An analysis of fluidic voltage statistical correlation for a diagnosis of PEM fuel cell flooding

The fault diagnosis is one of the most important topics on Proton Exchange Membrane Fuel Cell (PEMFC) stacks. Statistical methodologies for diagnosis are considered as one of the most relevant. This paper is dedicated to the diagnosis of flooding, using statistical methodology.In the literature, previous researches establish that in PEM (Proton Exchange Membrane) fuel cell stacks, single cell voltage variance increases when flooding conditions appear at the cathode side of the fuel cell stack. In this paper, these flooding conditions were completed by increasing the hygrometry of the inlet air flow at constant temperature of the fuel cell stack, or by decreasing the temperature of the fuel cell stack at constant hygrometry of inlet air flow. The obtained experimental results are here correlated with a fluidic model of flooding conditions. This new diagnosis approach directly provides the percentage of flooded channels in the fuel cell stack as a measure of the flooding of the fuel cell stack.

► Experimental tests of PEM fuel cell for diagnosis.
► A reverse fluidic model quantifies the flooding degree of a PEM fuel cell.
► The variance of the single cells' voltages can be used as a flooding sensor.