Prognostic Approaches for Early Detection of Thermal Runaway in Lithium‐Ion Batteries

ABSTRACT Thermal runaway in a lithium‐ion cell is an uncontrollable self‐heating process, which can result in a fire, smoke, or in some cases, an explosion. Since thermal runaway cannot always be prevented, there is interest in detecting it prior to its onset. This paper reviews experimental and modeling studies focused on the early detection of thermal runaway, including methodologies that incorporate voltage, current, impedance, state of charge, temperature, force, strain, pressure, acoustic, and gas‐emission indicators.

The paper then compares the methodologies based on their capability to detect thermal runaway triggered by different types of causes (electrical, mechanical, and thermal), the extent of advance warning provided for effective mitigation or evacuation, detection accuracy (minimizing false positives and false negatives), compatibility with different cell chemistries and form factors, implementation complexity, and system‐level practicality. Beyond detection, the paper reviews safety intervention strategies to deploy after anomalies are identified, including battery management system (BMS)‐based protective actions, active cooling approaches, thermal propagation mitigation techniques, and fire‐suppression methods.