Evaluation of Solid Oxide Fuel Cells Operating on Hydrogen Sulfide Contaminated Fuel

This research was conducted to investigate the effect of hydrogen sulfide on the performance of single solid oxide fuel cells. A test rig was designed and commissioned to test 5x5 cm2 cells (active area: 4x4 cm2). The test rig consists of a gas blender, a humidifier, a high temperature furnace, fuel and air manifolds and a control/data logging system. The characterisation techniques used in this project, include v-i measurement, EIS and SEM/EDX analysis. The first series of experiments were carried out to investigate the effect of time, hydrogen partial pressure and temperature on the performance of the cells operating on clean fuel. The results showed that the current of lowest resistance is independent of the operating temperature, however, depends on partial pressure of H2 and tends to increase as PH2 rises. The lowest resistance of the cell occurs at almost constant fuel utilization which was equal to 17 % in this research. In the second series of tests, the cells were exposed to a range of H2S concentrations i.e. 50, 100, 150 and 200 ppm. The composition of the fuel mixture was 0.1 nl/min (14.5 %) of H2, 0.567 nl/min (82.5 %) of N2 and 0.020 nl/min (3 %) of H2O (steam). All the contamination tests were carried out at 700 ˚C. The cells were exposed to H2S for 12 hours followed by a recovery period for 24 hours. The results revealed that the voltage drop at the end of the exposure period was similar for all degrees of poisoning. However, the performance at the end of the recovery, was different. The degree of recovery tended to decrease as the concentration of H2S increased. The SEM analysis of samples showed that H2S has caused the anode structure to change. This change occurred at the interface of anode functioning and support layers and was more severe at higher concentrations of H2S. In addition, two contamination models were developed based on the H2S degradation mechanism. The models considered the effects of time and H2S concentration. However, they could not predict the performance of the poisoned cells as the voltage drop at the end of exposure time was independent of the H2S concentration for the tested range.