Analysis of Factors Affecting the Oxygen Index of Thermal Insulation Materials

The oxygen index (limiting oxygen index, LOI) is one of the most widely used quantitative indicators of the flame retardancy of thermal insulation materials such as rigid polyurethane, polystyrene, phenolic, and elastomeric foams. This article provides a systematic analysis of the material, formulation, structural, aging, and testing variables that influence oxygen index results, with reference to ISO 4589-2, ASTM D2863, and GB/T 2406.2. Understanding these factors helps insulation manufacturers optimize formulations, control production quality, and demonstrate compliance with building fire safety requirements.

Technical Overview

The oxygen index, also known as the limiting oxygen index (LOI), is defined as the minimum concentration of oxygen, expressed as a volume percentage in a flowing mixture of oxygen and nitrogen, that will just support candle-like combustion of a vertically mounted specimen after ignition with a small external flame. It is calculated as follows: LOI = V(O2) / [V(O2) + V(N2)] x 100%. Because ordinary air contains approximately 20.9% oxygen, the oxygen index directly indicates whether a material will continue to burn in air under the conditions of the test.

• LOI below 21%: the material burns readily in normal air.
• LOI between 21% and 26%: the material burns slowly or marginally in air.
• LOI of 26% or above: the material is generally considered flame retardant or self-extinguishing.
• LOI of 32% or above: the material is generally classified as highly flame retardant and is required by many building and transportation specifications.

Oxygen index testing of insulation materials is performed in accredited third-party laboratories and in-house quality control laboratories according to ISO 4589-2, ASTM D2863, GB/T 2406.2, and JIS K 7201-2. However, the measured oxygen index is not a fixed material constant. It is influenced by the polymer matrix, flame retardant system, foam density, cellular structure, specimen geometry, production process, aging history, and the details of the test procedure itself. A clear understanding of these factors is essential for interpreting results correctly and for troubleshooting unexpected values.

Key Factors Influencing the Oxygen Index

The following sections analyze the principal variables that determine the oxygen index of thermal insulation materials, grouped into material-related factors, structure-related factors, and test-related factors.

Polymer Matrix and Chemical Composition

The intrinsic flammability of the base polymer sets the baseline for the oxygen index. Different insulation polymers differ substantially in their decomposition temperature, char-forming tendency, and halogen content.

• Polyethylene and polypropylene foams have a typical LOI of only 17% to 19% because polyolefins thermally degrade into combustible volatile fragments with little char residue.
• Polystyrene (EPS and XPS) has a base LOI of approximately 18%, since it depolymerizes to styrene monomer, which burns readily.
• Rigid polyurethane foam typically shows an LOI of 21% to 23% without flame retardants, due to its moderate char yield.
• Polyisocyanurate (PIR) foam achieves a higher LOI, commonly 26% to 32%, because the thermally stable isocyanurate ring structure promotes char formation.
• Phenolic foam exhibits an LOI of approximately 38% to 45%, the highest among organic insulation foams, owing to its highly aromatic, char-forming crosslinked structure.
• Elastomeric NBR/PVC foams reach an LOI of 28% to 38% because the chlorine content of PVC and the acrylonitrile content of NBR release flame-inhibiting species during decomposition.

In general, polymer structures with high aromaticity, high crosslink density, high char yield, or built-in halogen or nitrogen content display higher oxygen index values. Mineral wool products are non-combustible inorganic materials, and their fire performance is assessed by non-combustibility tests such as ISO 1182 rather than by oxygen index.

Flame Retardant Type, Loading, and Dispersion

For most organic insulation materials, the flame retardant system is the single most controllable factor affecting the oxygen index. Flame retardants operate through different mechanisms, and their efficiency and loading level determine the achievable LOI.

• Halogenated flame retardants act mainly in the gas phase by scavenging the radical species that sustain the combustion chain reaction, often used together with antimony trioxide as a synergist.
• Organophosphorus flame retardants, such as phosphate esters commonly used in polyurethane foam, act both in the gas phase and in the condensed phase by promoting char formation.
• Nitrogen-based agents such as melamine contribute to intumescence and dilute combustible gases with non-combustible decomposition products.
• Inorganic hydroxides such as aluminum trihydrate and magnesium hydroxide decompose endothermically and release water, but typically require loadings of 50% to 65% to achieve a significant LOI increase.
• Intumescent systems based on ammonium polyphosphate, pentaerythritol, and melamine form an expanded insulating char layer and benefit from strong phosphorus-nitrogen synergy.
• Expandable graphite is a highly efficient synergist for PIR and phenolic foams, forming a worm-like protective char that can raise the LOI by several points at relatively low loading.

Reactive flame retardants that are chemically bound into the polymer network, such as brominated polyols in PIR formulations, provide more durable flame retardancy than additive agents, which can migrate or evaporate over time. The LOI increases with flame retardant loading up to an effective plateau; excessive loading plasticizes the polymer, reduces closed-cell content and mechanical strength, and may even reduce the practical fire performance. Uniform dispersion is also critical: poorly mixed flame retardant produces local weak zones that lower the measured LOI and increase scatter between specimens.

Density and Cellular Structure

The foam structure strongly influences the burning behavior. Higher bulk density generally produces a higher oxygen index, because a greater mass of polymer and flame retardant is available per unit volume and heat penetrates more slowly. Typical rigid PU foam, for example, may gain several LOI points when density increases from 30 kg/m3 to 60 kg/m3 with an unchanged formulation.

• A high closed-cell content and a fine, uniform cell size promote the formation of a coherent char layer that blocks heat and mass transfer, raising the oxygen index.
• Large, irregular, or open cells allow oxygen to penetrate and volatiles to escape freely, which lowers the oxygen index.
• Foam anisotropy matters: cells elongated in the rise direction give different LOI results when specimens are cut parallel or perpendicular to the rise direction, so orientation must be controlled and reported.
• The dense skin of a foamed board has a higher effective density than the core; removing the skin during specimen preparation can measurably reduce the oxygen index.

Specimen Geometry and Thickness

The oxygen index is a thickness-dependent property. Specimens of different thickness from the same board can produce noticeably different results because thickness changes the heat balance at the flame front and affects melt dripping and edge burning. ISO 4589-2 specifies rigid specimens typically 80 mm to 150 mm long and 10 mm wide, with thicknesses of 4 mm or 10 mm, while ASTM D2863 commonly uses specimens 70 mm to 150 mm long, approximately 6.5 mm wide, and about 3 mm thick. Results obtained at different thicknesses, or by different standard methods, are not directly comparable, and the specimen thickness and test direction must always be stated in the report.

Production Process and Blowing Agent

Processing parameters influence the final LOI through their effect on the chemical structure and cell morphology of the foam.

• In PIR systems, a higher isocyanate index increases the isocyanurate content and raises the oxygen index; well-controlled trimerization catalysts and adequate post-curing are required to complete the reaction.
• Hydrocarbon blowing agents such as cyclopentane remain trapped in the cells and act as auxiliary fuel; freshly cut pentane-blown foam often shows a lower LOI that increases over days or weeks as the blowing agent diffuses out.
• Water-blown systems using carbon dioxide avoid this fuel contribution and give more stable LOI values.
• Incomplete curing, incorrect mixing ratios, or excessive water content can leave residual polyol or urea structures that reduce char quality and lower the oxygen index.
• Foaming conditions that produce coarse or collapsed cells reduce the LOI even when the formulation is correct.

Aging and Service Temperature

The oxygen index is not necessarily stable over the service life of the insulation. Additive phosphorus flame retardants are volatile and migrate toward the surface, so thermal aging can reduce the LOI by one to several points. Hydrolysis, UV exposure, and oxidation can also degrade both the polymer and the flame retardant. In addition, the oxygen index decreases significantly with increasing ambient temperature, because the decomposition of the polymer accelerates as less external oxygen is needed to sustain flaming. ISO 4589-3 defines the determination of the temperature index, which characterizes this behavior and is relevant for applications where insulation operates at elevated temperatures. Accelerated aging followed by repeated LOI measurement is therefore recommended for products requiring long-term flame retardancy.

Instrument and Procedure Variables

Finally, the measured LOI depends on the correctness of the test execution itself.

• The oxygen and nitrogen gases must be of high purity, the flow meters must be calibrated, and the total gas flow must be maintained at the standard linear velocity of approximately 40 mm/s through the chimney.
• Specimens must be conditioned, typically at 23 plus or minus 2 degrees Celsius and 50 plus or minus 5% relative humidity, for at least 88 hours before testing.
• The ignition flame size, the application time, and the height of the igniter above the specimen top must conform to the standard, since over-ignition artificially lowers the apparent LOI.
• The judgment of the burning criterion, which in ISO 4589-2 and GB/T 2406.2 is burning for more than 180 seconds or over a length of more than 50 mm after removal of the igniter, must be applied consistently by the operator.
• Periodic verification with certified reference materials such as PMMA, which has a nominal oxygen index of about 17%, is necessary to confirm instrument performance.

Testing Samples

Rigid Polyurethane and PIR Foam

Samples are cut from production boards or buns, removing the surface skin unless it is specifically being evaluated. Typical LOI values range from 21% for unmodified PU foam to above 30% for optimized PIR formulations with expandable graphite. Specimens are usually tested at 10 mm thickness, and the rise direction must be recorded.

Expanded and Extruded Polystyrene (EPS and XPS)

Unmodified polystyrene foam has an LOI near 18%, while flame-retarded and graphite-modified grades reach 26% to 32%. Specimens must be cut cleanly, ideally with a hot wire or sharp blade, because melted or crushed edges distort the burning behavior.

Phenolic Foam

Phenolic foam samples typically show LOI values between 38% and 45%. Because the material is friable, specimens must be handled carefully to avoid crumbling at the edges, which can accelerate ignition and lower the result.

Flexible Elastomeric NBR/PVC Foam

Elastomeric insulation foams used on pipes and ducts generally reach LOI values of 28% to 38%. Since the material is flexible, specimens are supported on the small frames or needles specified in ISO 4589-2 for non-self-supporting materials.

Polyolefin Foams and Mineral Wool

PE and EVA foams range from about 18% without flame retardants to 26% to 30% for treated grades. Mineral wool products are inorganic and non-combustible; their fire classification is based on ISO 1182 and heat release testing rather than oxygen index, although the LOI of any organic binder content may occasionally be evaluated.

Testing Items

Oxygen index testing is usually combined with complementary fire performance tests to fully characterize an insulation product:

• Room-temperature limiting oxygen index to ISO 4589-2, ASTM D2863, or GB/T 2406.2.
• Temperature index and elevated-temperature oxygen index to ISO 4589-3.
• Horizontal and vertical burning classification to UL 94 or GB/T 2408.
• Heat release rate and total heat release by cone calorimetry to ISO 5660-1.
• Smoke production to ISO 5659-2 or GB/T 8627.
• Single-flame source test to EN ISO 11925-2 and single burning item test to EN 13823 for EN 13501-1 Euroclass classification.
• Fire performance classification of building products to GB 8624 or EN 13501-1.
• Ignitability and glow-wire tests to IEC 60695-2 series for electrical applications.

Testing Methods

The oxygen index test uses a vertical glass chimney in which the specimen is clamped in a holder at its base. A controlled mixture of oxygen and nitrogen flows upward through the chimney at a fixed linear velocity, and the top of the specimen is ignited with a small gas flame. The operator observes whether the specimen continues to burn after removal of the igniter.

• Prepare specimens to the dimensions required by the selected standard, condition them in a standard atmosphere, and record thickness and orientation.
• Select an initial oxygen concentration based on the expected flammability of the material.
• Ignite the top of the specimen, remove the igniter, and record whether the burning time exceeds 180 seconds or the burning length exceeds 50 mm.
• Adjust the oxygen concentration in steps, using smaller increments of about 0.2% oxygen near the final value, applying the bracketing procedure defined in the standard.
• Determine the LOI as the oxygen concentration at the transition between sustained burning and self-extinguishing behavior.
• Report the LOI value together with the standard used, specimen thickness, orientation, conditioning, and any special observations such as dripping or afterglow.

When performed correctly in accredited third-party laboratories with calibrated equipment, the method shows good repeatability, with inter-laboratory differences typically within about 1% oxygen. Deviations beyond this range usually indicate differences in specimen thickness, orientation, aging state, or formulation rather than true material improvement.

Applications

• Formulation development and flame retardant screening for PU, PIR, EPS, XPS, phenolic, and elastomeric foam producers.
• Incoming material and production line quality control, including verification of flame retardant dosing accuracy.
• Compliance demonstration for building insulation products requiring B1 grade under GB 8624 or specific Euroclasses under EN 13501-1.
• Specification and acceptance testing for pipe insulation, HVAC ducting, refrigerated transport, and marine applications.
• Aging studies to verify that flame retardancy is retained after thermal cycling, humidity exposure, and long-term storage.
• Failure analysis when insulation materials show unexpected burning behavior in the field.

FAQ

What oxygen index value is required for flame-retardant insulation?

An LOI of 26% or above is generally considered flame retardant, while many building specifications and procurement standards require 28% to 32% for organic insulation foams. The exact requirement depends on the applicable classification standard, such as GB 8624 or EN 13501-1, and on the end-use scenario.

Why does the oxygen index of my foam decrease after storage?

The most common causes are migration or evaporation of additive phosphorus flame retardants and progressive oxidation of the polymer. In hydrocarbon-blown foam, the opposite may also occur: the LOI can increase during the first weeks as flammable blowing agent diffuses out of the cells.

Does higher density always mean a higher oxygen index?

Density and LOI are positively correlated for a given formulation, but density is not the dominant factor. A high-density foam with an inefficient flame retardant system can have a lower LOI than a lower-density foam with optimized PIR chemistry and synergists.

Can LOI results from different laboratories or thicknesses be compared?

Only results obtained with the same test standard, specimen thickness, orientation, and conditioning state are directly comparable. Comparing a 3 mm ASTM D2863 specimen with a 10 mm GB/T 2406.2 specimen can lead to incorrect conclusions.

How can the oxygen index of polyurethane insulation be increased?

Effective strategies include raising the isocyanate index to increase isocyanurate content, using reactive brominated polyols for durable flame retardancy, adding expandable graphite or phosphorus-nitrogen intumescent synergists, increasing density, and improving cell uniformity through better processing control.

Is oxygen index alone sufficient to prove fire safety?

No. LOI is a valuable screening and quality control parameter, but full fire safety assessment also requires heat release, smoke production, flame spread, and dripping evaluation using tests such as ISO 5660-1, ISO 5659-2, EN 13823, and EN ISO 11925-2 within the relevant classification framework.