Kinetic modelling of thermal decomposition in lithium-ion battery components during thermal runaway

This study presents kinetic models for the thermal decomposition of 18650-type lithium-ion battery components during thermal runaway, including the SEI layer, anode, separator, cathode, electrolyte, and binder. The decomposition kinetics were sourced from the literature. The approach used inverse modelling, employing a Genetic Algorithm to estimate kinetic and stoichiometric parameters. Experimental thermogravimetry data from the literature served as the reference benchmarks. The optimisation errors ranged from 0.039 % to 1.531 % , and the algorithm performed well in terms of reaction temperatures, with errors between 0.51 % and 11.07 % . The models were validated for calculating the mass loss of a full cell at 100 % state of charge during thermal runaway. The early stages of thermal runaway, including the decomposition of the anode and separator, were considered in an electrochemical-thermal simulation of charge/discharge cycling using PyBaMM solver. The results showed that these decompositions could advance temperature and voltage profiles by 0.07 C over 20 cycles, aiding early prediction of thermal runaway in battery management systems. This work introduces novel models to calculate mass losses, identify reactions, quantify heat release, and estimate thickness or volume reductions in battery components during thermal runaway. • Kinetic models generated for thermal decomposition of Li-ion battery components. • A Genetic Algorithm used to estimate kinetic and stoichiometric parameters. • Optimisation errors ranged from 0.039 % to 1.531 % . • Models validated for mass loss calculation during thermal runaway. • Models employed for simulating early thermal runaway stages using PyBaMM.