A Development of Pore Network Model Coupled Thermal Transport for Hydrophilic Porous Media of PEM Electrolyser

Hydrogen is a highly promising energy carrier vector. Water electrolysis process that provides splits of water to hydrogen and oxygen through the electricity has gained great recognition for hydrogen production because it provides high conversion efficiency, clean and purified hydrogen in recent years. A Polymer Electrolyte Membrane electrolyser that brings important advantages such as compactness, and the ability to work at low operating temperature, is an electrochemical device. From the gas channels in the cathode, the water enters the hydrophilic porous transport layer (PTL), also known as the gas diffusion layer (GDL), to the catalyst layer (CL) in which the electrolysis reactions occur. At high current densities (higher than 1 A/cm 2 ), the PEM electrolyser is also favorable to produce a high amount of hydrogen; however, management of heat and mass transfer in the GDL becomes crucial as the two-phase flow occurs, and can be lead to hot spots. Moreover, the mass transport overpotentials become important issues at high current densities and lead to be 25% losses in the overall electrolyzer overpotential. A development of model is crucial to understand and enhance the heat and mass transport mechanisms in CL-GDL interface. Here, pore network model (PNM) has gained a significant interest as it offers strong tool to simulated multiphase flow in three-dimensional (3D) porous media for electrochemical energy conversion. The principle behind of this modelling approach is that a porous medium behaves as a network of pores that connected by throats. Through this framework, the discretization approach to transport modelling becomes dramatically simple, and the computational cost and time are reduced among the continuum modelling approaches (e.g., Lattice-Boltzmann, CFD). In this work, to the best knowledge of authors, the first pore network model coupled heat transfer model is developed to characterize available GDL. Here, the pore network is extracted from the micro CT image of a hydrophilic GDL through the Snow algorithm employed in OpenPNM software (An open-source Pore Network Modeling Package) and reconstructed in this environment, while the geometry of catalyst layer is created randomly. The effective two-phase 2D model is developed to predict water transport in the GDL and CL through the percolation algorithm. For this purpose, the sink and source terms are employed for some pores in the network (especially near GDL-CL interface) according to electrolyser