Every spec sheet says the same two words: flame retardant. The word alone proves nothing. It is not a certification, and it carries no test number behind it. In a hospitality or retail fit-out, the fire inspector and the insurer do not read adjectives. They ask for the standard, the class, and the report number. That is the line that separates a plant that was treated from one that was tested, and it is the line procurement should be policing.

A word without a test is marketing

In the trade, the phrase covers three completely different realities. First, the material was treated with a retardant and nobody measured the result. Second, the supplier ran an internal test with no independent witness. Third, the batch was tested to a named standard, BS 5867 or UL94, and an independent laboratory issued a traceable report. The wording on a spec sheet does not tell you which of the three you are looking at. That is the trap: one word, three different grades of evidence.

The economics are lopsided. A compliant plant costs a few percent more per unit. A single fire event in a venue traces back through the risk assessment, the spec sheet, and the purchase order. If the only evidence in that file is the word, the chain of liability lands on you. The word is cheap in procurement. The report is what has value.

The two tests the spec sheet should name

Two standards dominate. BS 5867, the UK fire standard for soft furnishings, comes in two parts. Part 1 uses a smouldering cigarette and a match flame as ignition sources and measures afterflame time and char length. Part 2 steps up to a calibrated propane burner, a larger ignition source, and is the one hospitality risk assessments typically cite for panels and large soft goods. The pass line is quantitative: afterflame within a fixed number of seconds, char length under a fixed number of millimetres.

UL94 is the common route in US commercial work. The V-series applies a ten-second vertical flame to a specimen and grades the material as V-0, V-1, or V-2. The difference is total flame time and whether the specimen drips flame. V-0 is the strictest: no flame beyond the limit, no flaming drips. In either system the deliverable is not the grade itself. It is a test report number that a third party can verify and that is tied to a specific batch.

A compliant spec sheet therefore reads differently: standard named, class named, report number present, and the report traceable to the batch you are buying. Miss one of those four and the certificate is a photo of a paper, not a claim.

Three questions for the purchase order

First: which standard and class does this batch meet, BS 5867 Part 2 or UL94 V-0 or V-1, and what is the report number? A supplier with a real process answers without blinking. A supplier selling a word does not.

Second: is the report tied to this batch, or is it a one-off from a year ago? Material, formulation, and retardant lot all change. A report from a previous lot is evidence about a different product.

Third: will you accept a spot check on the delivered lot? Take one leaf, apply the ignition source for a timed interval, measure the afterflame. Ten minutes, one leaf, no charge. A supplier with a real process says yes, because their QC already does it. A supplier without one makes a polite exception.

Ask all three and the answers sort themselves into two camps. Flame retardant on the spec sheet is either a claim backed by a test number, or it is a marketing word. The plant in front of you looks identical either way. Only the file behind it tells you which.