京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第17890号 / 工博第3799号 / 新制||工||1581(附属図書館) / 30710 / 京都大学大学院工学研究科物質エネルギー化学専攻 / (主査)教授 江口 浩一, 教授 安部 武志, 教授 阿部 竜 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

No description available.

Hybrid solid oxide fuel cell / gas turbine (SOFC/GT) systems offer high efficiency power generation, but face numerous integration and operability challenges. This dissertation addresses the application of hardware-in-the-loop simulation (HILS) to explore the performance of a solid oxide fuel cell stack and gas turbine when combined into a hybrid system. Specifically, this project entailed developing and demonstrating a methodology for coupling a numerical SOFC subsystem model with a gas turbine that has been modified with supplemental process flow and control paths to mimic a hybrid system. This HILS approach was implemented with the U.S. Department of Energy Hybrid Performance Project (HyPer) located at the National Energy Technology Laboratory. By utilizing HILS the facility provides a cost effective and capable platform for characterizing the response of hybrid systems to dynamic variations in operating conditions. HILS of a hybrid system was accomplished by first interfacing a numerical model with operating gas turbine hardware. The real-time SOFC stack model responds to operating turbine flow conditions in order to predict the level of thermal effluent from the SOFC stack. This simulated level of heating then dynamically sets the turbine's "firing" rate to reflect the stack output heat rate. Second, a high-speed computer system with data acquisition capabilities was integrated with the existing controls and sensors of the turbine facility. In the future, this will allow for the utilization of high-fidelity fuel cell models that infer cell performance parameters while still computing the simulation in real-time. Once the integration of the numeric and the hardware simulation components was completed, HILS experiments were conducted to evaluate hybrid system performance. The testing identified non-intuitive transient responses arising from the large thermal capacitance of the stack that are inherent to hybrid systems. Furthermore, the tests demonstrated the capabilities of HILS as a research tool for investigating the dynamic behavior of SOFC/GT hybrid power generation systems.

Mixed ionic electronic conductors (MIEC) have gained importance recently due to their roles in energy conversion devices such as solid oxide fuel cells (SOFC). Recent experimental data have shown that an increased vacancy concentration in a MIEC changes its elastic modulus and causes volumetric expansion. Since the MIEC in a device is constrained mechanically, the volumetric changes can induce substantial mechanical stresses. Such stresses not only lead to premature failure but can also alter the electrochemical performance of the device. In order to fully understand the interactions between stresses and defect transport a coupled theory is needed. This thesis develops a framework to study stress-defect transport interactions. The framework is based on a proper construction of the stress dependent electrochemical potential by introducing two material properties, namely the coefficient of chemical expansion (CCE) and the open system elastic constants (OSEC). The CCE characterizes the strains due to non-stoichiometry while the OSEC represents the stoichiometry dependent elastic stiffness. In this work these parameters are determined using atomistic simulations. The system of equations that govern the coupled electrochemical and mechanical fields is solved using a combination of numerical and analytical techniques. The developed solutions are analyzed to provide insights into the nature and the extent of the interactions. It is found that the non-stoichiometry-induced stress is in the same order of magnitude or even higher than that induced by thermal mismatch in a typical SOFC. In the vicinity of material flaws (cracks, voids, etc.), such stresses are further intensified which may cause fracture of the MIEC. In addition, non-stoichiometry-induced stresses can significantly alter the distribution of point defects, thus affecting the electrochemical performance of the ionic device. Furthermore, the non-stoichiometry induced stresses increases the thickness of the surface charge layer. The thermodynamic framework and the computational algorithms developed in this work provides effective methodologies and tools to analyze stress-defect transport interactions in ionic solids for designing and reliability analysis of ionic devices such as fuel cells, oxygen pumps, chemical sensors, etc.

A reversible solid oxide fuel cell (RSOFC) system could buffer intermittent electrical generation, e.g. wind, wave power by storing electrical energy as hydrogen and heat. RSOFC were fabricated by thermoplastic extrusion of (La₀.₈Sr₀.₂)₀.₉₅MnO[subscript(3−δ)] (LSM) ceramic support tubes, which were microstructurally stable with 55% porosity at 1350°C. A composite oxygen electrode of LSM-YSZ was applied, providing a homogeneous substrate for a 20 μm - 30 μm thick YSZ electrolyte. A dip-coated 8YSZ slurry, and a painted commercial 3YSZ ink gave sintered densities of 90% and nearly 100% at 1350°C, respectively. A porous NiO/YSZ fuel electrode was also painted on. A Ag/Cu reactive air braze was unsuccessful at forming a void-free joint between the RSOFC and a 316 stainless steel gas delivery tube, as the braze did not penetrate the oxidation layer on the steel. Two alumina-based ceramic cements failed to fully seal the cell to an alumina gas delivery tube, due to thermal expansion coefficient mismatches and porosity after curing. Therefore, the maximum open circuit voltage (OCV) obtained during RSOFC testing was 0.8 V at 440°C. LSM-YSZ symmetrical cell performance measurements with oxygen pressure showed a diffusion polarisation, which was assigned to dissociative adsorption and surface diffusion of oxygen species. A collaborative RSOFC system software model showed ohmic and activation losses dominated the RSOFC, and diffusion losses were insignificant. Pressurisation from 1 to 70 bar increased the RSOFC Nernst voltage by 11% at 900°C, and reduced the entropy of the gases, reducing heat production and increasing electrical efficiency. A 500 kg Sn/Cu phase change heat store prevented the system overheating. Over a 16 h discharge-charge RSOFC cycle in the range 5 mol.% - 95 mol.% hydrogen in steam, at 20.4 A per cell or 3250 A m⁻², the electrical energy storage efficiency was 64.4%.

This study concerned the preparation, characterisation and evaluation of different proton conductors for the Reverse Water Gas Shift Reaction (RWGS) and their evaluation as electrolytes for Solid Oxide Fuel Cells (SOFC) under H₂ and O₂. Materials with both catalytic and conductive properties are of a great interest, as their use in electrocatalytical systems may be very important. Sr₃CaZr₀.₅Ta₁.₅O₈.₇₅ (SCZT), BaCe₀.₉Y₀.₁O₂.₉₅ (BCY10) and Ba₃Ca₁.₁₈Nb₁.₈₂O₈.₇₃ (BCN18), were the initial materials in this study. Thermogravimetric analysis under different atmospheres (5%H₂/Ar, Ar, 5%CO₂, etc), were performed on SCZT and BCN18; with both materials being shown to be stable under reducing and oxidising conditions. Conductivity measurements by two terminal a.c. impedance were also conducted on SCZT and BCN18 under oxidising and reducing atmospheres. As found in literature, BCN18 showed mixed conductivity; with electronic conductivity at high temperatures and pure ionic conductivity below 550ºC, This behaviour was shown in chapter 3, where the change on conduction process was observed upon different gas feeds. Its maximum conductivity values for the different atmospheres were: 8.50x10⁻⁵ S/cm (Dry 5%H₂/Ar at 200ºC), 4.24x10⁻⁴ S/cm (Wet 5%H₂/Ar at 500ºC) and 4.48x10⁻³ S/cm (Air at 900ºC). SCZT proton conducting behaviour was also measured (wet and dry 5%H₂/Ar). Exhibiting an order of magnitude higher in total conductivity upon hydration of the gas feed (σdry=1.01x10⁻⁶ and σwet=1.07x10⁻⁵ at 450ºC). The doping of barium cerate with Zr and Zn by Tao and Irvine, lead to a more stable and easily sinterable BaCe₀.₅Zr₀.₃Y₀.₁₆Zn₀.₀₄O₃ (BCZYZ). Following this work, the introduction of ZnO as a sintering aid to SCZT and BCN18 resulted in Sr₃CaZr₀.₄Ta₁.₅Zn₀.₁O₈.₇₅ (SCZTZ), and Ba₃(Ca₁.₁₈Nb₁.₇₀Zn₀.₁₂)O₈.₅₅ (BCNZ); with higher final densities (above 90% dense). As with BCN18, BCNZ also exhibited mixed conductivity; with higher total conductivity values than BCN18. A maximum of total conductivity of 1.85x10⁻³ S/cm at 900ºC for BCNZ was measured against 6.99x10⁻⁴ S/cm at 900ºC for BCN18. A change in conductivity process was observed when using air or wet 5%H₂/Ar, achieving a maximum of 3.85x10⁻⁴ S/cm at 400ºC when under wet hydrogen. All materials (as powders) have shown catalytic activity for the reverse water gas shift (RWGS) reaction, with the lowest conversion temperature onset at 400ºC for SCZT and a maximum conversion of CO₂ to CO of 42%, with circa 0.52 and 0.59 mmol/s.m² of CO produced at 900ºC by BCN18 and BCZYZ, respectively. No relation between mechanisms for the RWGS and σ for these materials were expected below 10% conversion, as no correlation was found between their activation energies. BCY10 as shown a partial decomposition when exposed to the RWGS reaction, for what BCZYZ After fuel cell testing under H₂ and O₂ both SCZTZ and BCNZ showed mixed conductivity. SCZTZ under different hydrogen partial pressures, exhibited a behaviour close to a pure proton conductor, however, when exposed to both reducing and oxidising conditions, its behaviour did not follow the theoretical values. On the other hand, BCNZ behaves as a pure ionic conductor below 500ºC; with increasing influence of the electronic conductivity on temperature increase. However, as seen for BCNZ conductivity data from 2 terminal a.c. impedance, below 650ºC wet 5%H₂ exhibited the highest conductivity values. This, in additions to the pure ionic conductive behaviour below 400ºC (from the effective ionic transport number), suggests that BCNZ becomes closer to a pure proton conductor with temperature decrease.

This thesis explores the development of efficient engineered composite alternative anodes and the use of ethanol as a fuel for Solid Oxide Fuel Cells. SOFCs can in theory operate with fuels other than hydrogen. However, this requires the design of efficient alternative anode material that do not catalyze carbon formation and that are tolerant to redox cycles. An innovative concept has been developed that relies on the impregnation of perovskites into porous YSZ structures to form the anode functional layer. Catalysts are added to provide sufficient catalytic activity. Cells with anodes containing LSCM and Ce/Pd have displayed excellent performance. At 800°C, and with a 65 μm thick electrolyte, the power outputs were above 1W/cm² in humidified hydrogen and 0.7 W/cm² in humidified methane. These electrodes have shown the ability to reduce CO₂ electrochemically with an efficiency that is similar to that which can be achieved for H₂O electrolysis and the anodes could operate on pure CO₂. The importance of incorporating an efficient catalyst was demonstrated. The use of 0.5 wt% of Pd is sufficient to dramatically improve the performance in such electrodes. The microstructure of impregnated LSCM-YSZ composites plays an important role in the high performance obtained. A layer of LSCM nanoparticles covering the YSZ is formed upon reduction, offering a great surface area for electrochemical reactions. The fabrication method presented in this thesis is a powerful tool for designing microstructures in situ. Among the various fuels under consideration for SOFCs, ethanol offers outstanding advantages. Half cell measurements have been performed to characterize the performance of different types of anodes when operated on ethanol/steam mixtures. Steady performance was achieved on LSCM-CGO anodes. Carbon deposits from gas phase reactions have been evidenced and were found to be responsible for the performance enhancement when the cell is operated in diluted ethanol as compared to hydrogen. At high steam content, polarization resistances of LSCM-CGO-YSZ anodes in ethanol/ steam mixtures were shown to be below 0.3 Ω.cm² at 950°C.

The high temperature oxidation of ferritic stainless steel interconnects results in chromium poisoning of the solid oxide fuel cell (SOFC) electrodes, which is a limiting factor for their utilisation as SOFC interconnects. Chromium-rich spinel materials were studied as electrode supports that would be situated at the interface between interconnect and electrode, in order to reduce the effect of chromium poisoning of the electrodes. The main goal of this thesis was to find chromium-rich spinel materials with good electrical conductivity (σ ≥ 0.1 S∙cm⁻¹) in air and reducing atmosphere, chemically and mechanically stable in SOFC testing conditions. The structure and properties of newly formulated chromium-rich spinels, such as Mn₁₊ₓCr₂₋ₓO₄ (x = 0, 0.5), MnFeₓCr₂₋ₓO₄ (x = 0.1, 1), MgMnCrO₄, MnLiₓCr₂₋ₓO₄ (x = 0.1) and MgMₓCr₂₋ₓO₄, (M = Li, Mg, Ti, Fe, Cu, Ga) were studied aiming at their application as electrode support material for solid oxide fuel cells. Cation distributions were determined by Rietveld refinement from X-ray diffraction (XRD), within the limits of XRD precision and correlated with electrical properties determined experimentally. The chemical stability in reducing conditions was studied and the reduction effects upon materials were evaluated by XRD phase analysis and microstructure analysis. It was found that MnMₓCr₂₋ₓO₄ materials have a limited stability to reduction, only MnCr₂O₄ proved to have good stability when reduced, with negative influence for its p-type semiconductor conductivity. Even though MnFeCrO₄ had limited stability to reduction, in reducing conditions the conductivity changed from p-type to n-type semiconductor. A similar behaviour to reduction was observed for MgFeCrO₄. Also the mechanical and chemical compatibility of some spinels with YSZ was studied in terms of thermal expansion coefficient (TEC/K⁻¹), sintering step and possible chemical reactions. Lithium titanate spinels, starting with LiCrTiO₄, were investigated in terms of structure, properties and spinel - ramsdellite phase transition temperature also with the purpose of new material development. For these materials positive results were obtained in conductivity and chemical stability in reducing conditions. The performance of MnFeCrO₄ and MgFeCrO₄ as electrode support materials was investigated when used alone or impregnated with (La₀.₇₅Sr₀.₂₅)₀.₉₇Cr₀.₅Mn₀.₅O₃, La₀.₈Sr₀.₂FeO₃, Ce₀.₉Gd₀.₁O₂, CeO₂ or Pd. Composite anodes for SOFC were prepared by aqueous infiltration of nitrate salts into porous MnFeCrO₄ and MgFeCrO₄ scaffolds and studied by electrochemical impedance spectroscopy (EIS) in symmetrical cell configuration. The performance of the composite anodes was evaluated in humidified 5%H₂/Ar in order to understand their stability and performance at 850 °C or lower temperature with respect to the porous substrates. It was found that all the impregnated phases adhere very well to the spinel and considerably enhance performance and stability to a level required for SOFC applications. An interesting next step in this work would be to apply such spinel materials on steel interconnects, integrate them into testing SOFC devices and evaluate their effect upon chromium poisoning of the electrodes.

The decarbonization of the energy system is one of the most complex and consequential challenges of the 21st century. Meeting this challenge will require the deployment of existing low carbon technologies at unprecedented scales and rates and will necessitate the development of new technologies that have the ability to transform variable renewable energy into high energy density products. Reversible Solid Oxide Cells (RSOCs) are electrochemical devices that can function both as fuel cells or electrolyzers: in fuel cell mode, RSOCs consume a chemical fuel (H₂, CO, CH₄, etc.) to produce electrical power, while in electrolysis mode they consume electric power and chemical inputs (H₂O, CO₂) to produce a chemical fuel (H₂, CO, CH₄, etc.). As such, RSOC systems can be thought of as flexible “energy hubs” that have unique potential to bridge the low power density renewable infrastructure with that of high energy density fuels in an efficient, dynamic, and bidirectional fashion. This dissertation explores the different operational sensitivities and design trade-offs of a methane based RSOC system, investigates the optimum operating strategies for a system that adapts to variations in the hourly spot electricity and fuel prices in Western Denmark, and provides an economic analysis of the system under a wide variety of design assumptions, operational strategies, and fuel and electricity market structures. In order to perform such comprehensive analyses, a 0-D computational model of a methane based RSOC system was developed in Python. In fuel cell mode, the system generates power by consuming natural gas, while in electrolysis mode the system generates synthetic natural gas (SNG) by electrolyzing steam and catalytically hydrogenating recycled CO₂ into CH₄ downstream of the RSOC. The model's flexibility enables the simulation of “part-load” operation, allowing the user to assess the changes in output, efficiency, and operating cost as the system is operated across multiple points. The model has the ability to evaluate the impact that changes in design choices and operating parameters (Area Specific Resistance, temperatures, current density, etc.) have on the system as it interfaces with time varying exogenous factors such as fuel and electricity prices. As such, one of the main contributions of this model is the ability to run simulations in which the operating strategy of the RSOC system responds and adapts to varying market signals. The computational model is used to develop a series of hourly optimizations for finding the optimal operating strategy for an RSOC system that can buy or sell electricity and gas in the spot electricity and natural gas markets in Western Denmark. After receiving an electricity and gas price signal, the optimization determines the operating mode (fuel cell, electrolysis or idle) and operating point (e.g., current density) that maximize the operating profits every hour for the given electricity and gas price pair. In order to avoid the speculation associated with traditional energy storage simulations, the system is “opened” at both ends, allowing it to instantaneously buy and sell any electricity or gas that is generated. Thus, the system never stores any of the products and it buys and sells them at the instantaneously available market price. By assuming that market prices reflect all existing information, this design choice removes the necessity of having to speculate about the future in order to determine the optimum operating strategy. This approach is one of the innovations presented in this work. The optimizations aim at maximizing the operating profits at each hour of the year, and decisions of operating mode and point are based on marginal operating costs for each electricity and natural gas price pair. The full economic analysis, however, requires the understanding of how design choices (e.g. operating limits, heat management, gas recycling systems, etc.) affect the investment costs, and therefore a Total Plant Cost (TPC) model is developed. For each design choice, the TPC model is used to compute a cost of the system per m² of active electrode area or kW of output. This value, assumed to be a sunk cost that does not affect the operating decision, together with the operating profits resulting from the optimization is used to assess the overall profitability of the system. For a system with 100m² of active electrode area, conventional costing metrics suggest that the balance of plant (BoP) components for managing the system's heat (Heat exchangers, evaporators, condensers) are the main cost drivers and represent roughly 50% of the TPC. The cost of the electrochemical RSOC stack, assembly, power inverter and piping represent 35% of the cost, with the other 15% coming from pumps, compressors and the methanation system. Twenty different optimization scenarios are developed in order to quantify the effect that system design choices, operating limits, and market prices have on the operating profile and on the overall economics of the system. The first 12 case studies are based on real hourly spot electricity and natural gas prices for the years 2009-2014 in Western Denmark. For the last 8 scenarios, a forecasted hourly time-series for electricity in the Danish grid for the year 2050 and two fixed SNG prices (high and a low) are used. The 2050 prices, which assume a fossil fuel free system, are used to understand the role and value that RSOC systems can offer in deeply decarbonized energy systems. For each optimization, different parameters such as the initial ASR and the operating limits (maximum current densities for each mode of operation) are varied in order to find the impact that these changes have on the system's design (balance of plant components), hourly operating mode, investment costs, hourly operating profits, and overall plant profits. For the 2009-2014 optimizations, it is found that the sale of electricity (fuel cell mode) and fuel (electrolysis mode) is not large enough to cover the fixed costs associated with the plant. Fuel cell mode dominates the operation (61% of the time) with electrolysis representing only ~ 4% of the operating hours. ASR is found to have an important impact on the system's economics, due to the fact that a lowering of the ASR leads to a reduction in the size of the heat management system, which in turn reduces the Total Plant Cost. For the 2050 dataset, it is found that under the high gas price scenario electrolysis mode dominates (50% of the time), and fuel cell operation represents 15% of the hours in the year. For the low SNG price, electrolysis still dominates (48% of the time), and fuel cell operation increases to 30% of the operating hours. Furthermore, for the high SNG scenario, the sale of fuel and electricity are large enough to cover the system's fixed cost, making the system attractive from an investment perspective. For the low SNG price, the system also becomes profitable when using ASR values of 0.4 ASR or below.

Solid oxide fuel cells (SOFCs) are high temperature (600C-1000C) composite metallic/ceramic-cermet electrochemical devices. There is a need to effectively manage the heat transfer through the cell to mitigate material failure induced by thermal stresses while yet preserving performance. The present dissertation offers a novel thermodynamic optimization approach that utilizes dimensionless geometric parameters to design a SOFC. Through entropy generation minimization, the architecture of a planar SOFC has been redesigned to optimally balance thermal gradients and cell performance. Cell performance has been defined using the 2nd law metric of exergetic efficiency. One constrained optimization problem was solved. The optimization sought to maximize exergetic efficiency through minimizing total entropy production while constraining thermal gradients. Optimal designs were produced that had exergetic efficiency exceeding 92% while maximum thermal gradients were between 219 C/m and 1249 C/m. As the architecture was modified, the magnitude of sources of entropy generation changed. Ultimately, it was shown that the architecture of a SOFC can be modified through thermodynamic optimization to maximize performance while limiting thermal gradients. The present dissertation highlights a new design methodology and provides insights on the connection between thermal gradients, performance, sources of entropy generation, and cell architecture.