As energy storage deployments grow, so does regulators' interest in what happens when a battery unit is fully involved in fire: does the fire spread within the unit, to the next unit, or through a fire-rated barrier? Large-scale fire tests answer that question, and several manufacturers have now published "burn chamber" tests of complete containers.
Look closely at those tests and the approaches, metrics and goals differ widely. Most fall into one of two types: the closed-door burn and the open-door burn.
The closed-door burn
A closed-door burn keeps the container doors shut. Thermal runaway is triggered inside and the system is expected to manage the fire with the protection it was designed with: ventilation that removes flammable gas and deflagration vent panels that relieve pressure. Vent sizing follows NFPA 68 and ventilation follows NFPA 69, usually supported by calculation and simulation.
The test therefore rests on a chain of assumptions. The vent area calculation depends on the estimated maximum gas generation rate and composition. The ventilation's effectiveness depends on the actual gas release rate and distribution — which in turn depend on how many modules go into thermal runaway at once and where the trigger is. In real conditions, these can differ from the design case, and so can the peak pressure.
Real failures are also not always single failures. If several faults occur together and ventilation or venting does not work as designed, the outcome of a closed-door scenario becomes much less certain.
The open-door burn
An open-door burn deliberately removes those design measures. The container doors are opened (sometimes the vent panels removed), the fire gets unlimited air, active suppression is switched off, and the system is typically at 100 % state of charge with neighbouring units at reduced spacing. The only thing left between the fire and its neighbours is passive fire protection: the enclosure structure, insulation, barriers and fire stopping.
Its purpose is to show that the system can resist a fully developed fire and stop propagation to adjacent units without any external intervention — the most severe scenario short of a site-wide failure.
Side by side
| Open-door burn | Closed-door burn | |
|---|---|---|
| Oxygen | Doors open, unrestricted burning | Enclosed, ventilation-limited |
| Active suppression | Disabled | May be retained, or its failure simulated |
| Explosion vents | Not relied on | Required; designed and checked to NFPA 68/69 |
| Heat to neighbours | Direct flame radiation to adjacent units | Reduced by the enclosure |
| What it proves | Passive fire protection; inherent resistance | Performance of the whole protection design |
Which should you run?
They answer different questions. The closed-door burn shows that the design works as intended: ventilation, venting and enclosure together. The open-door burn shows the product's own endurance when everything else has failed.
From a regulator's point of view, the open-door burn has the appeal of simple, explicit conditions: fewer assumptions about gas generation, ventilation and failure modes stand between the test and the conclusion. It is also harder to pass. The closed-door burn is more sensitive to its boundary conditions, so those conditions need to be justified — ideally by a deflagration test and simulation before the burn.
There is no single right answer. The choice depends on the product, the market, what the authority having jurisdiction will accept and what you want to be able to claim. Whichever you choose, the details decide the result: battery arrangement, connecting components, fixtures, instrumentation, and the safety management of the test itself.
How we can help
We plan and run both types at our Hebei large-scale fire test site and Jiading indoor fire hall: scenario design, passive fire protection and material selection, simulation before the burn, deflagration testing at our Nantong site, and the test itself. Tell us about your system.