TY  - CONF
AU  - Hammacher, Linus
AU  - Karyofylli, Violeta
AU  - Kuppa, Raman Ashoke
AU  - Danner, Yannik
AU  - Kungl, Hans
AU  - Karl, André
AU  - Jodat, Eva
AU  - Eichel, Rüdiger-A.
TI  - Elucidating Parasitic Currents in Proton-Exchange Membrane Electrolytic Cells Via Physics-based and Data-driven Modeling
M1  - FZJ-2025-01768
PY  - 2025
N1  - Acknowledgment: Financial support was provided by the German Federal Ministry of Education and Research (BMBF) within the H2Giga project DERIEL (grant number 03HY122C).
AB  - Proton-exchange membrane (PEM) water electrolysis plays a crucial role in green hydrogen production. To accelerate commercial deployment, it is pertinent to use efficient computational models which capture the inherent non-linearities and aid to system optimization. This poster presentation focuses on understanding degradation mechanisms, particularly the impact of parasitic currents on the performance of a PEM electrolytic cell (PEMEC) through macro-scale modeling and uncertainty quantification (UQ) [1]. Parasitic currents due to electron conduction through the membrane are a frequently observed but not fully understood degradation effect, leading to lower Faradaic efficiency. One possible cause of these parasitic currents is mechanical damage in the membrane-electrode assembly (MEA) [2]. To specifically address the effect of such parasitic currents on Faradaic efficiency and cell performance under varying design parameters, we present a one-dimensional steady-state physics-based model for PEMECs. A comprehensive dataset from this model is generated and used to train a machine learning (ML) surrogate model. Its performance is analyzed to assess the potential of ML in accurately and efficiently predicting the effects of parasitic currents in PEMECs. The chosen ML algorithm, eXtreme Gradient Boosting (XGBoost), excels in predicting the polarization behavior while significantly reducing computational demands. Using this ML surrogate model, UQ and sensitivity analysis (SA) [3] are applied to investigate the dependence of PEMEC performance and Faradaic efficiency on the electronic conductivity of the PEM, especially when electronic pathways are existent within the membrane and operating at low current densities.References:[1] V. Karyofylli, K. A. Raman, L. Hammacher, Y. Danner, H. Kungl, A. Karl, E. Jodat, R.-A. Eichel, Accepted by Electrochemical Science Advances on 01/2025[2] S. P. S. Badwal, S. Giddey, F.T. Ciacchi, Ionics 12 (2006), 1, 7-14 [3] V. Karyofylli, Y. Danner, K. A. Raman, H. Kungl, A. Karl, E. Jodat, R.-A. Eichel, J. Power Sources 600 (2024), 234209
T2  - 21st Symposium on Modeling and Experimental Validation of Electrochemical Energy Technologies
CY  - 11 Mar 2025 - 12 Mar 2025, Karlsruhe (Germany)
Y2  - 11 Mar 2025 - 12 Mar 2025
M2  - Karlsruhe, Germany
LB  - PUB:(DE-HGF)24
UR  - https://juser.fz-juelich.de/record/1039720
ER  -