Cette étude est dédiée au développement d’une anode pour une pile à combustible SOFC monochambre. Ce dispositif ne présente pas de séparation physique entre les compartiments anodique et cathodique, contrairement à une pile conventionnelle. Un mélange contenant de l’oxygène et un hydrocarbure est directement injecté sur l’ensemble du dispositif comprenant électrolyte, anode et cathode. La cathode doit être sélective à la réduction de l’oxygène et l’anode à l’oxydation de l’hydrocarbure. Ce dispositif permet donc de s’affranchir des problèmes d’étanchéité des dispositifs conventionnels mais les matériaux d’électrode doivent répondre à des critères catalytiques restrictifs. L’étude a été réalisée avec une anode de type cermet composée de nickel et d’oxyde de cérium gadolinié (CGO). L’hydrocarbure choisi est le propane. L’objectif du travail est de comprendre les phénomènes physico-chimiques se produisant à l’anode afin d’optimiser les conditions de fonctionnement de la pile monochambre, sous mélange O2 / C3H8. Pour cela, les propriétés catalytiques des poudres de nickel et de CGO ont été déterminées en fonction de la température et du rapport O2 / C3H8. Après avoir optimisé les paramètres de préparation des anodes par sérigraphie sur électrolyte support (CGO), des mesures de conductivité électrique par méthode Van der Pauw ont été réalisées également sous différents mélanges O2 / C3H8 en fonction de la température. La confrontation des résultats (propriétés catalytiques et électriques) a permis de mettre en évidence l’existence d’une température limite de fonctionnement, qui dépend du rapport O2 / C3H8. En-dessous de cette température limite, le nickel s’oxyde et l’anode catalyse principalement l’oxydation totale du propane. Au-dessus de cette température limite, le nickel reste sous la forme métallique et catalyse préférentiellement l’oxydation partielle du propane. Ces résultats sont également corroborés à des calculs thermodynamiques, qui mettent en évidence que la réaction préférentielle à basse température est l’oxydation du nickel. Des phénomènes d’oscillations de la conductivité électrique et de la température ont permis de proposer un mécanisme réactionnel basé sur le dépôt et l’oxydation de carbone. Enfin, des dispositifs complets monochambre ont été élaborés en utilisant une cathode BSCF (Ba0,5Sr0.5Co0,8Fe0,2O3) et testés. / This work is devoted to the development of a single chamber SOFC. Unlike a conventional SOFC, this type of fuel cell has no physical separation between the anodic and cathodic compartments. An oxygen and hydrocarbon mixture is injected directly on the overall fuel cell, including the electrolyte, the anode and the cathode. The cathode must have a high selectivity to the oxygen reduction, and the anode to the hydrocarbon oxidation. This device allows to avoid the conventional devices sealing problems, but the electrodes materials must fit with restrictive catalytic criterions. The study has been conducted with a cermet anode type, composed of nickel and gadolinia doped ceria (CGO). The selected hydrocarbon was the propane. The goal of this work is to understand physico-chemical processes taking place at the anode in order to optimize the operating conditions of the fuel cell, in an O2 / C3H8 mixture. The catalytic properties of nickel and CGO powders were determined as a function of temperature and O2 / C3H8 ratio. After optimization of the anode screen-printing conditions preparation on a CGO support, electrical measurements were done using the Van der Pauw method under different O2 / C3H8 mixtures as a function of temperature. The results of catalytic and electrical measurements show a limiting operating temperature, which depends on the O2 / C3H8 ratio. Below this limiting temperature, the nickel is oxidized and principally catalyses the total oxidation of the propane. Above this limiting temperature, nickel remains into its metallic state and preferentially catalyses the partial oxidation of the propane. These results are also corroborated to thermodynamics calculations, which show that at low temperature, the oxidation of nickel is the most favourable reaction. Some oscillations of the conductivity have been observed. Oscillations of electrical conductivity and temperature allowed proposing a reaction mechanism based on carbon deposition and oxidation. Finally, complete single chamber devices have been elaborated using a BSCF (Ba0,5Sr0.5Co0,8Fe0,2O3) cathode and tested.

Thesis (PhD (Science) (Viticulture and Oenology. Wine Biotechnology))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: The production of glycerol by Saccharomyces cerevisiae under anaerobic conditions is essential for maintaining the intracellular redox balance thereby allowing continuous energy generation through conversion of sugars into ethanol. In addition, glycerol can act as an osmolyte and is synthesized to maintain turgor pressure under hyperosmotic conditions. The production of ethanol from sugars can be a redox-neutral process, where the NAD+ (nicotinamide adenine dinucleotide) that is consumed during the glycolytic conversion of glyceraldehyde-3-phosphate to pyruvate is later regenerated by the reduction of acetaldehyde to ethanol. However, in particular the redirection of metabolic flux of pyruvate to biomass formation leads to excess NADH formation. The intracellular redox balance in these conditions is then primarily maintained through formation of glycerol which is control by two main enzymes, namely Gpd1p and Gpd2p. Deletion of the genes coding for these two proteins leads to accumulation of NADH and renders the cells incapable of maintaining their fermentative ability and growth under anaerobic conditions. The goal of this study was to investigate the growth, fermentative ability and metabolite synthesis of various gpd1Δgpd2Δ double mutant (DM) strains in which the redox balancing potential was partially restored through expression of native or heterologous genes. Strains were constructed by introducing alternative NADH oxidizing pathways or manipulating existing pathways to favour the oxidation of excess NADH. More specifically, the modifications included (i) sorbitol formation; (ii) establishing a pathway for propane-1,2-diol formation; and (iii) increasing ethanol formation. Apart from genetically manipulating the gpd1Δgpd2Δ double mutant, the addition of pyruvate during growth was also investigated. The experiments were carried out under oxygen limited conditions in a high sugar medium and the fermented product was analyzed for total sugar consumed, biomass and primary and secondary metabolites formed by the different strains. The relationships between sugar consumption, growth and metabolite production by different strains were investigated by comparing the data generated from the different strains by using multivariate data analysis tools. Analysis of the pathways involved in the production of primary (acids, ethanol and other metabolites) and secondary metabolites (aroma compounds) were also carried out in order to establish flux modification in comparison to the wild type (WT) strain. The results revealed that these manipulations improved the fermentative capacity of the gpd1Δgpd2Δ double mutant, suggesting a partial recovery of NAD+ regeneration ability, albeit not to the extent of the WT strain. As expected a significant correlation was found between sugar consumption and ethanol and biomass formation. Ethanol yields but not final concentrations were increased by the genetic manipulations. Sorbitol by DM(srlD) and DM(SOR1) strains and propane-1,2-diol by DM(gldA, GRE3, mgsA) strain were formed in significant amounts although at lower molar yields than glycerol. Furthermore, by genetic manipulation the yield of secondary metabolites (isobutanol, isoamyl alcohol, 2-phenyl ethanol and isobutyric acid) was increased whereas the ethyl acetate concentration and yield decreased. The results indicate that aroma compound properties of wine yeasts could be favourably changed by manipulating the glycerol synthesizing pathway. The addition of pyruvate during the growth of gpd1Δgpd2Δ double mutant contributes to excess NADH re-oxidation through additional ethanol formation. / AFRIKAANSE OPSOMMING: Die produksie van gliserol deur Saccharomyces cerevisiae onder anaërobiese toestande is noodsaaklik vir die onderhouding van die intrasellulêre redoksbalans en maak dus ononderbroke energie-ontwikkeling tydens die omsetting van suikers in etanol moontlik. Daarbenewens kan gliserol as ‘n osmoliet optree en word dit gesintetiseer om turgordruk onder hiperosmotiese toestande te onderhou. Die produksie van etanol uit suikers kan ‘n redoksneutrale proses wees, waar die NAD+ (nikotinamiedadenien-dinukleotied) wat tydens die glikolitiese omskakeling van gliseraldehied-3-fosfaat na piruvaat verbruik word, later deur die reduksie van asetaldehied na etanol regenereer word. Die nasending van die metaboliese vloeiing van piruvaat na biomassavorming lei egter na ‘n oormaat NADH-vorming. Onder hierdie toestande word die intrasellulêre redoksbalans dan hoofsaaklik deur die vorming van gliserol onderhou. Laasgenoemde word veral deur twee ensieme beheer, naamlik Gpd1p en Gpd2p. Die delesie van die gene wat vir hierdie twee proteïene enkodeer, lei tot ‘n akkumulasie van NADH en veroorsaak dat die selle nie hulle gistingsvermoë en groei onder anaërobiese toestande kan onderhou nie. Die doelwit van hierdie studie was om die groei, gistingsvermoë en metabolietsintese van verskeie gpd1Δgpd2Δ dubbelmutant (DM) rasse te ondersoek waarin die redoksbalanseringspotensiaal gedeeltelik herstel is deur die uitdrukking van inheemse of heteroloë gene. Rasse is gekonstrueer deur alternatiewe NADH-oksiderende weë in te voer of deur bestaande weë te manipuleer om die oksidasie van oormaat NADH te bevoordeel. Meer spesifiek het die modifikasies die volgende ingesluit: (i) sorbitolvorming; (ii) die vestiging van ‘n weg vir propaan-1,2-diol-vorming; en (iii) die verhoging van etanolvorming. Buiten die genetiese manipulering van die gpd1Δgpd2Δ dubbelmutant, is die byvoeging van piruvaat tydens groei ook ondersoek. Die eksperimente is onder suurstofbeperkte toestande in ‘n hoë-suiker medium uitgevoer en die gegiste produk is ondersoek vir totale suikerverbruik, biomassa en primêre en sekondêre metaboliete wat deur die verskillende rasse gevorm is. Die verhoudings tussen suikerverbruik, groei en metabolietproduksie deur die verskillende rasse is ondersoek deur die data wat deur die verskillende rasse gegeneer is deur middel van meerveranderlike data-analise te vergelyk. Analise van die weë wat in die produksie van primêre (sure, etanol en ander metaboliete) en sekondêre metaboliete (aromaverbindings) betrokke is, is ook uitgevoer om die verandering in vloei te bepaal in vergelyking met die wildetipe (WT) ras. Die resultate het gewys dat hierdie manipulasies die gistingsvermoë van die gpd1Δgpd2Δ-dubbelmutant verbeter het, wat ‘n gedeeltelike herstel van NAD+- regenerasievermoë voorstel, hoewel nie tot dieselfde mate as in die WT-ras nie. Soos verwag, is ‘n beduidende korrelasie tussen suikerverbruik en etanol- en biomassavorming gevind. Etanolopbrengs is deur genetiese manipulasies verhoog, maar nie die finale konsentrasies van etanol nie. Sorbitol is in beduidende hoeveelhede deur die DM(srlD) en DM(SOR1)-rasse gevorm en propaan-1,2-diol deur die DM(gldA, GRE3, mgsA) -rasse, hoewel teen laer molare opbrengste as gliserol. Verder is die opbrengs van sekondêre metaboliete (isobutanol, iso-amielalkohol, 2-fenieletanol en isobottersuur) deur genetiese manipulasie verhoog, terwyl die etielasetaatkonsentrasie en -opbreng verlaag is. Die resultate dui aan dat die aromaverbindingseienskappe van wyngiste voordelig verander kan word deur die gliserolsintetiseringsweg te manipuleer. Die byvoeging van piruvaat tydens die groei van die gpd1Δgpd2Δ-dubbelmutant dra by tot uitermate NADH-reoksidasie tydens die bykomende vorming van etanol.

L'oxydation de l'éthylène, du propane et du propène a été étudiée dans un large domaine de température (900 à 1200 Kelvin), de pression (0,1 à 1 MPa) et de richesse (0,15 à 4) au moyen d'un réacteur auto-agité par jets gazeux. L'analyse chromatographique d'échantillons du milieu réactionnel permet de suivre la réaction d'oxydation dans son ensemble. Un mécanisme cinétique détaillé comportant 278 réactions et faisant intervenir 47 espèces chimiques, a été développé grâce à l'utilisation d'une technique d'analyse de sensibilité, et permet de modéliser correctement nos résultats expérimentaux.

The environmental concerns in recent times have grown especially after signing Montreal protocol. In the last ten years, researchers have focussed mainly on understanding the boiling and condensation phenomena of HFC refrigerants in minichannels. As global warming concerns are growing day by day, due to high global warming potential, HFCs are not the ultimate option. In the near future, HFCs will probably be replaced by environmentally friendly refrigerants. Therefore, to find the potential replacements of HFCs and also to get a deeper understanding of the boiling phenomena in minichannels, more and more fluids having low GWP (Global Warming Potential) and ODP (Ozone Depletion Potential) should be tested. Recent efforts to protect the environment have led to a growing interest for natural refrigerants. However in the literature, flow boiling data of natural refrigerants in minichannels are scarce. To meet the environmental concerns and to understand the behaviour of natural refrigerants in minichannels and the performance compared to HFCs, flow boiling experiments in single circular vertical minichannels of internal diameters of 1.70 and 1.224 mm were performed using ammonia and propane as working fluids. Flow boiling heat transfer results of ammonia and propane with 1.70 mm channel showed that the heat transfer coefficient was a function of heat flux and the effect of mass flux was insignificant. The heat transfer coefficient of ammonia in 1.224 mm was dependent on heat flux at low vapour qualities then a clear dependence of the heat transfer coefficient on the mass flux was observed at higher vapour qualities. The heat transfer results of ammonia and propane were compared with well known correlations and among them Cooper (1989) correlation in case of ammonia and Liu and Winterton (1991) and Cooper (1984) pool boiling correlations in case of propane best predicted the experimental heat transfer data. Results of the two phase pressure drop studies of ammonia and propane showed that the two phase pressure drop increased with the increase of mass flux, with the increase of heat flux and with the decrease of saturation temperature. The comparison of the two phase pressure drop experimental data with well known predicting models showed that none of the correlations predicted the ammonia data well and that Müller Steinhagen and Heck (1986) was well in agreement with the propane data. Dryout of propane in 1.70 mm and 1.224 mm internal diameter channels was also investigated. Dryout heat flux was observed to increase with the increase of mass flux, with the decrease of vapour quality and with the increase of internal diameter. The effect of saturation temperature on the dryout heat flux was insignificant. The experimental dryout data were compared with macro and micro scale correlations and among them Bowring (1972) and Callizo et al. (2010a) gave best predictions. The heat transfer and pressure drop results of ammonia and propane and dryout results of propane were compared with R134a data taken on the same test rig by Owhaib (2007) and Ali (2010). The comparison of heat transfer showed that the heat transfer coefficient was a function of heat flux and the effect of mass flux was insignificant in all tested conditions except ammonia in 1.224 mm tube where the heat transfer coefficient was dependent on heat flux at lower vapour qualities and a clear dependence of mass flux was observed at higher vapour qualities. The heat transfer data of ammonia, propane and R134a were compared with correlations and among them Cooper (1989) correlation gave best predictions. The comparison of pressure drop results showed that the two phase pressure drop of all fluids was increased with the increase of mass flux, with the increase of heat flux and with the decrease of saturation temperature. At equal heat flux and mass flux, the two phase pressure drop of ammonia was increased with the decrease of internal diameter but the diametric effects of R134a were unclear. Müller Steinhagen and Heck (1986) and Zhang and Webb (2001) best predicted the experimental data of two phase pressure drop of ammonia, propane and R134a among the correlations considered for comparison. The dryout data of propane were also compared with dryout data of R134a and it was observed that the dryout heat flux of propane and R134a increased with the increase of mass flux, with the decrease of vapour quality and with the increase of internal diameter. The effect of saturation temperature on the dryout heat flux of propane and R134a was insignificant. The correlation of Bowring (1972) for conventional channels and the microscale correlation of Callizo et al. (2010a) were among the correlations which gave best predictions of experimental data of dryout. / QC 20120210

Enormous quantities of heat are available in air, soil, water, exhaust air from buildings, and in waste water of any kind. However these heat sources are use-less for heating purposes since their temperatures are lower than the tempera-ture required for heating. Heat pumps can be used to extract heat from these sources with a small expenditure of additional energy and up-grade and deliver the energy as useful heat for room heating. The heat pump cycle employs the well-known vapour compression cycle. The amount of heat delivered by a heat pump is equal to the amount of energy extracted from the heat source plus the heat equivalent to the compression work of the heat pump. Heat pumps, of course, are being generally accepted as outstanding energy saving units due their coefficient of performance (COP). Heat pumps for house heating have been used extensively in many countries and are especially common in Sweden. The annual growth rate of heat pump usage in Sweden is the same as in rest of Europe. According to the Swedish heat pump association, between 1986 to August 2003, the number of installed heat pump units in Sweden was 332,309. The demand for heat pumps started to increase from the year 1995 and in the year 2002, approximately 40,000 heat pump units were installed. Among the many types available, single-family heat pumps providing heating capacity of about 5 kW are widely popular. The main drawbacks of heat pumps are the complexity of the systems, high cost, need of technical knowledge, safety hazards and environmental effects of certain refrigerants, etc. An efficient heat pump with small refrigerant charge would have less of some of these drawbacks and could be a competitive alterna-tive to other heating processes. In this study, methods of refrigerant charge minimisation without reducing the performance of a small capacity (5 kW) heat pump have been investigated. Work has been focused on finding refrigerant charge distribution in different components of the heat pump, on finding out the solubility of refrigerant (pro-pane) with different compressor lubrications oils, on testing different types of compact heat exchangers, on constructing new minichannel heat exchangers and on finding correlations for calculating the heat transfer of minichannel heat exchangers. The results included in this thesis have been presented in four con-ference papers and five journal papers of which two were published and three were submitted for publication. / QC 20100707

Increasing environmental concerns as well as diminishing fossil fuel reserves call for a new generation of energy conversion technologies. Fuel cells, which convert the chemical energy of a fuel directly to electrical energy, have been identified as one of the leading alternative energy conversion technologies. Fuel cells are more efficient than conventional heat engines with minimal pollutant emissions and superior scalability. Proton Exchange Membrane Fuel Cells (PEMFCs) which produce electricity from hydrogen have been widely investigated for transportation and stationary applications. The focus of this study is on the Direct Propane Fuel Cell (DPFC), which belongs to the PEMFC family, but consumes propane instead of hydrogen as feedstock. A drawback associated with DPFCs is that the propane reaction rate is much slower than that of hydrogen. Two ideas were suggested to overcome this issue: (i) operating at high temperatures (150-230oC), and (ii) keeping the propane partial pressure at the maximum possible value. An electrolyte material composed of zirconium phosphate (ZrP) and polytetrafluoroethylene (PTFE) was suggested because it is an acceptable proton conductor at high temperatures. In order to keep the propane partial pressure at the maximum value, interdigitated flow-fields were chosen to distribute propane through the anode catalyst layer. In order to evaluate the performance of a DPFC which operates at high temperature and uses interdigitated flow-fields, a computational approach was chosen. Computational Fluid Dynamics (CFD) was used to create two 2-D mathematical models for DPFCs based on differential conservation equations. Two different approaches were investigated to model species transport in the electrolyte phase of the anode and cathode catalyst layers and the membrane layer. In the first approach, the migration phenomenon was assumed to be the only mechanism of proton transport. However, both migration and diffusion phenomena were considered as mechanisms of species transport in the second approach. Therefore, Ohm's law was used in the first approach and concentrated solution theory (Generalized Stefan-Maxwell equations) was used for the second one. Both models are isothermal. The models were solved numerically by implementing the partial differential equations and the boundary conditions in FreeFEM++ software which is based on Finite Element Methods. Programming in the C++ language was performed and the existing library of C++ classes and tools in FreeFEM++ were used. The final model contained 60 pages of original code, written specifically for this thesis. The models were used to predict the performance of a DPFC with different operating conditions and equipment design parameters. The results showed that using a specific combination of interdigitated flow-fields, ZrP-PTFE electrolyte having a proton conductivity of 0.05 S/cm, and operating at 230oC and 1 atm produced a performance (polarization curve) that was (a) far superior to anything in the DPFC published literature, and (b) competitive with the performance of direct methanol fuel cells. In addition, it was equivalent to that of hydrogen fuel cells at low current densities (30 mA/cm2).

Cette étude est dédiée au développement d'une anode pour une pile à combustible SOFC monochambre. Ce dispositif ne présente pas de séparation physique entre les compartiments anodique et cathodique, contrairement à une pile conventionnelle. Un mélange contenant de l'oxygène et un hydrocarbure est directement injecté sur l'ensemble du dispositif comprenant électrolyte, anode et cathode. La cathode doit être sélective à la réduction de l'oxygène et l'anode à l'oxydation de l'hydrocarbure. Ce dispositif permet donc de s'affranchir des problèmes d'étanchéité des dispositifs conventionnels mais les matériaux d'électrode doivent répondre à des critères catalytiques restrictifs. L'étude a été réalisée avec une anode de type cermet composée de nickel et d'oxyde de cérium gadolinié (CGO). L'hydrocarbure choisi est le propane. L'objectif du travail est de comprendre les phénomènes physico-chimiques se produisant à l'anode afin d'optimiser les conditions de fonctionnement de la pile monochambre, sous mélange O2 / C3H8. Pour cela, les propriétés catalytiques des poudres de nickel et de CGO ont été déterminées en fonction de la température et du rapport O2 / C3H8. Après avoir optimisé les paramètres de préparation des anodes par sérigraphie sur électrolyte support (CGO), des mesures de conductivité électrique par méthode Van der Pauw ont été réalisées également sous différents mélanges O2 / C3H8 en fonction de la température. La confrontation des résultats (propriétés catalytiques et électriques) a permis de mettre en évidence l'existence d'une température limite de fonctionnement, qui dépend du rapport O2 / C3H8. En-dessous de cette température limite, le nickel s'oxyde et l'anode catalyse principalement l'oxydation totale du propane. Au-dessus de cette température limite, le nickel reste sous la forme métallique et catalyse préférentiellement l'oxydation partielle du propane. Ces résultats sont également corroborés à des calculs thermodynamiques, qui mettent en évidence que la réaction préférentielle à basse température est l'oxydation du nickel. Des phénomènes d'oscillations de la conductivité électrique et de la température ont permis de proposer un mécanisme réactionnel basé sur le dépôt et l'oxydation de carbone. Enfin, des dispositifs complets monochambre ont été élaborés en utilisant une cathode BSCF (Ba0,5Sr0.5Co0,8Fe0,2O3) et testés.

En cas d'accident impliquant des produits chimiques, il peut être fait appel à des experts pour évaluer les effets générés par cet accident. Ces experts fournissent des distances d'effets à l'aide de modélisations informatiques et sont confrontés à une difficulté majeure : peu d'éléments à leur disposition pour caractériser la situation.L'objectif de cette thèse est de proposer une méthodologie permettant de prendre en compte les incertitudes relatives aux données d'entrée dans les modélisations effectuées en situation d'urgence tout en restituant de manière explicite cette incertitude au gestionnaire de la situation d'urgence.Une première phase a consisté à évaluer, pour une situation à même de générer un nuage toxique ou explosible, la dispersion des résultats des modélisations. Il a été établi une hiérarchisation des variables d'entrée en fonction de leur influence sur le résultat final. Cette phase a été réalisée au moyen d'une analyse de sensibilité dont la stratégie a été spécifiquement développée.Une seconde phase a eu pour but d'établir une méthodologie d'estimation des distances d'effets en situation d'urgence en tenant compte du niveau d'incertitude des variables d'entrée. Une méthodologie de classification opérationnelle des données d'entrée a été réalisée. Elle s'appuie sur deux critères : la sensibilité du modèle au paramètre d'entrée et l'incertitude sur sa valeur (imprécision ou variabilité). Sur cette base, une nouvelle manière d'utiliser ces variables a été proposée. Enfin, différentes façons de restituer de manière opérationnelle les résultats des modélisations ont été proposées.

Les problèmes liés à l'environnement et plus particulièrement à la couche d'ozone et à l'effet de serre sont fortement d'actualité. Depuis 1987, le protocole de Montreal impose une nouvelle réglementation en terme d'utilisation des fluides frigorigènes pour les pays industrialisés. Pour appliquer ce protocole, dans de bonnes conditions techniques et économiques, l'industrie a besoin de données thermophysiques concernant les nouveaux fluides frigorigènes. Afin de réaliser des mesures d'équilibre liquide vapeur sur les fluides frigorigènes, deux techniques expérimentales ont été choisies et utilisées. Il s'agit d'une technique statique-analytique avec échantillonnage des phases (analyse par chromatographie en phase gazeuse) et d'une technique synthétique avec une cellule à volume variable. Quatre systèmes binaires et un système ternaire ont été étudiés: 1,1,1,2,3,3,3-heptafluoropropane (R227ea)+ propane (R290); difluorométhane (R32) + R290; R32 + R290, dioxyde de carbone (CO2) + R227ea et R32+R290+R227ea. Pour réaliser le traitement des données et la modélisation, un progiciel a été développé: THERMOPACK. Ce progiciel a permis de comparer les résultats expérimentaux obtenus à des valeurs disponibles dans la littérature, de calculer la ligne des points critiques et de calculer la position des azéotropes.

Increasing environmental concerns as well as diminishing fossil fuel reserves call for a new generation of energy conversion technologies. Fuel cells, which convert the chemical energy of a fuel directly to electrical energy, have been identified as one of the leading alternative energy conversion technologies. Fuel cells are more efficient than conventional heat engines with minimal pollutant emissions and superior scalability. Proton Exchange Membrane Fuel Cells (PEMFCs) which produce electricity from hydrogen have been widely investigated for transportation and stationary applications. The focus of this study is on the Direct Propane Fuel Cell (DPFC), which belongs to the PEMFC family, but consumes propane instead of hydrogen as feedstock. A drawback associated with DPFCs is that the propane reaction rate is much slower than that of hydrogen. Two ideas were suggested to overcome this issue: (i) operating at high temperatures (150-230oC), and (ii) keeping the propane partial pressure at the maximum possible value. An electrolyte material composed of zirconium phosphate (ZrP) and polytetrafluoroethylene (PTFE) was suggested because it is an acceptable proton conductor at high temperatures. In order to keep the propane partial pressure at the maximum value, interdigitated flow-fields were chosen to distribute propane through the anode catalyst layer. In order to evaluate the performance of a DPFC which operates at high temperature and uses interdigitated flow-fields, a computational approach was chosen. Computational Fluid Dynamics (CFD) was used to create two 2-D mathematical models for DPFCs based on differential conservation equations. Two different approaches were investigated to model species transport in the electrolyte phase of the anode and cathode catalyst layers and the membrane layer. In the first approach, the migration phenomenon was assumed to be the only mechanism of proton transport. However, both migration and diffusion phenomena were considered as mechanisms of species transport in the second approach. Therefore, Ohm's law was used in the first approach and concentrated solution theory (Generalized Stefan-Maxwell equations) was used for the second one. Both models are isothermal. The models were solved numerically by implementing the partial differential equations and the boundary conditions in FreeFEM++ software which is based on Finite Element Methods. Programming in the C++ language was performed and the existing library of C++ classes and tools in FreeFEM++ were used. The final model contained 60 pages of original code, written specifically for this thesis. The models were used to predict the performance of a DPFC with different operating conditions and equipment design parameters. The results showed that using a specific combination of interdigitated flow-fields, ZrP-PTFE electrolyte having a proton conductivity of 0.05 S/cm, and operating at 230oC and 1 atm produced a performance (polarization curve) that was (a) far superior to anything in the DPFC published literature, and (b) competitive with the performance of direct methanol fuel cells. In addition, it was equivalent to that of hydrogen fuel cells at low current densities (30 mA/cm2).