• A multi-scale PEMFC cold-start performance and catalyst degradation model. • Model validated against FCEV start-up data from −18 °C to 35 °C. • The spatio-temporal results reveal intra-cell heterogeneities in PEMFC during start-up. • Sub-zero starts show ∼100 times higher per-start ECSA loss than warm starts. • Predictive framework supports diagnostics and optimization of cold-start durability. Proton exchange membrane fuel cells (PEMFCs) face durability challenges during cold start-up, particularly at sub-freezing ambient temperatures. This study introduces a multi-domain and multi-scale modeling framework to predict PEMFC cold start-up behavior and degradation at various ambient temperatures. The framework couples a detailed system-level electrochemical model with a mechanistic catalyst degradation model that resolves platinum dissolution and carbon support corrosion in space and time. Validated against Toyota Mirai data from −18 °C to 35 °C, simulations show ice-induced heterogeneities causing severe local hydrogen starvation and high electrode potentials, leading to catalyst electrochemical surface area (ECSA) degradation rates with peak instantaneous values of about two orders of magnitude higher than in warm start-ups. When integrated over the first 200 s, the per-start cumulative ECSA loss at sub-zero temperatures is approximately 100 times higher than for warm start-ups and is consistent with recent stack and short-stack cold-start ECSA degradation diagnostics. By resolving the spatio-temporal distribution of reactions and degradation hotspots within the cell, the model clarifies why colder start-ups accelerate catalyst degradation and underscores the importance of ambient temperature and control strategy in predicting fuel cell catalyst longevity. The framework, therefore, enables predictive diagnostics of cold-start performance and provides a virtual testbed for mitigation and start-up protocol optimization, supporting improved PEMFC durability under realistic operating environments.
Building similarity graph...
Analyzing shared references across papers
Loading...
Davor Rašić
Andraž Kravos
Tomaž Katrašnik
Energy Conversion and Management
University of Ljubljana
Building similarity graph...
Analyzing shared references across papers
Loading...
Rašić et al. (Wed,) studied this question.
www.synapsesocial.com/papers/69a134fbed1d949a99abe733 — DOI: https://doi.org/10.1016/j.enconman.2026.121254