances, trends and new perspe~tive via enthalpy-based finite element formulations for applications to solidification problems, Int. J. Numer. Methods Eng., 1990,30: 803-820. Therdthianwong A, Pantaraks P, Therdthianwong S. Modeling and simulation of circulating fluidized bed reactor with catalytic ozone decomposition reaction, Powder Technology, 2003, 133 (1-3): 1-14. Tischer S, Correa C, Deutschmann O. Transient three-dimensional simulations of a catalytic combustion monolith using detailed models for heterogeneous and homogeneous reactions and transport phenomena, Catalysis Today, 2001, 69 (1-4): 57-62. Trp A. An experimental and numerical investigation of heat transfer during technical grade paraffin melting and solidification in a shell-and-tube latent thermal energy storage unit, Solar Energy, 2005,79: 648-660. Turn S, Kinoshita C, Zhang Z, Ishimura D, Zhou J. An experimental investigation of hydrogen production from biomass gasification, International Journal of Hydrogen Energy, 1998, 23 (8): 641-648. van der Drift A, van ·Doorn J, Vermeulen JW. Ten residual biomass fuels for circulating fluidized-bed gasification, Biomass & Bioenergy, 200 1, 20:45-56. van der Drift A, Boerrigter H. Synthesis gas from biomass for fuels and chemicals, ECN report, 2006. VoIler V, Cross, M. Accurate solutions of moving boundary problems using the enthalpy method~ Int. J. Heat Mass Tran., 1981,24: 545-56. 43 VoIler V, Cross M. An explicit numerical method to track a moving phase change front, Int. J. Heat Mass Tran., 1983,26: 147-150. VoIler V, S waminathan CR, Thomas BG. Fixed grid techniques for phase change problems: a review, Int. J. Numer. Methods Eng., 1990,30: 875-898. VoIler VR. Fast implicit finite-difference method for the analysis of phase-change problems, Numerical Heat Transfer Part B-Fundamentals, 1990, 17(2): 155-169. Vorres KS. Coal. In Concise encyclopedia of chemical technology, 4th ed.; Kirk, R. E., Othmer, D. F., Eds.; Wiley: New York, 1999; pp 458-462. Walawender WP, Hoveland DA, Fan LT. Steam gasification of pure cellulose. 1. uniform temperature profile, Indus tria 1 & Engineering Chemistry Process Design and Development, 1985, 24 (3): 813-817 . Wang LJ, WeIler CL, Jones DD, Hannab MA. Contemporary issues in thermal gasification of biomass and its application to electricity and fuel production, Bioma~s and Bioenergy, 2008, 32: 573-581. Weaver JA, Viskanta R. Freezing of liquid saturated porous media. J. Heat Mass Transfer, 1986,33: 2721-2734. Wilber RA, Boehman AL. Numerical modeling of the reduction of nitric oxide by ethylene over Cu-ZSM-5 under lean conditions, Catalysis Today, 1999,50 (1): 125132. Wurzenberger JC, Wallner S, Raupenstrauch H, Khinast JG. Thermal conversion of biomass: Comprehensive reactor and particle modeling, AIChE Journal, 2002, 48 (10): 2398-2411. Yin XL, Wu CZ, Zheng SP, Chen Y. Design and operation of a CFB gasification and pow~r generation system for rice husk. Biomass & Bioenergy, 2002,23: 181-187. Xu X, Zhang YP, Lin KP, Di HF, Yang R. Modeling and simulation on the thermal performance of shape-stabilized phase change material floor used in passive solar buildings, Energy and Buildings~ 2005, 37: 1084-1091. Yang YB, Sharifi VN, Swithenbank J. Substoichiometric conversion of biomass and solid wastes to energy in packed beds, AIChE Journal, 2006, 52 (2): 809-817. 44 Zerkle OK, Allendorf MD, Wolf M, Oeutschmann O. Understanding homogeneous and heterogeneous contributions to the platinum-catalyzed partial oxidation of ethane in a short-contact-time reactor, Journal ofCatalysis, 2000,196 (1): 18-39. Zhang OZ, VanderHeyden WB. High-resolution three-dimensional numerical simulation of a circulating fluidized-bed, Powder Teehnology, 2001, 116 (2-3): 133-141. Zuberbuhler U, Specht M, Bandi A. Gasification of biomass: an overview on available technologies, RENEW -1 st European summer school on renewable motor fuels, 2005. (http://www.conference-biomass.comlpdfs/Visual-Presentations WednesdayFriday.pdD Zygourakis K, Aris R. Heat transfer. in the array of passages of a monolith reactor, in: Proeeedings of the AIChE 75 th Annual Meeting, Los Angeles, CA, 1982. Zygourakis K. Transient operation of monolith catalytic converters: a two-dimensional reactor model .and the effects of radially nonuniform flow distributions, Chem. Eng. Sei., 1989, 44(9):2075-2086. 45 Chapter 3 Simulating the Dynamics of Gas-Solid Flows in a Multichannel Micro-Circulating Fluidized Bed Yi-Ning WangI, Faïçal Larachi 1*, Shantanu Roy2 'Department of Chemical Engineering, Laval University, Quebec (QC), G1K 7P4, Canada. 2Department of Chemical Engineering, Indüm Institute of Technology (IIT) - Delhi, New Delhi 110016, India. Abstract The dynamics of gas-solid flows and distribution in monolithic multichannel microcirculating fluidized-bed reactors was analyzed using a computational fluid dynamics (CFD) modeling approach. A 2-D Euler-Euler multiphase model with the kinetic theory of granular flow has been solved for the detailed monolithic packing geometry. The assemblage of monolithic structured packings with through-flow gas-particulate flows is globally considered in the simulation to capture the dominant mechanisms contributing to the final overall aero/granular dynamics. Due to the complex nature of the interactions between gas and particulate phases and the stationary monolith backbone, one of the challenges in the design and operation of the monolith reactors is the prevention of flow maldistribution. The work presented in this paper forms the basis for a comprehensive reactor-scale model for exploring the intriguing possibilities that the proposed process intensification concept offers for chemical reactions of energy/environmental relevance such as biomass gasification and combustion. Keywords: CFD simulation; monolithic structured reactor; maldistribution; hybridization of gasification/combustion; biomass. * To whom correspondence should be addressed. Tel.: +14186563566; fax: +14186565993. E-mail: [email protected] (F. Larachi). 46 3.1 Introduction Biomass is one of the important pnmary and renewable energy sources. With evidence of depleting fossil fuel sources as weIl as the evolving global warming issues, the need for utilization of biomass for energy is very seductive, particularly because it is believed that energy obtained from biomass has a carbon-neutral cycle. This situation calls for the development of a biomass-based but energy efficient and environment friendly processes with better environmental acceptability and economic viability.l,2 Gasification to produce biosyngas is regarded as one of the most promising options for biomass conversion and utilization. However, thermodynamics and intrinsic kinetics dictate that endothermic biomass gasification reactions have to be carried out at high temperatures, which demands efficient heat supply and recovery policy. The concept of allothermal gasification offers an attractive solution for implementing high-temperature reactions by coupling strongly endothermic reactions with exothermic reactions? However, implementing the concept in practice is not straightforward. In the present work, we will first propose an innovative biomass gasification process concept in which the coupling of gasificationlcombustion process with high-temperature phase-change-material will be implemented in a monolithic structured reactor and intensified by periodic operation mode. To effectivelydesign and optimize this novel process, knowledge from different important fields (including biomass gasification, monolith reactor engineering, high-temperature phase change material, and fluidized bed reactor) is required. In addition, modeling and understanding of gas-solid (biomass particles) flow hydrodynamics in monolithic structured reactor is very important, in view of the complexity of two-phase flow within such confined micro-structured packings. Specifically, due to the complex nature of the interactions between gas and particulate phases and the stationary packing, one of the major challenges in the design and operation of the monolith reactors is prevention of flow mal-distribution. In order to overcome the limitations posed by this phenomenon, flow distribution characteristics in this type of reactors need to be quantitatively studied and understood. In this work, following the introduction of the novel process concept, the gas-solid two-phase flow distribution characteristics in a monolithic structured reactor have been investigated using 47 a computational fluid dynamics (CFD) simulation approach. A two-dimensional multifluid Euler-Euler CFD model with closure laws according to the kinetic theory of granular flow has been solved. To effectively characterize flow distribution, an assemblage of structured monolithic section with non-structured packed-bed sections is fully considered in our simulation, allowing comprehensive capture of various possible mechanisms contributing to the final overall aero/granular dynamics. The packed-bed sections are treated as porous media by imposing radial porosity distribution and interphase interactions through user defined functions (UDFs). 3.2 Hybridization of Gasification/Combustion Processes in Monolithic Structured Reactors Steam gasification of sol id carbonaceous fuels is highly endothermic, which demands input of additional heat to drive reactor conversion. This poses a major challenge because the input of energy reduces the maximum process efficiency. There are a number of potential problems 3 which could be encountered in developing process concepts for biomass gasification with steam: (i) If the biomass is reacted with both air and steam in one reactor, then nitrogen is present in the product stream and is costly to remove; (ii) If one attempts to avoid this problem by using oxygen instead of air, then a source of pure oxygen would be needed, which is again a costly option; (iii) It is possible to circumvent the separation issues by running the "oxygen-Iess" gasification and the combustion reactions in different locations, but then transferring heat from one location to the other is accompanied with heat losses; (iv) AIso, in all of these schemes, potential rapid cooling of the product gases leads to tar formation, .which adversely affects the process stability and efficiency as weIl. To avoid this, the product gases must be kept hot for an optimal duration of their residence time, which allows the tars to crack into lower molecular weight compounds. In the present work, a process concept which involves time-segregated hybridization of biomass gasification/combustion reactions in a monolithic structured reactor is proposed, as illustrated in Figure 3.1. In this process, the monolithic micro-circulating 48 • ste.am COZ +H2;O C fi> i Tibiom .3ss lim e ~ ~ E o o ai r CO +Hz +biom ass Figure 3.1 Proposed process concept fluidized bed is used as the reactor unit for gasificationlcombustion of biomass (Figure 3.1). At -the heart of the proposed process is a monolithic reactor through which gassolids cocurrent flow occurs, much as in a conventional circulating fluidized bed reactor. However, the presence of the numerous monolith channels serves to segregate the gassolids flow into these individual cells, which helps to intensify the process. Both the exothermic combustion step and the endothermic gasification step are undertaken in the same monolithic reactor. The process intensification by periodic operation mode is used to temporally segregate the gasification and combustion steps. This is made possible by coating the walls of the monolith channels with high-temperature phase-change-materials (PCM)4,s serving for successive heat storage and heat release in a cyclic operation. The biomass is supplied to the monolithic reactor aft.er fine granulation and subsequent pneumatic conveying. Hence, the process intensification is achieved both by temporal segregation of gasification and combustion as well as the use of a monolithic microfluidized bed reactor with walls coated with PCM. The proposed novel process is supported by the recent advances in: (1) the development of micro-fluidized bed concept 49 for biomass conversion;6 (2) the development of high-temperature PCM and its application in biomass gasification;4,s and (3) the pioneering experimental investigation in flow hydrodynamics of gàs-solid two-phase mixture in monolith. 7,s The proposed concept incorporates the diverse notions proposed by the above referenced papers onto a single platform. Testing the efficacy of the concept through modeling and experimentation is part of our ongoing work and the present contribution is a summary of our first full set of results in addressing a key enabling technology for the concept. The design and optimization of this novel hybrid process requires accurate understanding of not only the phenomena of biomass thermochemical conversion but also the two-phase hydrodynamics behavior in the monolithic micro-fluidized reactor which are highly complex in nature. In this regard, the flow distribution characteristics of gassolids two-phase hydrodynamics in monolithic structured reactor are significantly important for prediction of gasification/combustion performance and examination of strategies for process operation. In the following sections, the development of EulerEuler CFD multifluid simulation as weIl as its application for exploring maldistribution of two-phase flow in monolithic packing will be discussed in details. 3.3 Representation of Nonuniform Porosity Distribution for Packed-bed Sections The reactor computational geometry considered in the present work consists of three sections: upstream random-packing fixed-bed distributor section, central monolithic section and downstream random-packing fixed bed section (details are given in Section 7 3.5). This corresponds to the system geometry reported in the literature by Ding et a1. ,s, in which the monolith section is sandwiched between the two packed-bed sections. For the fixed bed randomly packed with solid particles with low D/dp ratios, the flow is remarkably affected by radial porosity distribution which is function of bed diameter (D), and particle diameter (dp ) and shape. Therefore, it is essential to define and implement porosity distribution in the simulations to capture the radial distribution characteristics in randomly packed beds. Experimental and computational investigations have shown that in low D/dp ratio beds the porosity is high near the vicinity of the wall and it oscillates 50 significantly in the near wall region, by following a damped oscillatory function until it reaches a constant value about 5 particle diameters from the wall. Mueller (1992) developed a correlation for radial variation of porosity, as a function of particle diameter and bed diameter, which has the following form: 9 E ( r) = E B + (1- E B ) Jo ( ar * ) e-br (la) where 12.98 8,2, 43 (D/d p -3.156) for 2.61~ D/d p ~ ]3.0 (lb) a= ( 7.383 b 2.932 (D/d p -9.864) for D/d p ~13.0 = 0.304- 0.724 (lc) D/d p r * r (Id) D and Jo is' zero th -order Bessel function of the first kind. This correlation represents the available experimental data with reasonable ' accuracy and is widely used. Figure 3.2 presents the simulated radial porosity variations for our numerical geometry which is characterized by low column-to-particle ratio D/dp = 5 (i.e., column of 50mm and packing particles of 10mm). As compared to the correlation proposed by Giese et al.,l0 the Mueller' s correlation is adopted in our work since it captures the wall-induced damped oscillations. 51 1.0 0 .8 - 0- - 0- Mueller(1992) Giese (1998) 0.2 0.0 +------.--..,-------.--..,----.----,.-----.----,.-----.---1 1.5 2.0 2.5 0.0 0.5 1.0 r/dp [-] Figure 3.2 Radial variation ofbed porosity in packed-bed sections 3.4 Eulerian-Eulerian Multifluid Model for Gas-Solid Flow in Monolithic Structured Reactor An Eulerian-Eulerian model with the kinetic theory of granular flow is used to model the hydrodynamics of gas-solid flow in the three-section monolithic reactor. The equations employed are a generalization of the Navier-Stokes equations for interacting continua, and aIl phases are considered to be continuous and fully interpenetrating. The model goveming equations for the gas and solid phases are as follows: 3.4.1 Continuity and momentum conservation equations 3.4.1.1 Mass conservation equations of gas and particulate phases Mass conversation equation for each phase (q=g,s) is described by: ~(p dt qa q)+V(pqaq~q )=O (2) Each computational cell is shared by the interpenetrating phases, the sum over aIl volume fractions is therefore unit y: (3) 52 3.4.1.2 Momentum conservation equation of gas and particulate phases Momentum conservation for the gas phase is written as: (4) Momentum conservation for the particulate phase can be expressed as :/ (aA ~' )+ v( a,p,v,v,)= -a,Vp + vp, + V . ~, +a,p, g- fJg , (v g where = Tg and = Ts -v, )+s, (5) are the phase stress tensors for gas and sol id phases, respectively; and /3gS is the drag coefficient between phases . . 3.4.2 Kinetic theory of granular flow equations Closure of the particulate phase momentum equation requires constitutive relations for calculating solid pressure, Ps ' solid shear viscosity, fl s ' and solid bulk viscosity, Àç , which can be derived from the granular kinetic theory.l1 The kinetic energy of fluctuation is accounted for by defining a granular temperature, (}ç : () - -1 ( v'2) s 3 s (6) where v~ is the particulate fluctuating velocity. The granular temperature conservation equation is: (7) where (- p) +;s):v~s is the generation of energy by the solid stress t~nsor, v' (kB,V (}ç) is the diffusion of energy, and YB, is the collisional dissipation of energy. The solid pressure, Ps , is composed of a kinetic term that dominates in the dilute flow regions and a collision contribution that is significant in the dense flow region: Il 53 (8) where ess is the coefficient of restitution for particle collisions. The radial distribution function, go' is a correction factor that modifies the probability of interparticle collisions. The solid shear viscosity, J.1s , Îs calculated by12,13 (9) The solids bulk viscosity, À:~ , is expressed as: Il (10) 3.4.3 Closure relationships for interphase interactions The interaction coefficient between the gas phase and flowing particulate phase can be described by a combination of Wen and Yu 14 and Ergun 15 equations. The final drag coefficient for this combination is expressed as f3 . = Erg lin ~c a~ag Pg I~s - ~g 1a4 dl' D 2 fJw"- y,, = 150 :< ~~ + 1.75 g The drag coefficient CD = l .\ CD ~ . [1+0.15(a a Re~ g 2 .65 g a: g (11) Iv, - Vg 1 S is evaluated by ReJ o687 ] Res < 1000 g Re, 0.44 ~ (12) 1000 with the relative Reynolds number, Res ,defined by I~s -~g l = _Pg d_s~_-,- Re s (13) J.1 g 54 To avoid discontinuity from the two equations, Gidaspow ·(1994) 13 introduced a switch function that gives a smooth but rapid transition from one regime to the other: arctan[150x1.75x(0.2-a )] f/Jgs = s (14) +0.5 J[ Thus , the interaction coefficient between fluid particle phases is finally expressed as (15) The interaction between the gas phase and the stationary packing phase, i.e., lower and upper packed bed sections, can be expressed using an Ergun-type equation 15 (16) In literature there are a few attempts made to evaluate the interaction force between powder and packing particle phases. 16- 18 In this work, the interaction between the flowing suspended phase and the stationary packing phase is expressed by:17,19 (17) where (18) " 2d p (l-ë) D=--- (19) 3ë (20) where ()~ = Vs . (a~ E) is the superficial suspended solids velocity vector. 55 . ~ 3~4.4 Definition of maldistribution quantities To quantify the flow non-uniformity of two-phase flow distribution in each channel, the flow factor, ri, which is the estimated ratio of the actual flow rate to the theoretical flow rate at uniform distribution,20 is here used: (21) where mi,o mi denotes the mass flow-rate of the qth phase in ith channel in actual cases, while q is the theoretical uniform flow-rate of the qth phase in ilh channel. In actual cases, the value of ri may be greater than or less than 1.0, representing a jlow excess or jlow starvation state in each channel, respectively. Besides the flow factor, the maldistribution factor, Mf ' which was first introduced by Hoek et al. 21 and modified by Marcandelli et al.,22 is also adopted in this work to determine the flow distribution over the entire cross-section of the monolith block: (22) where N ell is the number of ·channels. uniformly; and MJ MJ equals 0.0 when the flow is distributed approaches 1.0 when the flow is highly selective to one single channel. 3.5 Computational Geometry, Boundary Conditions and Numerical Solution The monolith geometry reported in the literature by Ding et al. 7 ,8 is considered in our simulation, which has a length of 600 mm and a (square) cell size of 3 mm (cell size represents clearance without cell wall thickness), see Table 3.1. As far as the cold unit simulations are concerned, this selected block length is judged representative for highlighting the gas and solids flow maldistribution issues. Should it be necessary and 56 depending on the reactions to be hosted in the future studies, the monolith length is extensible to make it compatible with the reaction characteristic times. The monolith section is connected to an upstream packed bed (length: 300 mm, sphere diameter: 10mm) which serves as distributor. In addition, a packed-bed section (length: 100 mm, sphere diameter: 10 mm) is hyphenated downstream to the monolith section. In the present study, the global assemblage of the three sections is taken into account. This low column-to-particle-ratio is chosen as the base condition as it reflects an experimental setup representative of that studied by Ding et al. 7,8 Larger ratios could mn the risk of inducing depth filtration and capture of particles in the pre- and post-distributors, which may not be desirable. As a first approximation, a two-dimensional symmetric domain is considered, providing a simplified scenario to get insight into the packed-bed-induced maldistribution flow characteristics in monolithic structured reactor. The computational geometry is schematically shown in Figure 3.3. solid ph<rse Gas phase Fi gure 3.3 T wo-dim ensional computational geometry with the assemblage of three-secti on stru cturedln on- structured packings (yellow . line, 2D symmetri c pl ane) 57 The solid volume fraction at the inlet is given by13, assuming homogeneous flow: (23) where u g is the superficial gas velocity, and G.I' is the solids mass flux. Flat velocity profiles are set as inlet boundary conditions for gas and suspended phases, which are calculated as follows: (24) (25) where u)s the axial interstitial solids velocity, and u g is the axial interstitial gas velocity. A no-slip condition is used for aIl the impermeable walls. The model equations are solved in steady state using commercial software Fluent (version 6.3). The porosity distribution model and the interphase momentum exchanges are implemented via user defined functions. The second-order upwind scheme is used for the convection terms of momentum equations. The velocity-pressure coupling is treated using the SIMPLE algorithm. 3.6 Results and Discussion In this work, an attempt is made to investigate gas-solid two-phase flows through the aforementioned composite monolith geometry. The gas continuous phase considered is Table 3.1 Basic simulation conditions used in thi s work Parameter Value Operation pressure (Pa) l .lE5 Gas phase (air): - Density (kg/m 3) 1.225 - Vi scosity (kg/m-s) 1.7894E-5 - Inlet velocity (m/s) 2.]4 Solid ph ase (biomass particles): - Den sity (kg/m 3) 450.0 58 - ParticJe diameter (m) 55E-6 - !nlet velocity (mis) 2.14 - !nlet solid volume fraction 6.4269E-4 Upstream packed-bed section: -Length (m) 0.300 -Colurnn diameter (m) 0.050 -ParticJe diameter (m) 0.010 Central monolith section: -Length (m) 0.600 -Diameter (m) 0.050 -Channel size (m) 3.0E-3 -Pitch (m) 3.3E-3 Downstream packed-bed section: -Length (m) 0.100 -Co]urnn diameter (m) 0.050 -Particle diameter (m) 0.010 air and the sol id suspended phase is biomass particles. The size of solid particles is 55 J.lm. The basic simulation parameters used in this work are listed in Table 3.1. 3.6.1 Modeling of two-phase flow behavior in monolith structured packings Figure 3.4 shows the radial solid mass fluxes of the suspended phase in different packing sections (z (m) < 0: lower packed bed; 0 < z < 0.6: monolith block; z > 0.6: upper packed bed). It can be seen that the suspended particles are distributed unevenly across the monolith assemblage cross-section. The highest biomass solids flux takes place at r = 0.0165m, i.e., 0.66 x column radius. It is very close to the value reported experimentally by Ding et al. 8 for glass beads using a positron emission particle tracking technique. In their work, it was found that the dominant peak of solid concentration occurs in the annular region around rIR ~ 0.7. Although the suspended particles (biomass particles) used in this simulation differ from those used by Ding et al. 8, it can be conservatively concluded that the model has the ability to capture non-uniform distribution features of solid phase flow in monolithic structured packing, definitely within what may be regarded as qualitative approximations. In addition, it can also be seen that for the packed-bed sections, there is a dominant peak at . a position close to the wall, which corresponds to a main feature of the measurement in gas-solid flow through packed bed. 23 ,24 This suggests that our model can be used to capture the main features from the experimental findings. 59 10.-------~------~------~------~----__, ~ 8~ ····· · · ····· · ····· ··~ · · · · · · · · · ··· · · · · · · ····~ · ·· ···· .. .• • •• . • . . • .. :.- . •• . ...... ~ .......• .:. . . .. . . . . .. ....... .. .. ·1 NE 0, ~ x - - - z=-0.25m --z=-0.10m +. z= 0.40m --z=0.65m 6 :::l ~ CfJ CfJ cu E 4~ · .. ·.. ·.... .. · ...... :.. ·.. ·.. · .... ·.. .... .. ; .. ·····.. .... ·.. ·.. ·..; .. ·.. · .. ·1 Q) CfJ cu ..c 2~ .. ·· .. · ·· · .. ·· ~ .. ··.... ·.. ·.. ·.. .. ·.. ~ .. ·........ r · :~ · .. ..:.. · ,' "" ,~,,· ~ · D~·11 ·~ .. · .... ·· .... ·· .... 1 0... ~ o (J) 0.005 0.000 0.010 0.015 0.020 0.025 Radial coordinate (m) Figure 3.4 Solids biomass flux of su spended phase in different packing sections 16 Ci) 14 N E 0, ~ x 12 -1 .. .... .. ... .... en en cu E z=-0.25m ---- z=-0 .1Dm z= 0.40m ----....-z=0 .65m 10 :::l ~ . --- 8 6 Q) en cu 4 ..c 0.. <il cu <.9 2 0 0.000 0.005 0.010 0.015 0.020 Radial coordinate (m) Fi gure 3.5 Gas mass flux es mirroring Fi gure 4 simulations 60 0.025 Figure 3.5 shows the gas mass fluxes in different packing sections. It is shown that there exists a very strong near-wall channeling for the gas flow in the lower fixed-bed random packing. Due to the block effect of monolith structured geometry and no-slip effect from wall, the packed-bed induced maldistribution for gas phase is reduced to ·a great extent in the monolith section. However, the gas-phase maldistribution in different channels is still remarkable. The channel adjacent to the column wall is responsible for significant transport of the gas phase. Figure 3.6 compares the radial profiles · at z = +0.4 m of solid and gas velocities as weIl as the variation of solid volume fractions inside the monolith channels. It can be seen that the non-uniform distribution characteristics for the gas and solid phases are completely different. The non-uniform distribution of the solid phase is evident, as reflected by both the sol id velocity and the solid volume fraction. For the solids phase, 0.012 6 0.010 - gas-phase velocity -<>- soli d-phase velocity -0- Cf) Q. 0.008 ~ < o C- solid volume fraction 0.006 ~ 0.004 g n> o :::J 0.002 o -+---M-....--~-----r4l~~-~-~-~-...---l~-~ 0.000 0.000 0.005 0.010 0.015 0.020 0.025 Radial coordinate (m) Figure 3.6 Channel dependence of gas-phase veJocity, solid velocity, and solid holdup (z= D.4m) the highest solid velocity can be found in two channels which are located two-channel away from the column wall. It is interesting to note that these two channels also correspond to the highest solid volume fraction, contributing to the largest solid transport 61 capacity. It is also observed that the velocities of the gas phase are generally several times , higher than those of the solid phase. 3.6.2 Comparison of gas-solid two-phase flow with single~phase flow It is of interest to compare the two-phase and single-phase flow behaviors. For comparison purposes, two simulation cases for single-phase flow are considered in this work. In the first case, the random packing is not taken into account. That is, no packing Îs arranged in the upstream and downstream fixed-bed sections. In the second case, the packings in fixed-bed sections are enabled. Compared to the two-phase flow, the singlephase flow corresponds to a two-phase flow with zero solids volume fraction. Figure 3.7 shows the variation of gas-phase velocity with the radial coordinate at different axial locations. It can be seen from Figure 3.7(a) that for disabled random-packings (empty 5.0 - - z =-O.25m -z=-O.10m .... z= 0.40m 4.5 en 4.0 Ê 3.5 .è 3.0 ID 2.5 ID 2.0 'u 0 > en ~z=O . 65m .~ E 1.5 ~ 1.0 cu W, 0.5 0.0 0.000 0.005 0.010 0.015 0.020 Radial cbordinate (m) (a) single-phase si mulati o n wüh out rando m packin gs 62 0.025 30 r---------------------------~ 25 U> E è 20 ·0 o ~ -z=-O.25m -z=-O.10m __ A,.">.~_ z= 0.40m 15 ID - - z = O.65m CI) .~ 10 E cu ~ êi5 5 0.000 0.005 0.010 0.015 0.020 0.025 Radial coordinate (m) (b) single-phase simulation with random packings Figure 3.7 The gas-phase velocity in single-phase f10w simulation pipe), the .distribution of gas velocity in the monolithic _channels is almost uniform, justifying that in most cases only one single-channel simulation is performed in literature to predict/represent the behavior of whole monolith. However, when the empty parts of the upstream and downstream sections are packed with non-structured packings of spheres, uniformity is broken. As shown in Figure 3.7(b), the monolith gas flow distribution is susceptible to the upstream maldistribution. As a comparison, Figure 3.8 shows the gas-phase velocity profiles under singlephase/two-phase simulation conditions. From this figure, single-phase flow with random packings brings about most serious anisotropic characteristic of flow in the monolith channels. And the introduc;tion of solid phase can mitigate seriousness of maldistribution to sorne degree; however, the problem of solid-phase flow distribution will arise (as shown in Figures 3.4 and 3.6). Compared to single-/two-phase flow with non-structured packing, the disabled packings (empty pipe) offer much uniform distribution of gas-phase flow in the monolith channels, except sorne mal distribution in the channel closest to the colurnn wall due to the empty pipe velocity profile. 63 16 en E 12 ~ two-phase flow (with nonstructured packirY,;)s) ---v- single-phase flow (with nonstructured packings) 8 4 o ,,,',.w_ single-phase flow (without nonstructured packings) t\À Al\ &~~f~l;\ / \j -ptW 'f 0.000 0.005 , r~~<O~~l , (ij~ 0.010 0.015 ,~~~ë~~~~h~Zr 0.020 0.025 Radial coordinate (m) Figure 3.8 Comparison of gas-phase velocities under single-phase/two-phase simulation conditions (z=OAm) with and without the nonstructured packjngs 3.6.3 Effeet of downstream-seetion paeking mode on flow distribution in monolith It is also of interest to investigate whether or not the downstream-section packing mod~ has any noticeable impact on flow distribution in monolith section. In this work, comparative simulations have been performed to examine three different cases with different packing modes for the downstream section. The first case is default one in which the downstream section is packed with particles of 10mm in diameter. The second case is a null packing (empty pipe) which allows the clear fluids to pass in this section. The last case corresponds to a non-homogeneous composite packing mode. In this case, while maintaining the larger packed particles (diameter=10mm) in the upstream section, smaller particles of 5.0mm in diameter are considered in the downstream-section packing, being entailed with higher flow resistance as opposed to the aforementioned two cases. It is noted here that in aIl these cases the upstream packing modes are kept same ( i.e., 64 Ci) 5 ~downstream section (dp=10mm particle) - ;)- downstream section (empty pipe) ---6-downstream section (dp=5mm particle) E è o 4 ~u Q) > 3 en crs 0') Q) en 2 .~ E crs ~ (jj 04--4~--~~--~~-r~~--~6---~~~~~~~~ 0.000 0.005 0.015 0.010 0.020 0.025 Radial coordinate (m) (a) Comparison of the gas velocities .-------------------------~------------------------, ~downstream section (dp=10mm particle) ---6-downstream section (empty pipe) -<>-downstream section (dp=5mm particle) 0.008 -downstream section (dp=10mm particle) --A.-downstream section (empty pipe) ,,-,""·'~· downstre am section (dp=5mm particle) 0.006 Cf) Q. ë2 en en crs E Q) 0.004 o< C 3 CD 2 en 0) crs Q. .c o· Q. ~ 0.002 ~ "0 Cf) 0.000 0.005 0.015 0.010 0.020 0.025 Radial coordinate (m) (b) Compari son of the solid mass fluxe s a nd the solids holdups Figure 3.9 Compalison of monolith-section flow distribution characteri stics (z=OAm) with and without the nonstructured packing in the downstream section 65 randomly packed with lOmm-in-diameter particles). Figure 3.9 shows the detailed comparison between the three packing modes for the gas and suspended solid phases. As in Figure 3.9a, there are sorne appreciable differences in gas-velocities of the three cases for the two channels near the column wall. However, the overall change of gas velocities for the whole multichannel system is generally insignificant. In addition, the comparison of solid-phase mass fluxes and solids pynamic holdups is further performed. Figure 3.9b depicts a fairly high-degree matching of solid flow distribution characteristic between the three cases. These simulated results indicate that the effect of downstream-section packing mode on monolith maldistribution characteristics is generally negligible under our simulation conditions. 3.6.4 Erreet of partiele size or nonstruetured paekings on flow eharaeteristies in monolith The size of the particles in the packed-bed sections affects .not only the near-wall channeling phenomena but also the pressure drop of reactor system. To reasonably select the particle sizes for a composite monolith, the effect of particle size in the nonstnlctured packing sections on the flow characteristic in monolith section needs to be understood. In this section, the influence of particle size of nonstructured packings on the monolith maldistribution characteristic and the pressure drop is systematically investigated. In our simulation, two scenarios characterized by nonuniform and uniform radial porosity distributions are taken into account, as shown in Table 3.2. For the nonuniform-porosity scenario, the radial porosity distribution is assessed by Mueller' s correlation and three particle sizes (diameter= 1Omm, 5.0mm, and 2.5mm) are considered to be packed in the upstream and downstream sections. Here, the three simulation cases are referred to as DplO_Mueller, Dp5.0_Mueller, and Dp2.5_Mueller for brevity (referring to Table 3.2). In our uniform-porosity scenario, a mean porosity is used as input, which is determined by averaging the radial porosity distribution based on Mueller' s correlation (for dp=lOmm). To further gain insight into the contribution mechanisms in interphase momentum interactions, the inclusion and exclusion of the phase interactions between the stationary packing and the flowing solid (or gas ) phase are considered here for decoupling mechanism contribution purpose. As in Table 3.2, three simulation cases 66 (labeled as DplO_GlS l, DplO_GlSO, and DplO_GOSO) are considered for the uniformporosity scenario. It is noted here that the number 'l ' and '0' denote inclusion and exclusion, respectively. And 'G' and 'S ' den ote interphase interaction between gas phase and packed phase and interphase interaction between suspended solid phase and packed phase, respectively. Table 3.2 presents the detailed comparison of pressure drop and flow maldistribution in monolithic channels for the two scenarios (nonuniform radial porosity distribution and uniform radial poro sity di stribution ). As shown in thi s table , the press ure drop contributions from different packing sections, the monolith maldistribution factor and flow factors for each phase are calculated as the comparison indexes. The relevant details of the channel locations and the centerline-based pressure sampling in the three-section monolith system are graphically illustrated in Figure 3.10. The effect of particle size is 0.7 ! '! 0 .6 ti) t~ i l' ~ '1' t? i 9 ~il i i i 0 .5 l i 1 ê) 1 ! ! i ! ! ! 1 E 0.4 i Q) êâ c =0 0 0 0.3 -'c 0 ëa c =0 :ê 0.2 0 0 0 M 0 0 i -'c Q) (J.) (J.) c cu c cu .c. ! ! .c. 0 0 0 0 -'c -'c (J.) c cu .c. ~ 0 0 i ~ & Ç•) 1 1 i i i i ! ! 1 -0.1 i -0 .2 ,,~ 0 Q) c cu .c. o I~ •• cc I~ • ! 1 i i ~ i {!\ 0 1 ~ 1 1 i A'.~ ," ~~. -0 .3 0 -'c :i,...0, 1 t!} ~' c cu .c. w o i 1 0.1 '& c ...J (J.) c cu .c. c cu .c. N -'c -'c (J.) ~ 0 0 ~ Il) 0 .005 0.0 15 0.01 0.02 0.025 Radial coordinate (m) Fig ure 3.10 Detai ls of the channel locati ons and centerl ine-based pressure sampling in the th ree-section monolith system 67 evaluated in case of nonuniform radial porosity distribution. It can be seen from Table 3.2 that that the reduction of particle size results in the increase of bed pressure drops in the upstream and downstream packed sections, as expected. The pressure drop in the . monolith section is found to increase as the particle size decreases. Due to the near wall channeling, the differences in pressure drop between the near-wall region and the bulk region can be observed in the three packing sections. These differences are magnified in case of employing packed particle of larger size, corresponding to the lower column-toparticle ratio. The . decrease of particle size will bring a positive contribution to the improvement of the overall flow maldistribution for the gas phase, as indicated by the decreasing trend iQ gas-phase maldistribution factor. The gas-phase flow factors in monolith channels demonstrate specifically the contribution of reducing particle size in suppressing the severity of the near-wall channeling. As compared to the gas phase, the effect of decreasing particle size on the maldistribution behavior is not so evident for the solid phase. With decreasing the particle size, the value of solid-phase maldistribution factor increases first and then decreases again, showing a non-linear variation relationship. The solid-phase flow factors in monolith channels also disclose the details of the nonlinear change, including both the change in peak magnitude and the migration in peak location. Note that reduction in particle size in the upstream and downstream packed beds would increase the risk for these sections to plug with biomass particles. Since the filtration ability of the beds was not included in the model analysis, it is believed that smaller particle beds would exhibit different maldistribution behaviors as the permeability of the bed could evolve with biomass particle capture. Averaging the radial porosity distribution of the DplO_Mueller case leads to a mean porosity value of 0.4439, which is used as the input of porosity for the uniform-radialporosity simulation cases (DpIO_GISI, DpIO_GISO, and DpIO_GOSO). It is interesting to make a direct c~mparison between the DpIO_GISl and DplO_Mueller simulation cases. In practice, the two cases represent different methodologies in treating the nonuniformity of radial porosity distribution. From physical viewpoint, the latter takes into account the radial nonuniformity in porosity distribution while the former neglects this kind of nonuniformity by simplifying it as flat distribution. It can be seen in Table 3.2 that the DpIO_GISI simulation case with the uniform-radial-porosity assumption can 68 Tab le 3.2 Effect of particle size and radial poros ity distribution of nonstructured packings on the tlow characteristics in monolith Nonuniforrn radial porosity distribution Uniform radial porosity distribution Dp lO_Mueller Dp5.0_Mueller Dp2.5_Mueller DplO_GIS I DplO_G l S0 DplO_GOSO Empty-pipe -5143.8 -15417. 1 -42 157.9 -7065.4 -6481.3 -23 .8 -24.2 Pressure drop (Palrn) Average pressure drop - upstrearn section - mono li th secti on -750.8 -770 .2 -949 .8 -855.4 -856.9 -840.2 -840.2 - downstream section -5297.5 - 15335 .0 -42338.8 -7035 .0 -6460.0 -31.3 -30.0 -4743.3 - 15300.0 -42256 .7 -7086.7 -6500.0 -36.7 -36.7 Wall-adjacent pressure drop - upstream section - monolith section -898,3 -831.7 -948 .3 -843 .3 -846.7 -835 .0 -835 .0 - downstream section -5100.0 - 15360.0 -42380.0 -7060.0 -6480.0 -30 .0 -30.0 - solids phase 0.298 0.353 0.220 0.077 0.065 0 .030 0.024 - gas phase 0.1 2 1 0.109 0.080 0.031 0.027 0.009 0.007 Maldistribution Factor Flow Factor - solids phase: channel_OO 0.822 0.729 0.760 0.973 1.006 1.052 1.062 channel_Ol 0.522 0.654 0.746 0.992 0.992 1.136 1.094 channel_02 0.628 0.653 0.731 0.992 1.005 1.058 1.008 channeL03 0.678 0.690 0.776 0.995 1.026 0 .972 0.957 channel_04 2.122 1.072 0.976 l.018 1.099 0 .931 0.958 1.0 16 ch annel_05 2.538 3.400 2. 180 1. 136 1.240 0 .967 chann el_06 0.388 0.602 1.639 1.338 1.04 1 1.021 1.028 chann el_07 0.301 0.200 0.191 0.556 0.591 0 .863 0.877 channel_OO 0.865 0.896 0.956 0.997 0.990 0 .989 0.985 channel_OI 1.069 0.946 0.966 0.991 0.994 0.962 0.974 channel_02 1.034 0.965 0.969 0.99 1 0.991 0.983 1.000 channel_03 0.897 0.934 0.975 0.990 0.984 1.010 1.016 - gas phase: channel_04 0.7 35 0.815 0.928 0.983 0.963 1.023 1.014 channel_OS 0.699 0.655 0.810 0.948 0 .925 1.007 0.990 channel_06 0.9 13 1.098 0.858 0.899 0.976 0.985 0.983 channel_07 1.787 1.692 1.540 1.202 1.177 1.042 1.037 • These values are calcu lated/a veraged using the multichannel centerline sampli ng data (see Figure 10 for the geometri cal detail s) 69 . -~ render a remarkably improved flow distribution characteristics for both the phases, as opposed to its counterpart (Dp 1O_Mueller). However, judging from the distribution of two-phase flow factors , it is found that the near-wall higher loading transport for the gasphase as weIl as the appearance of dominant peak in the solid-phase mass flux are still noticeable even in this ideal case. To further get insight into the phenomena, decoupling interphase interaction is attempted in this work to observe the evolutionary change in ' 2 .2 ·.'~?1.' '~.8.2; .i", 3.3 5 .' :?.,'~ 4 .6 ;~ ..;:.,. 1.7 ,ù ' 1.1 ~w 0.7 0 .6 0 .0 OA 0 .0 .," 1.5 40 ::H'I n lA 0 .0 ~ ,::}: 2.3 ""'<' 2 .2 1.6 :,",;'.',.:.·.·,.'.'•• :.·•.'.·1 · ; : ~ .; 2 .5 ~:~~..; il· ~~;; ~ . ~ ,.:, ~,:. :.r,:. 1 ~: ~ 7 ,":;': 0. 0 .5 ·,:,:':;, 0 .3 0 .2 0. 0 :.• .,:'~.I g ~: !jll . 1.9 2 .2 1 1 1 1 1 illi :\1 .• ••..•.,'••.,.• :.,. •. Ji 0 .3 0 .0 " ':,:.•'.,'::': '".•••.." . 1.5 1.3 1.7 1.0 0 .8 1 i} ';:::) J'; ~.~ 0 .2 0 .0 Uniform radial porosity distribution Nonunlfo,m ,ad"" pomsll. d'st"buUon Fi gure 3.11 Effect of parti cie size and porosity radi al di stributi on on the solid mass flux distribution in the composite monolith system flow distribution. To this end, the additional simulation cases (Dp 10_G 1SO and Dp10_GOSO) are introduced for comparison purpose. By comparing the three simulation cases, one may conclude that the interphase interaction mechanisms have an 'incremental' contribution in affecting the flow maldistribution; and . the graduaI exclusions of the interphase interactions lead to improved flow distribution for both the phases. FinaIly, we present the global contour comparison of the solid-phase mass flux in the composite monolith for aIl these simulation ,c ases (with/without considering nonuniformity in radial porosity distribution), as shown in Figure 3.11. The null-packing case 70 (empty pipe) is also presented here. Comparing with Dp10_00S0, the null-packing case can be regarded as a variant of Dp 10_OOSO with the porosity value of 1.0. As indicated in Figure 3.11, the nonuniformity of flow distribution in monolith structured packing section is to a great extent imprinted with the unique flow characteristics in the upstream nonstructured (random) packing section. 3.7 Conclusions In the present work, a process concept for biomass gasification is firstly proposed which involves the hybridization of gasificationlcombustion reactions in a monolithic structured reactor by using high-temperature phase-change material to intensify the process heat management. Following the proposaI of this concept, a computational fluid dynamics model is developed to investigate the gas-solid (biomass particles) two-phase flow distribution characteristics in monolithic structured packings. This model is based on Eulerian-Eulerian multifluid modeling approach with closure laws according to the kinetic theory of granular flow. An assemblage of monolithic structured packings with non-structured packed-bed sections is fully considered in our simulations with a view to effectively characterizing the flow maldistribution. The non-structured random packedbed sections are treated as porous media by implementing the non-uniformity of radial porosity distribution and the interphase interactions through user defined functions (UDFs). The numerical investigation was carried out to systematically explore the two-phase flow distribution in the three-section composite monolith system. The simulation results indicate that there exists a very strong near-wall channeling for the gas-phase flow in the fixed-bed random packings. The suspended particles are distributed unevenly across the monolith assemblage cross-section and the highest biomass solids flux takes place at a dimensionless column radius 0.66. The effect of downstream-section packing mo.des on monolith maldistribution characteristics can be generally consideredas negligible under our simulation conditions. The reduction of particle size in the non-structured packing sections results in an increase of pressure drop in the monolith section. In addition, 71 decreasing particle sizes leads to a positive improvement in the overall flow distribution for the gas phase and a nonlinear variation trend in flow maldistribution for the solid phase. Compared to the nonuniform-radial-porosity assumption, the uniform-radialporosity assumption provides considerably improved flow distribution characteristics for both the gas and solid phases. The interphase interaction mechanisms are found to exhibit "incremental" contributions in affecting the flow mal-distribution. GraduaI exclusion of the interphase interactions leads to an improved flow distribution for both the phases. The non-uniformity of flow distribution in monolith structured packing section is shown to be imprinted to a great extent with the unique flow characteristics in the upstream nonstructured packing section. The simulation results indicate the ability of CFD models to capture the non-uniformities of the flow pattern in monolithic structured packing, which we believe will further aid our development of the process concept. Acknowledgement Financial support from the "Chaire de recherche du Canada en procédés et matériaux pour des énergies durables" of the N aturai Sciences and Engineering Research Council (NSERC) is gratefully acknowledged. Nomenclature a constant in Eq.(l) ag volume fraction of phase g aq volume fraction of phase q as volume fraction of phase s b constant in Eq.(l) or coefficient in turbulence model Cf) drag coefficient dp diameter of packed particle, m d~ diameter of suspended parti cIe, m 72 D diameter of column, m D* hydraulic diameter of a packed bed, m e,fS restitution coefficient of particle collisions fk interaction coefficient between suspended phase and packed particles F,. Froude number g gravitational constant, m2 /s go radial distribution function of interparticle collisions G,~ flux of suspended phase, kg/m s 1 unit tensor 12D second invariant of the deviatoric stress tensor Jo zero th order Bessel function k e, diffusion coefficient for granular energy, kg/s m 2 mi . theoretical uniform flow-rate of the qth phase in ith channel, kg/s miq realistic mass flow-rate of the qth phase in th channel, kg/s M fq maldistribution factor of the qth phase in ith channel N eil number of channels Nq number of phases p fluid pressure, N/m 2 Ps solid pressure, N/m r radial coordinate, m r* dimensionless radial coordinate, [-] Re~ relative Reynolds number Sg source term due to gas-packing interaction O 2 73 s~ source term due to particulate-packing interaction time, s Ug axial interstitial gas velocity at the inlet, mis Us axial interstitial solids velocity at the inlet, mis Üs superficial suspended solids velocity vector Vg superficial' gas velocity Vg velocity of gas phase, mis Vq velocity of phase q, mis Vs velocity of solid phase, mis v~ particulate fluctuating velocity, mis Greek letters f3 combined coefficient of interphase momentum exchange, kg/m 3 s f3gs coefficient of interphase momentum exchange, kg/m 3 s f3Erglln fluid-solid interaction coefficient of the Ergun equation, kg/m s f3wen -: YII fluid-solid interaction coefficient of the Wen-Yu equation, kg/m s E voidage of packed bed EB constant in Eq.(l) qJgs switch function YB, collisional dissipation of energy,kg/s 3 m li flow factor of the qth phase in {h channel À\ bulk v.iscosity of solid phase, Pa s f.1 g shear viscosity of gas phase,Pa s 3 3 74 J1 s shear viscosity of solid phase, Pa s (}ç granular temperature,m 2/S2 Pg density of phase g, kg/m 3 Pq density of phase q, kg/m 3 Ps density of phase s, kg/m 3 Tg gas stress tensor, N/m2 Ts solid stress tensor, N/m 2 Subseripts g gas phase i-th channel q q-th phase s solid suspended phase Literature Cited (1) Kobayashi, N., Guilin, P., Kobayashi, J., Hatano, S., Itaya, Y., Mori, S., A new pulverized biomass utilization technology, Powder Teehnology, 2008,180(3):272-283. (2) Florin, N. H., Harris, A. T., Enhanced hydrogen production from biomass with in situ carbon dioxide captùre using calcium oxide sorbents, Chem. Eng. Sei., 2008, 63(2):287 -316. (3) Levenspiel, O., What will come after petroleum?, Ind. Eng. Chem. Res., 2005, 44(14): 5073-5078. 75 (4) Pletka R, Brown RC, Smeenk J. Indirectly heated biomass gasification using a latent heat ballast.l:experimental evaluations, Biomass & Bioenergy, 2001, 20(4): 297305. (5) Pletka R, Brown RC, Smeenk J. Indirectly heated biomass gasification ' using a latent heat ballast. Part 2:modeling, Biomass & Bioenergy, 2001, 20 (4): 307-315. (6) Potic, B., Kersten, S. R. A., Ye, M., van der Hoef, M. A., Kuipers, J. A. M., van Swaaij, W. P. M., Fluidization with hot compressed water in micro-reactors, Chem. Eng. Sei., 2005, 60(22): 5982-5990. (7) Ding, Y. L., Wang, Z. L., Ghadiri M, Wen, D. S., Vertical upward flow of gas- solid two-phase mixtures through monolith channels, Powder Teehnology, 2005,153 (1): 51-58. (8) Ding, Y. L., Wang, Z. L., Wen, D. S., Ghadiri, M., Fan, X. F., Parker, D., Solids behavior in a di lute gas-solid two-phase mixture flowing through monolith channels, Chem. Eng. Sei., 2006,61(5): 1561-1570. (9) Mueller, G. E., Radial void fraction distributions in randomly packed fixed beds of uniformly sized spheres in cylindrical containers, Powder Technology, 1992,72: 269275. (10) Giese, M., Rottschafer, K., Vortmeyer, D., Measured and modeled superficial flow profiles in packed beds with liquid flow, AIChE Journal, 1998,44(2):484-490. (11) Lun, C. K. K., Savage, S. B., Jeffrey, D. J., Chepumiy, N., Kinetic theories for granular flow: inelastic particles in Couette flow and slightly inelastic particles in a general flow field, J. Fluid Mech., 1984,140:223-256. (12) Gidaspow, D., Bezburuah, R., Ding, J., Hydrodynamics of circulating fluidized beds, kinetic theory approach. In: Potter OE, Nicklin th DJ (Eds.), Fluidization VII, Proceedings of the 7 Engineering Foundation Conference on Fluidization, New York, 1992, pp.75-82. 76 (13) Gidaspow, D., Multiphase Flow and Fluidization: Continuum and Kinetic Theory Descriptions. Academic Press, New York, 1994. (14) Wen, C. Y., Yu, Y. H., A generalized method for predicting minimum fluidizationvelocity, AIChE Journal, 1966,12(3): 610. (15) Ergun, S., Fluid fIow through packed columns, Chemical Engineering Progress , 1952,48(2): 89-94. (16) Yamaoka, H., Mechanisms of hanging caused by dust in a shaft fumace, ISIJ International, 1991,31 :939-946. (17) Shibata, K., Shimizu, M ., Inaba, 5., Inaba, S., Takahashi, R., Yagi, J., Pressure loss and hold up powders for gas-powder two phase fIow in packed beds, ISIJ International, 1991,31: 434-439. (18) Van der Ham, A. G. J., Prins, W., Van Swaaij, W. P. M., Hydrodynamics of a pilot-plant scale regularly packed circulating fluidized bed, AIChE Symp. Ser. , 1993,89: 53-72. (19) Dong, X. F., Zhang, S. J., Pinson, D., Yu, A. B. , Zulli, P., Gas-powder flow and powder accumulation in a packed bed: II-numerical study, Powder Technology, 2004,149:10-22. (20) Boremans, D., Rode, S., Wild, G. , Liquid fIow distribution and particle-fluid heat transfer in trickle-bed reactors: the influence of periodic operation, Chemical Engineering and Processing, 2004,43: 1403-1410. (21) Hoek, P. J. , Wesseling J. A., Zuiderweg, F. J. , Small scale and large scale liquid maldistribution in packed columns, Chem. Eng. Res~ Des., 1986, 64:431-449. (22) Marcandelli, C. , Wild G. , Lamine, A. S., Bernard, J. R., Liquid distribution in trickle-bed reactors, oil and gas science and technology - rev, IFP, 2000,55:407-415. 77 (23) Ding, Y. L., Wang, Z. L., Wen, D. S., Ghadiri, M., Fan, X. F., Parker, D., Solids behaviour in a gas-so1id two-phase mixture flowing through a packed partic1e bed, Chem. Eng. Sei., 200S,60( 19):5231-5239. (24) Liu, S. Q., Ding, Y. L., Wen, D. S., He, Y. R., Modelling of the behaviour of gas-so1id two-phase mixtures flowing through packed beds,Chem. Eng. Sei., 2006,61: 1922 -193l. 78 Chapter 4 Conclusions and Recommendations 4.1 General conclusions Biomass is a very important renewable energy source and it holds great potential for sustainable energy conversion processes. With the depletion of fossil fuel sources as well as the concem over the evolving global warming, there is considerable worldwide interest in exploiting the utilization of biomass renewable energy sources, particularly because it is believed that energy obtained from biomass has a carbon-neutral cycle. This situation calls for the development of a biomass-based but energy efficient and environment friendly system with better environmental acceptability. Gasification to produce biosyngas is regarded as one of the most promising options for biomass beneficiation. However, thermodynamics and intrinsic kinetics dictate that endothermic biomass gasification reactions have to be carried out at high temperatures, which demands efficient heat supply and recovery policy. The concept of allothermal gasification offers an attractive solution for implementing high-temperature reactions by coupling strongly endothermic reactions with exothermic reactions. However, implementing the concept in practice is not straightforward. In this work, an innovative biomass gasification process concept is proposed, which involves coupling of the gasification and combustion processes in monoliths with hightemperature phase-change-material. The proposed process is implemented in a monolithic structured reactor and intensified by periodic operation mode. To effectively' design and optimize . this novel process, knowledge from different important fields (including biomass gasification, monolith reactor engineering, high-temperature phase change material, and gas-solids fluidization) is required. Among them, modeling and understanding of gas-solid (biomass particles) flow hydrodynamics in monolithic structured reactor is very important, in view of the complexity of two-phase flow in structured packings. This work relates to our understanding of the hydrodynamics, as deduced from Euler-Euler computational 79 fluid dynamics (CFD) modeling. A computational fluid dynamics model is developed to investigate the gas-solid (biomass particles) two-phase flow distribution characteristics in monblithic structured packings. This model is based on Eulerian-Eulerian multifluid modeling approach with closure laws according to the kinetic theory of granular flow. A three-region composite monolithic structured reactor is con'sidered in our simulations with a view to effectively characterizing the flow maldistribution. The non-structured packed-bed sections are treated as porous 'm edia by implementing the non-uniformity of radial porosity distribution and the interphase interactions through user defined functions (UDFs). The numerical investigation is carried out to systematically explore the two-phase flow distribution in the three-section composite monolith system. The simulation results indicate that there exists a very strong near-wall channeling for the gas-phase flow in the fixed-bed random packings. The suspended particles are distributed unevenly across the monolith assemblage cross-section and the highest biomass solids flux takes place at a position of 0.66 times the column radius. The effect of downstream-section packing modes on monolith maldistribution characteristics can be generally negligible under our simulation conditions. The reduction of particle size in the nonstructured packing sections results in an increase of pressure drop in the monolith section. In addition, decreasing particle size leads to a positive improvement in the overall flow maldistribution for the gas phase and a nonlinear variation trend in flow maidistribution for solid phase. Compared to the nonuniform-radial-porosity assumption, the uniform-radial-porosity assumption can render considerably improved flow distribution characteristics for both the gas and solid phases. The interphase interaction mechanisms are found to have an 'additive' contribution to affect the flow maldistribtution and the graduaI exclusions of the interphase interactions lead to an evolutionarily positive improvement in flow distribution for both the phases. The nonuniformity of flow di stribution in monolith structured packing section is shown to be imprinted to a great ex~ent with the unique flow characteri stics in the upstream nonstructured packing section. The simulation results demonstrate the ability of CFD models to capture the nonuniformities of the flow pattern in monolithic structured packing, which we believe will further aid our development of the process concept. It is suffice to say that the present 80 model has been an important enabling step in the direction of our research on the process concept. 4.2 Recommendations for future investigations The following recommendations are made regarding the logical progression and continuance of the present work: (1) Future work is recommended to address the coupling of the gas-solid flow dynarnics with transient biornass gasification/cornbustion process in rnonolithic structured reactor by using the multichannel flow model; (2) Future work is recommended to introduce the present 2D modeling methodology to explore a full-scale 3D simulation of the unconventional composite monolith geometry, using improved computational resource; (3) Future work is recommended to investigate the turbulence effect and its impact on multichannel flow distribution characteristics in monolith structured reactors. 81