Limiting Oxygen Index Test: ISO 4589-2 Method and Result Interpretation

The Limiting Oxygen Index Test: The Short Answer

The oxygen index (OI) test answers one question with one number: what is the minimum fraction of oxygen that will keep this material burning? A small vertical specimen is clamped inside a transparent glass chimney through which an oxygen/nitrogen mixture flows upward at (23±2) °C. The top of the specimen is ignited, and the operator steps the oxygen concentration up or down until the material barely supports flaming combustion — that percentage is the oxygen index, reported to one decimal (for example, OI 28.4 %). Two criteria decide whether a given run counted as “burning”: flaming for 180 seconds or longer, or flame spreading past the 50 mm mark on the specimen 14. The result is a screening and specification tool — not a fire class, and not a substitute for a hazard assessment, as the standard’s own introduction warns [1].

What the Test Measures — and What It Deliberately Does Not

ISO 4589-2:2017 specifies methods for determining the minimum volume fraction of oxygen, in admixture with nitrogen, that will support combustion of small vertical test specimens [1]. The scope covers self-supporting bars and sheets up to 10.5 mm thick, for solid, laminated and cellular materials of apparent density 100 kg/m³ or greater; lower-density cellular materials are directed to ISO 3582 instead [1]. The test runs at ambient temperature — its sibling ISO 4589-3:2017 covers elevated temperatures from 25 °C up to 150 °C for the same specimens 2.

The standard’s introduction is unusually direct, and quoting it in a specification saves arguments later. OI results “can provide a sensitive measure of the burning characteristics of materials under certain controlled laboratory conditions, and hence may be useful for quality control purposes”, but results “are not applicable to describe or appraise the fire hazard presented by a particular material or shape under actual fire conditions, unless used as one element of a fire risk assessment” [1]. Two more limitations follow from the specimen rules: results from specimens of different thickness or different ignition procedures are not comparable, and no correlation with behavior under other fire conditions is implied [1]. A datasheet that quotes “OI 29” without naming the specimen type and ignition method has rubbed off a third of the information before the conversation starts.

Air contains about 21 % oxygen, so the scale has a natural anchor. Materials near or below 21 burn readily in air; as the number climbs, the material needs an increasingly oxygen-rich atmosphere to sustain flame. Polypropylene sits near 17–18 %, while a well-formulated flame-retardant compound can exceed 30 % 6. Around 28 and above, materials are generally described as self-extinguishing in normal air, which is where most product-standard specification limits cluster.

The Rig: Glass Chimney, Two Gases, One Analyser

The apparatus looks simple, and every part of it is specified [1][2][4]:

Three rig details decide whether the data survives an accreditation audit. The exhaust velocity of at least 90 mm/s through the restricted outlet keeps the chimney atmosphere uniform and clears combustion products before they feed back into the specimen [2]. The direct oxygen analyser measures the actual concentration rather than inferring it from flow readings — gas measurement drift is the classic source of false LOI values on manual rigs [4][6]. And the conditioning environment (23±2 °C, 50±5 % relative humidity, at least 88 h per ISO 291 3) is part of the apparatus in the audit sense: a laboratory that conditions specimens in the corridor is producing uncontrolled data.

Specimens: Six Forms, Fifteen or More Pieces

ISO 4589-2 defines six specimen forms covering the plastics, foam and film universe [1][3]:

Sampling is the first place results go wrong. At least 15 specimens are needed for materials whose known oxygen index varies within ±2; unknown materials, or those showing erratic burning behavior, need 15–30 [3]. Each self-supporting specimen carries a 50 mm mark for judging burning length [3]. For flexible film (form V), the strip is clamped in a frame that keeps it vertical and stops thermal shrinkage from pulling it out of the flame zone; form VI film is rolled on a wire to make a self-supporting cylinder. Without those supports, the test measures holder behavior, not material behavior.

The Procedure: Up-and-Down Staircase

The heart of the method is an iterative staircase rather than a fixed number of runs [1][4]:

1. Condition the specimens at (23±2) °C and (50±5) % RH for at least 88 h per ISO 291 [3].
2. Set the initial oxygen concentration — a sensible starting point is near the expected value or around 21 % — and purge the chimney until the analyser reads steady.
3. Ignite. Method A (top surface ignition) applies the flame to the top face for up to 30 s, withdrawing every 5 s to check for sustained ignition; it suits rigid self-supporting materials. Method B (propagating ignition) runs the flame down the vertical side about 6 mm and is used for materials that melt, drip or shrink away from the flame in Method A [3].
4. Judge the response. If the specimen burns for 180 s or more, or flame crosses the 50 mm mark, record X; otherwise record O.
5. Step the concentration — typically 0.2 percentage points of oxygen — in the direction the response indicates: X means the mixture was richer than needed, O means leaner [4].
6. Compute the oxygen index from the final runs: OI = (oxygen volume / total volume) × 100, to one decimal, with the standard’s statistical treatment of the staircase data [1].
7. Short-procedure option: when a product standard specifies a minimum OI, three specimens are tested at exactly that concentration and at least two must give O responses [4] — a far cheaper check than a full staircase.

The staircase structure is why reports must always state the ignition method and specimen form: a PVC sheathing tested by Method B and a polystyrene sheet tested by Method A both produce “an LOI”, but the numbers answer different practical questions and cannot be compared.

Full Staircase or Short Procedure?

Choosing between the two operating modes is a budget decision disguised as a technical one. The full staircase consumes specimens and operator hours — every step is another conditioned bar, another ignition, another judgement — but it produces the index itself, which is what R&D and formulation work needs. The three-specimen short procedure consumes almost nothing: three bars, one concentration, two judgements — and delivers a pass/fail against a specification limit, which is what incoming inspection and production QC need. Laboratories that serve both audiences schedule them differently: development campaigns run staircases on new formulations, then freeze the result into a minimum-OI limit that the production line checks forever after with the short procedure. Instrument selection follows the same split, because the short procedure is unforgiving of oxygen drift: when the acceptance band is “≥ 28 %” and your analyser reads 27.6 on a passing compound, the argument is not statistical, it is commercial.

Reading an Oxygen Index Report

Four things separate a usable report from an unusable one:

A useful mental scale runs from ordinary olefins near 17–18 %, through PVC-based compounds in the twenties, to mineral-filled or highly FR-formulated systems above 35 % [6]. Specification limits seen in real contracts — sheathing compounds near 28 %, demanding engineering applications near 32 % [6] — come from the product standards that name the OI test as their screening gate, not from the LOI method’s own logic.

Where LOI Sits in the Reaction-to-Fire System

The most common misconception we encounter in inquiries: the limiting oxygen index is not part of the EN 13501-1 Euroclass system. No Euroclass from A1 to F is decided by LOI — classes A1/A2 come from ISO 1182 plus the bomb calorimeter, B–D from the SBI test, and E from the small flame test, as set out in our EN 13501-1 Euroclass guide. LOI lives one level down: it is the designated screening and quality-control method inside product standards — wire and cable compounds, electrical enclosure materials, and transport interior specifications — where a minimum OI limit defines acceptance. A supplier who answers a Class B requirement with “OI 29.5” has misread the specification; a supplier who answers a sheathing standard asking for LOI ≥ 28 with an SBI report has done the same work twice.

The flip side is the correct use: within its scope, LOI is very good at what it was designed for. It is sensitive to small formulation changes — an extra percentage point or two of flame-retardant loading shows up reliably — which makes it the standard tool for compounding QC, incoming material verification and FR system development [1]. It also travels well across industries: cable and wire, electrical enclosures, aerospace and rail interiors all specify minimum oxygen index limits, which is why a single LOI rig often serves several qualification files at once [6].

Regional Variants: Same Fire, Four Paper Trails

The same chamber serves the world’s major markets with slightly different paperwork:

Practical consequence for exporters: a Chinese report citing GB/T 2406.2-2009 was executed to the 1996 edition of the international method, and the national revision is already moving — confirm which edition your laboratory executed against before filing the paperwork [8]. And note the exact wording of ASTM’s equivalence: D2863 is technically equivalent to ISO 4589-2 only when using the gas measurement and control device with direct oxygen concentration measurement [7]; a flow-calculated value sits outside that equivalence.

Buying an LOI Tester: What to Check

Procurement conversations about oxygen index rigs are really conversations about gas delivery [2][4]:

Manual rigs are cheaper and dominate teaching laboratories; automatic closed-loop rigs dominate commercial and accreditation work precisely because the oxygen concentration — not the operator — decides the result. Our limiting oxygen index tester for ISO 4589-1/2 covers the full ISO route with closed-loop control, and the LOI tester configured for ISO 4589-2 and ASTM D2863 adds the ASTM procedure set for laboratories serving US-facing specifications.

Common Mistakes With the Oxygen Index Test

Frequently Asked Questions

What is the current version of the LOI test standard?

The current edition is ISO 4589-2:2017, the second edition, published 11 April 2017 by ISO/TC 61/SC 4 and confirmed after its 2022 systematic review. It replaces ISO 4589-2:1996 (with Amendment 1:2005). The European adoption is EN ISO 4589-2:2017, published 7 June 2017, which supersedes EN ISO 4589-2:1999. The companion parts are ISO 4589-1:2017, covering general requirements, and ISO 4589-3:2017, covering the elevated-temperature test.

What LOI value makes a material self-extinguishing?

Air contains about 21 % oxygen, so materials with an oxygen index near or below that burn easily in air, and the index rises as the material needs a richer atmosphere to sustain flame. Values around 28 and above are generally described as self-extinguishing in normal air, and many product standards place acceptance limits in the high twenties to low thirties. Practical reference points: polypropylene sits near 17–18 %, while a well-formulated flame-retardant compound can exceed 30 %.

How many specimens does an oxygen index test need?

A full staircase determination needs at least 15 specimens for materials of known, stable oxygen index, and 15 to 30 for unknown or erratically burning materials. The shorter comparison procedure, used when checking against a specified minimum OI, runs three specimens at that concentration and requires at least two to self-extinguish. Reports built on fewer specimens cannot support the statistical treatment the standard prescribes.

What are the 180 s and 50 mm criteria in ISO 4589-2?

They define the threshold between a burning response and a non-burning response. After the igniter is removed, a specimen counts as burning if flaming persists for 180 seconds or longer, or if flame advances past the 50 mm mark on the specimen. Both criteria exist because some materials extinguish quickly yet travel a long distance, while others burn for minutes without spreading far; a single criterion would rank those two cases identically.

Is ASTM D2863 the same as ISO 4589-2?

Technically equivalent when the rig uses the gas measurement and control device with direct oxygen concentration measurement — that is the equivalence note printed in the ASTM standard itself. Editions and documented requirements still differ: ASTM D2863-23e1 defines three procedures and sets its density scope at apparent density greater than 15 kg/m³, versus 100 kg/m³ in ISO. For paperwork, cite whichever standard the customer’s specification names.

Why does GB/T 2406.2 still cite an old edition?

GB/T 2406.2-2009 is an identical adoption of ISO 4589-2:1996, while the international method has moved to the 2017 edition. The Chinese standard was reviewed in March 2025 with a revision conclusion, and revision project 20250374-T-606 is in approval. Laboratories serving Chinese-market specifications should confirm which edition a report was executed to, because overseas customers increasingly write ISO 4589-2:2017 into contracts.

What equipment does a lab need to run ISO 4589-2?

The core chain is four items: a glass-chimney test assembly with specimen holder and igniter; closed-loop oxygen and nitrogen gas control with a direct O₂ analyser; a conditioning chamber meeting ISO 291 at 23±2 °C and 50±5 % relative humidity; and timing with data capture for the staircase calculation. Fume extraction completes the rig for routine sample volumes.

Conclusion: A Buyer’s Checklist for LOI Testing

1. Decide the minimum OI first — the staircase answers “what is the number?”, while the three-specimen short procedure answers “does it clear the limit?” at a fraction of the runs.
2. Fix the specimen form and ignition method per series; comparisons only hold within them.
3. Buy closed-loop gas control with direct O₂ readout and condition per ISO 291 — those two decisions decide whether your numbers are evidence or anecdotes.
4. For export work, track the editions: ISO 4589-2:2017, EN ISO 4589-2:2017, ASTM D2863-23e1 and GB/T 2406.2-2009 are four current documents describing one fire.

If you are specifying or building an oxygen index capability, BRT manufactures LOI testers for the ISO 4589-1/2 route and for combined ISO/ASTM operation, with the conditioning chambers and calibration tooling around them. Tell us which editions you will be audited against, and our engineers will match the configuration to your standards list.

About BRT: Shenyang Baoruitong Automation Equipment Co., Ltd. (BRT) manufactures fire and flammability testing equipment — including limiting oxygen index testers, non-combustibility furnaces, oxygen bomb calorimeters and complete reaction-to-fire laboratories — for customers in China and export markets. This guide was prepared by the BRT engineering team from the published editions of the standards cited above.