A rapid calculation approach for PEM electrolyzer stack simulation based on flow resistance network and 3+1D multi-physics model

Simulating large-scale proton exchange membrane (PEM) electrolyzer stacks poses significant computational challenges. In this study, a novel rapid simulation framework is proposed and validated on a 4-cell PEM electrolyzer stack with each cell having an active area of 25 cm2. The methodology innovatively integrates a flow resistance network model for resolving hydraulic distribution among parallel cells, with a domain-decomposed 3D+1D multi-physics model. This hybrid approach retains full 3D resolution for the complex anode gas-liquid flow while simplifying the membrane and cathode into 1D electrochemical-thermal elements, thereby drastically reducing mesh complexity. Validation against a full-scale 3D CFD benchmark demonstrated that the proposed model captures performance variations caused by flow maldistribution, maintaining prediction errors for pressure, temperature, and current density below 6.68%. Crucially, the method reduces computational time by over 80% compared to the full 3D model while preserving essential physical heterogeneities. Furthermore, the study elucidates the critical influence of manifold distribution on the stack's internal multi-physics fields and polarization performance. Results reveal that hydraulic maldistribution directly dictates the spatial gradients of pressure and temperature, leading to significant voltage variations across the stack layers.