The performance and degradation of Solid Oxide Fuel Cells (SOFC) were studied under severe operating conditions. The cells studied were manufactured in a small series by ECN, in the framework of the EU funded CORE‐SOFC project. The cells were of the anode‐supported type with a double layer LSM cathode. They were operated at 750 °C or 850 °C in hydrogen with 5% or 50% water at current densities ranging from 0.25 A cm–2 to 1 A cm–2 for periods of 300 hours or more. The area specific cell resistance, corrected for fuel utilisation, ranged between 0.20 Ω cm2 and 0.34 Ω cm2 at 850 °C and 520 mV, and between 0.51 Ω cm2 and 0.92 Ω cm2 at 750 °C and 520 mV.The degradation of cell performance was found to be low (ranging from 0 to 8%/1,000 hours) at regular operating conditions. Voltage degradation rates of 20 to 40%/1,000 hours were observed under severe operating conditions, depending on the test conditions.Data analysis revealed a critical cell voltage of ca 750 mV, above which the degradation rates were trivial, but below which they were significant.Some cells were also tested using a different procedure to that usually applied at Risø. This gave a different aging behaviour, indicating that the detailed test circumstances may be decisive to the outcome.

No abstract

Hydrogen is the fuel for fuel cells with the highest cell voltage. A drawback for the use of hydrogen is the low energy density storage capacity, even at high pressures. Liquid fuels such as gasoline and methanol have a high energy density but lead to the emission of the greenhouse gas CO2. Ammonia could be the ideal bridge fuel, having a high energy density at relative low pressure and no (local) CO2 emission. Ammonia as a fuel for the solid oxide fuel cell (SOFC) appears to be very attractive, as shown by cell tests with electrolyte supported cells (ESC) as well as anode supported cells (ASC) with an active area of 81cm2. The cell voltage was measured as function of the electrical current, temperature, gas composition and ammonia (NH3) flow. With NH3 as fuel, electrical cell efficiencies up to 70% (LHV) can be achieved at 0.35A∕cm2 and 60% (LHV) at 0.6A∕cm2. The cell degradation during 3000 h of operation was comparable with H2 fueled measurements. Due to the high temperature and the catalytic active Ni∕YSZ anode, NH3 cracks at the anode into H2 and N2 with a conversion of >99.996%. The high NH3 conversion is partly due to the withdrawal of H2 by the electrochemical cell reaction. The remaining NH3 will be converted in the afterburner of the system. The NOx outlet concentration of the fuel cell is low, typically <0.5ppm at temperatures below 950°C and around 4ppm at 1000°C. A SOFC system fueled with ammonia is relative simple compared with a carbon containing fuel, since no humidification of the fuel is necessary. Moreover, the endothermic ammonia cracking reaction consumes part of the heat produced by the fuel cell, by which less cathode cooling air is required compared with H2 fueled systems. Therefore, the system for a NH3 fueled SOFC will have relatively low parasitic power losses and relative small heat exchangers for preheating the cathode air flow.

The performance of anode‐supported cells with a composite LSM‐YSZ cathode and an LSM current collector was investigated. Over the first 48 hours, after the application of a constant current, the cell voltage was observed to increase by up to 20%. When the current was switched off, the cell resistance increased significantly over the next four days at open circuit conditions. Apparently, at OCV conditions cell passivation occurs. The cell gradually reactivates, once the current is switched on again. Part of this activation / passivation process is fast enough to influence the resistance of the cell during i–V measurements (over less than 1 hour) and a considerable hysteresis is observed in the cell voltage during these measurements.Impedance spectroscopy was used to investigate the activation / passivation process. It was found that the series resistance and the part of the polarisation impedance above approximately 100 Hz were not influenced by the activation / passivation process. The part of the polarisation impedance between 1 and 100 Hz was highly influenced by the activation / passivation process and during cell polarisation this part of the polarisation impedance was up to 40% lower than at open circuit conditions. This frequency range of the spectrum was also sensitive to the oxygen partial pressure at the cathode side, indicating that it is the cathode that activates and passivates.

No abstract