Evaluation of Gas Bubble Coverage in a Flow Field by Deep Learning-Based Analysis: The Impact of Gas Content on PEM Electrolyzer Performance

Journal of The Electrochemical Society, vol. 172, pp. 084501

Abstract

Oxygen gas generated at the anode of PEM electrolyzers turns the flow in the anodic flow field into two-phase regime. Along with upscaling the electrolyzers and the objective to reach higher current densities, the impact of the two-phase flow regime becomes an issue of major concern for their future design and operating conditions. To observe and quantify gas bubble formation and local gas content in the flow field channels during operation using a high-speed camera, a test rig for a 25 cm 2 scaled laboratory electrolyzer cell with a transparent flow field has been set up. Current-voltage curves were recorded, while monitoring at the same time the impedance and the gas evolution in the anode flow field. Gas bubbles were distinguished by a deep learning-based image processing algorithm, revealing enhanced gas area coverage and bubble size with increasing current density. Corresponding calculations from a two-phase flow mixture model show a similar trend for local gas content in the flow field. In the high current density region, a strongly increasing voltage along with changes in the impedance pattern coincide with the transition from the flow regime with individual gas bubbles to an interconnected gas flow regime in the majority of the cell.

Authors 9

1. Violeta Karyofylli corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

2. André Colliard-Granero corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Theory and Computation of Energy Materials (IET-3), Forschungszentrum Jülich, 52425 Jülich, Germany

3. Christoph König corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

4. Ali Javed corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

5. Michael Hermann Eikerling corresponding Aachen Chair of Theory and Computation of Energy Materials Faculty of Georesources and Materials Engineering

   Forschungszentrum Jülich · RWTH Aachen University

   Affiliation as printed

   Chair of Theory and Computation of Energy Materials, Faculty of Georesources and Materials Engineering, RWTH Aachen University, 52072 Aachen, Germany

   Theory and Computation of Energy Materials (IET-3), Forschungszentrum Jülich, 52425 Jülich, Germany

6. Hans Kungl corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

7. Eva Jodat corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

8. André Karl corresponding

   Forschungszentrum Jülich

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

9. Rüdiger‐A. Eichel corresponding Aachen Institute of Physical Chemistry

   Forschungszentrum Jülich · RWTH Aachen University

   Affiliation as printed

   Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich, 52425 Jülich, Germany

   Institute of Physical Chemistry, RWTH Aachen University, 52074 Aachen, Germany