A standard is a common bar: one set of safety requirements, identical for everyone, that every product must clear. First the rule, then the product is built to meet it. In the table of NFPA 2010 it works the other way around. Here is how, step by step.

1. The number people breathe

NFPA 2010 is the standard for condensed-aerosol fire-suppression systems. It contains a table that sets how much aerosol may be released into a room with people in it — in plain terms, how much of the substance a person will breathe. Today’s values:

Table 1. Maximum allowable design application densities for normally occupied spaces (NFPA 2010, 2025 ed., Table A.5.2.4).

A facsimile of the original table is reproduced at the end of this article.

2. The letters are not chemistry — they are owners

What matters most in this table is what the letters mean. They are not types of substance and not hazard classes. They are codes for individual manufacturers. In the lists of the U.S. Environmental Protection Agency (EPA), each letter decodes as one product of one company: D is Stat-X (the company Fireaway); E is FirePro; I is GreenSol.

The decoding is on the public record. The UK Marine Accident Investigation Branch states, in Report 9/2023 (§1.18.2), that EPA lists FirePro as Powdered Aerosol E and Stat-X as Powdered Aerosol D. The occupied-space status of each was granted by a separate EPA SNAP Determination: Stat-X (D) by Determination 29, 79 FR 62863 (21 Oct 2014); FirePro (E) by Determination 34 (4 Oct 2018); GreenSol (I), carrying the 156 g/m³ value, by Determination 38, 88 FR 61977 (8 Sep 2023).

So «109.2» is not a norm for everyone. It is the private number of the FirePro product. No competitor’s aerosol may be used against it. Each company has its own row and its own number — and no one else’s.

3. A screen is not a test

The newest value is 156 grams per cubic metre, in the 2025 edition. The committee’s stated basis for it is the EPA SNAP listing. But a SNAP listing, under Section 612 of the Clean Air Act, is a substitute-acceptability decision: a comparison of one agent against the alternatives, resting on the data the manufacturer submits, screened by the agency. It is a screen, not a test. A screen of submitted data is not an acute-inhalation toxicity test conducted at the design application density — the one measurement that would show whether a person can survive that concentration of agent in the air.

That test is what is missing. No publicly accessible inhalation test report at 156 g/m³ has been produced — not by the manufacturer, not by NFPA, not by the certifier. The proceedings that set the value were never placed in open access; the committee materials reach the public only when compelled under law. A number that governs the air people breathe in an occupied room sits in the standard with no test in the public record beneath it.

For scale, 156 g/m³ is 42.9% above the 109.2 g/m³ listed for the FirePro product. And the hazard of this class of agent is not theoretical. According to the official report of the British authorities (MAIB, Report 9/2023), the discharge of a FirePro condensed-aerosol system in the engine room of the fishing vessel Resurgam killed a man — there was no fire; he was poisoned by carbon monoxide from the «extinguishing» agent itself. Further deaths from accidental discharges of agents in this class, in occupied or confined spaces and with no fire, have been reported in press accounts (Bangkok, 2016; Jeddah, 2020). The Resurgam death does not prove that 156 g/m³ is lethal; it proves that an agent of this class, already set the highest occupied-space density in the table, demands a test basis that can be examined — and none is public.

4. A closed circle

To sell such a system you need a certificate (UL 2775). But the certificate does not test the number: it takes it from the manufacturer’s manual and requires only one thing — that it appear in the NFPA standard. NFPA refers to the EPA listing. EPA’s listing rests on the manufacturer’s submission. And the manufacturer is where the number began. The circle is closed. At no point in it does an independent occupied-space test appear in the public record — and at the start and the end of the circle stands the same manufacturer.

5. What NFPA says of itself

In the same document, NFPA says of itself plainly: it «does not approve, inspect, or certify any installations, procedures, equipment, or materials nor does it approve or evaluate testing laboratories.» Consider that alongside what the table does: it sets a maximum concentration of agent for a room with people in it — a limit on human exposure in an occupied workplace. In the United States, binding limits on human exposure to harmful substances are set by the government occupational-safety authority, OSHA, by rulemaking on an evidentiary basis. By its own words, NFPA neither tests nor certifies; listing a value in a standard it sells is not, in itself, a validation that anyone can rely on.

6. We asked the certifier; the basis was withheld

UL 2775 is the certification standard that makes such a system saleable, and it incorporates NFPA 2010. So we put two plain questions to two people at UL Solutions: to the member of the NFPA AEG-AAA Technical Committee that adopted the 156 g/m³ value, who is at the same time UL’s Global Principal Engineer for Fire Extinguishing Systems; and to the UL market-surveillance engineer responsible for certification under that standard. The questions were narrow: did the committee member vote for the 156 g/m³ value, and were tests under UL 2775 carried out for the generator at that density? The first was deflected to «NFPA staff.» The second was answered, verbatim:

«Due to our contractual obligations and longstanding commitments to protect client intellectual property and sensitive data, we are unable to provide the information requested.»— UL Solutions, written reply, 24 June 2026

That reply does not prove that no test exists. It establishes that the basis of the mark was not made available for independent review. The one piece of information that bears on whether a person survives the air in an occupied room after discharge has been placed beyond examination, on grounds of a client’s confidentiality.

7. Bureau Veritas: a documented example

How a missing test basis surfaces at the point of certification can be read off a single document. On 31 January 2023, at Piraeus, Bureau Veritas issued marine Type Approval Certificate No. 31670/C0 BV for the FirePro condensed-aerosol system (generator FP-5700) for machinery spaces. In the certificate’s Schedule, under «Test Reports,» two of the cited items are these.

Item 3.2, the U.S. Coast Guard report. This is a document that runs against the certificate it supports. The U.S. Coast Guard R&D Center report (CG-D-03-06), on full-scale tests of three aerosol systems — FirePro among them — against IMO MSC/Circ.1007, records in plain words that «not one of the three systems successfully met the requirements of the MSC/Circ. 1007»; that only one of the 14 Class A fires was extinguished; and that the carbon-monoxide concentrations measured were significantly higher than the level the U.S. National Institute for Occupational Safety and Health (NIOSH) defines as Immediately Dangerous to Life or Health.

Item 3.3, the toxicology document that is not a test. The only toxicology document the certificate lists is a KEMA report. On the certificate it is dated 18 October 2002; in the manufacturer’s own published dossier the very same report carries the date 27 September 2002. One report, two dates. Both (2002) predate the registration of FirePro Systems Limited (Cyprus, HE 137692) on 17 April 2003. On its face it is a conditional «Health and Environmental Aspects» assessment, written throughout in the conditional, not an acute inhalation toxicity test; and the composition it describes is not that of the certified product.

We have written to Bureau Veritas asking it to reconcile these points; that correspondence is pending. The certified generator type is the FP-5700 — the same type involved in the Resurgam fatality (MAIB, Report 9/2023). We set these documents side by side and invite the explanation of the parties named.

8. The blank rows are for the next in line

The table also has empty places: letters with no number. They are filled in the same way 156 was. There is room for new manufacturers.

9. What this really is

Put it together. Every value in this table is agent-specific and originates in a manufacturer’s submission; the committee’s stated basis for the 156 g/m³ value is the EPA SNAP listing, which is a substitute-acceptability decision rather than an occupied-space test; no publicly available inhalation test report supporting that value has been produced; and the certifier declined to provide testing information for review. A per-product number, resting on a circular set of references with no public test beneath it, is not a common bar that a product is built to meet. It is a number written to fit one product, to the order of its owner. A standard in reverse. We call it a VIP standard.

10. Seven months without an answer

For seven months we have asked NFPA one question: what is the publicly auditable test basis for the number that people breathe? There is no answer — no test report in the public record, and from the certifier, a refusal on grounds of confidentiality. A life-safety value should rest on a traceable, independently reviewable test. On this value, that test has not been made public.

References

NFPA 2010, 2025 edition — Table A.5.2.4 and clause A.3.2.1. ANSI/UL 2775 (2022 edition). EPA SNAP Determinations 29 (79 FR 62863, 2014), 34 (2018) and 38 (88 FR 61977, 2023). UL Solutions, written reply of 24 June 2026. Bureau Veritas Type Approval Certificate No. 31670/C0 BV (Piraeus, 31 January 2023). U.S. Coast Guard R&D Center Report CG-D-03-06 (February 2006). KEMA report No. 50151217-KPS/TPE 02-6144. UK MAIB Report 9/2023 (FV Resurgam). Correspondence with UL Solutions (24 June 2026) and Bureau Veritas (June 2026) on file.