الرئيسية/المصادر/المعرفة/Testing&Analysis/Thermal Runaway Testing for Li-Ion Cells: 6 Methods

Thermal Runaway Testing for Li-Ion Cells: 6 Methods

Canrud November 17, 2026 42

Thermal runaway testing for lithium-ion cells forces a cell into a self-heating failure under controlled conditions and records what happens. The six common methods are accelerating rate calorimetry (ARC), external heating, nail penetration, crush, overcharge and external short circuit. Each answers a different question, so the right choice depends on what you need to learn.

Quick answer

  • Need onset temperature and heat release data? Use ARC.
  • Need to see how the cell reacts to an internal short? Use nail penetration or crush.
  • Need to know what happens when charging or wiring goes wrong? Use overcharge or external short circuit.

 

What does thermal runaway testing measure?

A thermal runaway test records how and when a cell fails, and what the failure releases. The main outputs are:

  • Onset temperature: when the cell starts heating itself faster than it loses heat.
  • Maximum temperature and rate of rise: how violent the event is.
  • Venting and flame: when the cell opens, and whether it ignites.
  • Gas: volume and composition, which matter for enclosure and vent design.
  • Pressure: in a closed test vessel, how much pressure the event builds.

Labs use thermocouples, pressure and flame sensors, infrared cameras synchronised with video, and optional gas analysers. Cell chemistry changes all of these results; our article on why LFP is more thermally stable than NMC explains one of the biggest differences.

1. Accelerating rate calorimetry (ARC)

ARC places a cell in a calorimeter that keeps it close to adiabatic, so almost no heat escapes. The instrument heats in small steps, waits, and checks whether the cell is warming on its own. Once self-heating is detected, the calorimeter follows the cell temperature until runaway.

What it tells you: the temperature at which self-heating begins, the runaway onset temperature and the temperature rise. These are the most useful numbers for comparing chemistries and for thermal models.

Limits: ARC shows behaviour without heat loss, which is harsher than most real installations. Large cells may need a large calorimeter.

2. External heating (oven or heater)

External heating raises the cell temperature with an oven or a heater pad until it fails. Heater pads are also the common way to trigger one cell inside a module during propagation testing.

What it tells you: the temperature at which the cell fails under outside heat, and how it vents.

Limits: the result depends on heater power, position and contact, so record those settings.

3. Nail penetration

A conductive nail is driven through the cell at a set speed, connecting the electrodes inside and creating an internal short circuit.

What it tells you: whether a sudden internal short leads to runaway, and how severe the result is.

Limits: a nail creates a larger and different short than a small manufacturing defect, and results can vary with nail shape, tip and speed.

4. Crush or indentation

A plate or shaped indenter deforms the cell until it shorts internally or reaches a set force or deformation.

What it tells you: how the cell responds to mechanical damage, such as in a vehicle crash.

Limits: the indenter shape and the direction of crush strongly affect the outcome.

5. Overcharge

The cell is charged beyond its rated voltage or capacity at a set current, as if charge control had failed.

What it tells you: whether the cell can tolerate a charger or battery management fault, and how it fails if not.

Limits: results depend on charge current and the voltage limit of the power supply.

6. External short circuit

A low-resistance connection is made across the cell terminals, and current and temperature are recorded.

What it tells you: how much current the cell delivers into a short and how hot it gets.

Limits: many cells survive an external short without runaway, so it is often a screening test rather than a runaway trigger.

Which method answers which question?

Method

Question it answers

Key outputs

Main limit

ARC

At what temperature does the cell start to run away?

Self-heating onset, runaway onset, temperature rise

No heat loss; harsher than real use

External heating

How does the cell fail under outside heat?

Failure temperature, venting

Depends on heater setup

Nail penetration

What happens after a sudden internal short?

Temperature, venting, flame

Not identical to a real defect

Crush

How does the cell respond to mechanical damage?

Force, deformation, short onset

Indenter shape matters

Overcharge

What if charge control fails?

Voltage, temperature, venting

Depends on current and limits

External short

What if the terminals are shorted?

Current, temperature

Often no runaway

 

If you need several of these methods on one cell design, Canrud's electrical and safety testing lab can plan them as one programme so results are comparable.

Which test conditions change the result?

Thermal runaway results are only comparable when the conditions match. Agree these with your lab before testing:

  • State of charge: fuller cells release more energy, but runaway can still happen at low charge.
  • Cell size and format: larger cells release more energy and may need different fixtures.
  • Fixture and constraint: whether the cell is clamped changes swelling and venting.
  • Sensors: where thermocouples sit, whether gas is collected, and whether video and infrared are recorded.
  • Atmosphere and enclosure: open chamber or sealed vessel changes pressure and burning.

A cell-level result is an input to pack design, not a pack approval. Thermal runaway tests release fire and toxic gas, so they must only run in a purpose-built chamber operated by trained staff.

For a widely used EV abuse test procedure, see SAE J2464.

How do you read a thermal runaway test report?

A useful report lets you compare cells and feed results into pack design. Look for:

  • Test conditions: state of charge, cell age, fixture, atmosphere and trigger settings, stated clearly.
  • Temperature curves: cell surface temperature over time, with self-heating onset, venting and peak marked.
  • Event timeline: when venting, smoke, flame or rupture happened, linked to video.
  • Mass loss: cell mass before and after, which shows how much material was ejected.
  • Gas data: volume and composition, if gas was collected.
  • Repeats: how many cells were tested and how much the results varied.

One cell rarely tells the whole story. Results can vary between cells of the same type, so plan repeats for any number you will use in design decisions.

What should you agree with the lab before testing?

  1. The question the test must answer, and which method answers it.
  2. Number of cells, including repeats and spares.
  3. State of charge and any pre-conditioning, such as ageing or cycling.
  4. Trigger settings: heating rate, nail type and speed, crush shape, charge current.
  5. Sensors, gas collection and video.
  6. The report format and raw data you will receive.

Frequently asked questions

What state of charge should cells be tested at?

Most thermal runaway tests use fully charged cells, because stored energy is highest and the result is closest to worst case. But low charge does not mean safe: Stress Engineering has reported very active runaway events at states of charge as low as 3%. Test at the charge levels your question actually needs.

What is the difference between thermal runaway and propagation testing?

Thermal runaway testing looks at one cell: when it starts to self-heat, how hot it gets and what it releases. Propagation testing looks at whether that event spreads to neighbouring cells in a module or pack. Propagation results depend heavily on the pack design, spacing and barriers, not just the cell.

Why measure the gas released during thermal runaway?

The gas affects venting design, toxicity and fire risk. Labs can measure gases such as hydrogen, carbon monoxide and carbon dioxide, and use detector tubes for hydrogen fluoride and hydrogen cyanide. Gas data helps pack designers size vents and plan for flammable or toxic release inside an enclosure.

Is nail penetration a good internal short-circuit test?

It is widely used because it is simple and dramatic, but it is not a perfect copy of a real internal short. The nail creates a larger, different short than a small defect would, and results can vary with nail shape and speed. Use it alongside other methods, not as the only safety evidence.

Which standards include thermal runaway tests?

Several do, at different levels. Examples include IEC 62660-2 and SAE J2464 for EV cells and batteries, GB 38031 in China, UN ECE R100 for vehicles, and UL 9540A for energy storage systems. Which one applies depends on your product, market and customer requirements.

 

Not sure which method answers your question?

Tell us your cell format, chemistry and what you need to learn, and talk to our test engineers. Testing is quoted after a short scope call, all costs listed upfront. More about Canrud's lab.