A review of portability of fault diagnosis methods from PEM fuel cells to PEM water electrolyzers

With the global push for clean energy and carbon neutrality, proton exchange membrane (PEM) fuel cells and PEM electrolyzers are becoming essential technologies for sustainable energy systems. Fault diagnosis is crucial for their reliable operation. However, while fault diagnosis methods for PEM fuel cells have been extensively studied, research on PEM electrolyzers remains limited. This paper explores the portability of fault diagnosis methods from PEM fuel cells to PEM electrolyzers. Analyzing the structural similarities and operating principles of PEM fuel cells and PEM electrolyzers reveals common fault modes, such as water, thermal, gas, and electrical management issues, which provide a solid theoretical foundation for the portability of diagnostic methods. Moreover, both systems can acquire essential sensor data—including voltage, current, temperature, and flow rate—offering crucial data support for the transfer of these fault diagnosis methods. Specifically, data-driven black-box model methods, feature extraction, feature transformation, and data preprocessing techniques developed for PEM fuel cells can serve as valuable references for PEM electrolyzers, enhancing diagnostic accuracy and real-time performance. This work fills a research gap by offering a systematic framework for assessing diagnostic method transferability and contributes to improving PEM electrolyzer reliability and advancing the hydrogen economy.