Thermal runaway is the failure mode that defines battery fire testing: a cell's internal temperature rises, its chemistry decomposes exothermically, it vents hot flammable gas, and - unless the design stops it - it heats its neighbours into the same failure. Testing this is not one test but a family, organised by the scale of the specimen. This post maps the family and the logic that connects it.
Level zero: making a cell fail on purpose
Every runaway test starts with a controlled insult. The common triggers are heating the cell (in an oven, with a film heater, or in an abuse chamber), mechanical penetration (the nail test family), overcharge, and external short circuit. Abuse chambers - the AEC and EPT configurations in the Motis range - exist to apply these insults inside a contained, instrumented volume and record the cell's response: venting temperature, gas release, self-heating rate. This cell-level characterisation is the raw material for everything above it: a cell that cannot be driven into runaway, or that vents little energy, changes the whole system assessment.
UL 9540A: the propagation ladder
UL 9540A structures the system question into levels. At cell level it establishes runaway behaviour and vent gas properties. At module level it watches whether runaway propagates cell to cell inside a module and what the module emits. At unit level the complete energy storage unit is the specimen: an initiating cell is driven into runaway and the unit's fire behaviour is measured, centrally by heat release calorimetry under a collection hood - the configuration on our UL 9540A calorimetry page. At installation level the question becomes the room: spacing, venting and suppression. The method produces data rather than a pass stamp; installation codes and the authority having jurisdiction decide what the data must show.
Fire exposure: the fire from outside
A second family asks the opposite question: not the battery as ignition source, but the battery as victim. The ECE R100 fire exposure test (the BFET configuration in the range) exposes a complete vehicle battery to an external fuel fire for a defined period and observes whether it explodes or propagates the event - a requirement written for electric vehicle type approval. Furnace and flame exposure of packs and enclosures follow similar logic in stationary standards.
Reading a runaway test report
Three things to look for. The trigger method - a heater-driven runaway and a nail-driven runaway are different events, and a report is only comparable within its trigger. The propagation outcome - how many cells followed the first, and over what time. And the heat release record - peak rate and total, which is what spacing and ventilation decisions are actually made from. A report that gives only "passed, no explosion" has thrown away most of its value.
Vent gas: the measurement that connects the levels
Heat release gets the headlines, but vent gas analysis is what links cell behaviour to installation safety. When a cell runs away it ejects a hot mixture - hydrogen, carbon monoxide, hydrocarbons, hydrogen fluoride among them - whose volume, composition and flammability decide whether a module or room fills with an explosive atmosphere before anything visibly burns. Test programmes increasingly sample this gas (FTIR and dedicated analysers are the standard tools) at cell and module level, because the gas data feed the deflagration and ventilation calculations that installation approval actually turns on. A laboratory equipping for battery work should treat gas sampling as core infrastructure, alongside calorimetry, from the first quotation.
Safety of the test itself
Battery fire testing inverts the laboratory's usual problem: the specimen is an energetic device actively trying to fail. Test cells and abuse chambers are therefore containment engineering as much as measurement - pressure relief, remote operation, gas extraction sized for a full specimen event, and procedures for the specimen that did not fail today but is damaged and unstable tonight. Reputable programmes write the post-test quarantine and disposal route before the first run. When you read a quotation for abuse or unit-level equipment, the extraction, containment and monitoring provisions are not accessories; they are the difference between a test facility and an incident.
How to read supplier claims in this field
Battery fire testing is young enough that marketing runs ahead of method in places. The checks that separate instruments from theatre are the same as elsewhere in fire testing, applied firmly: which published method, at which level, does this configuration execute - cell, module or unit; is the calorimetry genuinely oxygen-consumption based with a calibration routine, or a temperature measurement wearing a calorimetry label; and what containment and extraction engineering is included for the event the specimen is supposed to have. Ask those three questions of any battery test quotation, including ours, and compare the answers in writing.
Terminology worth pinning down
Three terms are used loosely in sales documents and precisely in methods, and confusing them corrupts comparisons. Thermal runaway is the cell-level event: self-heating decomposition that the cell cannot stop. Propagation is the system behaviour: runaway spreading to neighbouring cells, modules or units - a property of the assembly, not the cell. Vent gas is what the cell ejects, before or during runaway, and its flammability is a separate hazard from the heat. A test report that states its trigger, counts its propagation and measures its gas and heat has used all three correctly; one that announces a product "passed thermal runaway" has used none of them.
The one-page summary
Thermal runaway testing is a ladder: abuse insults characterise the cell, propagation tests judge the assembly level by level, calorimetry prices the event in kilowatts and megajoules, and exposure tests check the product against fires that start elsewhere. UL 9540A organises the ladder for energy storage; ECE R100 covers the vehicle battery exposure case. Equip for the rung your customers must stand on, and insist that every report state its trigger, its propagation count and its heat release record. With those three fields present, results from different laboratories and different years can be compared honestly - which is the entire commercial point of a standardised test method. Everything else in this field - the chambers, the hoods, the gas analysis - exists to make those fields trustworthy when the specimen is one that nobody in the room manufactured or can open.
Equipping a battery fire laboratory
Start from the level your customers must evidence. Cell and module work needs abuse chambers and gas handling; unit level needs the calorimetry hood and its gas analysis; exposure work needs the burner rig. The battery sector page sets out how the levels connect, and the battery fire testing category gathers the instruments.