Expert Presentations - Thursday 9th June 2022 (2/2)

9 June 2022 · Professor Luke Bisby (Fire Safety Engineering Expert), Counsel to the Inquiry · 2:12:19
▶ Watch on YouTube Open in interactive viewer

Professor Luke Bisby presents experimental work on cladding fire behavior, demonstrating how ACM PE panels and insulation materials contributed to fire spread at Grenfell Tower

Key moments

Full transcript

00:00:28 yes mr millet yes mr chairman good afternoon good afternoon members of the panel uh this afternoon professor luke bisbee will give an oral presentation based on his experimental work

00:00:39 based on his experimental work before we call him we should give a trigger warning for those who are watching several of the slides in professor bisbee's presentation contain images and videos of fires

00:00:50 videos of fires so it's important to warn anyone watching who would prefer not to view that content so without further ado may i please now call professor bisbee yes thank you very much

00:01:09 good afternoon professor bisbee welcome back uh you're going to make the affirmation i think i am the words are on the screen then would you read them up please

00:01:16 up please i do solemnly sincerely and truly declare and affirm that the evidence i shall give shall be the truth the whole truth and nothing but the truth thank you very much would you sit down make yourself comfortable

00:01:34 yes mr military thank you mr chairman professor bisbee welcome back to the inquiry and thank you very much for attending uh again to answer questions and present your

00:01:46 to answer questions and present your expert opinions um it's right i think that you've produced six reports for this phase two of the inquiry three of which we will deal with next week when you come back to give evidence and three of which i

00:01:58 to give evidence and three of which i think we'll be hearing from you about this afternoon during your presentation on the experimental work that you've conducted um can we please go to w lbywp

00:02:09 1 7 0 2

00:02:13 l b y w p 1 7 0 2

00:02:18 7 0 2 now that is a report entitled work package 1 regimes of burning date of the 15th of march 2020 and if we go please to at the bottom of

00:02:29 and if we go please to at the bottom of the page we can see a signature or middle of your screen is that your signature yes it is uh and i think it was updated on the 1st of june 2020 as you can see if we go to

00:02:41 of june 2020 as you can see if we go to page 17 of that report you can see there section 2.5 statements and if you go to page 18 we can see a signature is that your signature if we

00:02:52 signature is that your signature if we turn the page is that your signature there yes it is um

00:02:57 um these statements um are these your statements as to this report yes sir and that is your signature as you've confirmed have you read this report recently i have and can you confirm that the facts and matters

00:03:09 you confirm that the facts and matters the factual matters set out in the report are true to the best of your knowledge and belief i can and can you confirm that the opinions that you've given in this report are your honestly held professional opinions i can

00:03:20 held professional opinions i can and is it true that you provided your expert opinion in this report to the inquiry in the same way that you would have provided it to an english court that's correct thank you now let's go next please

00:03:31 next please to

00:03:32 to lbywp 2701

00:03:37 lbywp2701

00:03:40 this is your report entitled work package 2 system interactions and there's a signature above the date of the 15th of december 2021 is that your signature yes it is and if we gave these to page 14 of the report you'll see

00:03:52 to page 14 of the report you'll see section 1.5 at the foot of the screen and there are statements there set out and then they continue over the page onto page 15. um are those your statements there as to

00:04:03 um are those your statements there as to this report yes sir and have you read this report recently yes i have uh and uh there's a signature at the bottom i'm assuming that that's yours yes correct yes and um can you confirm that the facts and matters set out in this report

00:04:15 facts and matters set out in this report are true to the best of your knowledge and belief yes i can and that the opinions that you've given in this report are your honestly held professional opinions yes they are and is it true also that you've provided your expert opinion in this report in

00:04:27 your expert opinion in this report in the same manner that you would have done had you been asked to provide it to an english court that's correct yes and lastly then please if we can turn to l b y m t

00:04:36 y m t seven zeros two l b y m t seven zeros two

00:04:43 two this is a report entitled materials testing report 1 dated the 24th of february 2019 updated on the 1st of june 2020 with a signature in the middle is that your signature yes it is uh and i

00:04:55 that your signature yes it is uh and i think that report is um to be read in conjunction with your final phase one experts report is that right yes and if we go to page nine in this report please you'll see section 1.5

00:05:06 you'll see section 1.5 that's also bearing the title statements and you can see the statements there in four paragraphs 48-51 with the signature below it is that your signature yes it is uh and there's your statements yes

00:05:17 is uh and there's your statements yes there and have you read this report recently i have and can you confirm that the facts and matters set out in this report are true to the best of your knowledge and beliefs yes and that the opinions contained in it are your honestly held professional opinions yes

00:05:30 honestly held professional opinions yes and is it true again that you have provided this expert opinion to the inquiry in the same way that you would have done if asked to do so to an english court that's correct yes can you also confirm that none of the opinions

00:05:41 also confirm that none of the opinions and conclusions in your reports that i've shown you have changed since they were produced that's correct and today i think you're going to give us a presentation on these three reports can you confirm now that the facts and

00:05:53 can you confirm now that the facts and opinions that you express in your presentation uh are true so far as the facts are concerned and your honestly held opinions so far as they are matters of opinion that's correct uh thank you

00:06:04 of opinion that's correct uh thank you very much

00:06:05 very much professor we look forward to hearing your presentation thank you very much

00:06:15 when you're ready yep

00:06:18 good afternoon sir members of the panel um thank you very much for inviting me to give evidence today uh and good afternoon mr millet to begin my evidence today i'll i'll deliver a presentation that i've

00:06:29 deliver a presentation that i've prepared based on my phase two instructions from the inquiry i've scripted my presentation so i'll be reading from a script so as to keep myself to time and to avoid me getting a bit too interested

00:06:40 avoid me getting a bit too interested perhaps

00:06:42 perhaps or dwelling for too long on some of the finer technical points it's my expectation that this should take me about an hour and a half to two hours to complete perhaps slightly longer if there are any immediate questions that need to be

00:06:53 immediate questions that need to be dealt with as we go you've already received copies of my phase two reports these can be broadly characterized as my experimental reports which i'll discuss today and my regulatory and testing reports which i

00:07:06 regulatory and testing reports which i understand will deal with next week my experimental reports consist of my materials testing report one my work package one regimes of burning report and my work package to system

00:07:18 report and my work package to system interactions report my regulatory testing report regulatory and testing reports consist of my regulatory testing and the path to grenfell report my bre reconstruction report and my br

00:07:31 my bre reconstruction report and my br 135 desktop assessment report

00:07:35 now the time available for my presentation today doesn't allow me to cover all of the ground contained within these six reports obviously it's it's my intention this afternoon to focus on my experimental reports

00:07:48 to focus on my experimental reports and i'll do this in three parts as outlined here i hope to summarize illustrate and explain in brief what i've done why i did it what i found and what i think it means

00:07:59 and what i think it means i understand that mr millet may then have some questions for me both on this work and on my other reports uh next week

00:08:06 week i hope it goes without saying that i'd welcome questions or discussion at any point during this presentation so please don't be afraid to interrupt me if something does come up that causes confusion

00:08:15 confusion rather than waiting until i've finished

00:08:19 now uh i'll move firstly to part one of my presentation um now as it's been almost four years uh since i last delivered a presentation to the inquiry i'd first like to summarize

00:08:31 the inquiry i'd first like to summarize some important background and context for my work my phase two instructions asked me to prepare reports on the performance and respective contributions of the materials and products which formed parts of the exterior of grenfell tower

00:08:44 parts of the exterior of grenfell tower to the initiation growth and spread of the fire

00:08:47 the fire this was to include a program of experimentation aimed at understanding and quantifying the respective roles of the various materials and products that made up the clouding system at grenfell tower under a range of relevant fire

00:08:58 tower under a range of relevant fire conditions and system geometries i'd like to begin my presentation by quickly summarizing what i understood when i am when i embarked on this work back in the second half of 2018

00:09:10 back in the second half of 2018 at that time my understanding was based on evidence already available from phase one of the inquiry as well as on my own knowledge and professional experience i understood the fire at grenfell tower

00:09:21 i understood the fire at grenfell tower started on the fourth floor in the kitchen of flat 16. i understood that the ventilated rain screen over clouding system that had been installed on grenfell tower incorporated a large number of combustible materials and products

00:09:34 combustible materials and products most notably this included the aluminium composite material or acm rain screen product renibon pe with a polyethylene polymer core foil faced pir and phenolic foam polymer

00:09:47 foil faced pir and phenolic foam polymer insulation products and aluminium faced window infill panels with an extruded polystyrene core i understood that burning of materials and products within the external wall assembly and within grenfell towers

00:09:58 assembly and within grenfell towers architectural crown led to the rapid and uncontrolled fire spread up down and around the building on the night of the fire i understood the fire exited the kitchen of flat 16 and spread to the external

00:10:10 of flat 16 and spread to the external cladding i expressed an opinion at that time that the precise manner in which the fire spread from inside the kitchen out and into the cladding was most likely by a combination of two roots

00:10:24 during my june 2018 presentation i showed you this slide which shows the kitchen window of flat 13

00:10:31 13 after the fire and the geometry of the external wall arrangement here outside flat 13 as far as we can know is essentially the same as that which was present outside the kitchen window of flat 16 at the time of the fire

00:10:44 of flat 16 at the time of the fire i took a horizontal section through the external wall at that location shown here with the dashed line and i zoomed in on the highlighted region here where the windows meet the columns

00:10:56 windows meet the columns on that basis i developed this sketch of the materials products and geometry that exist at this location and i noted that one potential route for fire spread to the cladding was through the side of the window and into the back

00:11:08 the side of the window and into the back of the cladding cavity this could occur due to parts of the internal window surround and external clouding system being penetrated by the fire essentially with the fire entering the cladding around the side of the

00:11:20 the cladding around the side of the window frame

00:11:23 in this case the fire enters the cladding from the rear but then finds a layer of polymer foam insulation facing a renabon pe acm rain screen with a ventilated cavity in between them

00:11:35 ventilated cavity in between them i also developed a sketch through this vertical section of the kitchen window of flat 16 highlighting this region of the external wall

00:11:48 the resulting sectional sketch that i developed on this basis is shown here where the other most likely route of fire spread can be seen essentially like this

00:11:57 this this consists of fire exiting the kitchen window or the extract fan panel which was within the kitchen window with flames and hot gases impinging on the cladding directly above the window

00:12:08 cladding directly above the window again as the fire enters the cladding it finds a layer of polymer foam insulation facing a renovant pe acm rain screen with a ventilated cavity between them

00:12:19 with a ventilated cavity between them at that time i was unable to definitively say which of these was the dominant route for fire spread to the cladding

00:12:25 cladding my opinion remains that some combination of these two fire spread routes caused the initial ignition that led to sustained burning and escalation of the cladding fire

00:12:36 during my phase one work for the inquiry i closely examined the materials and products within the external wall arrangement at grenfell tower i concluded that the available evidence strongly supported a hypothesis that the

00:12:47 strongly supported a hypothesis that the presence of acm panels with a polyethylene core material over large areas of the external surface of the tower was the primary cause of upward downward and horizontal fire spread i observed that the renavan pe

00:12:59 i observed that the renavan pe incorporated a highly combustible polyethylene polymer which melts drips and flows whilst burning at elevated temperature i stated that this polyethylene material should be expected to release large

00:13:11 should be expected to release large amounts of energy during combustion to rapidly lose its mechanical properties on heating and to cause separation of the acm rain screen panels i also considered the evidence

00:13:22 i also considered the evidence concerning the role of the combustible insulation products in terms of their contribution to fire spread on the exterior of the building i stated that the available evidence was inconclusive regarding the extent to which the presence of combustible

00:13:34 which the presence of combustible insulation products within the rain screen cavity contributed to the rate and extent of upward downward and horizontal fire spread i found that pyrolysis and charring of the insulation products had occurred during the fire and that exposed

00:13:47 during the fire and that exposed surfaces of polymer foam insulation were present within the cladding system before the fire without wishing to preempt the remainder of my evidence i'd like to confirm that my views on these matters have not changed

00:13:58 changed at that time i considered it more likely than not that that the presence of the insulation products contributed to some extent although i wasn't able to quantify this contribution whilst i was satisfied that the renovant

00:14:10 whilst i was satisfied that the renovant pe was primarily responsible for the fire spread i had outstanding questions about the roles and contributions of the other products used within the external wall arrangement the polymer foam insulation products in particular

00:14:23 during my phase two work for the inquiry i undertook to better understand and demonstrate the potential contributions of the various products to the rate and extent of fire spread i performed a

00:14:34 extent of fire spread i performed a series of experiments in two work packages which i've referred to as work package 1 regimes of burning and work package 2 system interactions first

00:14:46 first in work package one i performed experiments to demonstrate and understand how these products responded to heating and burn in isolation this was necessary both to underpin my

00:14:57 this was necessary both to underpin my own understanding of the thermal and mechanical properties of the products under a range of representative heating conditions and to ensure that the inquiry is given rigorous scientific data upon which to base its eventual conclusions

00:15:09 conclusions second in work package 2 i performed further experiments aimed at understanding how these products burned as part of a system and how their individual burning behaviors might lead to interactions

00:15:21 behaviors might lead to interactions within a ventilated rain screen cladding system

00:15:25 system all of the work that i'll present today was completed under my supervision at the university of edinburgh i was assisted by a team of colleagues at edinburgh dr rory hatton dr simone

00:15:36 at edinburgh dr rory hatton dr simone santamaria mr mikhail krachevich mr david morrissette and dr angus law i'll mostly use first-person singular language in my presentation today but it's very important for me that i

00:15:47 it's very important for me that i highlight the significant contributions that were made by all of these individuals under instruction from the inquiry much of this work was also performed in close consultation with professor jose torreira

00:15:58 torreira who also witnessed many of the experiments and participated in our analysis of the results

00:16:05 in performing this work i was faced with a choice

00:16:09 a choice should i experiment with products that were taken from grenfell tower after the fire

00:16:15 sorry should i experiment with products that were taken from grenfell tower after the fire or should i use pristine materials and products provided to me by core participants to the inquiry i was concerned that products from the fire scene might have been damaged

00:16:27 fire scene might have been damaged during installation the fire itself or removal and that obtaining a sufficient quantity of undamaged products could present a challenge i was also aware the metropolitan police service were intending to use products

00:16:39 service were intending to use products taken from grenfell tower for the purposes of both standard fire tests and fire reconstructions and that the availability of undamaged products from the tower might become limited

00:16:49 limited i decided with the support of the inquiry to seek pristine proxy product samples directly from core participants i'm grateful to these core participants

00:17:00 i'm grateful to these core participants for providing the requested specimens in a timely fashion to satisfy myself that the products provided to me by the core participants were in fact the same in all material ways as the products

00:17:11 ways as the products that had been installed on greenfield tower i undertook my own due diligence diagnostic testing this was done by comparing against small samples of undamaged reference material of products taken from the tower and

00:17:23 of products taken from the tower and these were provided to the inquiry by the metropolitan police service these diagnostic tests confirmed the similarity of the proxy products that i was using to those present on grenfell tower at the time of the fire

00:17:42 so in the first half of my presentation today i'll describe and discuss the comparatively straightforward work package 1 and its results before moving on to what i consider to be the more interesting and complex

00:17:54 be the more interesting and complex results of work package 2. i began work package 1 with an overall expectation of how each clang product would perform under heating this was based on my knowledge experience and a

00:18:06 based on my knowledge experience and a physical understanding of the reaction to fire behavior of various types of materials and products work package one focused on the behavior of the following four products and i'll

00:18:18 of the following four products and i'll explain why i chose these four products as i go

00:18:22 as i go renabond p e uh acm or aluminium composite material which is shown here in basic terms this product consists of 2.5 millimeter skins of aluminium either

00:18:34 2.5 millimeter skins of aluminium either side of a three millimeter thick core of polyethylene celatex rs 5000 insulation board which is shown here i'll refer to this simply as rs 5000 throughout this presentation

00:18:46 as rs 5000 throughout this presentation this product consists of a of a polyisocyanurate or pir polymer foam core faced on both sides with a thin unperforated aluminium foil which is only a fraction of a millimeter

00:18:58 which is only a fraction of a millimeter thick

00:18:59 thick i've performed experiments on both the 80 millimeter thick rs 5080 product and the 100 millimeter thick rs-5100 product

00:19:11 kingspan cooltherm k15 insulation board shown here i will refer to this simply as k15 as i go forward this product consists of a phenolic polymer foam core faced on both sides

00:19:22 polymer foam core faced on both sides with a thin aluminium foil which again is only a fraction of a millimeter thick the aluminium foil facer in this case has small perforations in it which if you look very closely at this picture

00:19:33 you look very closely at this picture you can see a small dots on the foil's surface

00:19:39 and finally i'll u-glaze window in-fill panels shown here and this product consists of two flat aluminium sheets each 1.5 millimeters thick with an extruded polystyrene or

00:19:50 thick with an extruded polystyrene or xps polymer foam core which is 25 millimeters thick

00:19:55 so these are the four main products that made up the refurbishment cladding system at grenfell tower obviously there were other products involved but these are the four main products that i've considered it's worth noting that pir and phenolic

00:20:07 it's worth noting that pir and phenolic foams are both thermosetting polymers whereas polyethylene and xps foam are both thermoplastic polymers and i've discussed those issues before and i'll demonstrate the key differences between

00:20:18 demonstrate the key differences between these two classes of polymers during my presentation today

00:20:24 i've provided more detailed descriptions of all four of these products in my phase two reports so i'm going to avoid any further description here and i've we've been talking about them for for many years now

00:20:37 in my first presentation to the inquiry and indeed this morning uh professor torreira also covered many a lot of this ground

00:20:44 ground i described some fundamental fire science concepts and terminology including detailed discussions of pyrolysis thermal inertia flammability ignition

00:20:54 ignition heat of combustion heat release rate and flame spread and those are all really centrally important concepts as professor tarero illustrated this morning i'm not going to rehearse those concepts uh here again today

00:21:07 concepts uh here again today but the concepts that i covered during my presentation in 2018 are all immediately relevant to what i will present today

00:21:15 i believe that the most useful and illustrative way uh for me to present the outcomes of my experiments to individuals who are not fire safety specialists and who may not have a deep technical understanding of

00:21:26 have a deep technical understanding of all of the terminology that i've just listed

00:21:29 listed is by walking through some videos of selected experiments and describing my observations as we go and i'll do quite a lot of that today starting in a few minutes

00:21:40 starting in a few minutes however

00:21:41 however before presenting the experimental apparatus and the procedure that forms the core of my work package one experiments i think it's useful to frame the discussion in terms of some of these fundamental reaction to fire properties

00:21:52 fundamental reaction to fire properties or behaviors of the for clouding products of

00:21:56 interest now when we think about the reaction to fire of a material or product we need to consider a range of interrelated questions how easily is it ignited

00:22:09 how easily is it ignited under what conditions will it continue to burn once ignited will it spread flame over its surface and under what conditions how much energy i.e heat is released

00:22:21 how much energy i.e heat is released when it burns under the conditions of interest

00:22:25 interest and at what rate is that energy released answers to any one of these questions in isolation cannot tell the whole story of the fire hazards presented by a material or product

00:22:39 indeed this is why i routinely use the word bernie as a catch-all term describing hazardous with respect to fire

00:22:46 fire none of the other descriptors combustible flammable ignitable etc adequately describe fire hazard on their own

00:22:53 own and this is also why my work package one report and indeed my other experimental reports contain results from a number of experiments conducted at various scales and under a wide range of heating conditions

00:23:06 so first i'd like to consider the heats of combustion of the polymer core materials

00:23:11 materials from my four products of primary interest

00:23:18 as i explained in my 2018 presentation to the inquiry the heat of combustion of a material is a measure of the total energy that can be released per unit mass of the material when burning under optimal

00:23:30 material when burning under optimal conditions for combustion the heat of combustion thus provides an upper limit on the amount of energy that could be released by the burning of a material

00:23:39 material you'll hopefully have noticed that i'm using the word material determinantly in this context rather than product products do not have heats of combustion unless of course a product is made entirely of one material

00:23:51 entirely of one material and so that's why i'm using that terminology only materials have heats of combustion this slide shows a comparison of the heats of combustion that were measured for polymer core materials from each of

00:24:02 for polymer core materials from each of the four products of interest they were measured using a bomb calorimeter and it's clear that the polyethylene core material from the renaband pe the leftmost bar

00:24:13 leftmost bar has the highest heat of combustion and that's comparable to diesel fuel this has been mentioned numerous times throughout the inquiry and that the phenolic foam from the k15 product has the lowest of the products i'm showing

00:24:24 the lowest of the products i'm showing here

00:24:25 here however it's important to recognize that these values are given in a per unit mass of the core material basis rather than a per unit area of the product that incorporates it basis

00:24:37 incorporates it basis if i multiply these heats of combustion by the densities of the respective core materials and then by the thicknesses of core material within each product i get the maximum possible energy contribution per unit area of each

00:24:50 contribution per unit area of each product which is actually a more useful metric i.e the maximum energy that could be released if the product were to burn under optimum conditions and this comparison is shown here

00:25:02 and this comparison is shown here we can see that on this basis the story changes because it's not simply a question of how much energy is in the material per unit mass but also how much mass of material is present within the

00:25:14 mass of material is present within the product by this comparison we can see that the maximum amount of available energy per unit area becomes more comparable between the renabond and the polymer foam insulations and that the

00:25:25 polymer foam insulations and that the aluglaze window infill panel appear to be sorry the ali glaze window infill panels appear to be less significant

00:25:38 but we then need to ask another question how much of each product in terms of square meters of surface area was present on grenfell tower before the fire

00:25:48 fire based on the design drawings for the external wall arrangements i've calculated the total surface area of each product on a typical single story of grenfell tower as shown by this highlighted area

00:26:01 highlighted area we then end up with an approximation of the maximum possible energy contribution from each of the four main cladding products that were actually installed on grenfell tower

00:26:14 here i'm assuming is a worst case scenario that all four products burn entirely and that this and this burning occurs under optimal conditions for combustion which

00:26:25 optimal conditions for combustion which is actually not a fair comparison i should say here we can see that the result of this calculation for a single story of grenfell tower if we assume that only celatex rs 5000 insulation was present i.e there was no

00:26:37 insulation was present i.e there was no kingspan present and it was all the celatex product but we would see an almost identical story if we assume that only kingspan k-15 was present and i've outlined that

00:26:48 k-15 was present and i've outlined that graph in my work package one report in any case this plot shows that the maximum potential energy contribution from the alu glaze window infill panels is small when compared with the renabon

00:27:00 is small when compared with the renabon pe or the polymer foam insulations and for this reason i'll not discuss iuglay's window infill panels much more within this presentation however my work package one report does contain results of experiments on

00:27:12 contain results of experiments on iuglay's window infill panels and i do consider that these are relevant to clouding fire safety issues particularly for other buildings whose exterior walls and window sets may contain similarly

00:27:23 and window sets may contain similarly hazardous products now of course the comparison in this plot is only fair if we assume that all of the energy that is available within each of these products is actually released during a

00:27:34 products is actually released during a fire

00:27:35 fire and this is extremely unlikely in a real fire both because the combustion environment in a real fire is not optimal

00:27:42 optimal and because of the differing physical responses of the polymer core materials that are involved the thermoplastic core materials in renamon's pe acm and in aluglaze window infill panels

00:27:54 infill panels will melt and drip whereas the thermosetting polymer core materials of rs 5000 and k15 will char they do not melt and drip i'll show you how these differences affect their respective products burning

00:28:05 affect their respective products burning behaviors later on in this presentation in any case the maximum potential energy contributions on their own don't provide any information regarding the rate at

00:28:16 any information regarding the rate at which energy would be released after a product is ignited or indeed the total amount of energy that would actually be released in a real fire nor do the preceding comparisons account for the heat transfer processes that

00:28:27 for the heat transfer processes that lead to ignition the material geometry or the physical changes that may occur all of which will also affect a product's burning behavior to illustrate and understand these other factors we have to perform other types

00:28:39 factors we have to perform other types of experiments

00:28:45 for my work package one experiments we designed and built a bespoke experimental apparatus this was designed to investigate the mechanical behaviors of the products when subjected to a range of heating

00:28:56 when subjected to a range of heating conditions their characteristic burning behaviors the impacts of sample scale and localized heating the impacts of fixing and support conditions and to establish the heat

00:29:07 conditions and to establish the heat flux causing ignition of the various products of interest when tested under specific condition sorry when tested under specific conditions of heating and fixing

00:29:17 fixing i'll now give a basic description of my work package 1 apparatus this was used to heat samples essentially of the four products of interest under carefully controlled and

00:29:28 interest under carefully controlled and repeatable conditions and to study their thermal and mechanical responses

00:29:37 now this photo shows the heater that we used in our experiments it consists of an array of radiant panels that are fueled by a precisely controlled flow of propane gas

00:29:49 precisely controlled flow of propane gas mixed with air the radiant panel array was mounted onto a motorized actuator which allowed the distance between the panels and the samples to be varied as i'll show you in a moment

00:30:01 so it moved in that direction this allowed us to control the severity of heating which we call the heat flux as professor terrero explained this morning and which is measured in units of kilowatts per meter squared so units

00:30:13 of kilowatts per meter squared so units of energy per time per unit area to give you a sense of scale this rig that i'm showing here is about one meter tall

00:30:25 experimental specimens were mounted onto a specimen onto a specimen holder as shown here and were held above a tray which was used to catch any falling debris

00:30:38 used to catch any falling debris both the specimen holder and the debris tray were placed on balances essentially digital scales and these were placed there to measure the masses lost or gained by the

00:30:49 the masses lost or gained by the products throughout the experiment or mass is lost by the product or mass gained by the debris tray the experimental arrangement also included other features that were aimed

00:31:00 included other features that were aimed at improving consistency and repeatability of our experiments including a pilot flame a heat shield and a movable heat flux gauge these are all shown here and are described within my work package one report

00:31:14 the experimental specimens were square flat samples of their respective products of interest with exposed edges of core materials and these samples were either 100 millimeters square or 500

00:31:25 either 100 millimeters square or 500 millimeter square renault pe samples were tested in three different configurations namely

00:31:32 namely a rear fixing condition shown here with the sample supported from its rear aluminium skin thus allowing for downward mobility of the front face of the sample on heating

00:31:43 the front face of the sample on heating and hopefully become clear what i mean when i show you some of these experiments the front view in this slide on the left shows the heated face of the sample so the the face on the left in

00:31:54 sample so the the face on the left in that left left most photo is showing you the face that was heated during the experiment this photo also shows that we installed thermocouples on and within the samples to measure the

00:32:06 on and within the samples to measure the temperatures of the aluminium phasers and the polyethylene core material and that supporting rods were welded to the to the samples aluminium skins as shown here in the right hand image

00:32:20 shown here in the right hand image we also used a front fixing condition shown here with the sample supported from its front aluminium facer thus allowing for downward mobility of the rear face of the sample on heating

00:32:38 and finally a through fixing condition was also used with the samples drilled through and supported from both of their aluminium skins the front face and the rear face

00:32:49 skins the front face and the rear face as shown here thus preventing downward mobility of either of the aluminium skins and also restricting to some extent their mechanical separation on heating these three fixing conditions were

00:33:00 these three fixing conditions were chosen so as to shed light on some of the observed differences in the performance of acm products when exposed to fire in either riveted or cassette configurations that we've heard something about during this inquiry

00:33:11 something about during this inquiry and i've discussed that in some detail in my work package one report for instance to better understand the physical reasons why ac mpe products tend to perform better for instance in

00:33:22 tend to perform better for instance in single burning item tests when they're tested in a riveted rather than a cassette configuration

00:33:31 insulation product specimens in our experiments for work package one were held in place as shown here using two steel rods that were inserted into the rear face of the samples and in some cases again as shown here

00:33:44 and in some cases again as shown here the foil facer on the exposed face of the sample was removed prior to the experiment we did experiments both with and without foil this was done because during phase one i had identified locations within the

00:33:55 had identified locations within the cladding system where cut edges of insulation products were left exposed within cavities in the cladding system at grenfell tower

00:34:05 a large number of measurements were recorded throughout these experiments these included multiple videos framed from different vantage points the masses of the specimen and any

00:34:16 the masses of the specimen and any debris falling into the tray beneath the specimen as i've mentioned the heat release rate or hrr as we refer to it

00:34:23 to it which was measured using a technique called oxygen consumption calorimetry which i won't go into today and temperatures as i've already mentioned which were recorded using thermocouples

00:34:35 thermocouples in a few minutes i'll show you some videos of typical experiments performed using my work package one experimental apparatus

00:34:43 apparatus as i've described it here

00:34:47 and i hope that uh when i show you these videos this will become a bit clearer i can understand it's probably difficult to get your head around what i'm talking about at the moment so now let's turn to some of my main experimental results

00:35:00 before a product can burn with a flame it needs to ignite ignition is simply the initiation of sustained flaming weather ignition of a solid material occurs depends both on the severity of heating which which as

00:35:12 the severity of heating which which as i've mentioned we call the heat flux and the duration of heating in my 2018 presentation to the inquiry i also explained that flame spread can be thought of if you like as a series of

00:35:24 thought of if you like as a series of ignitions

00:35:26 ignitions thus in attempting to quantify the relative contributions of the products making up the external wall arrangement at grenfell tower it's important to understand and quantify their respective ignition characteristics

00:35:38 i performed a range of experiments to characterize the ignition behavior of the four products of interest some of the key outcomes of which are summarized in this plot on the vertical axis

00:35:50 on the vertical axis on this slide i'm plotting the heat flux again the severity or intensity of heating in kilowatts per meter squared now to put those values in context for you from 0 to 100 kilowatts per square

00:36:03 you from 0 to 100 kilowatts per square meter

00:36:04 meter at the low end exposure to a heat flux of 5 to 10 kilowatts per square meter is likely to result in skin burns in less than about 30 seconds at the high end a heat flux exposure of

00:36:15 at the high end a heat flux exposure of 75 to 100 kilowatts per square meter is in the range of the severities of heating that would be expected over the first one to two meters above the hearth in a bs 8414 large large-scale cladding test as professor torreira mentioned

00:36:27 test as professor torreira mentioned this morning it's noteworthy that all four of the products listed in this plot are metal-faced polymers but that the thermal and physical properties of both the metal phasers and

00:36:38 properties of both the metal phasers and the polymer cores differ substantially and you can see i'm plotting results both with and without the aluminium facers

00:36:50 if i highlight the renabond results we can see that the renault pe product will ignite at comparatively low heat fluxes this means that in the configuration that we tested it would be expected to

00:37:02 that we tested it would be expected to ignite in most any cladding fire scenario

00:37:06 scenario we can also see that the removal of the aluminium phaser moving from the the red data

00:37:13 data to the blue without

00:37:16 uh makes the renabon pe only slightly more susceptible to ignition so not much of an effect the same is true for the alu glaze window infill panels albeit us with a slightly higher critical heat

00:37:28 with a slightly higher critical heat flux

00:37:29 flux both with and without the aluminium facers in place however so sorry there's the alu glaze however the ignition behavior of the polymer foam insulation products celatex

00:37:41 polymer foam insulation products celatex rs-5000 and kingspan k15 is drastically reduced

00:37:47 reduced by removal of the foil facers again red with the foil facers and blue without these products are both much more difficult to ignite when their foil faces are present as we should expect

00:38:00 faces are present as we should expect the foil faces reflect radiation they alter the convective heat transfer conditions at the surface of the samples and to some extent they mitigate ignition and burning of the flammable pyrolysis products that are generated

00:38:11 pyrolysis products that are generated during heating professor i i'm going to interrupt you for a second i think the stenographer is finding it quite struggling to keep up with you

00:38:20 with you um so if you could slow down a little bit i think she'd i will certainly try do you mind i'm sorry yes apologies

00:38:34 okay yes so the beneficial effect of the foil facers is less pronounced for the kingspan uh cool therm k15 insulation

00:38:45 i consider it likely that this is at least partly due to the presence of the perforations in the k-15 foil facers that are not present for the rs-5000 i should also say here that i consider

00:38:57 i should also say here that i consider that the presence of perforations in the foil phasers of a polymer foam insulation should self-evidently be expected to influence by which i mean worsen

00:39:08 worsen its reaction to fire behavior i should also note that this plot fails to tell the whole story because whilst it tells us uh what's a variety of what severity of heating will lead to piloted

00:39:19 severity of heating will lead to piloted ignition eventually it doesn't tell us how rapidly this ignition will occur at a given heat flux

00:39:28 in general terms all other factors being equal it takes considerably longer to ignite the polyethylene core of an acm product than it does to ignite the polymer foam insulation without a foil facer

00:39:41 insulation without a foil facer we're talking minutes for the acm versus seconds for the polymer foam insulation when we take the foil away and of course all of this only says that a product will ignite and start to burn

00:39:53 a product will ignite and start to burn doesn't tell us whether it will continue to burn nor indeed whether it will burn quickly or slowly to more fully understand the reaction to fire behavior of these products we need to study their responses when subjected

00:40:05 to study their responses when subjected to intensities and distributions of radiant heating that are representative of those to which they might be exposed during a cladding fire so now i'd like to show you

00:40:16 so now i'd like to show you uh some of our experiments so that you can see some of these interesting behaviors for yourselves it's a shame that the chairman members of the panel and council to the inquiry weren't able

00:40:28 weren't able mostly thanks to the pandemic to come to edinburgh and witness any of these tests as i really do believe there's no substitute for observing such things in person

00:40:41 for renabon pe i knew that samples would be comparatively difficult to ignite as i've just shown but that under high heat fluxes they would be certain to ignite eventually

00:40:53 would be certain to ignite eventually i knew that once ignited the polyethylene would melt drip and flow most likely while it's burning i therefore knew that molten polyethylene could form pool fires

00:41:04 polyethylene could form pool fires i knew that mechanical failure or separation of the aluminium faces at comparatively low temperatures when not mechanically restrained in place for instance by rivets would govern the fire hazard

00:41:16 would govern the fire hazard and my experiments illustrated all of these expected behaviors so there were no real surprises there and i found that our well-controlled experiments were extremely repeatable which was also comforting

00:41:27 which was also comforting i'll show you a brief video of a typical experiment on a sample of renabond pe so that you can see how these experiments unfolded i've made videos of all of my experiments available via the inquiry

00:41:39 experiments available via the inquiry and i've also produced a number of composite comparative videos that are available alongside my work package one report and indeed are referenced within that report

00:41:51 mr chairman the the video that i'm about to show uh probably warrants a trigger warning um given its stark depiction of the fire hazards associated with these products so perhaps i'll just pause for a moment

00:42:09 so this video actually two videos will show an experiment on a 100 by 100 millimeter square sample of renabond pe acm supplied by arconic

00:42:21 acm supplied by arconic this was the rainscreen product installed on grenfell tower of course the sample in this case is supported from its rear face and has thermocouples welded to its front and rear faces

00:42:33 welded to its front and rear faces you can see the sample the radiant panel array

00:42:36 array the pilot flame just above the sample and the debris tray below the sample in these two images and i'll continue to show both the close-up view and the more distance view

00:42:47 close-up view and the more distance view as i let this play forward

00:42:51 in this particular experiment the sample was exposed to a heat flux of 50 kilowatts per square meter and you will have noticed in my report that we used heat fluxes of 25 50 and 75 kilowatts per square meter

00:43:04 25 50 and 75 kilowatts per square meter in our experiments this is essentially an experiment chosen at random to show you this heat flux 50 kilowatts per square meter is broadly representative of that which would be experienced about 1.5 to

00:43:16 which would be experienced about 1.5 to 2 meters above the hearth in a typical bs841 for a large scale cladding test

00:43:24 i'll explain to you first what we will see and then i'll show it to you and explain it as we go so we will see once i let this run we will see the radiant panel array which is to the right in these images

00:43:37 which is to the right in these images will move toward the sample our timer will start at the moment that the radiant panel array comes to rest in front of the sample after about 30 seconds we'll see pyrolysis gases visible above the

00:43:49 pyrolysis gases visible above the specimen

00:43:51 specimen between 30 seconds and 1 minute and 18 seconds the rate of production of pyrolysis gases will increase at one minute and 18 seconds ignition will occur at the top of the sample

00:44:02 will occur at the top of the sample just a few seconds later the front aluminium phaser will completely fall away from the sample and land whilst burning in the debris tray the rear aluminium facer which is

00:44:13 the rear aluminium facer which is restrained on the on the holding rods will remain in place and it will stay mounted on the sample holder as will the majority of the polyethylene core and that will remain attached to the rear

00:44:24 that will remain attached to the rear face

00:44:27 flaming will then uh continue on the front aluminium facer now in the debris tray and discontinuous flames will become visible over the front surface of the specimen flaming will then become well

00:44:39 flaming will then become well established over the front surface of the exposed polyethylene core that remains on the sample and burning droplets of polyethylene core material will begin to fall from the base of the specimen and land in the debris tray

00:44:52 specimen and land in the debris tray the rate of burning droplet production will increase until there's a near continuous stream of burning polyethylene core material falling into the debris tray and after about two minutes the dripping will cease as all the polyethylene will

00:45:05 will cease as all the polyethylene will be gone and burning will continue only in the debris tray as a pool fire eventually all flaming will cease and the experiment will be halted so i'll now let it play

00:45:16 so i'll now let it play and it should pause periodically so that i can point some things out

00:45:56 and i should say that the the computer that we're running this presentation off is struggling to keep up with the high res videos so occasionally the digital timer does something a little bit weird but you'll you'll get the impression i hope

00:46:09 so you can see now there's some smoke pyrolysis products coming off the sample in the right hand view

00:46:18 and the rate of production of pyrolysis gases appears to increase as we step forward in time as we would expect

00:46:44 just keep that playing at one minute and 18 seconds we'll see ignition occur at the top of the sample

00:46:53 there it is and a few seconds later at one minute and 22 seconds the front facer of aluminium falls away from the sample and lands in the debris tray and you can see

00:47:04 lands in the debris tray and you can see that it's now flaming in the debris tray the back face of the aluminium remains in place held on the sample holder and the majority of the polyethylene core material remains attached to that face

00:47:15 material remains attached to that face of the sample

00:47:21 flaming then continues on the front aluminium facer now in the debris tray and discontinuous flames become visible over the front surface of the specimen and you can see here we've switched to a low exposure view in the left hand side

00:47:33 low exposure view in the left hand side so that we can see things as the flames grow

00:47:36 grow by one minute and 30 seconds flaming is well established over the front surface of the exposed polyethylene and the first burning droplets of polyethylene begin to fall from the base of the

00:47:47 begin to fall from the base of the specimen and into the debris tray the rate of droplet production increases quite rapidly and a continuous stream of burning polyethylene falls into the debris tray

00:47:59 polyethylene falls into the debris tray and increases the size of the pool fire

00:48:13 it's probably also worth having a look at the left-hand image just the way the polyethylene is melting and flowing and dripping which you can sort of make out through the flames

00:48:28 at two minutes and nine seconds dripping ceases and flaming on the specimen then also ceases shortly thereafter between two minutes and nine seconds and four minutes and 48 seconds you'll see

00:48:41 four minutes and 48 seconds you'll see i'm running at 10 times speed now the length of the flames from the polyethylene which is still burning in the debris tray gradually decreases until extinction occurs

00:48:53 until extinction occurs the experiment is then terminated

00:49:03 i hope that we can agree that that looked like quite a lot of fire from actually a quite a small piece of material uh it's perhaps a bit tricky to visualize uh a 100 by 100 millimeter square sample of renibon pe i've brought

00:49:15 square sample of renibon pe i've brought a sample with me to put it into perspective this was the piece of renai bond that we tested in that video and that's about the size of a beer mat or a drinks coaster to put in

00:49:26 or a drinks coaster to put in perspective for you so it's a piece a piece like that so not very much material at all if you run the numbers which i've done you would find that a typical single story column cassette from grenfell

00:49:38 story column cassette from grenfell tower so a single panel of renabond pe taken from the tower from one of the column sections would contain about 270 of these in a single column cassette

00:49:52 now let's take a look at some of the resulting data from this experiment i'm going to focus on the data from this specific experiment and then provide you some some general conclusions and data afterwards

00:50:04 conclusions and data afterwards i mentioned previously that during these experiments we measured the heat release rate

00:50:09 rate due to burning i don't want to intimidate anyone with over technical language or data but i do think it's important for transparency if nothing else for me to show you the kinds of data that i've considered

00:50:20 data that i've considered in performing my analyses and drawing my conclusions

00:50:25 so this slide shows the measured heat release rate that was recorded during the experiment that i've just shown you you can see that we recorded negligible heat release up until the point of

00:50:36 heat release up until the point of ignition

00:50:38 ignition of the renault pe at which point the heat release rate increased rapidly to a peak value before then decreasing again as the sample ran out of polyethylene fuel and you can also see how quickly

00:50:49 fuel and you can also see how quickly that happened if we calculate the area underneath this heat release rate curve we can obtain an estimate of the total amount of energy that was liberated as that sample burned

00:51:01 that was liberated as that sample burned similar plots were developed and compared for all of the work package one experiments that we performed during our experiments we also as i've mentioned continuously monitored the

00:51:12 mentioned continuously monitored the amount of mass lost from the experimental sample and the amount of mass gained by the debris tray that was situated below the sample these data are shown

00:51:23 these data are shown here

00:51:24 here mass

00:51:25 mass lost on the left hand side and mass gained on the right hand side and by comparing the mass of the sample with the visual observation observations we observed a sudden decrease

00:51:36 we observed a sudden decrease in the sample mass at one minute and 22 seconds and this occurred when the front plate of aluminium fell from the specimen

00:51:43 specimen the recorded decrease of sample mass corresponds with the expected mass of the aluminium facer plate there's a corresponding increase in the mass of the debris tray at the same moment that you can see

00:51:54 moment that you can see and there's a gradual sample mass loss that follows this loss of the aluminium plate and that's associated with both burning and dripping of the polyethylene core material

00:52:05 core material during this period the debris tray gains mass due to dripping however all of the mass lost from the sample is not gained by the debris tray and clearly this is because some of the polyethylene has been consumed by burning before it

00:52:16 been consumed by burning before it reaches the debris tray similar plots again were developed and compared for all work package one experiments now i don't have time to rehearse

00:52:27 now i don't have time to rehearse all of our results from all of our experiments many dozens of experiments so i'll simply summarize what i found in the next few moments

00:52:38 of the four clouding products considered in work package one renaband pe has the largest potential to contribute to fire growth and spread it accounted for as much as 59 of the total potentially available

00:52:50 of the total potentially available energy on any single floor of grenfell tower and as much as 87 percent of the total potentially available energy within the architectural crown

00:53:02 of the four principal clouding products renamon p e is comparatively slow to ignite however once ignited it releases more energy more quickly than any of the other products

00:53:13 products in my work package one report i conclude that renovant pe was the largest contributor to energy release in the cladding at grenfell tower i found that the time to ignition of

00:53:24 i found that the time to ignition of renaband pe is not strongly influenced by the manner in which the panels are fixed

00:53:29 fixed i.e riveted or routed in a cassette format

00:53:33 format however the subsequent rate of burning is likely to be influenced by the fixings by which i mean a riveted versus a cassette configuration with riveted systems likely to perform better in

00:53:44 systems likely to perform better in general

00:53:46 general i also found that the balance between burning of the polyethylene core material in situ i.e at the location of the sample and its mobilization downward by dripping whilst burning

00:53:57 dripping whilst burning is significantly influenced by the method of forming and fixing

00:54:03 i've concluded that in most cases the majority of the polyethylene core material is likely to have mobilized downward during heating this is likely to have been accompanied by flaming of the downwardly mobile material and the

00:54:15 the downwardly mobile material and the formation of pool fires where it was able to collect on horizontal surfaces both within and outside the rain screen cladding system this corroborates my observations regarding fire spread mechanisms from

00:54:27 regarding fire spread mechanisms from the fire at grenfell tower that i made in my phase one evidence to the inquiry

00:54:33 i've i've shown that renaband's pe when locally heated directly on an exposed cut edge which i've not shown you the results of fails to ignite or spread flame at heat fluxes up to 75 kilowatts

00:54:45 flame at heat fluxes up to 75 kilowatts per square meter this was the most severe heating exposure that i used in my experiments and suggests that edge ignition under very localized heating is unlikely under most circumstances

00:54:58 one needs to more uniformly heat renaband pe over a more substantial area in order for fire growth and spread to occur

00:55:09 i performed similar experiments on the foil faced thermal setting polymer foam insulation products used within the external wall arrangement at grenfell tower namely the celetex rs 5000 and the

00:55:21 tower namely the celetex rs 5000 and the kingspan cool therm k15 i knew that these insulation products without their foil facers in place would be relatively easy to ignite due to their low thermal inertia

00:55:33 their low thermal inertia i knew that both products are charring thermoset polymer foams and i therefore expected that they would ignite quickly and experience a rapid rise in heat release up to a peak value but that they

00:55:44 release up to a peak value but that they would then experience reductions in heat release rate and may even extinguish or go out in the absence of a significant external heat flux because of their charring nature i knew that the foil faces on both

00:55:56 i knew that the foil faces on both products would drastically alter the heat transfer at their surfaces would interfere with ignition and would significantly complicate and improve their initial burning behavior

00:56:07 their initial burning behavior i'll now show you some videos of my experiments on the insulation products i'll show you two on celatex rs 5000 at a heat flux of 50 kilowatts per square

00:56:18 a heat flux of 50 kilowatts per square meter as for the acm a few moments ago both with and without a foil face are in place

00:56:24 place and two on kingspan cool therm k15 also at a heat flux of 50 kilowatts per square meter and again both with and without the foil facer in place these videos clearly demonstrate the criticality of the foil facer as regards

00:56:36 criticality of the foil facer as regards ignition and burning for these products i'd just like to reissue my my trigger warning as we'll again be watching videos of products that burned at grenfell tower

00:56:48 so this slide shows two experiments two different experiments on celatex rs 5000 this is the rs 5080 product both with and without the foil facer in

00:57:00 both with and without the foil facer in place obviously without the phaser on the left again the samples are being heated from the right hand side so the radiant panels are shining onto the the rightmost face on these samples

00:57:13 rightmost face on these samples before i press play i'd like to warn you to try and notice the speed with which the rs 5000 insulation ignites when the foil facer isn't present if you blink you might miss it

00:57:24 if you blink you might miss it after ignition we'll see that there's no obvious melting or dripping from the sample which is why i'm not showing the debris tray in these experiments nothing falls down

00:57:34 and i'll explain now what we're going to see

00:57:38 see we will observe the formation of a layer of char over the surface of the sample on the left-hand side this will result in a gradual reduction of the flow of flammable pyrolysis gases through the

00:57:49 flammable pyrolysis gases through the front face of the sample and this will result in a reduction of the severity of burning and eventually the flame will become intermittent on the surface and may even go out briefly we'll see the char will then crack

00:58:01 we'll see the char will then crack allowing flammable gases from deeper within the sample to exit the front face and flaming will then persist persist for several minutes before again extinguishing

00:58:12 extinguishing on the right with the foil facer present will not see any ignition at all for the full duration of the experiment i'd like to remind you that the foil facer on this product is only about 30 microns

00:58:23 this product is only about 30 microns thick that's 31 thousandths of a millimeter thick and yet its influence on the reaction to fire behavior is profound we will see the formation of bubbles

00:58:35 we will see the formation of bubbles behind the foil facer this is an indication of the generation of pyrolysis products being generated from the pi pir foam behind the foil and those pyrolysis products are unable to

00:58:46 those pyrolysis products are unable to escape thus preventing their mixing with oxygen and their burning

00:58:53 so i'll now let it run again stopping periodically

00:59:05 so you can see ignition at time zero in this case which is the moment that the radiant panels actually come to a rest in front of the sample so they ignite essentially instantaneously

00:59:18 after ignition there's no obvious melting or dripping again because these are thermal setting polymers we see continuous flaming in the formation of a char layer over the surface of the left hand sample

00:59:31 surface of the left hand sample this results in a gradual reduction of the severity of burning as we watch it play forward on the right in a moment we should see the formation of bubbles in the foil phaser

00:59:49 so there you can just make out the formation of a bubble in the foil facer

00:59:55 on the left we'll see the flaming reduce and become intermittent

01:00:26 so there we see the intermittent flaming and this continues until the char cracks at which point we still we see the flaming increase a little bit more so there you can start to see the formation of a crack

01:00:38 formation of a crack in the front face of the char in the left-hand video

01:00:46 so there's cracks formed in the charred surface

01:00:49 surface and flaming then persists for several minutes

01:00:51 minutes now i'm going to let the video run at 40 times the normal speed because these are 20 minute long tests and we can see that the sample which is

01:01:02 and we can see that the sample which is protected by a foil phaser doesn't ignite at all during the course of an otherwise identical experiment whereas the unprotected sample is gradually consumed as it runs forward

01:01:26 so now i'll show you the same comparison for kingspan cooltherm k15 insulation and you'll notice some differences in behavior

01:01:33 behavior again a trigger warning i'll show

01:01:39 i'll show two videos of experiments on kingspan cooltherm k15 insulation at the same heat flux 50 kilowatts per square meter both with and without the foil facer in place

01:01:50 place in this case on the left we will see that ignition doesn't occur as rapidly as was the case for the celotex rs5000 taking about 16 seconds for the sample without a foil facer

01:02:01 for the sample without a foil facer after ignition we'll see that there is no obvious melting or dripping from the sample as was the case for the rs-5000 and again

01:02:09 and again this is because phenolic foam like pir foam is a charring thermoset polymer we'll observe the formation of a char layer over the surface of the sample which will again result in a reduction

01:02:20 which will again result in a reduction in the flow of flammable gases through the front face of the sample this will again result in a reduction of the severity and severity of burning and eventually in this case extinguishment

01:02:32 eventually in this case extinguishment of the flame completely however

01:02:35 however we'll then observe that the char will begin to glow and that the sample will gradually be consumed by a combination of smoldering combustion and physical loss of material from the foam's surface

01:02:46 from the foam's surface on the right for the sample which is protected by a foil facer which in this case is only about 25 microns thick so 25 one thousandths of a millimeter again we won't see any

01:02:57 millimeter again we won't see any ignition

01:02:59 ignition we'll also notice that we won't see any bubbles forming in the foil facer for the k-15 product this is obviously because the foil facer is perforated and the flammable gases can easily get out of the sample so to speak so now

01:03:11 out of the sample so to speak so now i'll let it run

01:03:39 so there we have ignition at about 16 seconds beyond ignition there is no obvious melting or dripping from the sample as i've mentioned we can see the formation

01:03:50 i've mentioned we can see the formation of a char layer over the surface of the sample and the reduction of flaming quite rapidly in this case we get a reduction in flaming the char then begins to glow and the

01:04:01 the char then begins to glow and the sample is gradually consumed by smoldering and a physical loss of material from the foam's surface and once this has stopped and i've pointed to the glowing i'll let the

01:04:12 pointed to the glowing i'll let the experiment run at 40 times its normal speed so you remember that the rs 5000 did not glow like this

01:04:32 the k-15 sample which is protected by a foil facer on the right does not ignite the sample without the foil is gradually consumed and continues to smolder throughout this experiment and this behavior was fairly typical for

01:04:44 and this behavior was fairly typical for the kingspan product again i don't have time to rehearse all of the results from my experiments on both of these materials here so in a moment i'll just summarize what

01:04:55 so in a moment i'll just summarize what i've observed and concluded

01:05:00 you can see at the end of this experiment there's actually not that much material left on the left hand side and of course we could quantify the amounts of material consumed because these samples are on scales and we know how much mass is being lost throughout

01:05:12 how much mass is being lost throughout these experiments

01:05:19 so now to some results the pir foam core of rs 5000 and the phenolic foam core of k-15 have lower heats of combustion than the polyethylene core from renaband pe

01:05:32 polyethylene core from renaband pe of the four principle clouding products at grenfell tower assuming that the vast majority of the insulation installed was rs 5000 rather than k15 rs 5000 would make the second highest

01:05:44 rs 5000 would make the second highest contribution of available energy per floor

01:05:48 floor i found that with their aluminium faces in place rs 5000 and k15 are both comparatively difficult to ignite however i found that when the foil facer was present on both products kingspan

01:05:59 was present on both products kingspan k15 was easier to ignite than the celatex rs5000 and as i've already noted i consider that this is due to the presence of the perforations in the phaser

01:06:09 phaser i found that without their foil facers both insulation products are very easily ignited this is a consequence of their very low thermal inertia and indeed is precisely why these

01:06:20 and indeed is precisely why these products are otherwise so very attractive for providing thermal insulation in buildings once ignited sustained exposure to heat fluxes between 25 and 75 kilowatts per

01:06:33 fluxes between 25 and 75 kilowatts per meter squared resulted in gradual thermal decomposition of both types of polymer foam core

01:06:39 foam core and this was accompanied by charring and comparatively low heat release rates if you compare to the acm pe also the time scales over which this heat release occurred for the polymer foams

01:06:50 foams is longer than those of external fire spread when a renemont pe acm product is used

01:06:57 used by which i mean the acm would have burned away long before we would have observed these mass losses once ignited both insulation products have much lower heat release rates per

01:07:08 have much lower heat release rates per unit area than renabons pe once ignited neither of these products melts or drips although the kingspan k-15 does have a tendency to both lose section and smolder under sustained

01:07:19 section and smolder under sustained heating

01:07:20 heating it appears that flat samples of either of these products tested in isolation do not spread flame horizontally in the absence of an applied external heat flux neither of these products are therefore

01:07:32 neither of these products are therefore considered to have been primary or decisive contributors to heat release during the grenfell tower fire particularly within the time scales relevant to vertical fire spread up

01:07:43 relevant to vertical fire spread up i'll talk more on this last point in presenting my work package 2 experiments later this afternoon which is really focused on trying to really unpick that question

01:07:54 so that concludes the summary comments i wanted to make regarding my work package one experiments uh there is however a lot more information contained within my work package one report and obviously i'd encourage all

01:08:06 and obviously i'd encourage all interested parties to have a look at this

01:08:08 this next i'll move on to discuss my work package two experiments however looking at the clock i wonder if it might be a good time for a break i think you probably wouldn't before we start your next section so we'll take

01:08:20 start your next section so we'll take the break now slightly earlier than we usually do we'll resume please at 25 past

01:08:26 past three

01:08:28 three and uh in common with all the other witnesses i have to ask you not to discuss your evidence while you're out of the room yes thank you 25 past three thank you

01:08:43 [Music] thank you mr miller 25 past three please

01:26:20 would you ask professor bisbee to come back in please

01:26:37 all right professor visby well when you're ready uh we're ready for you thank you thank you

01:26:50 thank you mr chairman so before the break i discussed how the clouding products of primary interest responded to heating conditions that are representative of clouding fire scenarios

01:27:02 scenarios the heating scenarios that i used in my work package one experiments are most relevant to situations where you already have a big fire for instance a fire plume generated by a fully developed fire within a building

01:27:14 developed fire within a building compartment venting out of a window i illustrated some of the key burning behaviors of the main clouding products that were present at grenfell tower i also intimated that the mechanical responses of the products could also

01:27:26 responses of the products could also affect their burning behavior even when burning in isolation however those previous experiments the work package one experiments by themselves don't help us to fully

01:27:37 themselves don't help us to fully understand how these products might have interacted within the external wall arrangement at grenfell tower notwithstanding a host of complexities sorry notwithstanding a host of complexities including the presence of

01:27:49 complexities including the presence of various other materials and products cavity barriers etc as convincingly demonstrated by professor terrero this morning

01:27:57 morning the ventilated rain screen cladding system at grenfell tower was in essence a single outer surface and two surfaces facing one another across a ventilated cavity

01:28:08 cavity again as shown here in the highlighted area

01:28:12 area the development of fires in configurations like this is complicated because of the complex combustion environment within a ventilated cavity and because there are interactions between the surfaces on either side

01:28:25 between the surfaces on either side there is thermal feedback between the surfaces if one surface gets hot then this will heat the opposing surface and so on

01:28:32 so on there is also flow within the cavity and the question of how the energy released by burning is distributed within the cavity is heat lost out of the top of the cavity

01:28:43 cavity is heat retained and fed back to materials and products within the cavity thus promoting fire spread my word package 2 experiments were focused on this type of ventilated cavity configuration of products

01:28:56 cavity configuration of products again

01:28:57 again the time available to me will not allow me to cover all of the material contained within my work package 2 report

01:29:04 report so in the interest of time i'll be blunt about the fundamental questions that i again working with my team at the university of edinburgh and again in close consultation with professor terrero set out to answer in word

01:29:16 terrero set out to answer in word package two

01:29:20 did the rena bond pe acm burn so vigorously simply because of the presence of a cavity at grenfell tower or

01:29:29 or did the renabons pe acm burn so vigorously because it was very well insulated by the other products at grenfell tower or did the renault pe acm burn so

01:29:40 did the renault pe acm burn so vigorously because of additional burning of the combustible celatex or kingspan insulation products at greenfield tower to try to provide answers to these questions i needed to design another

01:29:52 questions i needed to design another bespoke experimental apparatus

01:29:56 the apparatus had to deliver consistent and repeatable results and it had to allow systematic exploration of the relative contributions of the different products the complexity and variety of

01:30:07 products the complexity and variety of real cladding geometry at grenfell tower would make systematic study of all possible permutations that existed on the building impractical my objective was not to make a

01:30:18 my objective was not to make a reconstruction of grenfell tower or to rate the products or systems but to study the physics that might be relevant to fire spread at the tower i wanted to understand rather

01:30:29 the tower i wanted to understand rather than to rate or reconstruct and it's very important that i make that distinction we decided therefore to create an experimental apparatus that featured two parallel surfaces with a ventilated

01:30:41 parallel surfaces with a ventilated cavity in between them parallel plates if you will we selected a geometry that we felt was relevant similar to the horizontal panel joint between rain screen cassettes of

01:30:52 joint between rain screen cassettes of renault pe on the columns at grenfell tower however again i must stress this is not a reconstruction on one side i placed a panel of renibon's pe acm rain screen and on the

01:31:05 renibon's pe acm rain screen and on the other side i placed a panel typically of insulation

01:31:10 i used a propane gas burner as an ignition source this was essentially a metal pipe with small holes in it the propane supply to the burner was carefully controlled using a mass flow

01:31:21 carefully controlled using a mass flow controller and this meant that i could control precisely and maintain precisely the size of the flame both within and between experiments i placed this propane burner on a slider

01:31:33 i placed this propane burner on a slider that allowed me to move it into and out of the rain screen joint as i've shown it in this diagram i recognized as i've already said that melting mobilization and pooling of the

01:31:45 melting mobilization and pooling of the polyethylene from the acm rain screen would have a significant impact on fire development i therefore knew that controlling the consistency of pooling of the polyethylene would be important in

01:31:56 polyethylene would be important in achieving the experimental repeatability that was essential if i wanted to isolate and study the influence of other experimental parameters

01:32:07 i placed a small drip tray shown here on the ledge under the burner this would allow me to collect dripping polyethylene in a consistent way between each of the experiments and to ensure

01:32:18 each of the experiments and to ensure experimental repeatability because i was interested in these experiments in escalation of the fire upwards the lower part of the rain screen was fabricated in solid steel

01:32:29 screen was fabricated in solid steel that's the steel shelf shown here rather than in acm and i constructed a series of supporting frames to hold this arrangement in place with open sides so that i could see or

01:32:41 with open sides so that i could see or look into the cavity as the experiments progressed so from the side the overall setup looks like this and again i'll show you with some photos in a moment that should make it a bit more clear

01:32:53 it a bit more clear and from the front which is perhaps even less clear as a drawing it looked like this you can see that the panel of acm in these experiments was about half a meter wide 500 millimeters by about one meter

01:33:06 wide 500 millimeters by about one meter tall about half the size of me if i were to stand facing you the apparatus was assembled as follows so i'll show you how it's built up first we have two load cells these are

01:33:19 first we have two load cells these are very precise digital scales we have aluminium frames that sit on top of the load cells one to hold the acm rain screen and one to hold the opposing face of the cavity

01:33:31 face of the cavity which was typically but not always as we'll see a sheet of polymer foam insulation

01:33:38 two more load cells that sit above the initial two load cells so that i could weigh both the samples and any falling debris independently of one another as the experiments progressed

01:33:50 the experiments progressed metal trays to catch any debris again one for the acm and another for the insulation or whatever was on the opposing face of the cavity

01:34:03 the components of the specific experiment being performed in this case i'm showing a renabond pe acm rain screen and an opposing face consisting of foil faced kingspan

01:34:14 consisting of foil faced kingspan cooltherm k15 insulation and you can see there's a cavity between the two of about 100 millimeters in these experiments

01:34:24 a velocity probe to measure the air flow velocity at the midpoint of the cavity

01:34:31 and a variable number of thermocouples to measure temperatures at various locations on the acm and within the insulation here you can also see the propane line

01:34:42 here you can also see the propane line burner on the left mounted on a chassis that sits upon a slider so that i could move it into and out of the joint below the acm and you'll see how this all plays out in a moment

01:34:53 how this all plays out in a moment this photo shows the line burner in a standby position and this photo shows the line burner in position within the rain screen joint i'll just go backwards and forwards a couple of times so you can see

01:35:05 couple of times so you can see how it moves

01:35:11 overall the apparatus looked something like this

01:35:16 like this and here you can see that we also used a large number of cameras to capture simultaneous videos from various vantage points during the experiments and i'll show you some of these videos in a moment

01:35:26 moment before i do that however i just want to show you uh how we cut and formed the renaband pe panels for our experiments because this is actually very important

01:35:38 this slide shows the rear face of a renabons pe acm panel mounted on its aluminium supporting frame before an experiment was conducted you can see that we have cut the edges of

01:35:49 can see that we have cut the edges of the panel and left the black polyethylene core exposed along the panel edges you can also see that we have routed the surface of the panel that faces into the cavity so you're

01:36:00 that faces into the cavity so you're looking at the surface of the acm that would face into the cavity looking at the opposing face and we've done this so as to fold it to create the horizontal ledge that you can see at its base

01:36:13 can see at its base we've also routed vertical lines down the sides of the acm panels so as to create free floating if you will plates of aluminium which are attached to the sample only via the acm's

01:36:24 to the sample only via the acm's polyethylene core otherwise the bolts that you can see down the side of the sample would have held that inside face in place and this is akin to the condition that would exist on the inside face of an acm

01:36:37 would exist on the inside face of an acm range screen cassette which is obviously why we've done it such as those used at grenfell tower

01:36:44 rather than explain every experiment in detail again i'll show you just one example experiment and i'll then focus the rest of my presentation on what i've found to the extent that i have time to

01:36:56 found to the extent that i have time to cover it and what i've concluded again i've made videos of all of my experiments available via the inquiry and have also produced a number of composite comparative videos that are

01:37:07 composite comparative videos that are available

01:37:08 available within

01:37:09 within and referenced within my work package to report

01:37:14 report this the example experiment i'll show you in detail which was selected essentially at random was on renaband's pe acm and kingspan cooltherm k15 insulation with the foil facers intact

01:37:26 insulation with the foil facers intact this is experiment number 21 in my work package 2 report

01:37:33 before showing you the video i'd like to walk you through the heat release rate data that we recorded during this experiment this helps to place key moments

01:37:43 moments that i'll highlight within the video within the context of the fire sizes and the fire hazards that resulted this graph shows the heat release rate hr

01:37:55 sorry that we measured from the beginning of experiment 21 starting at time 0 essentially on the left to the end of the experiment after 20 minutes and we ran all of these experiments for 20 minutes

01:38:07 20 minutes unless of course they ended much sooner than that or unless we had to stop for health and safety reasons the hr on the vertical axis is effectively telling us the size of the fire in terms of the amount of energy

01:38:18 fire in terms of the amount of energy it's releasing at any given moment in time so the higher that line the more energetic the fire is

01:38:25 at time zero you will see that the line burner is introduced into the clotting joint within the experimental rig and i'll show you a video of this in a moment you'll then see that the polyethylene

01:38:37 you'll then see that the polyethylene core from the acm begins to melt and to drip into the drip tray beneath the acm panel joint after four minutes you'll see that the line burner is removed

01:38:49 line burner is removed but that there is a small pool fire of burning polyethylene within the drip tray

01:38:54 tray you'll also notice that we leave the line burner burning it's a bit distracting in the left most view that you'll see there's a line burner there throughout but it's pulled away from the sample in that condition

01:39:06 sample in that condition the fire will then steadily grow during the course of about six minutes and will then observe a jet of flames from the top of the renault pe range screen from about 10 minutes into this particular experiment

01:39:18 experiment shortly after this we'll observe if we look very closely separation of the inside aluminium skin of the renabon to pe resulting in a very rapid and violent escalation of burning

01:39:31 violent escalation of burning and this is what i refer to as escalation to full involvement of the acm in my work package to report it then increases uh to a peak heat

01:39:42 it then increases uh to a peak heat release rate value the polyethylene from the renabon to pe gradually burns away over the course of a few minutes and the foil faced kingspan cooltherm k15 continues to smolder in flame in this

01:39:54 continues to smolder in flame in this case until the experiment is terminated after 20 minutes again

01:39:58 again a trigger warning is warranted here this video i personally find particularly shocking

01:40:14 okay so this image shows three different views of our experimental arrangement camera a is from the front you're basically looking at the outside

01:40:26 you're basically looking at the outside face of the renault bond pe this would be equivalent to standing outside the building and looking at its external surface you can see the propane line burner in the foreground it's in its standby

01:40:38 the foreground it's in its standby position here it has not yet been inserted into the rain screen panel joint in the test assembly

01:40:45 assembly camera b

01:40:46 camera b which is a bit dark admittedly in this view shows the apparatus from the side you can see the cut edge of the the pink k15 insulation at the right hand side of

01:40:57 k15 insulation at the right hand side of the camera b view and camera c also from the side but in this case it's a close-up of the location where the burner is inserted into the panel joint and again you can

01:41:08 into the panel joint and again you can see the insulation on the right hand side

01:41:11 side you can see that the foil facer if you look closely is still there in this case the experiments commence when the line burner moves into the acm range screen joint the burner is then held in place

01:41:23 joint the burner is then held in place as i've mentioned for four minutes before being removed and allowing the fire to progress or not under its own steam

01:41:30 steam i'll now talk through the video of this particular experiment and highlight as i've been doing previously for the other videos the various points of interest the experiment begins by moving the burner into the joint in the acm

01:41:49 you can see that when the burner is moved into the joint the flame is drawn into the cavity and begins to interact with the acm

01:42:02 again i'm showing this at 10 times speed it takes around 1 minute and 30 seconds of heating before the polyethylene starts to drip out of the acm and into the tray below you can view that particularly in the leftmost view

01:42:15 leftmost view here

01:42:16 here for about 90 seconds the rate of dripping increases now it's quite hard to see in this next video

01:42:23 video but at 3 minutes and 12 seconds there's a free-floating aluminium plate on the upper side of the aluminium ledge at the base of the acm and it drops into

01:42:35 at the base of the acm and it drops into the drip tray it's going to be in this view over here if you remember back to the photo i showed of the routing of the aluminium composite material the the aluminium

01:42:47 composite material the the aluminium plate along the top of that shelf is essentially free floating on a layer of polyethylene and when that polyethylene melts the plate drops off we've intentionally angled that shelf down at five degrees

01:42:59 down at five degrees because we found that we needed to do that in order to ensure consistency amongst the tests so we tried to make it so that that plate would consistently fall off if you will which gives you a sense of the complexity of the response

01:43:10 sense of the complexity of the response of these systems

01:43:19 so just there the plate dropped off

01:43:23 up to about four minutes a pool fire of burning polyethylene continues to grow within the drip tray the flames extend to about three quarters of the cavity height

01:43:34 quarters of the cavity height uh by about four minutes and remember this entire rig is about a meter high that sheet of acmpe at four minutes the line burner is removed

01:43:48 and after four minutes the flame height within the cavity continues to increase gradually as more molten pe mobilizes and exits the acm panel

01:43:59 and exits the acm panel at about 5 minutes and 19 seconds we'll see that the flames become visible above the experimental apparatus and you'll see this particularly in the leftmost view

01:44:11 and you remember now that the line burner has been removed so this is just burning polyethylene

01:44:19 so you can see

01:44:24 in the leftmost view flames above the rig between

01:44:28 between 6 minutes and 30 seconds and 8 minutes and 50 seconds the inside vertical aluminium face of the acm begins to separate from the polyethylene core

01:44:39 separate from the polyethylene core material and you'll start to see that over here in this rightmost view that line there is showing the routing line along the acm pe

01:44:51 flames become visible on the inside of the acm and the flames overall will now extend well above the apparatus at nine minutes and nine seconds if you watch the highlighted area at the top

01:45:05 watch the highlighted area at the top of

01:45:06 of this slide that i'll show you in a moment

01:45:09 moment you'll see that flaming becomes visible on the surface of the foil faced k-15 insulation so if you watch that highlighted circle near the end of this clip you'll see some flaming of the insulation

01:45:31 it's a bit clearer in a moment the flaming then continues at this location albeit only briefly

01:45:44 now i've previously noted the potential significance of the perforations in the foil facer for the k-15 product in particular and this allows the flammable pyrolysis products to exit the insulation directly

01:45:55 products to exit the insulation directly and enter the cavity whilst the facer is still in place at 10 minutes flames can be observed in the top leftmost picture it's easier to see once i let it run again but you can

01:46:06 see once i let it run again but you can observe that flames are now exiting the top of the acm panel itself and that'll be shown at the top left

01:46:13 left this will persist until about 10 minutes and 32 seconds when it appears that the entire inside face of the aluminium sheet facing into the cavity separates from the acm thus

01:46:24 the cavity separates from the acm thus fully exposing the remaining polyethylene core material and after this point we observe this violent escalation to full involvement of the acm

01:46:33 acm and i'm going to simply let this portion of the video play

01:47:12 and again apologies the the graphics card i think is having a hard time keeping up with these videos

01:47:42 even having witnessed this escalation in person

01:47:45 person now dozens of times i still find it pretty shocking indeed the comparatively small size of our experiments in general terms was dictated by the fact that these were the largest experiments that we felt we

01:47:57 largest experiments that we felt we could safely perform within our purpose-built fire labs these were some of the most shocking experiments i have ever witnessed after about 15 years of working in these labs and i feel at this point that i should

01:48:08 and i feel at this point that i should remind you that this is the experimental arrangement burning without any external stimulus the line burner has long since been

01:48:16 been removed this continues until about 13 minutes and 30 seconds and after this point we will be looking uh essentially directly at the k15 insulation through the renabond pe or where the renavans pe

01:48:29 the renabond pe or where the renavans pe used to be

01:48:48 the last visually obvious flaming from the polyethylene from the acm occurs at around 18 minutes when i let this run the kingspan k15 continues to smolder and flame until the experiment is

01:48:59 and flame until the experiment is terminated after about 20 minutes

01:49:45 so just to return to the heat release curve that i showed before we viewed the video this slide just recaps those observations insertion of the line burner

01:49:57 insertion of the line burner at zero minutes first strips of pe at about a minute and a half aluminium plate on the top of the acm pe shelf falls into the drip tray before the line burner is removed

01:50:09 before the line burner is removed the line burner is then removed after four minutes the fire then grows gradually over a few minutes there's some apparent partial separation near the base of the acm

01:50:20 the base of the acm as that fire is growing we then see visible flaming within the acm i.e between the two aluminium skins of the acm near its base

01:50:31 aluminium skins of the acm near its base we see loss of insulation foil facer just prior to escalation we then see a jet of flames that appears to be coming out of the top of the acm itself from between the two aluminium plates

01:50:43 between the two aluminium plates separation of the inside aluminium skin very rapid and violent escalation up to the heat release rate and then a slow decay over a period of a few minutes until the pe stops burning

01:50:54 few minutes until the pe stops burning and we finish the test after 20 minutes

01:51:01 now as i said i don't have time to go through all of our experiments that just gives you a flavor of the kind of things we observed and the way in which we tried to go through each experiment very carefully and to pick out

01:51:14 carefully and to pick out why we had observed what we had observed to understand the physical underpinning mechanisms for what we were observing in our data which is really important and so that going through the heat relation heat release rates curve looking at the

01:51:26 heat release rates curve looking at the changes in behavior going back to the videos and trying to articulate in physical terms or in or in heat transfer terms

01:51:34 terms why what we were observing was happening and in the end we undertook 37 experiments on basically this overall general arrangement varying parameters of interest so as to

01:51:46 varying parameters of interest so as to try to answer the three questions that i posed near the beginning of this presentation this section of the presentation so i'll now take you through my findings and for transparency i

01:51:57 my findings and for transparency i should say that the order in which i'm going to present my findings is not the order in which we performed the experiments i'm presenting the results in the order that i believe will be the easiest to explain

01:52:08 explain and hopefully to understand so just to remind you my key questions were

01:52:15 were essentially was it the cavity i.e was it the fact that we had the geometry of a cavity was it the insulation i.e the fact that the opposing face of the cavity was

01:52:27 the opposing face of the cavity was insulating keeping energy within the cavity or was it the combustibility of the insulation uh the fact that the opposing face of the cavity could burn thus

01:52:39 face of the cavity could burn thus contributing more energy to the system to better understand the role of the cavity specifically first we did what seemed logical and we performed an experiment without any

01:52:51 performed an experiment without any cavity

01:52:54 in this case an experiment only with the rena bonds pe panel configured and ignited identically to the other experiments that involved cavities so everything identical to what i've just shown you but with only a

01:53:06 i've just shown you but with only a sheet of acmpe no cavity no insulation

01:53:13 and i'm just going to show this for you in the interest of time at 10 times speed so you can see the burner gets inserted after about a minute and a half you see dripping

01:53:24 dripping we see the aluminium plate fall off of the little shelf on the acmpe we will see the line burner removed at four minutes

01:53:35 four minutes and we will see

01:53:39 that eventually this experiment goes out so we do not observe escalation to full involvement of the acm in this case

01:53:59 and eventually what you would see is that that small polyethylene pool fire gradually just burns itself out you'll have to perhaps in the interest of time maybe take my word for it

01:54:10 so this experiment with no cavity showed that if no insulation products were present to form the geometry of a cavity then escalation did not occur

01:54:22 however i was uncertain was this because the geometry of the cavity was absent or was it because the product from which i had formed a cavity for example

01:54:34 i had formed a cavity for example uh k15 insulation board in my experiment 21 that i showed you was highly insulating or because it was combustible to answer this question we created a non-insulating cavity

01:54:47 non-insulating cavity using a ren using renault pe on one side and a continuously water cooled steel plate on the other essentially a steel plate with a cascade of water running down its rear surface

01:55:01 the steel plate stayed at low temperature throughout the experiment and actually represented a heat sink that would remove energy from the system during the experiment so it creates the

01:55:13 during the experiment so it creates the geometry of the ventilated rain screen cladding system but without insulating it at all in fact taking energy away from it

01:55:21 from it again we found that escalation didn't occur

01:55:25 occur this meant that the presence of the cavity alone was not sufficient to cause escalation using our specific experimental configuration and procedure i consider this to be a really interesting result the geometry of the

01:55:37 interesting result the geometry of the cavity and the resulting flow conditions within the void by themselves did not lead to escalation of burning of the renabon pe to induce an escalation

01:55:48 to induce an escalation i needed the insulation product to either insulate the acm thereby retaining energy within the system or to burn thereby contributing energy to the system

01:55:59 to the system i'll first discuss what happens when i used a non-combustible but insulating product

01:56:05 product note that the product that i used which was a mineral wool product is deemed to be non-combustible from a regulatory perspective i.e it is a euroclass a1 product it does not however

01:56:18 euroclass a1 product it does not however actually have a heat of combustion of zero

01:56:21 zero which would make it actually non-combustible and i think that's an important distinction to make at this point

01:56:30 when this non-combustible insulation product was used i found that escalation occurred that is the renaband pe became fully involved in the fire

01:56:41 involved in the fire i consider this to be a very interesting result

01:56:44 result it suggests that the insulation product did not need to burn in order for an escalation to occur it was sufficient that the insulation product allowed heat to be retained

01:56:55 product allowed heat to be retained within the system having established that escalation could occur in the absence of a combustible insulation product the next question is therefore if the insulation product is

01:57:06 therefore if the insulation product is in fact combustible does it make matters worse by which i mean

01:57:11 mean faster or more energetic and so

01:57:15 and so and if so by how much so i undertook additional experiments with combustible insulation i removed the foil facers from celatex rs-5000 in an attempt to maximize the

01:57:27 rs-5000 in an attempt to maximize the possibility of an energy contribution from the pir foam to the fire

01:57:34 and in this experiment i also observed escalation from analyzing the data i concluded that this occurred slightly earlier than for the equivalent case with non-combustible

01:57:45 the equivalent case with non-combustible mineral wool insulation this may be because of an energy contribution to the system from the combustion of pyrolysis products coming from the insulation itself the insulation did ignite

01:57:58 the insulation did ignite did ignite rapidly and did spread flame over its surface in this configuration it may also be because the thermal inertia of the pir foam core from rs 5000 is considerably less than that of

01:58:11 5000 is considerably less than that of the mineral wool insulation i had used i.e it's a better insulator probably both factors are relevant i then undertook similar experiments

01:58:22 i then undertook similar experiments this time with kingspan's k15 phenolic foam insulation product again i removed the foil facing from the insulation

01:58:32 and again i found that the results were very similar to the case of pir with no foil phaser escalation occurred and again it occurred more quickly than for the

01:58:43 occurred more quickly than for the non-combustible insulation case i observed that the k-15 continued to smolder for some time after escalation and in some cases continued flaming for quite some time the celatex pir foam did not display

01:58:55 the celatex pir foam did not display this behavior

01:58:58 the fact that these polymer foam insulating products were combustible thus appeared to be of relatively little importance compared to the fact that all of these insulation products whether

01:59:09 of these insulation products whether combustible or not were very effective at retaining heat within the cladding system the insulating nature of these products rather than necessarily their combustibility

01:59:20 combustibility resulted in more rapid and widespread heating of the renavan pe more rapid mobilization of a larger volume of polyethylene core material more rapid separation of the inside

01:59:31 more rapid separation of the inside aluminium skin over a larger area and therefore more rapid escalation to full involvement of the renabon pe in the fire

01:59:39 fire the fact that the polymer foam insulation products could burn appeared to be of secondary importance as compared to their insulative properties

01:59:50 at this point i'd like to return again to the original questions that i posed at the beginning of this presentation did the renault pe acm burn so vigorously because of the presence of a cavity at grenfell tower

02:00:02 cavity at grenfell tower based on these experiments i would suggest that the answer to this question is no

02:00:07 is no the cavity by itself did not cause the renavan pe to burn so vigorously did the renault pe burn so vigorously because it was very well insulated by the other products at grenfell tower

02:00:19 by the other products at grenfell tower based on these experiments i would suggest that the answer to this question is yes to answer the last question did the renavant pe acm burn so vigorously

02:00:30 renavant pe acm burn so vigorously because of the additional burning of the insulation products at grenfell tower i wanted to perform some additional experiments using the combustible insulation products i also wanted to better understand the role of

02:00:42 wanted to better understand the role of the aluminium foil facers given that one would hope that these would typically actually be present in practice in one of these systems

02:00:51 now if you recall in the experiments that i've discussed thus far i chose to remove the foil faces from the insulation products of course we know that while there were exposed edges of

02:01:02 that while there were exposed edges of insulation within the cladding system at grenfell tower most of the insulation products whether rs 5000 or k15 had foil facers in place in the case of the kingspan product

02:01:14 in the case of the kingspan product again the foil facer is perforated i therefore repeated the above experiments but in each case i retained the foil

02:01:22 the foil sorry i therefore repeated the above experiments but in each case i retained the foil on the front face of the insulation including having foil on the front face of the combustible mineral

02:01:33 front face of the combustible mineral wool

02:01:34 wool insulation for the mineral wool insulation i found that the presence of a foil facer caused escalation to happen more rapidly when compared with the no foil case

02:01:46 when compared with the no foil case possible reasons for this which are quite complex are discussed and they're slightly counter-intuitive it has to be said they're discussed within my work package 2 report but i don't want to go into it here

02:01:58 for both the rs 5000 and the k15 however i found that the presence of the foil facers caused escalation to happen less rapidly when compared with the no foil cases this is probably unsurprising

02:02:11 cases this is probably unsurprising given that the foil slows the production and restricts the flow of flammable pyrolysis products from the polymer firms themselves i analyzed the heat released by all of these insulation products from the start

02:02:23 these insulation products from the start of each

02:02:24 of each experiment until the time at which the escalation of the fire occurred and i found that the contribution to the total heat released from the combustible insulations with foil facers up to the

02:02:36 insulations with foil facers up to the point of escalation was measurable but minor

02:02:39 minor in in the cases of combustible insulations without foil facers i found that the contribution to heat release up to the critical time i.e the point of escalation may be comparable to that of

02:02:50 escalation may be comparable to that of the acm itself i consider this to be a useful summary slide in some respects that gives an overview of the differences between the various

02:03:01 of the differences between the various configurations that i've investigated in my phase 2 work package 2 experiments

02:03:08 this plot shows the total amount of energy released during different experiments with so each bar represents a different experiment the total amount of energy released during different experiments on

02:03:20 during different experiments on different combinations of products with each experiment performed twice which is why you have two of each bar we ran every test in duplicate because we wanted to check the repeatability of our

02:03:31 wanted to check the repeatability of our results

02:03:33 results you can see that all cases that involve insulation of the cavity are broadly similar whether the insulation was combustible or not and whether foil facers were present or not

02:03:45 whether foil facers were present or not so for the rs 5080 the 5100 the k15 and the mineral wool with and without foil you end up with a total heat release that is

02:03:56 with a total heat release that is broadly similar

02:04:00 the highlighted bars here show the total amount of energy released during experiments involving renabons pe with celatex rs-508-o the 80 millimeter thick

02:04:11 celatex rs-508-o the 80 millimeter thick variant of rs-5000 you can see there are two experiments that involve foil facers those are the dark bars the two dark bars and two experiments that have uh sorry without

02:04:23 experiments that have uh sorry without foil facers and those are the lighter bars

02:04:27 bars despite the earlier escalation to full involvement of the renault pe when the foil faces were removed that i just mentioned

02:04:34 mentioned these data suggest in this plot very little difference in the total amount of energy released up until local burnout of the cladding system i.e during the 20 minutes of the experiments that we

02:04:45 minutes of the experiments that we performed

02:04:47 performed the same is true in broad terms for the celetex rs rs5100 product the 100 millimeter thick celotex product and for the kingspan cooltherm k15

02:04:59 and for the kingspan cooltherm k15 although in this case there does appear to be slightly more heat release in general

02:05:04 general this is likely a result of the ongoing smoldering and flaming that i commented on when we looked at that experiment a few minutes ago

02:05:14 and much the same result for the mineral wool insulation although in this case there is slightly less total heat release so you can see those bars are all slightly lower than we get for the polymer foam

02:05:26 than we get for the polymer foam insulations probably because this insulation contains very little fuel to release in the first place you can also see that whilst the presence of the foil facer might

02:05:37 presence of the foil facer might slightly change the time it takes for escalation to occur it does not drastically influence the total amount of energy that is eventually released for any of the insulation products

02:05:49 insulation products this is yet another indication of the dominance of the rena bond pe acm to this fire

02:05:57 finally in cases where there was no escalation i.e where the cavity was not insulated at all there was or there was no cavity at all the total heat released was very small

02:06:09 the total heat released was very small albeit there does appear to be a small influence from the presence of the cavity the experiments that had cavities if you watch the videos you'll see that the the pool fire is larger and burns

02:06:20 the the pool fire is larger and burns for longer in those cases hence a slightly higher heat release

02:06:29 now this brings me to the end of my main comments on my work package 2 experiments there were other experiments and other questions but i'll not rehearse these today as they're documented in great

02:06:41 today as they're documented in great detail in my work package to report instead i'll simply emphasize my conclusions

02:06:49 in all of my evidence to the inquiry to date i have consistently identified renabon pe as the primary cause of rapid fire growth and spread during the grenfell tower fire

02:07:00 grenfell tower fire my experiments have allowed me to quantify the relative contributions of the main cladding products to the rate and extent of fire growth and spread my experiments have highlighted the immense complexity

02:07:12 immense complexity of the heat transfer environment within a ventilated rain screen cladding system even for the relatively small and simple cladding system arrangements that i've tested

02:07:23 tested this suggests to me that an intuitive understanding of the factors that govern fire initiation growth and spread in such systems can very easily lead to incorrect assumptions as professor terrero

02:07:35 assumptions as professor terrero intimated as well this morning it remains my view that if the products and materials present in the external wall assembly of grenfell tower the renabon pe was primarily responsible for the rate and extensive fire spread

02:07:48 the rate and extensive fire spread with respect to the role of the insulation products i have found that it was primarily the thermal properties of the insulation products which caused the renaband pe to burn so vigorously

02:07:59 burn so vigorously the fact that these products are exceptionally good insulators made matters much worse for the combustible rain screen that was used i found that the combustibility of the

02:08:10 i found that the combustibility of the insulation played an obvious role only when large surfaces were unprotected by foil facers and only in so far as it accelerated escalation to full involvement of the renaband pe

02:08:22 renaband pe thus the combustibility of foil-faced insulation foams at grenfell tower appears to have played only a minor role in causing the polyethylene to burn so vigorously

02:08:33 vigorously whilst i've not discussed the results in my presentation today my experiments sorry my experiments have also highlighted the importance of the mechanical fixing details of the acm rain screen panels

02:08:46 details of the acm rain screen panels and their manner of routing folding and riveted on the fire growth and spread in ventilated rain screen systems incorporating these cladding products finally

02:08:57 finally it's noteworthy sorry finally it's noteworthy that i've not considered any smoke production issues as part of my investigations at all

02:09:05 all questions around toxicity and smoke production are relevant but are being addressed by other experts to the inquiry

02:09:16 given the context of this work and the evidence heard by the inquiry thus far i'd like to make a few general closing remarks if i could first

02:09:24 first it's clear that renaband's pe acm presents extreme fire hazards it is a very very burny material indeed all acmpe products present

02:09:37 indeed all acmpe products present extreme fire hazards there are few construction products intended for any use on the outside of a building that i have ever seen burn with such intensity every time we ran an experiment that escalated to full

02:09:48 experiment that escalated to full involvement of the acm i was surprised and alarmed once you've seen how these products behave under realistic fire exposures if you know anything about fire safety and anything about the testing methods

02:10:00 and anything about the testing methods that were being used to classify reaction to fire performance of cladding products in england you must surely know that these products are uniquely hazardous what you as a regulator a designer or an

02:10:11 what you as a regulator a designer or an approver then do with this knowledge is up to you

02:10:15 up to you second

02:10:16 second given the conclusions resulting from my work package two experiments in particular i want to emphasize that just because the physics tells us that a particular material or product may or may not have governed what occurred at grenfell tower

02:10:29 governed what occurred at grenfell tower this does not in my view mean that the inquiry shouldn't seek to understand as fully and completely as possible how these products came to be installed within grenfell tower in the external wall arrangement or the

02:10:40 in the external wall arrangement or the culture and context within which this occurred

02:10:43 occurred i want to finish my pre-prepared evidence by thanking the chairman the panel and the wider inquiry team for allowing me and my team the freedom tremendous freedom actually it has to be said to undertake our phase

02:10:55 it has to be said to undertake our phase two experiments as we saw fit thank you very much

02:10:59 very much well thank you very much professor bisbee it's been a fascinating afternoon and we're very grateful to you for conducting so many experiments with such care and for recording them in a way that enables us to

02:11:11 that enables us to understand what you've done and i should say again full credit to the team the wider team at edinburgh for that work of course yes thank you very much

02:11:19 much yes mr millet well that concludes professor bisbee's presentation for which we're all extremely grateful thank you professor uh and we will invite the professor to return on monday morning uh

02:11:30 professor to return on monday morning uh for some questions yes well we look forward to seeing you again on on monday at 10 o'clock and we'll see what questions there are for you thank you thank you very much indeed

02:11:47 [Music] thank you very much mr miller well that's it for the day it is yes mr chairman thank you and uh as i just indicated we'll resume at 10 o'clock on monday morning very good thank you very much thank you very much

02:12:18 you

↩ All hearings