Degradation analysis of dynamic voltage response characteristics of proton exchange membrane fuel cells for health evaluation under dynamic load

Abstract

Accurate health assessment for proton exchange membrane fuel cells (PEMFC) can facilitate early control/maintenance, thereby extending the lifespan and reducing costs. Existing health factors and online characterization of degradation mechanisms are mainly based on voltage or power. However, the dynamic voltage response under the dynamic load cycles and complex conditions poses significant challenges in assessing the degradation states. This paper provides an in-depth understanding of the degradation behavior in the dynamic voltage response. First, a dynamic voltage decomposition method is proposed, using differences in high-frequency impedance obtained from extended distribution of relaxation times (DRT) before and after load changes. Subsequently, a relative transient resistance is introduced to represent and quantify the degradation behavior of different components of the dynamic voltage response. The results show that platinum oxide and gas diffusion account for more than 80 % of the total contribution and they dominate the degradation of the dynamic response. The degradation behavior of each component is investigated in detail. Key determinants are identified and validated through a proposed semi-empirical mechanistic dynamic model (RMSE values are less than 0.0259). Furthermore, Pearson correlation between the polarizations and inductive processes is evaluated with a coefficient of 0.94. Finally, a perspective for extracting health factors considering dynamic responses is proposed. Compared to traditional voltage-based health factors, R_eq and R_ne extracted from the negative resistance equivalent circuit exhibit stronger monotonicity over time, a higher trend-to-variance ratio, and lower time-series complexity, offering significant advantages in health assessment.