Investigation of an Indicator for On-line Diagnosis of Polymer Electrolyte Membrane (PEM) Fuel Cell Flooding using Model Based Techniques

Abstract

The durability and reliability of producing high quality power for long periods of time have the potential to be the leading marketing factors for future hydrogen and fuel cell power sources. In the past few decades, several researchers have been devoted to investigating diagnostic techniques for fuel cell systems. However, in commercial fuel cell applications, on-line diagnosis is urgently required so that fuel cell degradation can be detected at its early stage, and mitigation strategies can be performed to recover fuel cell performance. In this paper, on-line diagnosis of fuel cell flooding is investigated. For this purpose, a generalised fuel cell stack model is developed, and water mass balance equations are used to study water balance inside the fuel cell stack. Moreover, with these equations, the flooding indicator is proposed and its relationship with liquid water inside the stack is evaluated. Results demonstrate that the proposed indicator is sensitive to the liquid water in the stack, thus can be used for flooding diagnosis. Furthermore, the expectation of the proposed indication in the cell flooding case is also presented. The advantage of this method is that parameters in the flooding indicator can be determined with measurements from tests, thus quick diagnosis can be made during the practical fuel cell operation.