A new concept of water management diagnosis for a PEM fuel cell system. (1st June 2023)

• Record Type: Journal Article
• Authors: Ziane, M. Ait; Join, C.; Benne, M.; Damour, C.; Steiner, N. Yousfi; Pera, M.C.
• Is Part Of: Energy conversion and management. Volume 285(2023)
• Publication Date: 2023-06-01
• Subjects: PEMFC water management diagnosis -- Membrane dehydration -- Water condensation -- Sensor fault detection; Direct energy conversion -- Periodicals; Energy storage -- Periodicals; Energy transfer -- Periodicals; Énergie -- Conversion directe -- Périodiques; Direct energy conversion; Periodicals; 621.3105
• DOI: 10.1016/j.enconman.2023.116986
• Languages: English
• ISSNs: 0196-8904

Abstract

Abstract: This paper presents a new diagnosis view of fuel cell water management. The faults related to water management in the fuel cell are addressed from the control point of view and their occurrence is considered as a consequence of a temperature sensor fault. To ensure proper operation of the polymer electrolyte membrane fuel cell (PEMFC), the stack temperature and inlet pressure are controlled by a model-free control. A fault in the temperature sensor that causes an imbalance in water content in the stack is detected by a new fault detection approach in model-free context. The validation of the proposed strategy is performed on a 1.2 kW fuel cell with real time detection of temperature sensor faults, leading to water condensation in the cells or membrane dehydration. Fuel cell control and diagnosis are achieved without any requirement of an accurate system analytical model and additional sensors. However, faults can be detected in quasi static operation only and not un transient state. Highlights: Model-free fault detection approach is applied to PEMFC water management fault diagnosis. A theoretical proof is presented for model-free fault detection. Sensor fault is the main cause of water condensation and membrane dehydration. Real-time fault detection in the back pressure, temperature sensor and water management.