Experimental Study on Dynamic Performance of a 50 kW PEM Water Electrolysis System for Hydrogen Production

Le résumé fourni par la source

With the acceleration of the global energy transition, hydrogen is increasingly considered a potential energy carrier for renewable-energy integration, large-scale energy storage, and industrial decarbonization. Proton exchange membrane (PEM) water electrolysis is well suited to variable renewable power because of its fast load response, wide operating range, and compact system structure. However, most existing studies focus on steady-state performance, materials, or model-based analysis, while system-level experimental data on the dynamic behavior of industrial-scale PEM water electrolysis systems remain limited. In this study, the dynamic performance of a 50 kW-class PEM water electrolysis system was experimentally investigated under stepwise load changes, pressure variation, and cold-start conditions. The responses of voltage, temperature, pressure, hydrogen-in-oxygen (HTO), oxygen-in-hydrogen (OTH), and system energy consumption were analyzed. The voltage followed current step changes within seconds, indicating a fast electrical response. In contrast, the thermal response was much slower, and the system required approximately 34 min to approach the rated thermal condition from a cold start. The gas-composition measurements exhibited minute-scale response delays and gradual settling after changes in operating conditions. When the operating pressure increased from 1.2 MPa to 2.9 MPa, the HTO content increased from 0.383% to 0.545%. When the current increased from 300 A to 1200 A, the OTH content decreased from 1001.77 ppm to 5.86 ppm. Energy-flow analysis showed that the total system power consumption under full-load operation was 69.7 kW, including the electrolyzer-related part and balance-of-plant consumption. These results clarify the different response time scales of electrical, thermal, and gas-composition variables in a 50 kW-class PEM water electrolysis system and provide experimental support for dynamic operation under variable renewable power input.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.