CFC2023

Towards sensitivity analysis and uncertainty quantification in PEM water electrolyzers

  • Karyofylli, Violeta (Fundamental electrochemistry (IEK-9), Institute of Energy and Climate Research, Forschungszentrum Juelich, 52425 Juelich, Germany)
  • Bauer, Alexander (Fundamental electrochemistry (IEK-9), Institute of Energy and Climate Research, Forschungszentrum Juelich, 52425 Juelich, Germany)
  • Eichel, Ruediger-Albert (Fundamental electrochemistry (IEK-9), Institute of Energy and Climate Research, Forschungszentrum Juelich, 52425 Juelich, Germany)

Please login to view abstract download link

This talk aims to highlight the development of a numerical electrochemistry model for a PEM water electrolyzer (PEMWE) cell within the scope of sensitivity analysis. The cell consists of five layers, i.e., the anode porous transport layer (aPTL), the anode catalyst layer (aCL), the membrane (MEM), the cathode catalyst layer (cCL) and the cathode porous transport layer (cPTL). The anode and cathode flow channels are adjacent to the aPTL and cPTL, respectively. The current density and its distribution over the five-layer PEMWE cell represents one important quantity of interest in our electrochemistry model. Based on the computational model and Faraday’s law, we can estimate the flux of oxygen into the anode flow channel and the flux of hydrogen into the cathode flow channel. This way, we can achieve the coupling between electrochemistry and the two-phase flow in the channels, which is our ultimate goal. The electrochemistry model is comprised of the charge balance equations for electrons and protons, the continuity equations for all species in our system, which are three of in total (e.g., hydrogen, oxygen and water), and the heat equation. The high-aspect-ratio PEMWE cell lends itself to a one-dimensional (1D) analysis, meaning the electrochemistry model consists of a set of 1D steady partial differential equations (PDEs). Specifically, we have conducted an initial sensitivity-analysis study using the libraries UQLab, Chaospy and Uncertainpy, in combination with the PEMWE solver, to determine which input parameters (i.e., conductivity coefficients and transport parameters) mainly influence the quantities of interest, e.g., the current density.