I have worked in a technical capacity in the UK Thermal Insulation Industry since 1979 and at the risk of sounding like a grumpy old man I have to say that I am appalled at the apparent lack of knowledge on the subject of Reaction to Fire performance for construction products in general and pipe insulation materials in particular. We all have ready access to lots of information these days via various AI assistants but all too often it seems that this information is parroted by people who seem to have little comprehension of what they are talking about.

           Don’t get me wrong I feel that the Building Safety Act is a good thing and a necessary response to laissez faire approach to Building Regulations that produced the Grenfell Tower tragedy, but I do think that it is important that the regulators and specifiers who are tasked with enforcing the new regime need to take the time to understand what they are talking about. In my industry I am regularly confronted by insulation manufacturers who claim that their products are non-combustible based upon passing a simple ignitability test whilst on the other hand you have insurance companies who refuse to cover the use of any insulation ‘foam’ product in any situation.

           There are no perfect insulation materials and so as with most things in life it is a case of ‘horses for courses’, we need to be able to be able to balance up the strengths and weaknesses of the options available to us in order to make the best choice. The purpose of this article is therefore to explain what the various Reaction to Fire tests used in the UK are, what they can and cannot tell us and how they can be applied to make better choices when it comes to pipe insulation materials.

           The basic hierarchy of Reaction to Fire tests is as follows:

Ignitability

Spread of Flame

Heat Release

Combustibility

           The easiest type of Ignitability to pass is what is called a Zero Irradiance Ignition Resistance test, in this type of test a small flame is applied to a specimen for a short period of time, the specimen can be in a horizontal or vertical orientation, usually horizontal tests are easier to pass. In the UK our Reaction to Fire tests used to be in the BS 476 series and Part 5 was the Ignitability test, a product which passed was designated as Class P and could be described as “Not easily ignitable”. You can take a fairly combustible product such as Polyethylene Foam incorporate a modest amount of flame retardant additive and achieve Class P.

           You can make Ignition Resistance tests more difficult to pass by applying background radiation another way of checking for greater ignition resistance is to test the product in an Oxygen enhanced atmosphere. Normal air contains approximately 20% oxygen, in the Oxygen Index Test you place the sample in a glass cylinder and gradually increase the concentration of oxygen until you get sustained ignition when a small flame is applied. The percentage of oxygen required to achieve ignition is called the Oxygen Index. As a Rule of Thumb a product with an Oxygen Index of 30 can be described as “Not Susceptible to Casual Ignition. You can also check the resistance of a product to ignition by placing it in an environment where the temperature can be increased to high levels, this is the basis of the so called Temperature Index test.

           The Oxygen Index and Temperature Tests are useful small scale screening tests mainly used for Quality Control purposes by organisations such as Thermal Insulation Manufacturers and Regulators with a good understanding of Reaction to Fire Tests such as the M.O.D and TfL.

           Before going further into this subject, I would like to say something about understanding the nature of the product you are testing and the limitations of the test you are applying. When it comes to polymeric materials it is important to know whether a product is thermoplastic or not. Products which melt can sometimes remove themselves from the ignition source by doing so and this can be helpful when it comes to passing a test. It also important to know how and where a flame is applied to a sample. The product which caused the rapid spread of flame from floor to floor in Grenfell Tower was a rain-screen cladding material with a solid polyethylene core and a thin aluminium sheet facing material. The flame was applied to the aluminium sheet, the product passed the fire test but when, in the real world, a fridge caught fire flames from it got on the edge of the panel causing it to melt, the polyethene contained little or no flame retardant and so went up like a 20m long candle which reached the top of the building in less than 20 minutes. This is what happens when saving money is the only goal and ignorant people try to undermine fire safety by calling it ‘red tape’.

           Returning to the subject, I think we can agree Ignition Resistance is not the whole picture but can be a useful starting point. These days we use a European harmonised set of Reaction to Fire tests. The Classification system is BS EN 13501-1 and the classes range from A1 to F. If a product is classed as F this means that no Reaction to Fire tests have been performed. The Ignition Resistance test is EN ISO 11925-2 this is a zero irradiance, horizontal orientation test, in order to achieve a Class E , the product is exposed to a flame for 15 seconds and the flame-spread must be less than or equal to 150mm after 20 seconds; although this classed as, an Ignition Resistance test we do see the introduction of Spread of Flame for the first time as a means of quantifying the degree of resistance.

           EN 11925-2 is also used as part of the Class D test regime, this time the sample is exposed to the flame for 30 seconds and the flame-spread must be equal to our less than 150mm after 60 seconds. If the product is just relying upon a flame retardant additive and is itself inherently flammable then there is more chance that the longer exposure time will uncover this weakness. It is at the Class D level that the SBI Test EN 13823 is also introduced for the first time. The Single Burning Item test is a larger scale test which is able to quantity several aspects of fire performance namely Surface Spread of Flame, Heat Release and Smoke Production. Heat Release in the SBI test is quantified by what is called a Fire Growth Index, which is usually referred to as the FIGRA Index. In order to achieve Class D a flat product such as Construction Board or a Thermal Insulation sheet material must have a FIGRA of 750 Watts per second of less. The is a separate Criteria for Linear Pipe Thermal Insulation Products the limit for FIGRA here is less than or equal to 2100W/s but an additional limit is also placed upon the Total Heat Release which must be less than or equal to 100 Megajoules at 600 seconds (viz. 10 minutes). The Ignition source in the SBI Test is a 30kW propane burner which is designed to have about the same amount of energy you would get if a wastepaper basket went up in flames.

           EN 13501-1was developed as part of the Construction Products Directive which required all Member States of the European Union to modify their national regulations so that we had a set of harmonised standards which were acceptable to all. Previously UK Building Regulations would have referred to the BS 476 series when it came to fire performance but by getting fire experts from across Europe to agree on what was important when it came to fire safety manufacturers were able to develop products which met the needs of a much bigger market. The UK was a major contributor to the development of EN 13501 Part 1 and Part 2 (which covers Fire Resistance) and we are still happy to work to them even though we are no longer in the EU.

           Having said this there are significant differences between the system which we previously operated in the UK and the harmonised system which we are using now. Under the old system BS 476: Part 7 was very clearly a Surface Spread of Flame test. Samples were placed horizontally in front of a gas-fired furnace and a pilot flame was switched on at the base of the sample edge nearest to the furnace. The radiant heat being produced at the end of the sample nearest to the furnace was 25kW/m2. The idea was that the pilot flame would ignite a flame-front across the surface of the sample and that the Fire Engineer would measure the initial and final flamespread. If both the Initial and Final flamespread were 165mm or less for 6 samples then the product was designated Class 1. This is a good and reliable flamespread test for a range of construction products including most thermal insulation materials but it does have its limitations. In some cases vertical flamespread can give a better indication of potential risk and there is an argument that in some cases a higher irradiance level is required to distinguish fire risk more accurately. One example of a more severe flamespread test is IMO A653(16) which has a similar sample size and the same sample orientation, but which uses an irradiation level of 50kW/m2 at the hot end.

           It is certainly true to say that flamespread alone cannot give you the whole picture when it comes to evaluating fire risk you need to know something about the Heat Release characteristics of a construction product because a high rate of heat release can cause nearby products to burn. In the BS 476 series of tests this was determined by BS 476 Part 6, samples of material 225mm square were tested in an apparatus which looked like a large toaster. The apparatus was run without any samples in to give a time/temperature curve and then 3 samples were tested on the same piece of equipment and the time/temperature curves were again plotted. The difference between the control curve and the sample curves gives an indirect measure of the rate of heat release which was expressed as an Index of Performance. The tests took place over 20 minutes and the area under the curve was divided into 3 sections. The results were averaged across the 3 samples, the initial index, i1, had to be 6.0 or less and the Total Index, I , had to be 12.0 or less. Under UK Building Regulations a product which achieved Class 1 to Part 7 and the Indices of Performance indicated above to Part 6 was designated Class 0.

           Products which achieved Class 0 were usually regarded as ‘fire safe’ by most regulators in the UK for many years. In most cases this was true but if you knew what you were doing it was not too difficult to ‘cheat the test’; for example one could take a thermoplastic insulation foam with an Oxygen Index of 22 to 24 and cover it with 20 microns of Aluminium Foil, this would achieve Class 0 nine times out of ten.

           No fire tests are perfect, some of our colleagues in Europe did not like certain aspects of our BS 476 series and equally we could point out the weaknesses in their tests, this how we ended up with EN 13501-1 and when it came to measuring Flamespread and Heat Release it was agreed to do them both together using EN 13823 the SBI test. EN 13823 is what is often referred to as an Intermediate Scale test, it is basically a scaled down version of a longer established method called the Corner Wall Test. The apparatus consists of two walls at right angles and a floor, all of which are made from non-combustible materials; this is mounted on a trolley made from metal sections; the whole thing can be positioned under an exhaust hood. The burner is a triangular arrangement in the corner.

           Using the SBI test it is possible to measure both horizontal and vertical flamespread, to get a direct measurement of heat release, to quantify smoke production and to check for the production of flaming droplets or particles. I have already mentioned the SBI Test above and given the criteria which are required to designate flat products as either Class E or Class D. Before going further I should make it clear that not all products are tested not Reaction to Fire characteristics as flat products in a vertical orientation. Flooring Products are tested in accordance with the same EN ISO 11925-2 Ignitability method for Class E but for Classes D to A2 a different apparatus known as EN ISO 9239-1 is used; this is a horizontal flamespread test where the heat source is radiant panel placed above the samples and where a Critical Flux which is varied from 3.0kW/m2 to 8.0kW/m2 is used to determine classification.

           There is also a version of the SBI test which is used to test so-called ‘Linear Products’ which usually means pipe thermal insulation products, these are placed on a series of metal pipes covering both wings of the walls with an air gap between each insulated pipe. I have already mentioned some of this above but I now want to revisit this subject in a bit more detail.

When it comes to Class D with the SBI Test you have to achieve a FIGRA of 750W/s or less for flat products but you are allowed up to 2100W/s for pipe insulation which nearly 3 times as much. The pipe insulation manufacturers argued that there is usually significantly less mass in a few pipes that in a continuous flat sheet and I think that this is usually a fair argument, especially with the backstop provided by the limit on the total amount of Heat Release but I do urge any specifier responsible for looking at Reaction to Fire tests to consider the actual numbers not just the Classifications and to remember the difference between Class D and Class Dl.

           When we move on Class C the FIGRA Limits are 250W/s for flat and 460W/s for pipes, there is the same Lateral Flame Spread limit for both configurations and the Total Heat Release for both at 600 seconds is limited to 15 Mega Joules, which is a significant reduction from the 100 MJ limit which is permitted for Class D. When I sat on the BSI Reaction to Fire Technical Committees, we were trying to get Class C to be roughly equivalent to Class 1 to BS 476 Part 7.

           The FIGRA limits for Class B are 120W/s for the general case and 270W/s for Bl; the Heat Release limit THR 600s for both is halved again to 7.5MJ. By the time we have reached this level we hope and expect that we are dealing with products which have low heat release and low flame-spread characteristics. For those who still think in terms of the old British standards we are at about the same sort of level as we aimed to have with Class 0.

           We are now approaching the highest performance level in Reaction to Fire testing namely Combustibility. In an ideal world everything would be non-combustible, right? Except that this is one of the most misunderstood and misused terms when it comes to fire performance. Back in the days when I first joined the thermal insulation industry we understood that for a material to be non-combustible it had to pass BS 476 Part 4 and that if it passed BS 476 Part 11 it was a material of Limited Combustibility, we also understood that the latter was still a high level of performance and probably meant that the product was about 92% inorganic.

 476 Part 4 is entitled ‘Combustibility test for Materials’, the apparatus is a type of muffle furnace and the pass criteria are Temperature Rise for the Furnace and the Specimen centre must not exceed 50 Deg. C or less and sustained flaming must not exceed 10 seconds. Some of you are probably thinking ‘If it is totally non-combustible why is there any temperature increase or flaming?’ The short non-technical answer is that if you can meet these limits in this apparatus the organic content is probably 3% or less and it can be regarded as non-combustible for any Construction Product application.

BS 476 Part 11is entitled “Method for Assessing Heat Emission” , it was first published in 1982 and by this time the fire testing community had decided that it was a good idea to explicitly acknowledge that nothing is completely non-combustible, hence the change of title. It is also a furnace operating at 750 C and in BS5422:2009 Annex E defines non-combustible materials as either passing Part 4 or “does not flame and does not cause any rise in temperature on either the centre (specimen) thermocouple exceeding 35 Deg. C and a rise exceeding 20 C on the furnace thermocouple”

By 2009 EN 13501-1 was on its way but had not replaced the BS 476 methods in the UK. The harmonised system retains a furnace method as a way of assessing non-combustibility but also introduces the idea of a Bomb Calorimeter. The Non-Combustibility test is now defined in EN ISO 1182 it now uses a tube furnace, it still operates at 750 C but has updated instrument control and proprietary analytical software. The performance requirements for an A2 and an A1 Ratting are defined by EN13501-1.

Some other members of the EU were not as keen on relying totally on furnace methods as we in the UK had been. They favoured what is called a Bomb Calorimeter as the best way of assessing the potential maximum total heat release of a product. The idea is that you burn a sample in pure oxygen in order to determine what happens when you burn it completely. The apparatus for achieving this objective is defined by EN ISO 1716- Determination of the heat of combustion.

You can achieve an A2 rating by passing either EN ISO 1182 or EN ISO 1716 but to achieve A1 you have to pass both; I think that this ‘belt and braces’ approach is a very good idea when it comes to fire safety. The details are a bit more complicated, but I will do my best to make them comprehensible.

For the A2 rating you still have to do the SBI test and achieve the same rating for Euroclass B as you would for each of the three types of Construction Product namely flat products excluding flooring, floorings and linear pipe thermal insulation products. For the Furnace test EN ISO 1182, the Temperature Increase must not exceed 50 C and at least 50% mass must be retained, and any transitory flaming must not last longer than 20 seconds. The reason for the mass loss limit is to prevent the use of large amounts of endothermic fillers to mask the underlying performance of the base material. 

When it comes to the Bomb Calorimeter test EN ISO 1716 limits are specified for PCS (Gross Heat of Combustion), the A2 limits are 3.0 Mega Joules per Kilogramme for Homogeneous Products and substantial components of non-homogeneous products, there are slightly different limits ranging from 3.0 to 4.0MJ/kg for any other non-substantial components of non-homogeneous products.

When it comes to A1 there is no need to do the SBI test but you must do both the EN ISO 1182 and the EN ISO 1716 tests with tighter limits set for them. With the furnace test the temperature increase must not exceed 30 C, at least 50% mass must be retained and no sustained flaming is permitted. For the bomb calorimeter test the Gross Heat of Combustion is limited to 2,0MJ/kg for homogeneous products and substantial components of non-homogeneous products with limits ranging from 1.4 to 2.0MJ/kg for any other non-substantial components of non-homogeneous products.

There are also additional classification criteria for smoke production, flaming droplets/particles in EN 13501-1. In the case of Flaming Droplets or Particles the first point at which a rating can be achieved is Class E. There are 3 levels of performance d0, d1 and d2. If a material ignites the paper in the small scale Ignition Resistance test EN 11925-2 it is Classified as d2, if it passes the test (no ignition) no formal classification is given but the fact of passing the test can be recorded on the manufacturer’s literature. If you pass this test and have a chance of achieving a higher primary classification then you would probably do an SBI test EN 13823 which also includes this additional criteria but, in an apparatus where the material is subjected to a larger flame source and radiant heat.

 In the SBI test a d2 rating is classified as not do or d1; to put it in plain English the sample produces flaming droplets or flaming particles. You can achieve d1 if no flaming droplets/particles persist for longer than 10 seconds during the 10 minute duration of the test. What this usually means is that the product is thermoplastic but that the flame retardant additive is still able to extinguish the polymer once it escapes the flames and radiant heat produced by the propane burner. In order to achieve d0 that material must produce no flaming droplets or particles in 600 seconds. My advice for what it is worth is never use a product graded d2 and only use a product graded d1 in low risk situations. I should mention at this point that these additional classifications are currently only advisory in the UK, they have to be reported but there is no requirement to restrict usage under current Building Regulations.

Smoke is a more tricky subject, how much is produced can depend upon product orientation, the availability of oxygen, the amount of radiant heat in the relevant area and the presence or absence of air movement. I am not going to get into any of these complexities here. It should also be noted that it is only the obscuration caused by the smoke which is measured, EN 13501-1 makes no attempt to measure or control the toxicity of combustion products.

Under the EN 13501-1 Classification System smoke is measured using the SBI test breaking down the amount of smoke produced in to 3 bands namely s1, s2 and s3. If a product is classified as s3 this means that it is unable to meet the limits set for either s1 or s2. The amount of smoke produced is classified according to a Smoke Growth Index or SMOGRA, the method seeks to give an indication of how quickly smoke is produced and how much is produced in total. In order to achieve s1 for a flat product other than Flooring SMOGRA must be less than or equal to 30m2 per second squared and the Total Smoke Production must be less than or equal to 50m2, for Pipe Insulation products the s1 limits are equal to or less than 105m2/s2 and TSP600S must be equal to or less than 250m2. I do not have much insight into what this would look like visually but my understanding is that those who are experienced in this area are happy that this is a reasonable limit in both cases for what one might call a ‘low smoke’ material.

The limits for s2 for flat products and pipe insulation are as follows: Flat: SMOGRA equal to or less than 180m2/s2 and TSP600S equal to or less than 200m2 and Pipes SMOGRA equal to or less than 580m2/s2 and TSP600S equal to or less than1600m2. These seem like quite broad ranges to me, and I understand them to go from quite low to moderately high in both cases. If I were the specifier here I would be inclined to look at not just the band a product falls into but also the actual numbers it achieves.

There are only two bands for smoke production when it comes to Flooring materials s1 and s2. For s1 Smoke must be less than or equal to 750% per minute and s2 just means anything above this limit. The Smoke Production is measured as part of the EN ISO 9239 Radiant Panel test. There is the same limit for all Classifications from A2 to D and no performance is determined for EF or DF.. I have even less insight in to what these two classifications mean for Flooring and could not tell you whether those in the sector regard s1 as a good performance level or not, all I can say is that Flooring Manufacturers claim that smoke is not usually a significant safety issue with flooring products and I can see how they might be correct provided that the primary Reaction to Fire characteristics of said product are good.

Building Regulations outside the UK specify requirements for the additional classifications particularly smoke in a number of countries. I can see the merits of doing this particularly in buildings with a high level of occupancy, this is of course still an option available to Consulting Engineers and other Specifiers on a Project by Project basis but as with most things in life better products usually cost more.

May I take this opportunity to congratulate those of you who have taken the time to read and understand the contents of this article. Doing so will not make you an expert on Reaction to Fire testing in the UK market but being a one-eyed man in the Kingdom of the Blind would make a significant difference if enough of the relevant people take the time to understand what the regulations mean not just what they say.

Les Johnson

Technical Director

AIM Ltd September 2025