PREDICTION FAILURE FOR PEM FUEL CELLS

A new conceptual methodology and some methods are used to predict failures that could potentially occur in Proton Exchange Membrane Fuel Cell (PEMFC) systems. The combination methods for prediction durability and safety for fuel cell design starting with matrices of technological process, function, components and requirements for PEM fuel cell systems. After input with characteristic date are applied adequate some methods like Failure Modes and Effects Analysis (FMEA), fuzzy method and Fault Tree Analysis (FTA) for prognostic and analysis failure for PEMFC system or/and components, like product or/and process. For application and solving objectives according to the methodology proposed, as a case study to consider the methods specified for fault prediction in a PEM fuel cell type, based on analysis of process parameters like pressure flow of hydrogen and oxygen (or air), electric voltage, electric current and the humidification of the proton exchange membrane. These variables determining the functioning of the fuel cell are adequately analyzed with Fuzzy Fault Tree method (FFT). Methodology algorithm is solved using LabVIEW software provided by the National Instruments. The proposed methodology is validated by specified references from scientific literature under experimental and modelling appearance