Towards an ageing model of a PEMFC for prognostics purpose

One of the main constraints of Proton Exchange Membrane Fuel Cell (PEMFC) in its breakthrough in the industrial world is its life duration [1], which is too short and not enough managed. From this point of view, the development of Prognostics and Health Management (PHM) technology [2] appears to be a relevant discipline; the estimation of the Remaining Useful Life (RUL) of a fuel cell system enables deciding for actions in order to try to mitigate the degradation and ensure longer life duration and availability. This paper addresses this topic and aims at proposing a basic FC ageing model [3][4] that will serve for prognostics issues. Here, the proposed ageing model is based on the combination of a static model with a dynamical one. A physical based model is considered built on a Butler-Volmer law where the voltage is function of the current density and of fuel cell's physical parameters. This analytical static model allows the separation of cathode and anode contribution. The dynamical model is based on an electrical equivalence; on which different internal components can be distinguished. This model is expressed in a state-space representation and defines the voltage depending on the current. Moreover, the model also provides the possibility of modelling the ageing of the Polymer Exchange Membrane Fuel Cell. Indeed, different internal parameters can be represented using time-dependent functions. The limits of this model and its suitability for prognostics are discussed. The model validation is based on actual long-duration experimental tests.