This work shows the application of artificial neural networks in terms of modeling and simulating the aging process and the degradation of proton exchange membrane water electrolysis stacks. It includes the training process based on extracted measurement data, the evaluation, and the extrapolation of the network. The fundamentals of the utilized artificial neural network and the training algorithm are clarified. Next, the principle degradation effects are presented as well as the methodology of the underlying measurements. The resulting degradation of the electrolysis stack for different operation conditions is shown.
This paper presents a numeric model for proton exchange membrane electrolysis, which describes the current‐voltage dependency. Besides physical constants and parameters characterizing material properties, it also contains parameters that cannot be measured ex situ, e.g., charge transfer coefficients and exchange current densities. To determine these parameters, the model includes an automated parameter calibration procedure that is able to find the best parameter combinations. An exemplary integration of the model in MATLAB/Simulink is presented, including a validation with dynamic measurement data.
To test commercially available water electrolysis systems under dynamic (close‐to‐real) operation a test area with three different electrolyzers is build up, covering the significant technologies alkaline water electrolysis, proton exchange membrane water electrolysis, and high temperature steam electrolysis using solid oxide electrolysis cells. The hydrogen outputs of the systems are in the range between 5 to 10 Nm3 h−1 hydrogen at delivery pressures between 10 and 35 bar. Additional balance of plant is installed to demonstrate and evaluate the utilization of hydrogen for different applications.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.