A resource-efficient in-situ thermal modeling method for second-life lithium-ion batteries for stationary applications

The growing demand for electric energy and the increasing number of retired electric vehicle batteries with remaining energy storage capabilities make second-life applications a key opportunity for circularity and profitability. Since these batteries have lost performance capabilities due to prior use, it is important to assess their current state and thermal behavior to design adequate Thermal Management Systems and ensure safety. As these batteries vary in chemistry and format, and cell accessibility for testing is rare, a robust characterization method that avoids disassembly or extra sensors is needed. This article presents an optimized, resource-efficient thermal modeling method that only requires BMS records and is applicable in-situ to batteries already in operation. This method builds a 0D lumped thermal model that simulates battery temperature with an RMSE below 1 °C across different currents, cooling conditions, and battery degradation states. While traditional models focus only on the heat generated by the Joule effect, this model also considers reversible heat. At currents below 0.5 C , such as those recommended for second-life batteries, the impact of reversible heat has been found to reach 60 % of the total heat generated. This considerably improves the model’s accuracy and enables more precise studies of battery safety. The method is applied to model two NMC battery modules with different cell formats, configuration and rated energy. However, due to its data driven nature and the promising results obtained, the method proposed could be extended to other cooling strategies or battery chemistries, which is regarded as future lines of this work.