Kei Hosoi

Development of Intermediate-Temperature SOFC Module Using Doped Lanthanum Gallate

An intermediate temperature solid oxide fuel cell (SOFC) module was developed using electrochemically active cells composed of (La, Sr)(Ga, Mg, Co)O 3 electrolyte, Ni-(Ce, Sm)O 2 anode, and (Sm, Sr)CoO 3 cathode. Seal-less planar type stack design was employed. The first generation module successfully provided the output power of I kW with thermal self-sustainability below 800°C. Maximum electrical efficiency obtained with this module was 43%[LHV] together with the corresponding fuel utilization of 78%. Dynamic performance tests demonstrated the capability of output power alteration from 0.6 to 1 kW while maintaining a high electrical conversion efficiency. Further testing and modification of the module for methane fuel utilization are in progress.

SOFC Module and System Development by Means of Sealless Metallic Separators with Lanthanum Gallate Electrolyte

The third-generation 1-kW e -class module was developed with an automatic control system. A conversion efficiency of 48% ac/lower heating value [ac/LHV] was achieved with an exhaust heat recovery unit. An endurance test using the third-generation 1-kW e module was done for over 1000 h and no degradation of the power generation performance was observed. In parallel, a single-cell unit, which includes one cell and two metallic separators, was tested for over 10000 h and the degradation rate of the terminal voltage was found to be 1-2%/1000 h. In the direction of scale-up, a triple-stack module of 3-kW e output was developed. A partial load as well as excess loads on the module were tested and the output power of 1-5 kW e was attained under thermally self-sustainable conditions. It was found that a high efficiency of 55% dc/lower heating value [dc/LHV] was obtained under stable operation. Ongoing research of the fourth-generation 1-kW e module has resulted in the conversion efficiency of 58% [dc/LHV].

Oxide ion and electronic conductivity in Co doped La0.8Sr0.2Ga0.8Mg0.2O3 perovskite oxidePresented at the 78th International Bunsen Discussion Meeting on ?Complex Oxides: Defect Chemistry, Transport and Chemical Reaction?, Vaals, The Netherlands, October 6?9, 2002.

Partial electronic and hole conductivity in Co doped LaGaO3 based perovskite oxide was investigated with the ion-blocking method. Typical S-shaped polarization curves were observed on La0.8Sr0.2Ga0.8Mg0.2−XCoXO3 (0 < X < 0.1). The oxygen partial pressure (PO2) dependence of the electronic and hole conductivity is estimated to be PO2−1/4 and PO21/4, respectively, at temperature higher than 1173 K. However, these decreased to PO2−0.12 and PO20.06 respectively at 873 K. It is considered that the electronic and hole conductivities, that are intrinsic to LSGM are dominant at high temperature, however, the extrinsic electronic and hole conductivity caused by doped Co becomes dominant with decreasing temperature. The estimated transport number of the Co doped sample was higher than 0.95 over the PO2 range from 1 to 10−30 atm, which is slightly higher than that estimated by the H2–O2 cell. The partial electronic and hole conductivities in Co doped LaGaO3 based oxide increased with increasing the amount of Co, in particular, increase in the electronic conductivity is significant at Co content higher than 8.5 mol% to Ga site. PO2 dependence for electronic and hole conductivity is much smaller than that of PO2−1/4 and PO21/4, respectively, suggesting that the electronic and hole conductivity which is extrinsic to LSGM is dominant with increasing Co amount and the specimens behaves like an intrinsic semiconductor. The estimated theoretical efficiency of the electrolyte reaches a maximum value of ca. 0.90 around a thickness of 100 μm in 5 mol% Co doped sample at 0.8 A cm−2 and 1073 K.

Residual Stress Analysis in Lanthanum Gallate-Based Cells before and after Fuel Cell Operation

The distribution of residual stress in cells with the practical size of 120 mm diameter was measured by X-ray diffraction utilizing synchrotron radiation at Super Photon ring 8 GeV (SPring-8). The residual stress was determined by the constant penetration depth method. When NiO-samaria-doped ceria (SDC) was sintered on (La,Sr)(Ga,Mg,Co)O 3-δ (LSGMC), the tensile stresses resided in NiO and SDC, and the compressive stress resided in LSGMC. The reduction of the anode by hydrogen decreased tensile stresses in the anode. The residual stresses in the electrodes did not change at any X-ray penetration depth. The sintering of the cathode slightly affected the residual stresses of the cells.

A Unique Seal-Less Solid Oxide Fuel Cell Stack and Its CFD Analysis

Combined Heat and Power (CHP) generation units based on intermediate temperature (600∼800°C) solid oxide fuel cell (SOFC) modules have been collaboratively developed by Mitsubishi Materials Corporation and The Kansai Electric Power Co., Inc. Currently, hydrocarbon fuel utilising units designed to produce modular power outputs up to 10 kWe-AC with overall efficiencies greater than 80% (HHV) are being tested. A unique seal-less stack concept is adopted to build SOFC modules accommodating multiple stacks incorporated of stainless steel separators and disk-type planar electrolyte-supported cells. In order to advance the current technology to achieve improved levels of efficiency and reliability, through design iterations, computational modelling tools are being heavily utilised. This contribution will describe the results of coupled computational fluid dynamics (CFD) analysis performed on our fourth-generation 1 kW class SOFC stack. A commercially available CFD code is employed for solving the governing equations for conservation of mass, momentum and energy. In addition, a local electrochemical reaction model is coupled to the rest of the transport processes that take place within the SOFC stack. It is found that the CFD based multi-physics model is capable of providing necessary and proper guidance for identifying problem areas in designing multi-cell SOFC stacks. The stack performance is also estimated by calibrating the computational model against data obtained by experimental measurements.Copyright