Fuel cell flooding diagnosis based on time-constant spectrum analysis

Abstract

Proton Exchange Membrane (PEM) Fuel Cell is a key component for the exploitation of hydrogen energy. Its diagnosis relies on several in-situ diagnosis tools. Electrochemical impedance spectroscopy is a major one. But this technic has some drawbacks, among which the complexity of the required equipment and the measurement time. An alternative method for diagnosis is developed in this work. A time-constant spectrum (or relaxation-time distribution) is extracted from the fuel cell voltage response to small current steps. This spectrum can be read as a distribution of series RC cell representing, such as an impedance spectrum, the whole dynamic of the voltage response to a given current excitation. In order to demonstrate the ability of the method to diagnose flooding, different simulations of voltage responses are first performed using appropriated parameters’ variations in a fuel cell model. The resulting time-constant spectrums are analyzed to underline the sensitivity to the flooding effect. Then, experiments are achieved to confirm the potentiality of this technic to diagnose flooding.