Chapter 5: Fire testing

Introduction

¶5.1 In this chapter we describe the reaction to fire tests which supported the regulatory framework governing the fire safety of external walls of buildings in England & Wales in the years leading up to the Grenfell Tower fire. They include the national and European reaction to fire tests¹¹⁴¹¹⁴ Reaction to fire must be distinguished from resistance to fire. They are different concepts and subject to different tests. and classification regimes, as well as the national large-scale testing and classification regime.

¶5.2 There are eight reaction to fire tests referred to in Approved Document B; four national test standards (BS 476 Parts 4, 6, 7 and 11) and four European test standards (BS EN 13823, BS EN ISO 1716 and 11925-2 and BS EN 1182). National test method BS 8414 and the criteria contained in a document published by BRE entitled Fire performance of external thermal insulation for walls of multi storey buildings, generally known as BR 135, relate to the large-scale testing of external wall cladding systems. A number of experts instructed by the Inquiry, including Dr Lane, Professor Bisby and Professor Torero, gave evidence about the background to those tests, their purpose, the methods by which they are conducted, and the information that can be derived from them. Here we summarise the key points arising from their evidence and the documents on which it was based.

National reaction to fire tests

¶5.3 Before considering each of the key national reaction to fire tests, it is useful to understand the basic stages of a fire in a compartment or room, since it is that "reference scenario" (i.e. the test conditions and related assumptions) which underpins many of the national and European tests. In broad terms, the development of a fire in a room can be divided into three key stages: ignition and growth, fully developed and decay.¹¹⁵¹¹⁵ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 6.

¶5.4 The phenomenon commonly known as "flashover" occurs at the transition between the second and third stages and is characterised by an exponential growth in the size of the fire caused by the rapid ignition of all fuel sources within the compartment.¹¹⁶¹¹⁶ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/136} at 12.4.18. It is an important indicator of the onset of a fully developed fire.¹¹⁷¹¹⁷ Lane {Day68/9:12-13}. Reaction to fire tests are used to characterise the capacity of products and materials used in construction to contribute to the initiation and growth stages of a fire, leading up to flashover,¹¹⁸¹¹⁸ BSI 476-10:2009 "Fire tests on building materials and structures" {BSI00001757/16} at 5.1. and therefore each of the tests in the BS 476 series is relevant when considering these stages of a fire in a room.¹¹⁹¹¹⁹ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 9. More specifically, the small and intermediate scale tests that are referred to in the guidance in Approved Document B on constructing external walls are associated with the early stage of fire growth before flashover has occurred.¹²⁰¹²⁰ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 9.

BS 476-4: non-combustibility

¶5.5 BS 476-4¹²¹¹²¹ {CTAR00000014}. is a test to determine non-combustibility. In paragraphs 91 to 101 of his report entitled Regulatory Testing and the Path to Grenfell¹²²¹²² Bisby, Phase 2 Report - Regulatory Testing and the Path to Grenfell {LBYP20000001/20}. Professor Bisby summarised the evolution of testing for combustibility and non-combustibility, both in the UK and internationally, including the publication in 1970 of BS 476-4.¹²³¹²³ With further amendments made in 1978 and 1983 and further information added to the foreword in 2014 - Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 21. The test involves a small cuboid sample¹²⁴¹²⁴ Measuring 40mm x 40mm x 50mm - Lane, Expert witness presentation - 10 November 2020: https://www. grenfelltowerinquiry.org.uk/hearings/expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 24 of the material being placed in an electric furnace which has been heated to 750°C for a minimum of 10 minutes before the test.¹²⁵¹²⁵ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/ expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, pages 21-29 and see Lane {Day68/15:9}-{Day68/17:25} including video played of the test procedure. Three specimens are tested separately and a record is made of the measurements made by two thermocouples placed at the centre of each specimen. Heat is provided in the furnace by metal coils and no direct flame impingement occurs. The sample is inserted into the furnace and a record is made of the temperature at the thermocouples for 20 minutes. The occurrence of any flaming in the furnace is noted. Air is allowed to flow through the furnace to enable the combustion of any fuel vapours released by the specimen to the extent that they are capable of forming a combustible mixture and are ignited.¹²⁶¹²⁶ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/120} at 12.3.11.

¶5.6 A material is deemed to be non-combustible if, during a test conducted in accordance with BS 476-4, none of the three specimens either (1) causes the temperature reading at either of the thermocouples to rise by 50°C or more above the initial furnace temperature or (2) is observed to flame continuously for 10 seconds or more inside the furnace. Otherwise, the material is deemed to be combustible.¹²⁷¹²⁷ {CTAR00000014/10}. As explained by Professors Torero and Bisby, it is important for the user of this test data to understand that a designation of "non-combustible" does not mean that the material is completely inert,¹²⁸¹²⁸ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/120} at 12.3.19. since the criteria are somewhat arbitrary, including the criterion of no continuous flaming for 10 seconds.¹²⁹¹²⁹ Bisby Phase 2 Report - Regulatory Testing and the Path to Grenfell {LBYP20000001/23-24} at 109-110 and 116.

BS 476-11: method for assessing heat emission

¶5.7 BS 476-11¹³⁰¹³⁰ {CTAR00000015}. is the method used for assessing the heat emitted by materials. It is the national test for assessing whether a material is of "limited combustibility", as defined in Approved Document B. The standard makes it clear at the outset that it is not intended to assess the effect of materials on fire growth.¹³¹¹³¹ {CTAR00000015/4} and see discussion by Torero at, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/123}. It was first published in 1982 and has not been amended since.¹³²¹³² Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 30. Some information was added to the foreword in 2014. The test apparatus is similar to that used in the BS 476-4 test.¹³³¹³³ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, pages 31-36 and Lane {Day68/18:1}-{Day68/21:18} including video played of the test procedure. A small¹³⁴¹³⁴ Specimens are 45mm in diameter and 50mm in height. cylindrical specimen, whose mass has been measured before the test, is inserted into a furnace which has been heated to 750°C. A thermocouple is inserted at the centre of the specimen and thermocouples are also provided for measuring the temperature of the furnace and the wall of the furnace. The temperatures measured by the thermocouples during the test are recorded and in addition visual observations are made. The procedure is repeated five times. The duration of each test is longer than under BS 476-4 and can be up to 120 minutes. Following each test, both the rise in the temperature of the furnace and the rise in the temperature of the specimen are calculated in relation to each of the five specimens and an average value is obtained. The average duration of sustained flaming is also recorded. Additional measurements are also made, including of the loss of mass of the sample. Other than the reporting of measurements and temperature differences, there is no classification or designation in the test standard itself. Such classifications and designations are to be found in other documents, including Approved Document B.¹³⁵¹³⁵ Torero, Phase 2 Report: Adequacy of the Current Testing Regime {JTOR00000006/124} at 12.3.40. It is a requirement of BS 476-11 that any report make it expressly clear that the results relate only to the behaviour of the specimens under the particular conditions of the test and that the results are not intended to be the sole criterion for assessing the potential fire hazard of the material in use.¹³⁶¹³⁶ {CTAR00000015/9} at 8(g).

BS 476-6: method of test for the fire propagation of products

¶5.8 BS 476-6, which was first published in 1968,¹³⁷¹³⁷ Three further versions were published with the most recent in 2009 {CTAR00000016} and in 2014 a change was made to its foreword – see Lane, Expert witness presentation - 10 November 2020: https://www. grenfelltowerinquiry.org.uk/hearings/expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 37. is the test method for assessing the fire propagation properties of products. Professor Bisby explained that it was developed in order to distinguish better between the fire hazards presented by different products when used within a compartment.¹³⁸¹³⁸ Bisby, Regulatory Testing and the Path to Grenfell {LBYP20000001/32-35} at 156-162. The standard makes clear that the result of the test is expressed as a fire propagation index which provides a comparative measure of the contribution to the growth of fire made by an essentially flat material, composite or assembly.¹³⁹¹³⁹ {CTAR00000016/5} at 1 Scope. The test is primarily intended for the assessment of the performance in a fire of internal wall and ceiling linings.¹⁴⁰¹⁴⁰ {CTAR00000016/5} at 1 Scope. This test, together with BS 476-7, is used to determine national Class 0 as defined in Approved Document B.

¶5.9 The test apparatus comprises a combustion chamber with a specimen holder fixed onto the front face. The combustion chamber contains a gas burner and two electrical heating elements and is surmounted by a removable steel chimney. The specimen holder is recessed and takes a specimen measuring 225mm x 225mm. Because the opening to the combustion chamber measures only 190mm x 190mm, the edges of the sample are not directly exposed to heat from the gas burner or electric heaters.¹⁴¹¹⁴¹ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/ expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, pages 39-50 and see Lane {Day68/21:19}-{Day68/33:8} including video played of the test procedure. Only the front face of the specimen is heated and the fire propagation index subsequently calculated reflects that surface heating.¹⁴²¹⁴² See also Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/127} at 12.3.54 and {JTOR00000006/129} at 12.3.64. The test runs for 20 minutes. During the test, the output from the thermocouples measuring the temperature of the flue gases in the chimney is recorded and specific temperature measurements are taken at prescribed intervals. For a composite material the outer surface is heated first and any rise in temperature recorded in the chimney reflects the combustion of that surface together with any combustion of the material behind it.¹⁴³¹⁴³ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 54. Visual observations are also made during the test, including of deformation or spalling of the specimen, which can result in the test being regarded as invalid in certain circumstances. Data obtained from the test is compared with data obtained when a calcium silicate board is tested. At least three, and no more than five, specimens must be tested. If more than three specimens are tested, three can be selected to provide the test result.¹⁴⁴¹⁴⁴ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/127} at 12.3.53. This is a depiction of the test apparatus, taken from Dr Lane's presentation:

¶Figure 5.1: Depiction of apparatus for BS 476-6 test.

¶5.10 The results of the test are expressed as a Fire Propagation Index, which is based on the difference in the rise in temperature of the effluent gases between the tested samples and the calcium silicate board. In respect of each sample, an average is calculated of the differences in temperature recorded during each period of time in which the rate of heating of the furnace varies. An average of the results is then calculated to obtain a single value for each sample (known as sub-indices i1, i2, i3), which are added together to give an overall Fire Propagation Index, referred to as I.¹⁴⁵¹⁴⁵ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/127} at 12.3.56; Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert-witness- presentation-dr-barbara-lane-10-november-2020, Part 1, page 52. That overall value, together with the subindices, are the critical results of the test and are referred to in Approved Document B. The key information which the test provides is an indication of how much heat the product is capable of releasing.¹⁴⁶¹⁴⁶ Torero {Day292/28:7-16}.

¶5.11 The contents of the test report are laid down in the standard, which provides that it shall include details of the form in which the specimens were tested (i.e. material, composite or assembly), together with the thickness of the specimen and, where appropriate, its orientation and backing material.¹⁴⁷¹⁴⁷ {CTAR00000016/11} at 11(f). The test report must also contain a statement that the test results relate only to the behaviour of the test specimens of the product under the particular conditions of the test and are not intended to be the sole criterion for assessing the potential fire hazard of the product in use.¹⁴⁸¹⁴⁸ {CTAR00000016/11} at 11(i). Annex B to the standard makes it clear that the influence of underlying layers on the assembly should be understood and that care should be taken to ensure that the results obtained on any assembly are relevant to its use in practice.¹⁴⁹¹⁴⁹ {CTAR00000016/18} left column; Lane, Expert witness presentation - 10 November 2020: https://www. grenfelltowerinquiry.org.uk/hearings/expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 43. The standard also provides that where the product is to be used in conjunction with a particular substrate, it should be tested in conjunction with that substrate.¹⁵⁰¹⁵⁰ {CTAR00000016/18} top right column; Lane, Expert witness presentation - 10 November 2020: https://www. grenfelltowerinquiry.org.uk/hearings/expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 44.

¶5.12 When ACM PE panels are tested using this method, the aluminium can shield the polyethylene, protecting it from exposure to a direct flame and allowing it to melt and flow away, rather than ignite within the panel during the test. If it does not ignite, the temperatures measured in the chimney are lower.¹⁵¹¹⁵¹ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, pages 58-59; Lane {Day 68/32:2-10}. It follows that if a product containing a core that is completely enclosed does not suffer a failure of the external material, the results of the test will not reveal the presence of any combustible material within.¹⁵²¹⁵² Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/129} at 12.3.64.

BS 476-7: surface spread of flame test

¶5.13 BS 476-7¹⁵³¹⁵³ {CTAR00000017}. was first published in 1971 and was revised in 1987 and 1997.¹⁵⁴¹⁵⁴ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/ expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 61. There was also a change to the foreword in 2014. It is a test method designed to determine the ease with which flame will spread across the surface of a product. It is the second main test used to support a national Class 0 classification. The test was developed as one of a new breed of small-scale fire tests which were intended to assess the fire hazard presented by a product when used in practical applications.¹⁵⁵¹⁵⁵ Bisby, Regulatory Testing and the Path to Grenfell {LBYP20000001/29-31} at 144-151. It measures the horizontal spread of flame across the surface of a product and provides data which is suitable for comparing the performance in use of flat materials, composites or assemblies used primarily as the exposed surfaces of internal walls or ceilings.¹⁵⁶¹⁵⁶ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, page 63.

¶5.14 The scale of apparatus used is very different from that used for the national fire tests described above; generally, this is a much larger test.¹⁵⁷¹⁵⁷ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, pages 64-72; Lane {Day68/33:9}-{Day68/40:24} including video of test procedure. The apparatus consists of a radiation panel measuring 850mm by 850mm, with a specimen holder protruding at a right angle to the radiating surface. The specimen holder comprises a water-cooled steel frame with water-cooled face-plates which overlap the specimen at the edges, thereby ensuring that the edges are not exposed during the test. The specimen is marked with reference lines at set distances before it is mounted into the test rig. Four vertical lines are marked at the distances corresponding to the classification limits for Classes 1–4. A minimum of six and a maximum of nine specimens are tested, which should be representative of the exposed surface of the product. If the product is normally used in conjunction with a substrate, it should be tested with that substrate.

¶5.15 During the test the radiant burner swings into place at a right angle to the specimen and a pilot flame is ignited in the lower corner of the specimen nearest the radiant panel. The pilot flame is extinguished one minute after the start of the test. The test involves recording the extent of flame spread along the face of the specimen after one and a half minutes and ten minutes, the latter being the duration of the test. Any flaming which occurs from material that has fallen or melted below the test rig is disregarded, although observations of such phenomena should be made. The test is terminated if the flame front reaches the 825mm line towards the end of the sample away from the radiant burner.¹⁵⁸¹⁵⁸ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 1, pages 71-72. During the test the radiant panel provides an external heat flux of 32.5 kW/m² at the face of its vertical edge closest to the panel, which drops to 5k W/m² at the distant end of the sample.¹⁵⁹¹⁵⁹ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/130} at 12.3.70. This is a picture of the test during operation taken from Dr Lane's presentation:

¶Figure 5.2: Image of BS 476-7 test in operation

¶5.16 Depending on how far the flame has spread after one and a half and ten minutes, the product will receive a classification ranging from Class 1 to Class 4, Class 1 being the best and Class 4 the worst.

National testing classification regime

¶5.17 Unlike Europe, the UK has no overarching system for classifying the reaction to fire of products and materials. Instead, the national classes rely on definitions contained in Approved Document B. We set out below a diagram from Dr Lane's presentation which sets out the relevant definitions in Approved Document B. We have concentrated on the three definitions central to our work: non-combustible, limited combustibility and national Class 0.

¶Figure 5.3: National framework - reaction to fire definitions

Non-combustible

¶5.18 Table A6 of Approved Document B sets out the "Use and definitions of non-combustible materials".¹⁶⁰¹⁶⁰ {CLG00000224/131}. The national class definitions of non-combustible materials are based principally on testing in accordance with BS 476-4 and BS 476-11. Any products classified as non-combustible in accordance with BS 476-4 are non-combustible pursuant to the definition.¹⁶¹¹⁶¹ {CLG00000224/131} sub-paragraph (d) in Table A6 national class. BS 476-11 itself does not provide any limits on temperature rise or duration of flaming, but Table A6 states that a material will be classified as non-combustible when tested to BS 476-11 if it does not flame or cause any rise in temperature on either the specimen or furnace thermocouples.¹⁶²¹⁶² {CLG00000224/131} sub-paragraph (a) in Table A6 national class. In addition Table A6 makes it clear that any totally inorganic materials such as concrete are also to be regarded as non-combustible.¹⁶³¹⁶³ {CLG00000224/131} sub-paragraph (b) in Table A6 national class – see (b) in Table A7 national class.

Limited combustibility

¶5.19 Table A7 of Appendix A sets out the "Use and definitions of materials of limited combustibility".¹⁶⁴¹⁶⁴ {CLG00000224/132}. The table is divided into different rows by reference to particular parts of the guidance in Approved Document B which refer to a requirement for materials of limited combustibility. Any material classed as non-combustible meets that definition¹⁶⁵¹⁶⁵ {CLG00000224/132} sub-paragraph (a) in Table A7 national class. and in some cases testing to BS 476-11 is relevant. For insulation materials in external wall constructions referred to in paragraph 12.7 of Approved Document B, Table A7 provides that any material with a density less than 300kg/m³ will satisfy the definition of limited combustibility if, when tested to BS 476-11, it does not flame for more than ten seconds and the rise in temperature on the specimen thermocouple is not more than 35°C and on the furnace thermocouple is not more than 25°C.¹⁶⁶¹⁶⁶ A more onerous requirement is set for materials of a density of 300/kg/m³ or more.

National Class 0

¶5.20 In the 2013 version of Approved Document B,¹⁶⁷¹⁶⁷ {CLG00000224} 2006 edition with 2007, 2010 and 2013 amendments. in a section headed "Internal linings", paragraph 13 of Appendix A provided the following definition of national Class 0:¹⁶⁸¹⁶⁸ {CLG00000224/122}.

"The highest National product performance classification for lining materials is Class 0. This is achieved if a material or the surface of a composite product is either: a. composed throughout of materials of limited combustibility; or b. a Class 1 material which has a fire propagation index (I) of not more than 12 and sub-index (i1) of not more than 6. Note: Class 0 is not a classification identified in any British Standard test."¹⁶⁹¹⁶⁹ {CLG00000224/122} at 13.

¶5.21 Two different ways of meeting the requirements for Class 0 are apparent from that definition. First, they are met if a material or the surface of a composite product is composed throughout of materials of limited combustibility. That takes the reader back to Table A7 of Approved Document B, which contains the definitions of limited combustibility. Item 6 in Table A7 refers to "Class 0 materials meeting the provisions in Appendix A, paragraph 13(a)". The definitions of materials of limited combustibility for that category are based on testing to BS 476-11 (or BS 476-4 if the material is non-combustible).¹⁷⁰¹⁷⁰ {CLG00000224/132}.

¶5.22 There is some ambiguity in the reference in Approved Document B to the "surface" of a composite product being composed "throughout" of materials of limited combustibility. This appears to contemplate that, provided the surface of a composite material is of limited combustibility, national Class 0 can be achieved regardless of the nature of the material behind that surface, or whether the surface is capable of encapsulating what lies behind. In addition, the reference to "surface" is uncertain. No definition of "surface" is provided in Approved Document B and it is not clear whether it is intended to refer to a paint surface or other coating only, or whether it includes a more substantial surface of the external wall, such as an outer skin or sheet which is itself bonded to another material.

¶5.23 In our view, therefore, there are difficulties with the interpretation of paragraph 13. On one reading of it, the author appears to have contemplated that the "surface" of a composite product might be composed "throughout" of materials of limited combustibility, but in this context the word "throughout" more naturally refers to the interior of a material or product than to its surface; and if the intention had been simply to refer to the whole of the surface area, that could have been achieved much more simply and clearly by referring to the "entire" surface of a composite product. We therefore think it unlikely that that is what paragraph 13 was intended to mean. It is more likely, in our view, that the wording is the product of a clumsy attempt to condense into one compendious expression the following distinct ideas: (i) an homogeneous material of limited combustibility,

¶(ii) a composite product composed throughout of different materials, all of limited combustibility and (iii) a composite product whose surface is composed of one or more materials of limited combustibility. It can, and in our view should, be read as referring to those three distinct situations.

¶5.24 The second way of meeting the requirements for Class 0 refers to test results from BS 476- 6 on fire propagation and BS 476-7 on surface spread of flame. The reference to a "Class 1" material comes from the BS 476-7 test¹⁷¹¹⁷¹ Class 1 means that the flame must not spread more than 165mm from the heated end after 1.5 minutes of the test and it should also not exceed this limit for the full duration of the test (10 minutes). and the reference to the "index (I) of not more than 12 and sub-index (i1) of not more than 6" refers to results from the BS 476-6 test. It is only by combining the results from both tests that a Class 0 classification can be obtained.

¶5.25 The history of how national Class 0 has evolved over time is set out in Chapter 6. As we explain in that chapter, the language used in the Approved Documents from 1985 onwards introduced ambiguities and a lack of clarity in relation to the definition of Class 0 which had not existed in the Building Regulations of the 1960s and 70s.

European reaction to fire tests

¶5.26 Some of the European reaction to fire tests are similar to the national tests and others are very different. Unlike the national classification system, the European classification system has been designed specifically for the purpose and contains carefully chosen classification standards. Each European reaction to fire test has its own number and is issued by the European Committee for Standardisation, an association which brings together the national standardisation bodies of 34 countries.¹⁷²¹⁷² Lane {Day68/61:1-5}.

BS EN ISO 1182: non-combustibility

¶5.27 BS EN ISO 1182 is a test method for determining the non-combustibility of materials. At the time of the Grenfell Tower refurbishment the fifth edition (2010) was current.¹⁷³¹⁷³ {BSI00001742}. The test apparatus is very similar to that used in BS 476-11,¹⁷⁴¹⁷⁴ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, pages 133-137; Lane {Day68/62:25}-{Day68/64:22} including video of the test procedure. including an electric furnace capable of maintaining a steady temperature of 750°C. Five cylindrical specimens are prepared¹⁷⁵¹⁷⁵ Each with a diameter of 45mm and a height of 50mm. and are tested individually, although data from all five is used to produce the final test result. The mass of each specimen is measured before and after the test. The specimen is inserted into the furnace, after which various temperature measurements are taken. The occurrence of any sustained flaming and its duration are also noted. The following information is recorded: (i) the percentage loss of mass of the specimen, (ii) the temperature difference between the maximum gas phase temperature and the temperature in the final minute before the test concludes and (iii) the total duration of any sustained flaming.¹⁷⁶¹⁷⁶ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/142} at 12.4.41 and see {BSI00001742/25-26} at 8.1-8.3. Despite the title of the test standard, it contains no criteria for determining whether a material can be classified as non-combustible. The purpose of the test is simply to provide data that can be used for the purposes of the reaction to fire classification system set out in EN 13501-1.¹⁷⁷¹⁷⁷ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/142} at 12.4.42. The standard makes it clear that a statement should be included in the report of the test explaining that the results relate to the behaviour of the product under the particular conditions of the test and are not intended to be the sole criterion for identifying the potential fire hazard of the product in use.¹⁷⁸¹⁷⁸ {BSI00001742/26} at 9(p).

BS EN ISO 1716: gross heat of combustion

¶5.28 BS EN ISO 1716 is a test method for measuring the gross heat of combustion (calorific value) of a product or material. The 4th edition (2010) version of this standard was current at the time of the Grenfell Tower refurbishment.¹⁷⁹¹⁷⁹ {BSI00001737/19} at 10(q). It is used to determine the gross heat of combustion of materials or products at constant volume in a piece of apparatus called a bomb calorimeter.¹⁸⁰¹⁸⁰ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, pages 139-146; Lane {Day68/64:23}-{Day68/67:10} including video of the test procedure. Essentially, a sample of material of a known mass is ground to a powder and burned in an atmosphere of oxygen. A flaming wire is used to ignite the material and the rise in temperature of the surrounding vessel of water is measured. Three specimens are tested, the result being an average of the values obtained expressed in joules or megajoules. The rise in the temperature of the water can be used to establish the amount of energy in the form of heat produced by the combustion of the test sample.¹⁸¹¹⁸¹ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/143} at 12.4.48. Metal powders are not suitable for testing in this way because they present a risk of explosion.¹⁸²¹⁸² Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, page 143. Again, the test report must make it clear that the results relate only to the particular conditions of the test and should not be the sole criterion for assessing the fire hazard in use.¹⁸³¹⁸³ {BSI00001737}.

BS EN 13823: single burning item test

¶5.29 BS EN 13823 is commonly referred to as the "single burning item test".¹⁸⁴¹⁸⁴ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, page 147; Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/148} at 12.4.65. The full title of the standard is "Reaction to fire tests for building products - Building products excluding floorings exposed to the thermal attack by a single burning item" - see {BSI00000119}.Show full note The 2002 version of this standard was relevant to some of the tests on ACM commissioned by Arconic.¹⁸⁵¹⁸⁵ {BSI00000119}. The 2014 version was current at the time of the Grenfell Tower refurbishment,¹⁸⁶¹⁸⁶ {BSI00000119}. although there are no material differences for present purposes. The single burning item test is known as a "scenario" test¹⁸⁷¹⁸⁷ Professor Torero categorises any tests which represent a scenario which is deemed to be realistic as a scenario test - Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/86} at 10.2.16. because it is intended to establish the performance of materials under conditions representative of a fire within a compartment. It is intended to act as a scaled-down version of the full-scale room corner test – ISO 9705 – and was developed in such a way that the results would be indicative of those that would be obtained in a full-scale test.¹⁸⁸¹⁸⁸ Torero, Phase 2 Report, Adequacy of the Current Testing Regime{JTOR00000006/148} at 12.4.66; Bisby, Phase 2 Report - Regulatory Testing and the Path to Grenfell {LBYP20000001/35} at 166.

¶5.30 The method involves testing a specimen forming a corner.¹⁸⁹¹⁸⁹ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, pages 147-162; {Day68/67:11-74:6} The specimen holder is a metal frame with a calcium silicate backing board. The intention of the test is to represent the end-use application of the product, so the specimen is fixed to the backing board in a manner consistent with the fixing conditions adopted in the end-use application.¹⁹⁰¹⁹⁰ {BSI00000119/15} at 5.2.2(b) and see Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/150} at 12.4.69. The specimen holder contains two wings, one long one (1m in length) and one short one (495mm in length), each 1.5m in height. Apart from the BS 8414 test, it is the largest specimen used in any of the reaction to fire tests, either in Europe or this country.

¶5.32 The test rig contains two burners, a primary burner and an auxiliary burner. The primary burner is located in the corner between the two wings and although it is offset by 40mm, flames given off by the burner can make direct contact with the outside face of the test specimen. It is a gas and sand burner which is calibrated to give a heat output of 30kW and intended to represent a waste paper bin on fire in the corner of a room. The auxiliary burner is located at a distance from the test apparatus; its only purpose is to run for a short period before the primary burner is ignited to provide a baseline average burner heat and smoke output at the start of the test. The contribution of both burners is subtracted at the end of the test to calculate the contribution of the specimen. The specimen is mounted on a trolley which is moved under the test rig. Above the rig is a hood which contains an exhaust system. During the test several measurements are taken in the duct through which the smoke is extracted, including the temperature and density of the smoke. This is a depiction of the test apparatus taken from Dr Lane's presentation:¹⁹⁴¹⁹⁴ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, page 153.

¶Figure 5.4: Depiction of test apparatus for BS EN 13823

¶5.33 The test runs for about 20 minutes during which the rate of heat production from the specimen is determined by means of measurements taken within the extractor hood. This information is expressed in a range of forms that include the average heat release rate, the total heat released, including in the first ten minutes, and the fire growth rate index "FIGRA".¹⁹⁵¹⁹⁵ The FIGRA is intended to represent, in a standardised manner, how fast a fire spreads over the sample. It does not relate to any physical phenomenon beyond the fact that if the fire reaches a large heat release rate in a short period of time it delivers a large FIGRA - see Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/153} at 12.4.88. The propensity for lateral flame spread is measured by a single visual observation of whether sustained flames reach the end of the long wing (1m) at any time during the test. The rate of progression of the flame is otherwise not recorded. The propensity to produce flaming droplets is recorded within the first ten minutes if droplets reach floor level outside the burner zone. Information about smoke production is also produced by means of measurements which include the average smoke production rate, the total smoke production and the smoke growth rate index "SMOGRA" which measures the rate of increase of smoke production.¹⁹⁶¹⁹⁶ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/150} at 12.4.72-12.4.75, 12.4.89.

¶5.34 Data is collected from three separate tests.¹⁹⁷¹⁹⁷ {BSI00000119/18} at 5.4. The results of the test are expressed primarily in a series of graphs which display both burning and smoke production behaviour.¹⁹⁸¹⁹⁸ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, page 156. Numerical values are calculated and recorded for some parameters, including the FIGRA and the total heat release rate ("HRR"). These figures are recorded in the test reports and then averaged over the three tests and expressed as a single figure. The average is relevant for classification to EN 13501, as discussed further below. If a specimen does not perform as expected in a single burning item test, it is possible to test up to two more specimens. If that occurs, the highest and lowest results are discarded and the mean value is calculated on the remaining three values to arrive at an overall average figure.¹⁹⁹¹⁹⁹ {BSI00001738/17-18} at 7.1-7.5.

¶5.35 The standard also identifies a range of information which must be included in the test report, including a statement that the test results are not the sole criterion for assessing the potential fire hazard of the product in use.²⁰⁰²⁰⁰ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, page 157. The rate at which the fire is able to grow due to ignition of the product being tested is considered indicative of the product's propensity to bring a fire from its initial stages to the point of flashover.²⁰¹²⁰¹ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/151} at 12.4.77.

BS EN ISO 11925-2: single-flame source test

¶5.36 BS EN ISO 11925-2 is known as the "single-flame source" test.²⁰²²⁰² {BSI00000125} the full title of the standard is "Ignitability of building products subjected to direct impingement of flame. Part 2 – Single-flame source test." The 2010 version was current at the time of the Grenfell Tower refurbishment. The test is designed to simulate a small flame being applied directly to the surface or to the edge of a material. The test standard makes it clear that the configuration of the test specimen should reflect the end-use application and if the product is installed with unprotected edges, tests should be performed on both covered and unprotected edges.²⁰³²⁰³ {BSI00000125/10-11} at 5.4.4 and 5.5.

¶5.37 The apparatus for the test comprises a Bunsen burner housed within an outer compartment called the combustion chamber.²⁰⁴²⁰⁴ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 2, pages 163-177; Lane {Day68/74:7}-{Day68/79:18} including video of test procedure. The test sample is 250mm long by 90mm wide²⁰⁵²⁰⁵ With a maximum permitted thickness of 60mm. and is suspended from the back wall of the combustion chamber with an aluminium tray placed below containing sheets of paper which can catch any flaming droplets. During the test the Bunsen burner applies a flame directly to the surface or edge of the material at an angle of 45°.²⁰⁶²⁰⁶ When a surface exposure is used the flame is applied 40mm up from the bottom of the specimen on the front face of the specimen. When an edge exposure is used the flame is applied on the bottom edge of the specimen. The flame is applied for either 15 or 30 seconds²⁰⁷²⁰⁷ If applied for 15 seconds, the total test duration is 20 seconds and if applied for 30 seconds, the total test duration is 60 seconds. That allows for a period of time after the flame has been removed to observe what happens when the flame is removed. depending on the classification the test sponsor wishes to obtain. Ignition and any vertical flame spread is observed and recorded, together with the amount of flaming droplets or particles and whether they ignite the paper below. The presence of any flaming on the specimen once the pilot flame is removed is also recorded.

¶5.38 The following picture of a sample about to undergo an ISO 11925-2 test is taken from Professor Torero's report on the adequacy of the current testing regime:²⁰⁸²⁰⁸ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/147}.

¶Figure 5.5: Picture of sample about to undergo ISO 11925-2 test

¶5.39 This is another "scenario" test, the purpose of which is to simulate the initiation of a compartment fire and to establish if a small, localised heat source can ignite a product or material within the compartment and sustain a flame away from the initiating flame.²⁰⁹²⁰⁹ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/146} at 12.4.54.

European reaction to fire test classification system

¶5.40 BS EN 13501-1 is the European standard which contains the European reaction to fire classification system. The 2009 edition was current at the time of the Grenfell Tower refurbishment.²¹⁰²¹⁰ {BSI00001738} BS EN 13501-1:2007+A1 2009 "Fire classification of construction products and building elements. Classification using test data from reaction to fire tests." See also the 2002 version at {BSI00000620}. It contained a reaction to fire classification for all construction products, including products incorporated into building elements. The system was designed to create a single system of classification that would capture a broad range of physical processes.²¹¹²¹¹ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/136} at 12.4.16.

¶5.41 Table 1 describes the European reaction to fire performance classes. There are seven classes, from A1, which is the highest classification, to F, which means that no performance can be determined.²¹²²¹² Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 3, pages 178-190. For each category the standard provides that the classification can be obtained only by undertaking the tests or the extended application process required for that particular product. This is a copy of Table 1:²¹³²¹³ {BSI00001738/40}.

¶Figure 5.6: Table 1 from BS EN 13501-1

¶5.42 As is apparent from the table, in order to achieve European classification A1 data obtained from the following two reaction to fire tests is required: (1) EN ISO 1716, the gross heat of combustion using the bomb calorimeter method and (2) EN ISO 1182, the non-combustibility furnace test method. In order to be classified A1, the material must achieve certain results in both.

¶5.43 There are two combinations of tests that can be used to determine European classification A2. In either combination the single burning item test (BS EN 13823) is required. In addition, the material should be tested either using the bomb calorimeter method (EN ISO 1716) or the non-combustibility furnace test method (EN ISO 1182). For Class A2, the requirements are less stringent than for A1: for example, if using the bomb calorimeter method a gross heat of combustion less than 3MJ/kg is permitted, whereas for Class A1 a gross heat of combustion of less than 2MJ/kg is required.

¶5.44 For European classes B to D, test combinations are required. Those classifications rely on the single burning item test and the single-flame source test only, with the requirements for each of the tests becoming less onerous as one moves down the classes. For example, a Class B material must achieve a FIGRA of less than 120W/s in the single burning item test, whereas a Class C material must achieve a FIGRA of less than 250W/s and a Class D material less than 750W/s. For classification E, only the single-flame source test is required.²¹⁴²¹⁴ There has to be flame spread less than 150mm within 20 seconds and no ignition of the paper below. Class F is appropriate when a product fails to obtain Class E and is applied when a product has no performance criteria.

¶5.45 The following summary of the European classes and the properties required to obtain them is taken from Dr Lane's presentation:²¹⁵²¹⁵ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 3, page 190.

¶Figure 5.7: Summary of the European classes

¶5.46 Each of the classes is also divided by reference to the volume of smoke and flaming droplets produced by the sample, which is designated by the notations s1, s2 and s3 for smoke production and d0, d1 and d2 for flaming droplets. Classification of smoke production is obtained from data taken during the single burning item test only, whereas classification relating to the production of flaming droplets is obtained by reference to observations taken during both the single burning item test and the single-flame source test.²¹⁶²¹⁶ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 3, pages 192-193.

¶5.47 EN 13501-1 provides that the field of application of any classification which has been achieved must be identical to the field of application resulting from the test(s) or from any extended application process.²¹⁷²¹⁷ {BSI00001738/38} at 15 "Field of application of the classification". It also makes it clear that a different classification may apply if different end-use applications are envisaged for the particular product.²¹⁸²¹⁸ Ibid. It also states that while the classification may be valid for products within the same family where the reaction to fire classification can be proven to be unchanged, it may be that the field of application is extended in an extended application report, separate rules for which are given in CEN/TS 15117.

The European classes cited in Approved Document B

¶5.48 The 2013 version of Approved Document B referred to the European classifications at various places in relation to the construction of external walls. In Table A6 dealing with the use and definition of non-combustible materials,²¹⁹²¹⁹ {CLG00000224/131}. any material which is classified as A1 under the European classification system is regarded as non-combustible. In addition, certain materials are regarded as Class A1 without the need for testing and are also regarded as non-combustible.²²⁰²²⁰ As defined in European Commission Decision 2003/424/EC 6 June 2003 e.g. cement, masonry cement.

¶5.49 In Table A7, which addresses the use and definitions of materials of limited combustibility,²²¹²²¹ {CLG00000224/132}. any insulation material used in an external wall construction of the kind referred to in paragraph 12.7 of Approved Document B must either be non-combustible, as specified in table A6, or be classified A2-s3, d2 or better in accordance with BS EN 135011. The reference to "s3, d2" in Table A7 means that no limits are set for the production of smoke or flaming droplets or particles.

¶5.50 In Diagram 40 of Approved Document B the external surfaces of walls of buildings over 18m in height were required to be either national Class 0 or European Class B-s3, d2 or better. That meant that, if one were seeking to adopt the European classification, the external surface had to have achieved certain results in the single burning item and single-flame source test, but that there were no minimum requirements for the production of smoke or flaming droplets or particles.

Large-scale testing to BS 8414 and classification to BR 135

¶5.51 Methods for large-scale fire testing of cladding systems are contained in BS 8414 Parts 1 and 2. The classification method and performance criteria are set out in BR 135.²²²²²² Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 3, page 222. At the time of the Grenfell Tower fire BRE was the UK's only testing laboratory with accreditation to undertake BS 8414 testing. There is no European equivalent of the BS 8414 test.

¶5.52 BS 8414 is a two-part British Standard. Part 1²²³²²³ BS 8414-1:2002 "Fire performance of external cladding systems – Part 1: Test method for non-loadbearing external cladding systems applied to the face of the building." {BSI00000163}. A second edition was published in 2015 - BS 8414-1:2015 {BSI00000253}. contains the method for testing external cladding systems applied to masonry walls; Part 2²²⁴²²⁴ BS 8414-2:2005 "Fire Performance of external cladding systems – Part 2 Test methods for non-loadbearing external cladding systems fixed to and supported by a structural steel frame." {BSI00000097}. A second edition was published in 2015 - BS 8414-2:2015 {BSI00000167}. contains the method for testing external cladding systems fixed to and supported by structural steel frames. The exposure to fire of the system being tested is intended to be representative of an external fire source or a fully developed (post-flashover) fire in a room venting through an opening such as a window and exposing the cladding to the effects of external flames.²²⁵²²⁵ See {BSI00000163/5} and {BSI00000097/5} under "1 Scope".

¶5.53 The system to be tested is formed of two walls, each at least 8m high. One, the main wall, contains the combustion chamber and has a wing wall set at right angles to it at one side. The composition and structure of the walls are determined by the person commissioning the test. The main wall contains an open-faced combustion chamber at ground level measuring 2m by 2m²²⁶²²⁶ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/ expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 3, pages 242-246 and see Lane {Day68/106:25}-{Day68/114:11}, including video of the test procedure. in which the fuel source is located. The fuel source is a timber crib comprising layers of softwood sticks arranged alternately; when ignited, it is designed to produce a total heat output of 4,500 megajoules with a peak heat release rate of 3 megawatts. This fire source is designed to produce flames which are typical of a fully developed building fire impinging on the façade.²²⁷²²⁷ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/100} at 12.2.15 and see BR 135 1st Edition {BRE00001077/5} bottom left column. The main wall of the test specimen must be a minimum of 2.6m wide and 6m in height (when measured from the top of the combustion chamber); the wing wall must be a minimum of 1.5m wide and 8m high. No guidance is given on how the system is to be constructed around the opening to the combustion chamber.²²⁸²²⁸ Torero, Phase 2 Report, Adequacy of the Current Testing Regime {JTOR00000006/100} at 12.2.9. The second editions of BS 8414-1 and 8414-2 provided that "the test specimen shall include all relevant components assembled and installed in accordance with the manufacturer's instructions".²²⁹²²⁹ {BSI00000167/13} under section 6 "Test specimen". The "external cladding system" is defined in 3.3 under Note 1 as including "for example, sheeting rails, fixings, cavities, insulation and membranes, coatings, flashings or joints" see {BSI00000167/7}.

¶5.54 Two types of data are recorded during the test: temperatures and visual observations. To measure temperatures a number of thermocouples²³⁰²³⁰ Thermocouples on the main wall are positioned on the centreline of the wall and then at 500mm and 1000mm either side of this centre line. Thermocouples on the return wall are positioned at 150mm, 600mm and 1050mm from the junction with the main wall. are located on the exterior of the cladding system at 2.5m (level 1) and 5m (level 2) above the test opening. These thermocouples do not make direct contact with the cladding system and are positioned at a distance of 50mm from the surface. Internal thermocouples are also placed within the cladding system in any combustible layers greater than 10mm in thickness. These internal thermocouples are positioned at level 2 only. A photograph and diagram from Dr Lane's presentation showing the location of the different thermocouples appears below.

¶Figure 5.8: Diagram and picture showing the location of thermocouples in BS 8414 test

¶5.55 The visual observations to be taken during the test include any change in flaming conditions or the mechanical behaviour of the cladding system, especially detachment of any part of the system (whether flaming or otherwise) or any other penetrations through fire stops incorporated within it.²³¹²³¹ See Second Edition of BS 8414-2:2015 at 8.4 Test Observations {BSI00000167/14}. A continuous audio-visual record of the full height of the test face is taken throughout the test and for steel-framed systems an audio-visual record is also required of the internal face of the cladding system, so that any burn-through can be assessed.²³²²³² {BSI00000167/12} at 5.9 and Lane, Expert witness presentation - 10 November 2020: https://www. grenfelltowerinquiry.org.uk/hearings/expert-witness-presentation-dr-barbara-lane-10-november-2020, Part 3, page 246.

¶5.56 Temperature and visual recordings are taken from five minutes before ignition until 60 minutes after ignition.²³³²³³ Lane, Expert witness presentation - 10 November 2020: https://www.grenfelltowerinquiry.org.uk/hearings/expert- witness-presentation-dr-barbara-lane-10-november-2020, Part 3, page 247. The crib is allowed to burn for 30 minutes, at which point it is extinguished and the test allowed to run for a further 30 minutes during which any observed flaming is allowed to continue. The test is stopped early if at any point flames extend above the test rig or if there is a risk to the safety of the personnel within the test facility.²³⁴²³⁴ The provision for early termination was contained in BS 8414-2:2005 but not BS 8414-1:2002. It was introduced to the Part 1 test in 2015 - see {BSI00000253/14} at 8.6 Early test termination criteria. The reference to early termination was also introduced into the 3rd Edition of BR 135 in 2013 - see {BRE00005555/27} at A2. A test report should be provided for each test undertaken, even if the test is terminated early; the standard sets out a list of information which should be included in the test report.²³⁵²³⁵ See example for Part 2 {BSI00000167/15} at section 10. The following photographs of a BS 8414 test in operation are taken from Dr Lane's presentation:

¶Figure 5.9: photographs of a BS 8414 test in operation

¶5.57 The second²³⁶²³⁶ {BRE00005554} Colwell, Martin 2003. and third²³⁷²³⁷ {BRE00005555} Colwell, Baker 2013. editions of BR 135 both contained two annexes each providing performance criteria relevant to the BS 8414 tests. The third edition, which is most relevant to the Grenfell Tower fire, contains three main performance criteria: external fire spread, internal fire spread and mechanical performance. In relation to external and internal fire spread the document contains failure criteria only. Failure due to external fire spread is deemed to have occurred if the rise in temperature over the initial ambient temperature of any of the thermocouples at level 2 exceeds 600°C for a period of at least 30 seconds within 15 minutes of the start time, which is defined as the time at which any thermocouple at level 1 equals or exceeds 200°C for a period of 30 seconds.²³⁸²³⁸ {BRE00005555} at Annex A, A2.1 and Annex B, B2.2. Failure due to internal fire spread is deemed to have occurred if the rise in temperature above the initial ambient temperature of any of the internal thermocouples at level 2 exceeds 600°C for a period of at least 30 seconds within 15 minutes of the start time.²³⁹²³⁹ {BRE00005555} at Annex A, A2.3 and {BRE00005555/33} at Annex B, B2.3.,²⁴⁰²⁴⁰ In addition, Annex B for steel-framed systems tested to BS 8414-2 provides that where system burn-through occurs so that flame reaches the internal surface, failure is deemed to have occurred if continuous flaming, defined as a flame with a duration in excess of 60 seconds, is observed on the internal face of the test specimen at or above a height of 0.5m above the combustion chamber within 15 minutes of the start time - see {BRE00005555/33} at Annex B, B2.3.Show full note No failure criteria are set in respect of mechanical performance, but continuing combustion of the system following extinction of the crib is to be included in the test and classification reports, together with details of any collapse, spalling, delamination of the system, the production of flaming debris or pool fires.²⁴¹²⁴¹ {BRE00005555/29} at A2.4 and {BRE00005555}/33} at B2.4. Although there are no criteria for mechanical failure, the nature of the mechanical performance should be considered as part of the overall risk assessment when specifying the system.²⁴²²⁴² {BRE00005555/29} at A2.4 and {BRE00005555/33} at B2.4.

¶5.58 The third edition of BR 135 made it clear that, in order for a system to be classified in accordance with BR 135, it must have been tested for the full duration provided for in BS 8414 without any early termination of the period of exposure to the full fire load.²⁴³²⁴³ {BRE00005555/27} at A2 and {BRE00005555/32} at B2. Parts 1 and 2 of BS 8414 stipulate that records must continue to be made for an additional 30 minutes up to a maximum duration of 60 minutes, unless no part of the system is still burning 30 minutes after ignition, in which case the test should be terminated.²⁴⁴²⁴⁴ Part 1 (2002) {BSI00000163/10} at paragraph 7.4; Part 2 (2005) {BSI00000097/11} at paragraph 8.5. The third edition of BR 135 also provides in terms (unlike the second edition) that the classification applies only to the system as tested and described in the classification report.²⁴⁵²⁴⁵ {BRE00005555/28} at A2, left column and {BRE00005555/32} at B2 right column. Both a BS 8414 test report and a BR 135 classification report are therefore necessary to demonstrate that the system has been fully tested and classified to the necessary standard.