There is no such thing as UL 9540A certified

I keep running into this phrase, and I have come to think it is one of the more consequential pieces of sloppy language in the storage business. Consequential because it converts a document a competent reader is supposed to interrogate into a checkbox nobody opens.

It is the same shape as several things I have written on this beat. A gigawatt of storage is half a number. A round-trip efficiency figure hides what it was measured under. In each case a rich technical fact gets compressed into a short phrase, and the compression is where the information dies. This one is worse than most, because the phrase being used does not exist.

Two documents, similar names, different jobs

There are two things here and they are constantly conflated.

UL 9540 is the Standard for Energy Storage Systems and Equipment. It is a system-level safety standard, and a product can be listed and certified to it. That listing is real and carries a real mark.

UL 9540A is the Standard for Test Method for Evaluating Thermal Runaway Fire Propagation in Battery Energy Storage Systems. UL Solutions describes it as the American and Canadian national standard for assessing fire propagation related to thermal runaway events in battery energy storage systems. It is a procedure. You run it, you get data.

Lucas Miller, writing for Mayfield Renewables, put the distinction about as plainly as it can be put:

UL 9540 defines a system-level listing standard. UL 9540A defines a specific test methodology. Despite how similar their names are, the two things are not the same and should not be conflated.

On the exact phrase in my headline, the same article is unambiguous: "There is no certification for UL 9540A. Certification to UL 9540 requires UL 9540A testing if indicated by the codes your AHJ follows."

The Sustainable Energy Action Committee's energy storage standards working group says the same thing in its informational bulletin on the two documents, in the flattest language I have found anywhere:

UL 9540A is a "Test Method" used to characterize the internal thermal runaway and fire propagation performance of a BESS as required in accordance with applicable fire codes and UL 9540. UL 9540A is NOT a certification standard, therefore there is no "listing" that results from performing UL 9540A testing.

So when a datasheet says "UL 9540A certified," the charitable reading is that the product holds a UL 9540 listing supported by 9540A test data. The uncharitable reading is that somebody ran some tests. Both readings are consistent with the words on the page, which is precisely the problem.

The detail that shows how separate they really are

Here is my favourite piece of evidence, and it comes from UL's own summary of the fifth edition.

Among the changes to the test report requirements, the cell-level test report "Removed requirement to include 'whether the energy storage system has achieved UL 9540 certification.'"

Read that twice. A field that used to sit inside the 9540A test report, asking whether the system had achieved 9540 certification, was taken out. The standards body itself pulled the cross-reference. These are two documents answering two different questions, and the people who maintain them keep making that more explicit, not less.

Four levels, and what each one is looking at

The method is a ladder. UL's own descriptions of the levels, condensed:

1. Cell level evaluates the thermal runaway characteristics of the cell along with the composition and flammability of the gases it vents.
2. Module level evaluates the tendency of cell thermal runaway to propagate, the heat and gas release rate of the module, and the potential danger of ignition or deflagration.
3. Unit level evaluates the likelihood of fire spread between modules, the unit's heat and gas release rates, and the potential for deflagration or re-ignition.
4. Installation level mainly evaluates the effectiveness of the fire protection system, along with the heat and gas release rate of the system and the danger of deflagration or re-ignition.

You climb the ladder only as far as you have to. There are performance criteria at each rung, and meeting them can end the sequence. At module level in the fifth edition those criteria are: cell vent gas is nonflammable, there is no spread of flame outside of the module, and the module exterior surface temperature does not exceed the cell venting temperature.

Here I want to be careful, because "there is no pass/fail" is itself an overstatement that gets repeated online. There are performance criteria inside the method. What there is not is a pass/fail badge at the end meaning "this battery is safe." Meeting the criteria at a given level determines how much further testing you have to do, and the accumulated data determines what installation conditions the product can be used under. That is a different kind of output than a seal, and a more useful one, provided somebody reads it.

Why the report is really a real-estate document

This is the part that gets lost, and the part with money attached.

The scope statement on the standard says the data generated supports the manufacturer's guidance on separation distances between systems and the fire protection required for installations complying with NFPA 855, NFPA 70, UL 9540 and related codes.

Separation distance is the operative phrase. NFPA 855 sets a default, and the way to move off the default is test data. From the clause as quoted by Mayfield:

Individual ESS units shall be separated from each other by a minimum of 3 ft (914 mm) unless smaller separation distances are documented to be adequate based on fire and explosion testing complying with 15.13.

So the economic function of a 9540A campaign is to buy back spacing. Every foot of separation you do not have to leave is enclosure area you do not have to lease, trench you do not have to dig, and cable you do not have to run. On a large site that is a real number.

Which means the test report is not a safety formality bolted onto the end of procurement. It is the document that sets your site plan. And a phrase like "UL 9540A certified" tells you nothing about it, because it does not say which levels were run, on which cells, in which enclosure configuration, or what the measured heat flux at the egress path turned out to be.

I help design the AI battery-cycling systems for a veteran-owned (HUBZone) energy-storage integrator; I do not own it and earn nothing from this link. Full policy here. I am flagging it because the questions further down are the ones I ask when a test report lands on my desk, and you should know I have a hand in that work when you read me recommending them.

The method changed twice in about a year

There is a second reason a bare "UL 9540A tested" claim is thin. The method has been moving.

UL's own e-book describes the fifth edition as released in March 2025, and lists a substantial set of revisions: newly defined terms including initiating cell, residential use, non-residential use, open parking garage and thermal runaway propagation; cell-level test methods added for lead-acid and nickel-cadmium; cell, module, unit and installation methods added for high-temperature sodium batteries; hydrogen now required to be measured with an appropriate sensor and total hydrocarbons with a flame ionization detector; and, for residential unit-level testing, the NFPA 286 test room replaced with an instrumented test wall.

Then a sixth edition followed. Tyson Bittrich of Mayfield Renewables gives its publication date as 13 March 2026, and describes the headline change as a fourth evaluation level involving intentional ignition of off-gases to assess whether a fire or explosion occurs and what it does to the surroundings, replacing what had been the installation-level test. UL's own service page notes that the sixth edition eliminates the unit-level test for most systems, while it remains required for residential.

That direction of travel was visible a year ahead. The fifth-edition e-book closes by saying a dedicated task group within the UL 9540A technical committee had developed a large-scale fire test method, that the committee convened on 29 April 2025 to review it, and that it was headed for a June ballot before the full 150-plus member committee. Paul Hayes, chair of that task group and vice president of energy infrastructure at fire protection consultancy The Hiller Companies, said at the time:

Our committee consists of diverse stakeholders vested in ESS safety, including leading global OEMs, ESS integrators, code authorities, insurance leaders, specialized consultants and testing organizations. We are actively collaborating with the NFPA 855 to align on final requirements. I am confident this task group is producing a quality proposal to revise UL 9540A to address increased requests for large-scale fire testing of energy storage systems that go beyond a single module test.

The code side moved in step. Reporting on the 2026 edition of NFPA 855, Andy Colthorpe of Energy-Storage.news wrote that where previous editions already encompassed the UL 9540A thermal runaway propagation test, the 2026 version adds large-scale fire testing, which the article characterises as entire units being intentionally set on fire and left to burn with all fire suppression equipment disabled, to simulate a worst case. An NFPA spokesperson told the outlet the 2026 version is significantly updated from the 2023 version, with new requirements covering project emergency response plans, thermal runaway propagation protection systems, explosion control, backup power supplies, and fire and explosion testing.

Put that together and the practical consequence is simple. "UL 9540A tested" without an edition year is now a materially incomplete claim. A 2023 report and a 2026 report are not describing the same battery of tests, and the newer one is harder.

The questions that replace the phrase

None of this requires you to become a fire protection engineer. It requires treating the test report as a document rather than a status.

1. Which edition was the testing performed to? Fourth, fifth or sixth. If that is not on the report cover, it is the first thing to fix.
2. Which levels were run, and where did the sequence stop? Stopping at module level because the criteria were met is a legitimate and good outcome. Stopping there for some other reason is a different fact. The fifth edition requires module, unit and installation reports to include a summary of the previous levels, so the chain should be visible inside the document.
3. Was the tested article the article you are buying? Same cell, same module, same enclosure, same suppression package, same mounting. Unit-level conditions are specific down to wall material: indoor floor mounted, indoor wall mounted, outdoor wall mounted, open parking garage, combustible rooftop, and installation beneath PV panels or other obstructions are all separately described scenarios in the fifth edition.
4. What separation distance does the manufacturer's guidance claim, and does it trace to this report? This is the one with money on it. If reduced spacing is justifying your site layout, the data behind it is in the report, and your authority having jurisdiction is entitled to see it.
5. What does the report say about the gas? Composition, flammability, hydrogen. Deflagration risk drives explosion control, and it is an area both the fifth edition and the 2026 NFPA 855 tightened.
6. Does the product also hold a UL 9540 listing? That is the separate question, and it is the one people think they are asking when they say "9540A certified."

What I could not confirm

The standard itself is a paid document. I did not buy it, and nothing here quotes a clause of UL 9540A. Everything attributed to its contents comes from UL's own published summary of the fifth edition, UL's public service and store pages, and two named third-party technical articles. NFPA 855 is likewise paywalled; the separation-distance clause above is reproduced as Mayfield quoted it, not as I read it in the code.

I also hit a real ambiguity on edition dates. UL's e-book states the fifth edition was released in March 2025. Mayfield gives 13 March 2026 for the sixth. The UL Standards store listing for UL 9540A carries a product identifier ending 20250312 while displaying the standard as edition 6 active, which I could not reconcile from the outside. I have attributed each date to the source that states it rather than picking a winner. If UL's store record turns out to be authoritative, the correction goes here.

Finally, I have no data on how often "UL 9540A certified" appears in real procurement documents versus how often the correct phrasing is used. This report argues the sloppy phrasing is common and consequential. The first half of that is an impression formed from reading a lot of datasheets, not a measurement.

The signal

Certification answers a yes-or-no question and hands you a mark. A test method answers a how-much question and hands you a stack of measurements. Storage safety, as it is actually regulated in North America, runs on the second kind of answer, and the phrase everyone uses describes the first kind.

That is the whole gap. Somebody paid a great deal of money to burn cells in a laboratory so a fire marshal would have real numbers to reason about, and the result gets summarised on a sales sheet as a few words that mean nothing. Ask for the report. It is the most informative document in the package, and it was written to be read.