PEMFC fuel cell diagnostic by electrochemical noise analysis: apply to water management

This thesis deals to optimize the performance of the PEMFC, through the development of diagnostic tools dedicated to water management. This water management is one of the key elements to increase the performance and life of fuel cells. The electrochemical noise technique was used to detect the phenomena related to the water behavior during the operation of the cell, under different operating conditions (relative humidity of the gases, temperature and current). The electrochemical noise recorded during these experiments was associated with mechanisms through an appropriate experimental approach and signal processing based on frequency and time analysis. The results of spectral and statistical analysis made it possible to detect the drying and flooding of the cell by the generation of the descriptors in the frequency and time domain, signatures of each operating state of the fuel cell. Finally, to differentiate between the drying of the membrane and the flooding of the cell, the electrochemical noise measurements were completed by measuring the anodic and cathodic pressure fluctuations of the PEMFC. (author)