Electro-Thermal Coupling Pathways Governing Catastrophic Failure in Parallel Lithium-Ion Modules

Lithium-ion batteries store remarkable amounts of energy in a compact form and survive thousands of charge–discharge cycles with relatively little degradation which make them the centre of modern electrification. However, when these cells are grouped together into modules, the neat boundaries that separate electrical and thermal behaviour in single-cell tests begin to blur. A fault in one cell may, under the wrong circumstances, spill into its neighbours. Thermal runaway remains the primary fear and although the chemistry behind these reactions has been dissected in great detail, the way electrical currents and heat flux interact within a parallel module is still not fully mapped out. Parallel configurations are widely used in high-capacity systems, yet they bring their own peculiar vulnerabilities: the moment a cell develops an internal short, its partner can feed it additional current, deepening the local heating and altering the sequence of events that follow. Much of the earlier work in this area offers useful clues, but many studies lean on assumptions that do not quite match what happens inside an actual module. In modelling studies, for example, cells are sometimes treated as electrically isolated or given fixed resistance values, even though a cell’s resistance can swing dramatically as temperature climbs. These simplifications tend to wash out the rapid changes in current paths and thermal gradients that occur during failure. Diagnostic approaches based only on surface temperature or voltage drift miss the faster transitions that signal the onset of electro-thermal imbalance. Without those details, our understanding of how heat accumulation, electrical asymmetry, and chemical degradation reinforce one another remains incomplete. The need for clearer insight becomes even more pressing when we look at how batteries fail in practice. A striking number of large energy-storage incidents happen not during idle periods but while the batteries are actively cycling. Discharge currents shift the state of charge and reconfigure internal reaction pathways in ways that may either buffer the system against runaway or, conversely, make it more fragile. However, controlled experimental work on operational modules—where electrical load and thermal abuse occur simultaneously are surprisingly rare. That absence leaves a gap in hazard assessments and, ultimately, in the design principles meant to keep these systems safe. To this account, new research paper published in Journal of Energy Storage and conducted by Professor Yuanhua He, Jiaxin Liang, Jianquan Jin, Zitong Li, and led by Dr. Jiang Huang from the Civil Aviation Flight University of China, researchers developed two complementary electro-thermal models that quantify heat generation during both active discharge and short-circuit conditions in parallel lithium-ion modules. One model evaluates Ohmic and reversible heat under varying discharge rates, while the other captures Joule heating during the rapid current surge following internal short circuits.

The research team constructed a parallel two-cell module using commercial prismatic lithium-ion cells and arranged it within an explosion-proof chamber equipped with high-speed imaging, infrared thermography, and multi-channel electrical sensing. Their setup carefully reproduced the close-contact, minimally insulated environments typical of battery modules, including a surface-mounted heating plate that initiated runaway in one designated “directly heated” cell. Thermocouples were embedded not only along cell surfaces but also within the electrode terminals, allowing the team to distinguish bulk heating from electrode-localized heating due to short-circuit currents. The authors performed the experiments under two broad conditions. In the non-operational case, the module rested at full charge while external heating forced the first cell into runaway. Under operational conditions, the module was discharged at rates of 1C, 2C, and 3C, and thermal abuse was applied simultaneously. This arrangement introduced electrical load balancing and evolving state-of-charge differences between the two cells, which the researchers expected would modify the runaway process.

In the non-operational module, the directly heated cell entered runaway at roughly 145 °C, accompanied by a striking spike in short-circuit current—nearly 498 A—which delivered intense Joule heating to the electrode tabs. This current surge accelerated internal reactions, producing a peak temperature above 430 °C. As heat migrated to the second cell, its safety valve opened and runaway was triggered at a considerably higher onset temperature. However, the severity of runaway in the second cell was muted: its peak temperature was more than 40 °C lower, and its temperature rise rate was only a fraction of the first cell’s. This asymmetry reflected the rapid discharge of the second cell into the first one during the short-circuit interval, which lowered its available stored energy before its own runaway began. The operational tests altered this pattern. When the module was already discharging, the short-circuit current generated after the first cell failed was reduced substantially—falling to 420, 357, and 309 A at 1C, 2C, and 3C respectively. The authors found the higher discharge rates depleted the state of charge more quickly, raising the runaway onset temperatures but lowering the total heat release. Interestingly, as discharge rate increased, the temperature gap between the two cells shrank dramatically. At 3C, the peak temperatures differed by only 7 °C, which suggest that the electrical energy exchange between the cells had weakened to the point where runaway propagation was less forceful. Moreover, a quantitative model incorporating Ohmic, reversible, and Joule heating supported these observations. Discharge-related heating rose steeply with discharge rate, but Joule heating from short-circuit currents declined. The model showed that at high discharge rates, short-circuit heat accounted for less than 5% of total runaway heat which is an important shift that helps explain the reduced propagation severity observed experimentally.

In conclusion, the new study by Professor Yuanhua He, Dr. Jiang Huang and colleagues provides a clearer picture of how intertwined electrical and thermal behaviours govern the escalation of failure within parallel lithium-ion modules. One of the most consequential insights is that the short-circuit current—typically assumed to be a dominant source of heat during propagation—does not act uniformly across all operating conditions. In modules at rest, the high state of charge in the neighbouring cell permits a large burst of current into the failing cell, sharply intensifying runaway. However, as the authors demonstrated, when the module is actively discharging, much of the stored energy has already been removed, and this dramatically limits both the magnitude and duration of short-circuit currents. The result is a form of “self-limiting” behaviour in which operational modules may, somewhat counterintuitively, experience less violent runaway propagation. We believe the new findings matters because safety standards have traditionally been written around worst-case scenarios derived from static testing, which do not always reflect how batteries fail in real installations. Large-scale energy-storage systems spend most of their time in cycling conditions, and the interactions described in this study indicate that system-level hazard models will need to incorporate discharge rate, dynamic resistance evolution, and state-of-charge gradients to avoid over- or under-estimating real risks. The authors’ measurements of electrode-localized heating also highlight the need for diagnostic schemes capable of detecting early shifts in current pathways rather than relying solely on surface temperatures, which may lag dangerously behind internal conditions.

Another implication lies in how future battery management systems could mitigate runaway. the research points toward strategies such as targeted current redistribution, deliberate pre-discharge during emergency shutdowns, or intelligent pack segmentation to avoid high-energy short-circuit loops by clarifying how discharge behaviour shapes the thermal field distribution within a module. The innovative electro-thermal model developed by the authors provides a useful foundation for such efforts because it disaggregates the contributions of discharge heating and Joule heating—two components that scale in opposite directions with discharge rate. More broadly, the new work suggests a path toward designing multi-modal warning systems that combine electrical and thermal indicators to identify deviations in real time. Because resistance, voltage decay, and temperature rise evolve in coupled fashion during the pre-runaway stage, monitoring them together could yield earlier and more reliable warnings than any single parameter alone. If implemented in large battery arrays, such systems might reduce the likelihood of cascading thermal events, which remain one of the most challenging hazards for grid-scale storage.