Fault Detection and Identification for Polymer Electrolyte Membrane Fuel Cell Stack by External Magnetic Field

Abstract

Many technological locks prevent the deployment of Proton Exchange Membrane Fuel Cell (PEMFC). Particularly for vehicle applications, cost and durability represent two of the most significant challenges to achieve clean, reliable, and cost-effective PEMFC systems. Solving these shortcomings will be a great opportunity for a decisive breakthrough towards mass-diffusion of the current PEMFC technology. The long-term electrical performance losses are associated to either the degradation of the MEA constitutive materials or due to specific operating conditions (flooding, drying, etc.). In perspective, local current density distribution may support stack-level degradation analyzes as well as effective stack control to mitigate the consequences of degradation due to load cycles (electrochemical, thermal, mechanical and humidity effects) or faults.External magnetic field measurement (MFM) shows a high potential in PEMFC non-invasive diagnosis. Thus, 2D and 3D current distributions and their anomalies in the fuel cell can be identified independent of the size of the causing fault.To perform this local characterization, the external magnetic field measurement will enable to model the current density in stationary conditions over the cell surface and for the different cells of the stack. This methodology requires the complementary development of magnetic field techniques and inverse models.In this work, this challenge is tackled by using tailored defective MEAs and specific operating conditions (flooding, drying, ...) to characterize, how local and overall performances of the MEA are affected and to identify the signatures of the various anomalies. Figure 1