@Article{ee.2026.085579,
AUTHOR = {Hangyu Sun, Yang Wang, Sixi Zha, Jianping Gao, Congfa Yang},
TITLE = {A Systematic Review of Water-Oxygen Transport Performance in Porous Transport Layers of PEM Electrolyzers Based on Pore Network Models},
JOURNAL = {Energy Engineering},
VOLUME = {},
YEAR = {},
NUMBER = {},
PAGES = {{pages}},
URL = {http://www.techscience.com/energy/online/detail/27636},
ISSN = {1546-0118},
ABSTRACT = {Proton exchange membrane water electrolyzers (PEMWEs) require efficient liquid-water supply and oxygen removal in the anode porous transport layer (PTL), where pore-scale heterogeneity strongly affects mass transport, interfacial contact, and high-current-density operation. Unlike previous PTL reviews that mainly focus on materials, fabrication routes, surface modification, durability, or cell-level performance, this systematic review specifically examines how pore network models (PNMs) translate PTL microstructural descriptors into quantitative water–oxygen transport predictions. The reviewed studies show that pore and throat sizes, porosity gradients, tortuosity, connectivity, anisotropy, and wettability jointly determine liquid-water permeability, capillary invasion, gas saturation, preferential oxygen pathways, and interfacial mass-transfer resistance. Recent PNM and complementary pore-scale studies have reported strong gas accumulation under high-current-density operation; for example, dual-scale PNM predicted gas-phase saturation values of 29% in the catalyst layer and 23% in the PTL at 2000 mA cm<sup>−2</sup>, while tailored porosity configurations were shown to reduce oxygen saturation and improve oxygen expulsion at 5 A cm<sup>−2</sup>. This review summarizes PNM construction strategies, single-phase and two-phase transport applications, multiphysics coupling frameworks, and experimental validation requirements. The analysis indicates that PNM provides a computationally efficient bridge between PTL microstructure design and transport-performance prediction, but its reliability remains limited by reconstruction uncertainty, simplified interfacial physics, static wettability assumptions, and insufficient operando validation. Future work should prioritize uncertainty-quantified pore-network construction, dynamic two-phase modeling, stronger multiphysics coupling, and direct validation using <i>in situ</i> or operando diagnostics.},
DOI = {10.32604/ee.2026.085579}
}