Analysis of flooding as a stochastic process in polymer electrolyte membrane (PEM) fuel cells by impedance techniques

Impedance measurements were conducted to gain insight into flooding of a single polymer electrolyte membrane (PEM) fuel cell. The stochastic character of the formation of water droplets and subsequent removal by gas flow is demonstrated to increase the standard deviation of impedance measurements, yielding a sensitive manner to detect onset of flooding. The increase in stochastic noise associated with flooding was more apparent at low frequencies, due to the closer match to the characteristic frequency associated with the growth and removal of water droplets. The onset of flooding was sensitive to the design of the gas diffusion layer.