Aujourd’hui, la transplantation est le seul traitement efficace proposé aux patients souffrant d’une insuffisance hépatique fulminante. La nécessité de disposer d’un système de suppléance hépatique transitoire apparaît donc indispensable. C’est dans cet axe que se sont développés les systèmes qualifiés de foies bio artificiels (BAL). Leur principale caractéristique est d’incorporer un bioréacteur hébergeant des cellules pouvant restaurer l’activité hépatiques dans son ensemble. A l’heure actuelle, les hépatocytes primaire humains (HEP) issus de foies de donneurs non transplantables sont considérées comme le meilleur choix. Cependant, leur utilisation reste limitée par leur faible disponibilité et la difficulté à les maintenir différenciés en culture in vitro. Pour remédier à ce dernier point, l’approche la plus prometteuse semble être une co-culture des hépatocytes avec les cellules non parenchymateuses afin de recréer un environnement proche des sinusoïdes hépatiques. Ce travail de thèse repose sur la mise en place d’une nouvelle approche de co-culture tridimensionnelle sous la forme de sphéroïdes, d’HEP primaires avec les principaux types de cellules non-parenchymateuses (les cellules de Kupffer, les cellules endothéliales et les cellules étoilées) selon des proportions spécifiques. Puis de leurs encapsulations dans des billes d’alginate et leurs cultures au sein d’un bioréacteur à lit fluidisé. Ce modèle s’est révélé pertinent et approprié à maintenir les fonctions hépatiques dans le temps. Bien que beaucoup d’optimisation reste à définir, ce travail exploratoire témoigne de l’intérêt de cette approche intéressante pour le progrès des systèmes BAL. / Liver shortage makes transplantation inapplicable to all acute liver failure patients. Bioartificial Iiver (BAL) devices represent a temporary solution for these patients which are thereby bridged tilt Iiver transplantation or regeneration BAL treatment offers blood purification and substitution of metabolic functions through the activity of hepatocytes (HEPs), which are integrated in the device within acclimating containers, so-called bioreactors. Primary human hepatocytes are the ideal cell type to use in BAL, but they are scarcely available and difficult to maintain in vitro. Co-culture of HEPs with supporting cells has been proposed as the most promising strategy for preserving HEP behaviors in in vitro conditions. In fact, assisting cells types hold their ability to influence functional responses of the HEPs by providing them with cues of the native organ.This PhD work proposed a novel approach of co-culture for the functional sustain and preservation of the HEPs in the environment of the fluidized bed bioreactor (designed in our Iaboratory). Definition of this model took inspiration from the cellular organization in the organ; therefore, it employed three major sinusoidal non-parenchymal cell populations (liver sinusoidal, Kupffer, and hepatic stellate cells) which, together with HEPs, were cultured with three-dimensional arrangement (spheroids) and according to specific proportions. The resulting model was characterized in terms of functional benefits for the HEPs, and then applied in the microenvironment of alginate beads, which provide cells with immunological and mechanical protection in the fluidized bed bioreactor. This spheroidal multi-cultured model revealed its potentiality in sustaining in vitro HEP behaviors over time. Although much remains to be refined, this model may represent an interesting approach for the progress of BAL

La nécessité de nouveaux modes de suppléance hépatique se fait clairement ressentir pour maintenir en vie les patients en attente d’une greffe. Des traitements, basés sur l’utilisation de cellules cultivées dans un environnement adéquat, pourraient même permettre le rétablissement de certains patients en insuffisance hépatique sévère et ainsi éviter la greffe et les traitements immunosuppresseurs associés. Une suppléance hépatique efficace pourrait aussi servir au rétablissement des patients récemment greffés ou ayant subi une hépatectomie. L’objectif de cette thèse a été de développer, créer et valider un nouveau foie bioartificiel avec une prise en compte des contraintes réglementaires des médicaments combinés de thérapie innovante. Ce BAL doit pouvoir inclure des bioréacteurs à lit fluidisé perfusés par du plasma et contenant des sphéroïdes d’hépatocytes encapsulés. Des microparticules de verre, ajoutés aux billes d’alginate pour les alourdir, permettent d’obtenir une fluidisation optimale dans du plasma pathologique sans que des effets délétères ne soient observables ni pour les cellules ni pour les propriétés mécaniques des billes. Une méthode de culture cellulaire utilisant un revêtement anti-adhérent sur des boites de Petri en verre permet de produire un nombre important de sphéroïdes viables in vitro. Ces sphéroïdes encapsulés peuvent être maintenus vivants et métaboliquement actifs dans un bioréacteur à lit fluidisé pendant au minimum 4 jours.Pour assurer l’efficacité du BAL et la sécurité des patients, une circulation extracorporelle complexe a été mise au point pour être compatible avec une machine d’épuration extracorporelle commerciale, la Prismaflex® de la société Gambro, déjà utilisée en soins intensifs. Une étude préclinique sur un modèle ovin a montré que le traitement était bien toléré en ce qui concerne les aspects hémodynamiques. La prochaine étape concerne la mise en place d’une étude dans un modèle porcin d’insuffisance hépatique, avant de pouvoir procéder aux premiers essais cliniques. / The need for new liver support devices is clearly felt to allow keeping alive patients waiting for a transplant. Treatments, based on the use of cells cultured in an adequate environment, may even allow the recovery of some patients suffering from acute liver failure and avoid graft and associated immunosuppressive therapies. A hepatic substitution could also be used to reestablish patients recently transplanted or who had underwent an hepatectomy.. The objective of the thesis was to design, create and validate of a new bioartificial liver with consideration for the regulatory requirements of the Advanced Therapies Medical Product (ATMP). This device has to include fluidized bed bioreactors perfused with plasma and hosting alginate-encapsulated hepatocytes spheroids. Microparticles of glass have been added to weight down alginate beads in order to have an optimal fluidization in pathological plasma without negative effects neither on cells metabolism nor on mechanical properties of the beads. A cellular culture method using non adhesive coating in Petri dish led to the production of a large amount of viable spheroids in vitro. These encapsulated spheroids can be kept alive and metabolically active in a fluidized bed bioreactor during a minimum of four days. To ensure the efficacy of the BAL and the safety of patients, a complex extracorporeal circulation was designed to be compatible with a commercial medical device, the Prismaflex® monitor of the company Gambro, already used in intensive care units. A preclinical study on sheeps has shown that the treatment was well tolerated in terms of hemodynamics considerations. The next step is the establishment of a study in a porcine model of liver failure, before we can proceed to the first clinical trial.

In gas-solid fluidized beds, the generation of electrostatic charges due to continuous contacts between fluidizing particles, and the particles and the fluidization vessel wall, is unavoidable. Industrial operations, such as the production of polyethylene, are susceptible to significant operational challenges caused by electrostatics including reactor wall fouling, a problem known as “sheeting”. The formation of particle sheets can require shutdown periods for clean-up which results in significant economic losses. To gain a better understanding of the underlying mechanisms of electrostatic charging in gas-solid fluidized beds, in an attempt to eliminate or minimize this problem, a pilot-scale pressurized gas-solid fluidization system was designed and built, housing an online electrostatic charge measurement technique consisting of two Faraday cups. The system permits the study of the degree of particle wall fouling at pressures and temperatures up to 2600 kPa and 100°C, respectively, and gas velocities up to 1 m/s (covering a range including turbulent flow regime). The system also allowed, for the first time, the measurement of the fluidizing particles’ mass, net charge and size distribution in various regions of the bed, especially those related to the wall coating under the industrially relevant operating conditions of high pressures and gas velocities. Experimental trials were carried out using polyethylene resin received from commercial reactors to investigate the influence of pressure and gas velocity on the bed hydrodynamics and in turn, the degree of bed electrification. Mechanisms for particle charging, migration and adherence to the column wall were proposed. The size distribution of the gas bubbles shifted towards smaller bubbles as the operating pressure was raised. Thus, higher pressures lead to greater mixing within the bulk of the bed and resulted in a higher degree of particle wall fouling. Moreover, the extent of wall fouling increased linearly with the increase in gas velocity and as the bed transitioned to turbulent regime, due to the increase in particle-wall contacts. Bipolar charging was observed especially within the wall coating with smaller particles being negatively charged. Overall, particle-wall contacts generated negatively charged particles resulting in a net negative charge in the bed, whereas particle-particle contacts generated positively and negatively charged particles resulting in no net charge when entrainment was negligible. The formation of the wall layer and its extent was influenced by the gravitational and drag forces balancing the image force and Coulomb forces (created by the net charge of the bed and the metallic column wall as the attraction between oppositely charged particles).

La combustion en boucle chimique (CLC) est un procédé du type oxy-combustion où des particules sont utilisées pour fournir de l'oxygène à la combustion. Des études sont nécessaires pour l'extrapolation et l'optimisation du procédé CLC, fonction des propriétés des particules du groupe B et de la technologie CFB. Les études hydrodynamiques ont été faites dans un riser de 18 m de hauteur. Des profils axiaus de pression, ainsi que les profils radiaux de flux et de quantité de mouvement ont été obtenus. Trois types de particules ont été utilisées ayant un diamètre de Sauter entre 250 et 300 μm et une densité entre 2600 et 3300 kg/ m3. Un impact de la sphéricité des particules sur la perte de charge a été révélé. Dans des conditions identiques, les billes de verre génèrent des pertes de charge d'environ 50% inférieures à celles du sable. Dans la zone d'écoulement développée, la présence du régime cœur-anneau a été détectée. Un modèle hydrodynamique 1D du riser qui est à la fois fondé sur des données expérimentales et sur les équations gaz-solide Euler-Euler, a été développé. Une nouvelle corrélation pour la force de traînée moyennée sur la section est proposée. Une nouvelle corrélation des conditions limite dans la partie inférieure du riser a aussi été établie. Le modèle 1D final est en mesure de prédire la perte de charge du riser pour différentes conditions opératoires et en tenant compte des propriétés des particules, comme la densité, la taille et la forme. Une étude sur la pertinence de l'utilisation du logiciel Barracuda CPFD® pour simuler des particules du groupe B en régime de transport a été réalisée. Il a été montré que le code sous-estime la perte de charge pour le sable / Chemical Looping Combustion (CLC) is an oxy-combustion like process where particles are used to supply oxygen to combustion. Further work is still needed for extrapolation and optimization of the CLC process, concerning properties of Group B particles and CFB technology. Hydrodynamic tests were made on a 18 m tall riser. Axial pressure profiles as well as radial flux profiles and radial momentum quantity profiles were obtained. Three types of Group B particles were used with Sauter mean diameters between 250 and 300 μm and densities between 2600 and 3300 kg/m²s. An important impact of particle sphericity on riser pressure drop has been revealed. At identical conditions, glass beads present about half the pressure drop generated by sand. In the developed region of the riser, the core-annulus regime has been found. A 1D model of the riser, based on experimental results and on the Euler-Euler gas-solid equations, has been developed. Moreover, a new cross section averaged drag force correlation is presented. A new boundary condition on the bottom of the riser has been investigated. The final 1D model is capable of predicted riser pressure drop from the operating conditions and it takes into account particle properties such as density, size and shape. A study on the adequacy of the use of the commercial CFD code Barracuda to simulate risers with Group B particles was made. It was shown that the code under estimates pressure drop along the riser for sand simulations

This thesis is focused on the possibilities of the utilization of the fludized fly ash in ceramic technology, especially in the field of brick manufacturing and ceramic tiles. There were used fly ashes from thermal power plants in Hodonín, Ledvice and Tisová. Fluidized fly ash can be used for manufacturing of building materials. There hasn’t been done extensive research in this area. One of the problem is sulphur dioxide leakage during burning. The work describes the behaviour of fluidized ash during firing. The attention is focused on air pollution of sulphur dioxide. The production of SO2 is controlled so that the aim of the work is to find a way how to avoid leaking. It was proved that addition of sodium ions causes the formation of minerals hauyne and nosean during firing. Also is described the behaviour of the fly ash-soil mixtures. The anhydrite decomposition temperature is lower than in pure fly ash. In the end is fluidized ash applied for the production of dry pressed ceramic tiles. This ash can be used in ceramic technology. However, despite the interesting results there are number of problems that need to be solved.

Biomass gasification using air with addition of oxygen can be the right way to increase the quality of the syngas produced. This thesis aims to characterize the gasification process using oxygen-enriched air and verify its effect on the gas. The theoretical part deals with description of the technology with an emphasis on various gasifying agents and different possibilities of increasing gas quality. The practical part contains the results of measurements on the atmospheric fluidized bed reactor and their evaluation. The experiment was performed with various oxygen enrichment taking into account the capabilities of the equipment used.

This thesis aims to burn raw material with a lower content of carbonate component for the production of Portland cement. Substitution of carbonate component is done by fluidized fly ash. The model cements are tested of basic technological properties and the hydration process is observed.

This thesis consists of two parts discussing modelling of heterogeneous catalytic reactors. In the first one, an industrial prototype of a fluidized bed reactor serving as a hydrogen generator based on endothermic decomposition of formic acid is studied. After initial determination of the main reactor characteristics a system of nine con- stituents is derived and, consequently, reduced to a three phase flow. The solid and bubble particles immersed in a liquid are modelled by the Basset-Boussinesq- Ossen equation. Furthermore, an averaging technique is used to derive a three phase Euler-Euler model. Finally, numerical computations with a verification towards the measurements and a CFD analysis are proceeded. The second part discusses interfacial transport phenomena between a bulk and catalytic surfaces of a reactor mediated via the boundary conditions. The constitu- tive relations, that by construction comply with the second law of thermodynamics, follow from the specification of suitable thermodynamic potentials together with an identification of the bulk and surface entropy productions. The derived model is suitable for further analysis providing clear guidelines for the incorporation of the Langmuir-type adsorption model as well as other sorption models. Keywords: Heterogeneous catalysis, multi-phase...

In der vorliegenden Arbeit werden verschiedene Simulationsansätze angewandt um die Hydrodynamik in einem neu entwickelten Wirbelschichtvergaser zu untersuchen. Die Ansätze umfassen a) entdimensionalisierter Ähnlichkeitskennzahlen und empirischer Gleichungen, b) 1D Simulationen mittels ASPEN Plus®, c) 3D CFD Simulationen mittels Ansys Fluent® zur detaillierten Abbildung der zu erwartenden Hydrodynamik. Vor- und Nachteile der jeweiligen Ansätze sowie Klassen von ermittelbaren Simulationsdaten werden diskutiert. Ein Schwerpunkt der Arbeit liegt in der Identifizierung geeigneter Experimente aus der Literatur, auf Basis von Ähnlichkeitskennzahlen, um die Simulationen zu validieren. Die Vergasersimulationen zeigen, dass sich erwartungsgemäß ein aus hydrodynamischer Sicht gestufter Prozess ausbildet. Die entstehenden Zonen lassen sich als Festbett, blasenbildende Wirbelschicht, Jet-Wirbelschicht mit Rezirkulationszelle und strähnenbildende, zirkulierende Wirbelschicht identifizieren und entsprechen demnach dem Verfahrensanspruch.:1 INTRODUCTION 1 1.1 Market Situation 1 1.2 Objective Work 3 1.3 Structure of this Work 4 2 FUNDAMENTAL CONSIDERATIONS 5 2.1 Fundamentals of Gasification and Gasifiers 5 2.1.1 Counter-Current Fixed-Bed Gasifiers 7 2.1.2 Fluidized-Bed Gasifiers 9 2.1.3 Entrained-Flow Gasifiers 10 2.1.4 Technology Development Trends 11 2.1.5 Conclusion 12 2.2 Fundamentals of Fluidized-Bed Systems 13 2.2.1 Particle Characterization 13 2.2.2 Types of Fluidized Beds and Key Parameters 15 2.2.3 Fast-Fluidized Beds 18 2.2.4 Jetting-Fluidized Beds 19 2.2.5 Spouted Beds 24 2.2.6 Conclusion 27 3 APPROACHES TO ASSESS FLUIDIZED BEDS 28 3.1 Empirical Simulation 28 3.1.1 Nondimensional groups 28 3.1.2 Conclusion 36 3.2 Simulation with ASPEN Plus® 36 3.3 CFD Simulation 38 3.3.1 Modelling Approaches for Numerical Simulation of Fluidized Beds 38 3.3.2 Two Fluid Model (TFM) 40 3.3.3 Kinetic Theory of Granular Flow (KTGF) 44 3.3.4 Conclusion 46 4 COORVED GASIFICATION CONCEPT 48 4.1 Concept of Staged Conversion 48 4.1.1 Drawbacks of Conventional Fluidized-Bed Gasifiers 48 4.1.2 Basic Concept COORVED Gasifier 49 4.1.3 COORVED – Fixed-Bed Zone 49 4.1.4 COORVED – Bubbling-Bed Zone 50 4.1.5 COORVED – Jetting-Bed Zone 50 4.1.6 COORVED – Fast-Bed Zone 51 4.1.7 Conclusion 51 4.2 Test Facility and Reactor Design 52 4.3 Cold Flow Test Unit 53 4.4 Reference Cases 54 4.4.1 Solids Characterization 54 4.4.2 Gas Phase Properties 54 5 COORVED REACTOR IN FLOW REGIME DIAGRAMS 56 5.1 Reh Diagram for the Reference Case 56 5.2 Reh Diagram for Experimental Campaigns and CFD Case 57 5.3 Regime Diagrams for the Jetting-Bed Zone 60 5.4 Conclusion 61 6 CFD SIMULATION OF COORVED REACTOR 62 6.1 Verification of Multiphase CFD Setup 62 6.1.1 Parallelization 64 6.1.2 Pressure Drop and Minimum Fluidization Velocity 65 6.1.3 Conclusion 67 6.2 Grid Study 68 6.2.1 Pressure Drop 69 6.2.2 Voidage Profiles 69 6.2.3 Velocity Profiles 71 6.2.4 Conclusion 72 6.3 Validation Experiment Bubbling Bed and Fast Bed 72 6.3.1 Experimental Setup Holland 73 6.3.2 Simulation Setup 75 6.3.3 Results 77 6.3.4 Conclusion 84 6.4 Validation Experiment Jetting Bed 85 6.4.1 Experimental Setup 85 6.4.2 Simulation Setup 87 6.4.3 Results 88 6.4.4 Conclusion 95 6.5 CFD Simulation COORVED 96 6.5.1 Computational Grid 97 6.5.2 Cold Flow, Single Phase Jet 97 6.5.3 CFD setup 99 6.5.4 Results 99 6.5.5 Conclusion 103 7 ASPEN PLUS® SIMULATION OF THE COORVED GASIFIER 105 7.1 Validation Experiment Bubbling Bed and Fast Bed 105 7.2 COORVED Simulation 107 7.3 Conclusion 108 8 SUMMARY 109 9 OUTLOOK 114 9.1 Modeling Tools 114 9.2 COORVED Development 114 10 APPENDIX 115 11 REFERENCES 120

Zincor, a refinery in South Africa, uses the conventional roast-leachelectrowinning process to produce zinc metal. The roasting process of ZnS concentrate makes use of four Lurgi fluidized bed roasters to produce calcine (contains ZnO and ZnFe2O4 as zinc products) and SO2 gas. The roasting plant consists of two 18 m2 and two 35 m2 cross sectional area roasters. Prior to 1996, Zincor utilized air as the only oxidant and fluidizing medium in its roasters. The maximum dry feed rates that the roasters could process were 6.5 t/d.m2. In an attempt to increase production, oxygen enrichment was first trialled and then introduced into the fluidizing air. The ability of oxygen enrichment to increase the rate of the ZnS oxidation reaction allowed higher feed rates to the roasters. This was successful and oxygen enrichment was permanently implemented. That enabled dry feed rates to be maintained at 7.0 t/d.m2 and 7.3 t/d.m2 for the small and big roasters respectively. Oxygen enrichment up to 26% in the fluidizing air is utilized. Due to the highly competitive nature of the zinc industry, innovative processing techniques are necessary to be competitive. The aim of this study is to determine whether oxygen enriched air can be reduced by introducing micro-pelletized concentrate into the roaster feed blend, whilst maintaining current roaster feed rates and calcine quality. This study was executed in four parts. Firstly, the role entrainment played in influencing average particle residence time. Secondly, a study of production methods for stable micro-pellets. Thirdly, a study of the influence of oxygen enrichment and particle size on the roasting of micropellets. The fourth part of the study was introduction of micro-pellets into the Zincor roasters to determine whether oxygen enrichment could be reduced. The particle size distribution of a blend of feed concentrate to the roasters is 50% passing approximately 48 μm. This leads to entrainment values between 87% and 91%. The micro-pelletization process reduces the –500 μm fraction from 87% to 10%, which degrades to 30% during roasting. This requires that approximately 48% of the concentrate needs to be micro-pelletized to restore the 70% designed entrainment target. It was determined that entrained particles spend on average 0.46 hr to 2.44 hr in the bed compared to particles in the overflow that have residence times between 3.93 and 4.00 hr. The calculated times for entrainable particles are somewhat higher and for the bed overflow lower compared to the values measured by Spira, 1970. The required reaction time for micropellets was found to be far below their residence time inside a Zincor roaster. With a load of 20% micro-pellets introduced into the feed concentrate, the oxygen enrichment could be reduced by 60%. The quality of the calcine produced was maintained above the target of 98.8% ZnS to ZnO conversion. The results of this study have shown that the use of micro-pelletization of concentrate at Zincor reduces entrainment of particles successfully. Manipulation of entrainment through micro-pelletization can be used successfully to reduce oxygen enrichment, whilst improving production and maintaining quality at Zincor. Copyright / Dissertation (MEng)--University of Pretoria, 2009. / Materials Science and Metallurgical Engineering / unrestricted