Skip to content

Thermal runaway trigger methods

What is this: an overview of the three recommended methods for triggering thermal runaway during the propagation verification test, plus the rule that lets a manufacturer use a different method.

"It is recommended to use C.5.3.3, C.5.3.4, or C.5.3.5 as optional triggering methods for the thermal propagation test. The manufacturer may choose one of these methods or select another method to trigger thermal runaway." (C.5.3.1)

Method Clause What it does
Needle penetration C.5.3.3 Steel needle pushed into the cell to short cathode and anode
External heating C.5.3.4 Planar / rod / film heater on the cell exterior
Internal heating plate C.5.3.5 Resistive plate built into the cell at the coil/stack centre

All three are recommended, not mandatory. A manufacturer can use a different method, but if it fails to trigger TR, the burden is to demonstrate that none of the three recommended methods would have triggered TR either (C.5.3.7, last paragraph; see runaway confirmation).

Choosing a trigger cell

"Thermal runaway triggering object: The battery cell inside the test object. Choose a battery cell near the center of the battery pack or one that is surrounded by other battery cells." (C.5.3.2)

The trigger cell sits where propagation has the most paths to spread — a cell with cell neighbours on all sides represents the worst case, because it has the most adjacent victims and the least access to the pack housing for heat rejection.

Tradeoffs at a glance

Aspect Needle External heating Internal heating plate
Speed to TR fast (seconds) slow (minutes, depends on power) medium
Repeatability medium (depends on penetration position) high high (when properly fabricated)
Intrusiveness to cell high (mechanical insult) low (cell unmodified) very high (cell must be re-built around the plate)
Mimics real failure? yes — internal short via mechanical insult no — external thermal stress yes — internal hot spot, similar to manufacturing defect
Pack-level integration hardest — pack housing must be penetrated; vent hole must be sealed (C.5.3.3 g) medium — heater must be installed before pack assembly or via a service hatch hardest — cell must be a custom-built unit with the plate inside, then assembled into the pack
Stop condition TR or 90 % cell-dimension depth (C.5.3.3 f) TR or monitor T = 300 °C (C.5.3.4) TR or monitor T = 300 °C (C.5.3.5.2)

What "another method" means in practice

C.5.3.1 explicitly permits other methods. Examples that have been used in industry:

  • Laser ignition — fast, point-source, but hard to integrate at pack level.
  • Overcharge to a single cell — in principle achievable through pack wiring but hard to isolate to one cell without modifying the BMS extensively.
  • Mechanical crush of a single cell in situ — closer to the bottom-impact failure mode.

Whatever method is chosen, it must produce TR confirmed by the (a OR b) AND c rule, and if it does not, the OEM owes the demonstration that none of the three recommended methods would have either.

Common ground for all three methods

Regardless of trigger method:

  • Sampling interval < 1 s for voltage and temperature (C.5.3.6 a). See monitoring and sensors.
  • TR confirmation by (a OR b) AND c per C.5.3.7. See runaway confirmation.
  • Test object SOC ≥ 95 % (externally chargeable) or ≥ 90 % (vehicle-charged only) per C.5.2 b).
  • Trigger cell chosen near the centre of the pack, surrounded by other cells (C.5.3.2).
  • Observation until all monitor points ≤ 60 °C, minimum 2 h (C.5.3.8).

Source: GB 38031-2025, Appendix C sections C.5.3.1 and C.5.3.2 (PDF p. 35).