Estimation of internal fuel cell temperatures from surface temperature measurements

In this paper, a method to monitor PEM fuel cells internal temperature from surface measurements is presented. The aim of this work is to monitor fuel cells to prevent damages due to internal overheating. The measurements are taken at the side of the bipolar plate, and heat flux and temperature at the border of the active zone are estimated. The method is based on sensitivity analysis and inverse problem algorithms. The mathematical formulation and algorithm are described. The model is a transient heat conduction model in two dimensions, the inverse problem is solved with an optimization method using adjoint equation. Numerical test cases are presented for graphite and steel bipolar plates. The results show that internal temperature can be correctly estimated. The response time of the method is limited by the heat transfer rate in the material. Therefore, the method is particularly appropriate to fuel cells made of steel bipolar plates.