Dans le contexte de la transition énergétique, la pile à combustible devient l'une des sources d'énergie alternatives les plus prometteuses. Récemment, la recherche a mis l’accent sur les piles à combustible, et plus particulièrement sur celles à membrane échange de protons (Proton Exchange Membrane Fuel Cell ou Polymer Electrolyte Membrane Fuel Cell ou PEMFC) qui est l'une des meilleures candidates pour les applications stationnaires et transport. Même si cette technologie évolue constamment, elle n'est pas encore prête pour un déploiement industriel à grande échelle en raison de sa durabilité et de sa fiabilité limitées. Le "Prognostics and Health Management" (PHM) est une approche récente pour gérer et prolonger la durée de vie des systèmes. Les techniques de pronostic sont capables de fournir une estimation de l'état de santé (State of Health ou SOH) des piles à combustible et une prédiction de leur durée de vie résiduelle (Remaining Useful Life ou RUL) afin d’aider les fabricants à améliorer les performances et à gérer leur durée de vie de ces systèmes.Ce travail a pour objectif de développer de nouvelles méthodes d’estimation de la durée de vie adaptée à la complexité des systèmes PEMFC. En effet, ces systèmes sont multi-échelle et multi-physique, et présentent divers défis sont à relever:1. La définition du SOH pour construire un indicateur de dégradation.2. La coexistence de phénomènes de dégradation à la fois réversibles et irréversibles.3. La prise en compte des différentes causes de détérioration et des effets des conditions opératoires.Dans la première partie, nous effectuons une analyse bibliographique de l’utilisation du PHM pour les PEMFCs, dans le but de proposer une définition de SOH et de construire un indicateur de dégradation. Etant donné que les mesures PEMFC sont peu nombreuses, nous avons également exploré l'état de l'art sur les batteries au lithium, qui sont d'autres cellules électrochimiques.Dans la deuxième partie, nous développons un algorithme de pronostic basé sur le filtrage particulaire utilisant la mesure de puissance de la PEMFC. Les premiers résultats montrent que l'algorithme de pronostic est perturbé par la dégradation réversible existante. L’ambiguïté peut être levée en estimant la dégradation irréversible grâce à des tests de caractérisation, tels la spectroscopie d'impédance électrochimique (Electrochemical Impedance Spectroscopy ou EIS), appliquée de temps en temps. Nous proposons donc un algorithme de pronostic étendu et adapté, prenant en compte deux indicateurs : la dégradation de la puissance et le SOH estimé à partir de la caractérisation EIS. La performance de l'algorithme proposé est évaluée par différentes indicateurs de performance, et les résultats montrent l'intérêt de cette approche.Dans la troisième partie, les problèmes sont abordés d'un point de vue plus théorique. En effet, l’évolution de la dégradation d'un système est souvent corrélée à des covariables internes et externes qui sont généralement difficiles d'accès en raison des coûts de mesure élevés. Par conséquent, nous avons d'abord développé une approche comprenant des inspections en ligne de la covariable de dégradation à un autre niveau, puis nous avons proposé une approche d’estimation de la RUL basée sur un ensemble de modèles en utilisant différentes sources à différents niveaux. Les RULs prédites par les deux modèles sont agrégées dynamiquement sur la base des performances évaluées sur les données historiques. Par conséquent, la précision de la prédiction est améliorée car les inconvénients des deux modèles ont été surmontés en tirant parti de leurs avantages. Dans la dernière partie, le problème est étendu au pronostic multi-niveaux qui ouvre de nouveaux aspects pour la recherche future sur le pronostic et la gestion de la PEMFC. / In the context of the energy transition, fuel cell becomes one of the promising alternative energy sources. Recently the spotlight is on fuel cell systems research, and more particularly on Proton Exchange Membrane Fuel Cell (PEMFCs) which is one of the best candidates for both stationary and transportation applications. Even if this technology is close to being competitive, it is not yet ready to be considered for a large scale industrial deployment because of its limited durability and reliability. Prognostics and Health Management (PHM) is a recent approach to manage and possibly extend life duration of technological systems. Prognostic techniques can provide an estimation of fuel cell State Of Health (SOH) and a prediction for their Remaining Useful Life (RUL) to help the manufacturers improving fuel cell performance and managing its lifespan.The objective of this work is to develop prognostic methodologies for the RUL prognosis adapted to the complexity of PEMFCs. Indeed, the PEMFC is a multi-scale and multi-physics system, and various challenges are faced:1. The definition of SOH to build a degradation indicator.2. The coexistence of both reversible and irreversible degradation phenomena.3. Taking into account different deterioration causes and effects of operating conditions.In the first part of our work, we conduct a state of the art analysis on PHM for PEMFCs, with the aim of proposing a SOH definition and building a degradation indicator for PEMFC prognosis purposes. And since PEMFC measurements are scarce, the state of the art on Lithium batteries, other electrochemical cells, is also explored.In the second part, we develop a particle filtering based prognostic algorithm for PEMFC, based on output power measurements. The first results show that the prognosis algorithm is disturbed by the existing reversible degradation. However, the irreversible degradation can be estimated thanks to characterization tests, such as Electrochemical Impedance Spectroscopy (EIS), which is applied from time to time. We propose thus an adapted & extended prognostic algorithm to take into account both health indicators: the output power degradation and the SOH degradation estimated from EIS characterization. The performance of the proposed algorithm is evaluated by different prognostic performance metrics, and the results show the interest of this approach.In the third part, the problem is addressed from a more theoretical point of view. Indeed, a system's degradation behavior is often correlated with internal and external covariates which are usually difficult to access owing to expensive measurement cost. Therefore, we first developed a prognostic approach with online inspections on the degradation covariate at a different level, and then we propose an approach for RUL prognosis based on an ensemble of models using different sources at different levels. The RUL predictions of both models are dynamically aggregated on the basis of prognostic performance evaluated on a set of historical data. Consequently, the prediction accuracy is improved by overcoming both models' drawbacks and leveraging their strengths. In the last part, we extend the problem to multi-level prognostics and explore new possibilities, which open new aspects for future research on PEMFC lifetime prognosis and management.

La première partie de cette thèse est dédiée à la modification de polyélectrolytes pour former des films de multicouche de polyélectrolytes (PEM) ayant des propriétés d’adhésion de protéine et de cellules bien contrôlées et modifiables par étirement. L’acide polyacrylique a été modifié avec des groupes latéraux phosphorylcholine (PC) à des taux de 25 % (PAA-PC) ou avec des chaînes oligo(éthylène oxyde) terminées par la biotine : (EO)nBiotine (n = 0, 3, 9 et 18) avec de taux de modification de 1, 5, 10 ou 25 %. Des PEM incorporant ces polymères lient spécifiquement la streptavidine et repoussent tout autre protéine. Les propriétés d’adsorption et la sélectivité de ces PEM ont été mesurées par microbalance à quartz. Sur un substrat de PDMS étirables, on a construit des PEM terminés par un PAA portant des RGD recouvert par deux couches contenant PAA-PC. Au repos, seuls les PC sont exposés et inhibent l’adhésion cellulaire ; sous étirement, les groupes RGD sous-jacents sont exposés et déclenchent l’adhésion de fibroblastes.La deuxième partie est consacrée à l‘étude d’acide polyméthacrylique modifié hydrophobiquement avec des chaînes alkyle liées par des esters à la chaîne principale. 3 chaînes différentes ont été greffées : -C12H25 ; -C18H35 et C4H8-OOC- C11H23 avec des taux de 1, 5 and 10 %. Ces polymères sont associatifs et forment des hydrogels dans des tampons physiologiques pour des taux de modifications de 5% et des concentrations supérieures à 4% en poids. Ces gels ont été caractérisés par des mesures rhéologiques. Leur incubation avec des lipases provoque une baisse de leur viscosité, interprétable par une coupure des esters. Quand les gels faits à partir du PAA-C12 sont incubés avec une culture de Pseudomonas aeruginosa, la viscosité baisse également, ce qui montre que les chaînes sont également coupées in vivo. / The first part of this thesis is dedicated to the modification of polyelectrolytes to form polyelectrolyte films with controlled and stretch responsive cell and protein adsorption properties. Poly(acrylic acid) (PAA) was modified with side phosphorylcholine groups (PC) at rates of 25 % or with oligo(ethylene oxide) chains ended by biotin ((EO)nBiotin, (n =0, 3, 9 and 18) at 1, 5, 10 and 25 % modification rates. Polyelectrolytes multilayer films (PEM) containing these polyelectrolytes bind selectively streptavidin but repel all other proteins. The adsorption properties and selectivity were measured by quartz crystal microbalance. On a stretchable PDMS substrate, we have built PEM ended by PAA bearing RGD, covered by two PAA-PC layers on the top. Under rest, only the PC groups are exposed and prevent cell adhesion; when the film is stretched, the underlying RGD groups are exposed, and trigger adhesion of fibroblasts.The second part was consecrated to the study of poly(methacrylic acid) hydrophobically modified with alkyl chains connected through an ester moiety to the main chain. Three different chains were grafted -C12H25; -C18H35 and -C4H8- OOC-C11H23 with a rate of 1, 5 and 10 %. These polymers associate in water and form hydrogels in physiological buffer, for modification rates higher than 5 % and polymer concentrations higher than 4 wt. %. The gels were characterized by rheology. Their incubation with lipases resulted in a decrease of their viscosity, which could be interpreted by the cleavage of the hydrophobic side chains, by rheological tests. When the gels with PAA-C12 were incubated with a culture of Pseudomonas aeruginosa, their viscosity decreased, which shows that alkyle chains are also cleaved in vivo.

Fuel cell performance was obtained as functions of the humidity at the anode and cathode sites, back pressure, flow rate, temperature, and channel depth. The fuel cell used in this work included a membrane and electrode assembly (MEA) which possessed an active area of 25, 50, and 100 cm2 with the Nafion® 117 and 115 membranes. Higher flow rates of inlet gases increase the performance of a fuel cell by increasing the removal of the water vapor, and decrease the mass transportation loss at high current density. Higher flow rates, however, result in low fuel utilization. An important factor, therefore, is to find the appropriate stoichiometric flow coefficient and starting point of stoichiometric flow rate in terms of fuel cell efficiency. Higher air supply leads to have better performance at the constant stoichiometric ratio at the anode, but not much increase after the stoichiometric ratio of 5. The effects of the environmental conditions and the channel depth for an airbreathing polymer electrolyte membrane fuel cell were investigated experimentally. Triple serpentine designs for the flow fields with two different flow depths was used. The shallow flow field deign improves dramatically the performance of the air-breathing fuel cell at low relative humidity, and slightly at high relative humidity. For proton exchange membrane fuel cells, proper water management is important to obtain maximum performance. Water management includes the humidity levels of the inlet gases as well as the understanding of the water process within the fuel cell. Two important processes associated with this understanding are (1) electro-osmotic drag of water molecules, and (2) back diffusion of the water molecules. There must be a neutral water balance over time to avoid the flooding, or drying the membranes. For these reasons, therefore, an investigation of the role of water transport in a PEM fuel cell is of particular importance. In this study, through a water balance experiment, the electro-osmotic drag coefficient was quantified and studied. For the cases where the anode was fully hydrated and the cathode suffered from the drying, when the current density was increased, the electro- osmotic drag coefficient decreased.

Mathematical modeling helps researchers to understand the transport and kinetic phenomena within fuel cells and their effects on fuel cell performance that may not be evident from experimental work. In this thesis, a 2-D steady-state cathode model of a proton-exchange-membrane fuel cell (PEMFC) is developed. The kinetics of the cathode half-reaction were investigated, specifically the reaction order with respect to oxygen concentration. It is unknown whether this reaction order is one or one half. First- and half-order reaction models were simulated and their influence on the predicted fuel cell performance was examined. At low overpotentials near 0.3 V, the half-order model predicted smaller current densities (approximately half that of the first-order model). At higher overpotentials above 0.5 V, the predicted current density of the half-order model is slightly higher than that of the first-order model. The effect of oxygen concentration at the channel/porous transport layer boundary was also simulated and it was shown the predicted current density of the first-order model experienced a larger decrease (~10-15% difference at low overpotentials) than the half-order model. Several other phenomena in the cathode model were also examined. The kinetic parameters (exchange current density and cathode transfer coefficient) were adjusted to assume a single Tafel slope, rather than a double Tafel slope, resulting in a significant improvement in the predicted fuel cell performance. Anisotropic electronic conductivities and mass diffusivities were added to cathode model so that the anisotropic structure of the porous transport layer was taken into account. As expected, the simulations showed improved performance at low current densities due to a higher electronic conductivity in the in-plane direction and decreased performance at high current densities due to smaller diffusivities. Additionally, the concentration overpotential was accounted for in the model; however it had little influence on the simulation results. / Thesis (Master, Chemical Engineering) -- Queen's University, 2008-09-19 12:14:29.079

The catalyst layer of a proton exchange membrane (PEM) fuel cell is a mixture of polymer, carbon, and platinum. The characteristics of the catalyst layer play a critical role in determining the performance of the PEM fuel cell. This research investigates the role of catalyst layer composition using a Central Composite Design (CCD) experiment with two factors which are Nafion content and carbon loading while the platinum catalyst surface area is held constant. For each catalyst layer composition, polarization curves are measured to evaluate cell performance at common operating conditions, Electrochemical Impedance Spectroscopy (EIS), and Cyclic Voltammetry (CV) are then applied to investigate the cause of the observed variations in performance. The results show that both Nafion and carbon content significantly affect MEA performance. The ohmic resistance and active catalyst area of the cell do not correlate with catalyst layer composition, and observed variations in the cell resistance and active catalyst area produced changes in performance that were not significant relative to compositions of catalyst layers. / Master of Science

Proton Exchange Membrane (PEM) fuel cells are becoming an increasingly important alternative to combustion engines as the fossil fuel reserves are depleted. Several papers have presented steady state analyses of the system, but few are known to present dynamic analysis of the flow and control of the hydrogen delivery process. This thesis presents the dynamic analysis of hydrogen delivery to a PEM fuel cell system. The hydrogen is delivered to the anode with use of an ejector for passive recirculation. The system to be studied consists of the manifolds, ejector, and pressure control valve. Models describing the elements of the anode delivery systems are formulated. The governing nonlinear equations are solved analytically and numerically, and the regimes of stable hydrogen delivery process are established. The linearized models are used for performance analysis and optimization of the hydrogen delivery process. The nonlinear model is used to improve the simulation of the dynamics of the PEM fuel cell system and validate the parameters at optimal linearized stability. Experiments are conducted to find the parameters used in the model, as well as validate the results. Both the linear and nonlinear models are implemented in Simulink and tested against the laboratory data from the PEM fuel cell system. The analysis showed that the models have the same time constant and dynamic behavior as the PEM system. The optimal parameters for stability and a faster response with no oscillations in the output are obtained. The redesigned valve and resulting dynamics of the PEM fuel cell system provides improved system performance.

L’électrolyse est une technologie qui permet de répondre à deux problématiques cruciales. D’une part, répondre au besoin en stockage d’énergie liée à l’intégration de sources intermittentes sur les réseaux électriques. D’autre part, répondre à la croissance de la demande en hydrogène, liée aux marchés naissants de l’hydrogène énergie. La nature des besoins liés au développement de la technologie d’électrolyse implique des sollicitations intermittentes dont les impacts quant au fonctionnement du système sont encore méconnus. En ce sens, et face aux manques de la littérature quant à la modélisation à l’échelle système de la technologie d’électrolyse PEM, un nouveau modèle est développé. Pour cela, le formalisme de modélisation graphique Bond Graph est utilisé, notamment pour sa capacité à représenter tout type d’échange énergétique de manière unifiée. Le modèle développé permet de représenter l’intégralité d’un système d’électrolyse PEM, ses différents composants et lois de contrôle associées. Il est validé sur la base du comportement dynamique d’une installation semi-industrielle disponible au CEA. Ce modèle est ensuite utilisé pour identifier et comprendre les enjeux liées à une sollicitation intermittente d’un système d’électrolyse PEM d’un point de vue de l’efficacité du système, de sa flexibilité et de sa capacité de suivi de charge, de sa fiabilité, de sa sûreté ou encore de sa durabilité. Différentes modifications de conception sont simulées et évaluées à la lumière de ces différents enjeux. Finalement, le modèle Bond Graph est exploité d’un point de vue de ses propriétés structurelles afin d’analyser les conditions de surveillabilité d’un système d’électrolyse PEM. / PEM Electrolysis is a technology which to enable to face two major challenges : (i) Fulfill the need of energy storage caused by the integration of intermittent energy sources on electricity networks; (ii) Cope with the growing need of carbon free hydrogen caused by the future market applications of hydrogen energy. These particular needs, regarding electrolysis technology development, involve an intermittent operating mode which impacts on the dynamic behavior of the system remain unknown. Modelling is a critical tool to understand these issues and provide a thorough analysis. State of the art of existing modelling works highlighted that only a few models take into account the dynamic of the whole system including Balance of Plant. Therefore a new dynamic and multiphysic model was developed under Bond Graph formalism. This graphical modelling formalism was selected especially thanks to its ability to represent any kind of power exchange in a unified way. The model enables to represent the whole system including balance of plant and associated control laws. It is validated on the dynamic behavior of an experimental device available in CEA. The model is then used in order to identify and understand the issues related to intermittent operation of a PEM electrolysis system. These issues are related to system efficiency, flexibility, reliability, safety and durability. Regarding these issues, some design changes are simulated and assessed. Finally, the Bond Graph model and its structural properties enable to perform diagnosis and monitorability analyses of a PEM electrolysis system.

<p>A model is developed for the Quanser torsion system available at Control Systems Research Laboratory at Chulalongkorn University. The torsion system is a laboratory equipment that is designed for the study of position control. It consists of a DC motor that drives three inertial loads that are coupled in series with the motor, and where all components are coupled to each other through torsional springs.</p><p>Several nonlinearities are observed and the most significant one is an offset in the input signal, which is compensated for. Experiments are carried out under feedback as the system is marginally stable. Different input signals are tested and used for system identification. Linear black-box state-space models are then identified using PEM, N4SID and a subspace method made for closed-loop identification, where the last two are the most successful ones. PEM is used in a second step and successfully enhances the parameter estimates from the other algorithms.</p>

In polymer electrolyte membrane (PEM) fuel cells one of the most important components is the flow field. The flow field distributes reactant gasses to the active area and also delivers electrons from the outer circuit so that the electrochemical reaction may be completed. Optimizing flow field design is extremely important in order to increase the overall power density of the fuel cell. It is particularly important to understand the ways in which the different portions of the flow field, namely the land and channel sections, interact with the gas diffusion layer (GDL), catalyst layer and membrane; this study focuses on those interactions. The most common type of flow field design currently used in PEM fuel cells is the serpentine flow field. It is used for its simplicity of design, its effectiveness in distributing reactants and its water removal capabilities. The knowledge about where current density is higher, under the land or the channel, is critical for flow field design and optimization. Yet, no direct measurement data are available for serpentine flow fields. In this study a fuel cell with a single channel serpentine flow field is used to separately measure the current density under the land and channel, which is either catalyzed or insulated on the cathode. In this manner, a systematic study is conducted under a wide variety of conditions and a series of comparisons are made between land and channel current density. Results show that under most operating conditions, current density is higher under the land than that under the channel. However, at low voltage, a rapid drop off in current density occurs under the land due to concentration losses. The mechanisms for the direct measurement results and general guidelines for serpentine flow field design and optimizations are provided. In addition the same technique is utilized to separately measure current density under the land and channel on a variety of serpentine flow field geometries. Each flow field is tested under a wide variety of operating conditions thereby providing guidance for the optimum design geometry. Experimental results show that generally flow fields with both thinner lands and thinner channels provide better overall performance. However, the optimal flow field designs are highly dependent on fuel cell operating parameters. Finally, it is critical not only to know where the current density is greater, under the land or under the channel, but to understand the fundamental mechanisms driving these differences. Resistance was measured, ex-situ, between the GDE and flow plate under the land of the flow field and under the channel separately. The contact resistance between the gas diffusion electrode (GDE) and the graphite flow plate were measured using an ex-situ technique. The resistance was measured under different land and channel widths. Cyclic Voltammetry tests were also conducted in order to determine if there is any different in electrochemically active area(ECA) under the land and under the channel and what the cause of this difference might be. Results show that the compression of the gas diffusion electrode not only affects the electronic resistance but the ECA as well and that these are key factors in current density variations under the land and channel.

An investigation was conducted to determine the feasibility of using the technique of ultrasonic nebulization of water into the anode gas stream for evaporative cooling of a Proton Exchange Membrane (PEM) fuel cell. The basic concept of this form of internal evaporative cooling of the PEM fuel cell is to introduce finely atomized liquid water into the anode gas stream, so that the finely atomized liquid water adsorbs onto the anode and then moves to the cathode via electro-osmotic drag, where this water then evaporates into the relatively dry cathode gas stream, carrying with it the waste thermal energy generated within the fuel cell. The thermal and electrical performance of a 50 cm2 PEM fuel cell utilizing this technique was compared to the performance obtained with conventional water management. Both techniques were compared over a range of humidification chamber temperatures for both the anode and cathode gas streams so as to determine the robustness of the proposed method. The proposed method produced only meager levels of evaporative cooling (at best 2 watts, for which a minimum of 30 watts was required for adequate cooling), but the average cell voltage increased considerably (as much as a 10% gain), and the technique increased the fault tolerance of the fuel cell (the Nafion™ membrane did not dry out even if cell temperature went well in excess of 70° C despite both anode and cathode humidification temperatures of 55° C). An interesting phenomena was also observed wherein the fuel cell voltage oscillated regularly with a period of tens of seconds, and that the amplitude of this oscillation corresponded inversely with the level of humidification received by the fuel cell.