Electrochemical modeling and parameter calibration for lithium-ion battery electrolyte degradation

Lithium-ion batteries, as the core energy storage medium for electric vehicles, energy storage systems, and 3C digital products, rely on internal electrolyte reactions to ensure both performance and safety. Existing methods for assessing electrolyte content face significant limitations: destructive testing prevents online application, indirect characterization does not capture the bulk-phase status, and multi-component monitoring suffers from signal interference. In contrast, electrochemical modeling can extract internal degradation information, providing theoretical support for full-life cycle monitoring of electrolyte depletion. However, conventional electrochemical models exhibit shortcomings in parameter identification and safety mechanism analysis, which limit their ability to track dynamic changes in electrolyte composition and affect the accurate assessment of battery performance and aging behavior. To overcome these challenges, this study develops a multi-scale coupled P2D-SEI-solvent consumption model, which explicitly establishes the mechanistic linkage between SEI growth and electrolyte depletion within a unified electrochemical framework. Based on this model, a full-lifecycle parameter calibration strategy combining a genetic algorithm and a support vector machine is proposed, enabling a non-destructive quantitative assessment of electrolyte depletion. The approach is further experimentally validated through electrolyte-filling gradient tests, allowing direct verification of the model predictions. The proposed framework enables mechanistic tracking and quantitative evaluation of electrolyte aging over the entire battery life cycle, providing a new pathway for electrolyte health monitoring and safety assessment in lithium-ion batteries.