Critical infrastructure preservation, resilience, and hardening intelligence.Preserve the asset. Protect the mission.
Readiness instrument

The UL 9540A evidence check

“UL 9540A tested” is not one thing. The method runs at four levels and they answer four different questions — only one of which is whether a fire in one enclosure reaches the next one. Say what your vendor actually handed you and this returns what it can be used for, what it cannot, and what to ask for next.

What do you actually hold?
Not what the brochure says — the level named in the document you can actually open.
An AHJ reads reports. Summaries are sales documents wearing a lab coat.
Changes which questions bite first, not whether the evidence is sound.
The four UL 9540A test levels Cell level characterises the gas and heat one cell releases. Module level shows whether runaway spreads inside an enclosure. Unit level shows what a whole enclosure does, including a test with the vent gases deliberately ignited. Installation level is the large-scale fire test that shows whether fire reaches the adjacent unit. Only the installation level answers the separation question. Cell what one cell vents, and how hot Module does it spread inside the box Unit what the whole enclosure does, vent gases deliberately lit Installation large-scale fire test: does fire reach the next unit — the separation question answers less about your site answers the question NFPA 855 asks

Swipe the diagram sideways to see all four levels.

And what is in the report?

These are the things an authority having jurisdiction is told to look for. Tick only what you have seen with your own eyes.

We make barriers, and a barrier is not self-justifying.

Preservation 2 is affiliated with Amidon (360 Ballistics, LLC), which manufactures protective construction materials, so read this page knowing Amidon sells one of the things it discusses. NFPA 855 does allow reduced separation where large-scale fire test data supports it and the AHJ approves — but the barrier does not carry itself. The evidence carries it, the test layout has to represent the wall as it will actually be built, and a fire protection engineer has to interpret the result.

We also cannot be the approved testing laboratory and we cannot be the registered design professional who interprets the data for your submittal. Both of those have to sit outside the supply of the thing being justified. Any supplier who offers to be both is telling you something about how they read the standard.

What this instrument will not give you. A separation distance. Not because we are being coy — because the number is a determination made by a fire protection engineer and an authority having jurisdiction, against the edition your jurisdiction actually enforces, on the evidence in front of them. Any tool that returns a setback from three dropdowns is guessing, and a guess is worth nothing in a plan review.

What it does give you. A straight answer about whether what you are holding is a submittal or a sales document, and an ordered list of what to ask for. That list is the useful part. It is written to be forwarded to the vendor as it stands.

The distinction the whole thing turns on

UL 9540 — a listing

A product can hold it

A product safety standard for energy storage systems. A unit is listed to UL 9540. That listing says the equipment met a set of construction and performance requirements. It does not describe what happens when a cell inside it goes into thermal runaway.

UL 9540A — a test method

Nothing is “9540A certified”

A method for evaluating thermal runaway fire propagation. It produces data — temperatures, heat flux, gas composition, heat release, smoke — for an engineer and an AHJ to interpret. It is not a pass or a mark. If a document says “UL 9540A certified”, whoever wrote it has not read it.

NFPA 855 — the installation standard

Only enforceable once adopted

The Standard for the Installation of Stationary Energy Storage Systems. It is the document that decides what the test data has to show. It is not law on its own; it becomes enforceable through adoption into a model code and then into state or local law, and adoption lags publication by years.

The large-scale fire test

The propagation question

Under the 2026 edition, Section 9.2.1.2.1, it must be conducted or witnessed and reported by an approved testing laboratory, and it must demonstrate that a fire involving one ESS unit will not propagate to an adjacent unit. That sentence is the whole submittal in miniature.

The vent-gas ignition test

Often the missing one

Section 9.2.1.2: where cell- or module-level testing shows flammable gases released during thermal runaway, an additional unit-level test is required in which the vent gases are intentionally ignited. A report that skips this because “it did not ignite on its own” has answered a different question.

The interpretation

A person, not a PDF

Section 9.2.2.2 calls for a supplemental report from a registered design professional with fire protection engineering expertise, interpreting the test data. Raw heat-flux plots are not a conclusion. Somebody with a licence has to say what they mean for this site.

Version
watch

The 2026 edition of NFPA 855 was released in October 2025 and supersedes the 2023 edition. Among the changes: a hazard mitigation analysis becomes the default for effectively all ESS installations rather than a conditional requirement, with narrow exceptions; the fire and explosion testing provisions moved to Section 9.2; emergency response plans must be reviewed annually with annual refresher training and notification to responders (4.3.3); a new Section 4.10 requires emergency power supply for critical safety systems; and the scope in Table 1.3 expands to further chemistries including iron-air, sodium sulfur, zinc-air and nickel-hydrogen.

Publication is not adoption. The 2024 International Fire Code still references the 2023 edition, and many jurisdictions sit further back than that. Confirm the enforced edition with your AHJ before relying on any section number here, including ours. This page states its own last review date below; if that date is stale, treat the page as stale.

Optional

Want the gap list sent over?

Print it or copy it and never give us an address — it is the same text either way. If it is useful to have a person send it across and be reachable for the follow-up, leave your details and the result travels with the message.

Section numbers on this page are from the 2026 edition of NFPA 855, Standard for the Installation of Stationary Energy Storage Systems, and the report checks follow the UL 9540A authority-having-jurisdiction checklist. Preservation 2 is not affiliated with NFPA, UL Solutions, or any code body, and nothing here is compliance advice, a legal opinion, an engineering judgement, or a substitute for the standard itself. The standard, your registered design professional and your authority having jurisdiction govern. Where a number would be a determination rather than a fact, this page declines to supply one.

Draft v0.01. Last reviewed against the standard 13 September 2026. Found something wrong? Tell us — corrections to this page are more useful to us than leads.