Extending a Multiphysics Li-Ion Battery Model from Normal Operation to Short Circuit and Venting

Abstract

Mitigation of Li-ion battery system fires consists of reliable fault detection and proactive, fast discharge control. Both require modeling of failure modes due to high temperatures and currents between normal operation and thermal runaway. In this work, we present a control-oriented, reduced-order, multi-physics model that captures the electrochemical, thermal, gas generation, mechanical expansion, and venting behavior of NMC pouch cells undergoing an external short circuit (ESC) from different initial state-of-charge (SOC). The model is parameterized through experiments by fitting the solid-electrolyte interphase (SEI) decomposition rate, the cell's thermal parameters, and the particle solid-phase diffusion parameters to capture the first venting timing, peak temperature, and diffusion-limited electrical behavior at high currents. Using a single parameter set, the multi-physics model can capture behavior during an ESC to predict whether a cell will generate gas and vent, predict the vent timing within 10 seconds of it occurring in the experiment, and maximum cell expansion pressure within 10 kPa for cells that did not vent. The model can also predict the SOC trajectory for cells with a high initial SOC within 6% SOC for the 15-minute discharge or until the cell vents.