Expert Presentations - Thursday 9th June 2022 (1/2)

Full transcript

00:00:29 good morning everyone welcome to today's hearing

00:00:33 hearing today we're going to hear evidence from two of the expert witnesses instructed by the inquiry

00:00:39 inquiry and we're going to hear it in the form of presentations yes ms granger yes good morning good morning mr term and good morning the panel members so this morning professor torreira will be giving a presentation

00:00:50 torreira will be giving a presentation based on his report entitled determining performance the adequacy of the current testing regime and he's going to be presenting uh this morning and then answering questions on

00:01:01 morning and then answering questions on that report later and he will be dealing today with the role of the fire safety strategy in particular in relation to a regulatory system that has functional performance

00:01:12 system that has functional performance requirements as opposed to prescriptive ones what performance objectives are relevant and should be considered how the relevant performance objectives can be understood and quantified scientifically

00:01:24 scientifically how the current testing regimes produce data and how that data can or cannot be used to understand the relevant performance objectives the testing of products and systems to produce data and

00:01:35 products and systems to produce data and who is responsible for fire safety in buildings and the competency they require to perform that appropriately so those are the topics he will cover um

00:01:44 um and this will be an oral presentation i do need to give a trigger warning several of the slides that professor torreira's presentation uh in his presentation contain images and videos

00:01:55 presentation contain images and videos of fires

00:01:56 of fires including fires inside homes and in tall buildings and there are also images of the interior of grenfell tower post fire so i give that trigger warning now and it's important to warn anyone watching

00:02:08 it's important to warn anyone watching who prefer not to view that content so could we perhaps call professor tarero now yes thank

00:02:20 you morning professor you're going to affirm i think is that right yes the words are there on the screen could you read them out okay i do solemnly sincerely and truly declare and affirm that the evidence i shall give shall be

00:02:32 that the evidence i shall give shall be the truth the whole truth and nothing but the truth thank you very much and welcome back to the inquiry thank you thank you

00:02:39 thank you yes

00:02:40 yes yes professor torreira thank you very much for attending the inquiry again and assisting us with our investigations is very much appreciated if we can please bring up your report that you're going to speak to today and

00:02:51 that you're going to speak to today and also

00:02:52 also next week it's at jtor706

00:02:59 before we look at that professor um i note you're still standing you know of course you're very welcome to sit down if you would like to um i understand that you might prefer to stand but if you want to switch between

00:03:10 stand but if you want to switch between standing and sitting um please feel free to do that as well thank you very much as long as there's a chair provided for you i'll find one well i'm sure we can do that yes

00:03:22 yes so if we could have the report up on the screen there we have it phase two grenfell tower inquiry adequacy of the current testing regime and we can see it's dated

00:03:33 and we can see it's dated january the 4th 2022 and there's a signature is that your signature yes it is and if we also look at page 7 of your report it's entitled statements and there are

00:03:44 it's entitled statements and there are various declarations there again is that your signature at the bottom of the page yes it is now have you provided that report in the same way that you would have provided a report to a court yes i have can you

00:03:57 report to a court yes i have can you confirm that the facts and matters set out in this report are true to the best of your knowledge and belief yes they are can you confirm that your report represents your true and complete professional opinion on the matters addressed within it yes i can confirm

00:04:10 addressed within it yes i can confirm and today you'll be giving a presentation about that report and then you'll be asked questions about it next week yes so please would you go ahead and give your presentation thank you very much

00:04:27 okay thank you very much so before i get started um as ms grange said uh this is about the report that uh that i wrote and it focuses on determining the

00:04:39 focuses on determining the determining performance and the adequacy of the current testing regime and i would like you to bear with me as i cover some background information that i'm sure you've heard many times already

00:04:51 i'm sure you've heard many times already but i do think it's necessary to set the context uh for the the the rest of the of the report and and i think that uh that it addresses uh this background information

00:05:02 addresses uh this background information addresses a number of general issues that pertain the fire safety strategy and the role of a fire safety engineer that i really think is important in trying to understand what expectations

00:05:13 trying to understand what expectations we must have in regards uh to the performance of systems such as a building envelope of grenfell tower so i think the the first thing that needs to be you

00:05:24 the the first thing that needs to be you know address you know it is a contextual issue so when we're talking about fire safety

00:05:31 safety the first thing that we really need to recognize is a very it's a very multi-disciplinary space uh you've heard evidence from architects on monday you heard evidence from mr lay who is an

00:05:42 heard evidence from mr lay who is an aerospace engineer you've heard everything from dr barbara lane who is a structural engineer i myself am a mechanical engineer and we have all found ways by which we have educated ourselves to try to understand

00:05:55 educated ourselves to try to understand you know the different aspects of fire safety engineering nevertheless fire safety does not have a common language and and therefore uh there is a

00:06:06 language and and therefore uh there is a number of elements of information that many times are very difficult you know to translate and therefore when i am describing certain things as a mechanical engineer an individual with a

00:06:17 mechanical engineer an individual with a mechanical engineering background i need to be extremely explicit in the way in which i present things to make sure that others can understand me appropriately now in

00:06:29 appropriately now in in in a similar way uh when we apply fire safety into buildings we have to recognize that buildings are going to outlive the designers and we are designing them to stay for a long

00:06:41 we are designing them to stay for a long period of time and therefore there has to be a very clear record of those elements that pertain you know to the to to to the fire safety strategy so that

00:06:52 to to the fire safety strategy so that uh that those elements can be used um in you know through the life of the building by the multiple stakeholders and that's the final point of context

00:07:03 and that's the final point of context which is the fact that there is multiple stakeholders and those multiple stakeholders range from the occupants that are are not required to have any

00:07:14 that are are not required to have any knowledge on fire safety matters you know to the fire brigades who have a certain aspect of knowledge pertaining to their uh activities you know to

00:07:26 to their uh activities you know to people that are doing far risk assessments and so forth so uh those multiple stakeholders again come from very different backgrounds and therefore required to have a very explicit you

00:07:37 required to have a very explicit you know representation of

00:07:40 of of the information that is available and this is why i i very strongly believe that documentation is absolutely essential in fire safety because it is

00:07:52 essential in fire safety because it is the way by which we not only transfer the information you know to the multiple stakeholders for a very long period of time but also the way in which we establish

00:08:02 establish the language that we're going to be using and everything that that language implies you know among people that might have very very different backgrounds now this pertains very

00:08:13 now this pertains very in a very critical manner to a slightly different aspect which is uh many times are complicated by regulatory frameworks

00:08:24 by regulatory frameworks when we designed uh a building to be fire safe uh one thing that is absolutely clear is that fire is a very complex problem

00:08:35 that fire is a very complex problem and in our design methods there are going to be a very significant assumptions and those assumptions need to be documented because they are not trivial and not only they are not trivial but

00:08:47 and not only they are not trivial but they are not absolute assumptions you know i can make a set of assumptions that brings me to an outcome but i could reach the same outcome with a completely different set of assumptions

00:08:58 different set of assumptions now when i change the assumptions one of the things that i actually change and i would like to emphasize this throughout my presentation is one of the things that i change are the dependencies

00:09:09 that i change are the dependencies so this is the second level uh that needs to be understood because if i make a set of assumptions i will create dependencies between different components of the fire safety strategy

00:09:22 components of the fire safety strategy that might be completely different to the dependencies that i create if i made a completely different set of assumptions so the relationship between assumptions and dependencies is extremely

00:09:33 and dependencies is extremely complicated and is one that really needs to be explicitly addressed particularly because many of those dependencies can lead to conflicts sometimes i can introduce certain type

00:09:44 sometimes i can introduce certain type of measures that are in conflict with other measures and will only work if for example they operate in a sequential way so but the moment the two

00:09:55 sequential way so but the moment the two of them are required to operate simultaneously then you might introduce a conflict a perfect example of that will be the use of sprinklers next to a smoke management tent

00:10:06 smoke management tent so the sprinkler performance is determined by the flow of air or the flow of hot smoke that passes the sprinklers so if i put a massive vent to evacuate the smoke which is a different measure

00:10:17 measure then what i will actually result is in a situation where i have created a conflict now if the sprinkler activates before i open the vent and i create that sequence both the sprinkler and the vent

00:10:30 sequence both the sprinkler and the vent will operate appropriately so the need to be able to understand the the sequence of assumptions leading to dependencies and potentially to conflicts is something that is

00:10:42 conflicts is something that is fundamental and is very specific to each unique fire safety strategy so without the privilege of having all that information explicitly written

00:10:53 explicitly written we are creating an enormous potential for confusion at the moment where different stakeholders have to make decisions so in essence i wanted to contextualize this

00:11:05 i wanted to contextualize this by saying that that i all the things that i'm going to discuss through this presentation i believe are things that should in principle always be documented and they should always be

00:11:17 be documented and they should always be documented and making sure that at each point we're explicitly clear about our assumptions and the language that we're using so i excuse myself for uh using the

00:11:29 so i excuse myself for uh using the language that pertains to my own background and uh and i will try to uh explain things in as much detail as possible but please if uh if at any point you believe that uh that i should

00:11:40 point you believe that uh that i should stop and explain something on the side i'll be more than happy uh you know to to do that because i do recognize that uh that i'm explaining things from my own personal background

00:11:53 so i think uh the first thing that needs to be understood you know from the beginning is that if we are going to design something you know we need you know to have a target so we need to be

00:12:06 know to have a target so we need to be able to understand explicitly what is the objective of our design and uh and in the case of fire safety the the real question that we need to

00:12:17 the the real question that we need to answer is what is a fire safe building now for certain systems that's a very trivial and simple question to ask you know if you're designing a gearbox for a car you might define that as a

00:12:29 for a car you might define that as a million cycles in the testing regime so you define a very clear testing regime and then you apply a million cycles in that and if the gearbox doesn't fail then your gearbox is appropriate so here

00:12:41 then your gearbox is appropriate so here you have a very clear performance performance criteria you know which is a million cycles in this specific testing regime you know there are other cases where it's a little bit more complex so if you're trying to design a seat and a

00:12:53 if you're trying to design a seat and a seat belt in a car then it becomes more complicated because now

00:12:59 now you have to cover a wide range of individuals so you know there's it's not the same thing you know to have an impact with a child sitting in the car they're having an impact with an adult you know sitting in the car so you're going to start

00:13:11 in the car so you're going to start creating in the testing regime the possibility of expanding you know your uh

00:13:19 uh the the way in which you conduct the test so that you could cover the potential range of users that you're going to use in the car nevertheless you are going to quantify

00:13:30 are going to quantify the way in which you conduct the test in a manner such that it represents in an appropriate way what you would expect you know will be a for example a collision in in in the case of of a

00:13:42 collision in in in the case of of a vehicle

00:13:43 vehicle and then you apply performance criteria so you will probably apply a performance criteria that is in no part of the dummy that is used you know for

00:13:54 of the dummy that is used you know for the test the pressure will exceed a certain threshold so when you're designing a system and you're testing against the system you have a very clear and well articulated

00:14:07 have a very clear and well articulated quantitative and explicit performance criteria

00:14:10 criteria so when we say what is a fire safe building the first question that we should be asking ourselves is what is the performance criteria

00:14:22 is what is the performance criteria and this is something that i struggle very much because obviously uh when we're trying to define a quantitative and explicit performance criteria and it re it pertains the matter of fire

00:14:36 and it re it pertains the matter of fire safety

00:14:37 safety we're clearly addressing the fact that we are dealing with people's lives so being explicit about what is the performance criteria

00:14:48 about what is the performance criteria is a very very difficult task and people will argue that if i make a very simplistic statement like nobody should be affected in a negative

00:15:00 nobody should be affected in a negative way by the fire that becomes an impossible task because within the variance of people and the conditions in which people are going to operate and so forth you know it is an impossibility to

00:15:12 forth you know it is an impossibility to guarantee that everybody you know will be unharmed by the fire so instead you know we could modulate that and take acceptable criteria that are

00:15:25 and take acceptable criteria that are used generally in regards you know to life safety and say nobody should be harmed nobody should be negatively affected you know when it comes to a fire

00:15:36 comes to a fire to a reasonable probability that could be defined quantitatively as 10 to the minus 6 10 to the minus 8 10 to the minus 9. you know but all these discussions um

00:15:49 discussions um are essentially bringing us to a position in which we do require to make an explicit statement of what is the target performance that

00:16:01 of what is the target performance that we are going to expect you know from our building

00:16:05 building now

00:16:06 now this is something that it is important for me to state from the beginning because clearly as we will see later this explicit performance requirement is not something that is written within our

00:16:18 not something that is written within our building regulations our building regulations have no explicit performance requirement which eventually leads people to try to use comparative performance requirements

00:16:31 use comparative performance requirements not as badass better than the dan or not as bad as is

00:16:39 is something that is mostly associated to guidance or prior examples or other applications so we're trying to compare what we're doing on a specific design

00:16:50 what we're doing on a specific design with something that was done before now obviously if that's what we're going to try to do if we don't understand explicitly what are the assumptions the

00:17:01 explicitly what are the assumptions the dependencies and the conflicts associated to those examples that are given in our building regulations as guidance

00:17:10 guidance then we will never be able to extrapolate that into our appropriate design you know which is effectively the definition of competency you know what we have is a situation by

00:17:21 you know what we have is a situation by which we are constantly extrapolating because we do not have a definition of safety you know what we have is just examples of what works and we are

00:17:33 examples of what works and we are requiring the professionals to make that extrapolation in in a space that is truly multi-disciplinary and is a space where language is very much confused

00:17:44 much confused and in a space where this information then needs to be conveyed to a group of stakeholders that range from stakeholders that have no responsibility for knowledge to stakeholders that in

00:17:56 for knowledge to stakeholders that in theory should have great responsibility for their competency so in principle we've created a space in which the responsibility on the competency of

00:18:07 responsibility on the competency of those professionals making those extrapolations is very high because they will define what is a fire safe building so

00:18:18 having cleared that up then we can start looking

00:18:21 looking into the process of how we actually establish

00:18:25 establish a fire safe building so i i'm i'm taking this information from my phase one report and uh and basically i'm gonna walk you know

00:18:37 basically i'm gonna walk you know through the concept of how we attempt to implement a fire safety strategy that delivers an adequate level of safety for a building of the nature of of of

00:18:48 a building of the nature of of of grenfell tower so

00:18:51 so if if if you look at this building it is a tall building and in this residential building is occupied by people the people who occupy this building are not

00:19:02 people who occupy this building are not deemed to be responsible to have any competency in matters of fire safety and therefore should be guided through the whole process you know there is an expectation that they will be guided through the process to a point where

00:19:14 through the process to a point where they are actually safe okay so how do we achieve this so

00:19:19 so basically if we have a fire the first thing that we try to do is make sure that we have alerted people that is our notification and that is the level of knowledge that

00:19:31 and that is the level of knowledge that we have imparted on the occupants that they should recognize an alarm

00:19:37 an alarm so that alarm is generally delivered by a detector and there's a technology for the smoke detector or the temperature detector that will deliver the alarm in a principle a timely manner

00:19:49 a principle a timely manner so

00:19:50 so once

00:19:52 once people

00:19:53 people have

00:19:55 have been alarmed and

00:19:57 and and they are in a position in which

00:20:00 in which they can start moving

00:20:02 moving out of the area of risk we are putting a whole series of measures to try to make sure that this is done in an appropriate way

00:20:13 appropriate way now this is uh what brings me to the next level of language that we tend to use in fire safety which is the language of the surrogate you know we never explicitly describe

00:20:27 you know we never explicitly describe what is what we are trying to quantify what we do is we use surrogate variables so what let me try to explain that uh a very commonly used surrogate variable

00:20:39 a very commonly used surrogate variable is the concept of maximum egress distances

00:20:43 distances so what is a maximum increased distance it's actually nothing real a maximum increase distance is a creation that represents in terms that an architect

00:20:54 represents in terms that an architect can understand when they are designing the layout of a building a particular problem so let me try to explain what that particular problem is so if you have a building or a

00:21:06 so if you have a building or a compartment there's an opening in here and you have a fire

00:21:12 a fire what is going to happen is that that fire is going to produce smoke and the smoke is going to start accumulating at the top of the compartment if you have people in the immediacy of

00:21:23 if you have people in the immediacy of that fire

00:21:24 that fire once the smoke detector or the temperature detector has activated these individuals are going to start displacing themselves okay and we can argue that there's

00:21:36 okay and we can argue that there's multiple stages to that displacement there will be what we call a pre-movement time which is the period where people gain awareness and they realize they have to move so there's a delay an initial delay then they're

00:21:48 delay an initial delay then they're going to start moving and they're going to move at a certain speed which is going to require them another time

00:21:54 time to be able to get to a safe place okay and those times can be accumulated but in essence what we're really talking about is an evacuation time

00:22:07 about is an evacuation time we're not talking about a maximum egress distance

00:22:11 distance the distance is a surrogate is velocity times time i know what the velocity is going to be and therefore if i multiply by the time the time then i get this distance

00:22:24 the time then i get this distance but this is where the big assumptions come in

00:22:27 come in because in reality a maximum interest distance could be infinity

00:22:33 infinity y could be infinity because if i put a massive extraction system in here and i take all the smoke out and i have an absolutely infinitely efficient suppression system i will get

00:22:45 efficient suppression system i will get rid of the problem take all the smoke out and nobody has to move okay so i have made my maximum egress distances infinity because there is no need for people to move as

00:22:57 there is no need for people to move as you know perfectly well another way by which i achieve this is by creating compartments if i box the fire then i can get people out of here i can put a barrier that prevents the smoke to

00:23:11 put a barrier that prevents the smoke to come out

00:23:12 come out okay and that effectively once i've crossed that barrier into a safe space i am safe

00:23:19 i am safe and the moment that i am safe then i have the capacity to stay there forever this is the principle behind stay put that you can actually stay there because you are protected by the compartment

00:23:31 you are protected by the compartment that is going to enclose the fire okay so in principle it is a time that we're talking here about and we're not talking about a distance the distance is the surrogate

00:23:44 distance the distance is the surrogate and these are the kinds of decisions that we make when we translate the information so that we can present this information to an architect so an architect can

00:23:56 to an architect so an architect can design a layout but of course there's a lot of embedded assumptions the first embedded assumption is going to be how long do i need

00:24:07 do i need to get people out but also how long do i have

00:24:12 have because this fire is going to grow and it's going to grow at a certain rate the rate of growth is going to be dependent on what materials i'm going to be using in the room the rate of growth is going to be

00:24:23 the rate of growth is going to be dependent on the size nature characteristics of the compartment and it's going to be dependent on other measures in other words other dependencies i am putting into the system

00:24:34 i am putting into the system so in principle what i am doing is creating a situation by which i have a

00:24:41 a an available time and that's the time to reach untenable conditions and that available time which i'll call the time of the fire okay has to

00:24:52 okay has to always be related to the egress time so if the time available is smaller than the time that i need then i have harmed the occupants of this

00:25:05 then i have harmed the occupants of this compartment but if the time available is actually much much greater than the egress time then i have fulfilled my objective

00:25:18 then i have fulfilled my objective of having these individuals remaining unharmed

00:25:21 unharmed so in essence when you think about these things what you realize is that

00:25:27 is that the a lot of the uncertainty that happens in these discussions about terminology like maximum egress distance is not necessarily related to the

00:25:38 is not necessarily related to the calculations that people are making and it's not necessarily related to the actual concepts themselves it is related to the assumptions because the reality is that if i double

00:25:49 because the reality is that if i double the height of this compartment okay which is part of the assumptions this value is going to change drastically so if i fix this maximum egress distance

00:26:01 so if i fix this maximum egress distance what i have fixed is a list a very very long list of assumptions so we can spend hours and hours arguing about whether it is 9.5 or 10.1

00:26:13 about whether it is 9.5 or 10.1 and if we don't list all the assumptions the conversation becomes meaningless and this is why it's so important once again to be absolutely explicit

00:26:25 again to be absolutely explicit and if we are not explicit we have to recognize

00:26:29 recognize that all those participating in the dialogue will have not only a common language but a common level of competency because only if they have the same level of competency and they can speak the

00:26:40 of competency and they can speak the same language they will be able to interpret without being told what is that very very long list of assumptions and this problem manifests itself pretty

00:26:51 and this problem manifests itself pretty much everywhere in the design of a fire safe building so this is sort of a sort of an initial example that i feel serves to illustrate to some extent uh you know the the process so i'll

00:27:03 uh you know the the process so i'll continue going through the different measures and i'll try to bring other examples to there so once we have allowed

00:27:13 allowed people to be aware that there is a fire then we expect them to move out of the area of risk now we are going to play with this idea of maximum egress distance so the

00:27:24 maximum egress distance so the comparison between the required safe egress time and the available safe increased time we are going to be playing with this by

00:27:34 by creating

00:27:35 creating a different layer of measures which is the compartmentation so

00:27:42 so if i

00:27:43 if i deem

00:27:44 deem that

00:27:45 that the whole path to a safe place is too long then i want to break that path creating a number of redundancies that enables me to protect people at each

00:27:56 enables me to protect people at each stage of the motion so

00:27:59 so in essence you know i set people to egress

00:28:04 egress once they understand that there is a fire

00:28:07 fire and then i introduce elements of a compartmentation now the compartmentation in a building like rainfall tower has fundamentally two purposes one is to contain the fire

00:28:19 one is to contain the fire in other words is basically boxing the fire so the rest of the units are unaffected and the second is to protect the means of egress to enable at all stages people to move into a safer place

00:28:33 to move into a safer place now

00:28:35 now we are going to consider for example what you see uh here in in yellow

00:28:41 yellow a first level of compartmentation and that first level of compartmentation you know that you can see in this diagram uh will have certain characteristics that

00:28:53 will have certain characteristics that help protect that corridor now this is again when we recognize what the assumptions are so we know that that compartmentation has vulnerabilities and those vulnerabilities are very clear

00:29:05 and those vulnerabilities are very clear the door

00:29:06 the door the door is going to open and the moment that it opens it might potentially allow smoke the door by its nature might not have the same capacity to control the fire or

00:29:18 the same capacity to control the fire or box the fire as the walls so i know i have a weak point so

00:29:24 so we know perfectly well that we can put other complementary measures like putting self-closing mechanisms but we know that those again are very vulnerable you know those self-closing

00:29:35 vulnerable you know those self-closing mechanisms are going to fail those explosive mechanisms are inside the unit so they have to be maintained you know and they have to be maintained with many times little control over how they're being maintained so we know that this

00:29:46 being maintained so we know that this are

00:29:47 are vulnerable points so we introduce redundancies into the system so this is a point where we put a second level of compartmentation

00:29:58 compartmentation so again we have a a second fire door now we have two layers we're only requiring people to move out of that corridor into the next level of safety you know which in this case becomes the lobby

00:30:09 lobby and uh

00:30:11 and uh and and again i'm gonna have not one door but now i have two doors that need to fail so i've created a level of redundancy and i've created a level of redundancy to enable

00:30:22 to enable a safe egress of individuals now

00:30:26 now there's another need for redundancy which is i have to establish how long people are going to be in each safe space

00:30:35 space so if i think of the corridor of that unit it's going to be a few seconds but if i think of the stair we recognize that if you are in the top floor and you have to descend the entire

00:30:47 floor and you have to descend the entire building the period is going to be very long so the stair will require another level of redundancy so i'm compartmentalizing the stair in such a way that i've added an additional

00:30:58 such a way that i've added an additional level of redundancy that protects my means of igress making the stair a safer place than the lobby and making the lobby a safer place than the hallway of

00:31:09 lobby a safer place than the hallway of the unit and so forth so i've introduced the concept of redundancies to make sure that my systems work now

00:31:18 now many times we find that we get into a position in which we have created a dependency because the dependency that we have here is that if the lobby gets compromised

00:31:30 is that if the lobby gets compromised you cannot get to the stair so you've created naturally a dependency so you have to find a way to try to compensate for that dependency and this is where you introduce things such as

00:31:42 is where you introduce things such as smoke management and i'm going to use smoke management as an example uh because i know you've heard uh in monday and tuesday a lot about smoke management so i can allow myself to um to explain uh using that as

00:31:55 myself to um to explain uh using that as an example uh of of how you introduce this other level of redundancy and how the dependencies start piling up as i change the nature of the redundancy so in principle

00:32:08 of the redundancy so in principle that fourth level of redundancy is there to resolve the dependency between lobby and stair

00:32:14 and stair and but also as another layer of protection to the system so now i have four levels of redundancy so if i think about how

00:32:24 about how [Applause] we can move a flow so let me just take a pipe so if i just take a simple pipe okay we can all recognize

00:32:37 okay we can all recognize that if i take a piece of paper and i push the air wind is going to be created what that paper is producing is a pressure on the flow and the pressure

00:32:48 pressure on the flow and the pressure induces a velocity in a similar way if i'm standing in the street and wind hits me i'm going to feel the pressure of the wind so there's a very clear relationship

00:33:00 so there's a very clear relationship between velocity and pressure the pressure is the force that creates the velocity and the velocity once it hits an object can be reverted back into a force or to a

00:33:13 be reverted back into a force or to a pressure now that relationship basically says that the velocity is going to be proportional to the square root of the pressure

00:33:21 pressure and that relationship is intrinsic to the way in which hydraulic systems work to the way in which smoke management works and all these different types of systems

00:33:32 types of systems so

00:33:33 so the velocity once again becomes a surrogate

00:33:37 surrogate well the pressure is the one who is driving the problem so the velocity is a result of having an appropriate pressure that creates that flow that is going to move

00:33:48 creates that flow that is going to move the smoke away from one place and put it in another okay so if i have a system like this where the flow goes in one direction all that i need is to create a pressure here

00:34:01 that i need is to create a pressure here one that is greater than a pressure here that is two so if p1 is greater than p2 then what happens is i introduce a velocity

00:34:13 happens is i introduce a velocity and that velocity is a function of the difference between the two of them so delta p

00:34:19 delta p is actually p1 minus p2 okay

00:34:23 okay now the problem with this type of systems

00:34:26 systems is that velocities are much easier to measure

00:34:29 measure so if i want to design a system and i want to make sure that it works

00:34:35 works it's going to be much more difficult in a lobby or in a unit to start measuring pressures so what i'm going to do is create a surrogate okay but the problem is that what drives

00:34:47 okay but the problem is that what drives the problem is not the surrogate what drives the problem is the pressures so

00:34:54 so this pressure is the one who is going to determine the flow and that's what's going to determine if i can extract the smoke so what happens if i change the parameters now let's go now to a

00:35:05 parameters now let's go now to a building

00:35:07 building so you are in a building and you have a set of lobbies and you have a stairwell

00:35:14 stairwell okay and you're gonna have floors

00:35:19 and you might have a fire here okay so this particular fire because of gravity bringing smoke hot smoke up and trying to push it away this

00:35:33 smoke up and trying to push it away this fire is going to create a pressure that's the pressure of the fire okay

00:35:39 okay now

00:35:40 now then i'm going to have a duct

00:35:44 and that that because this is going to be hot smoke

00:35:49 hot smoke this duct is going to have a temperature of the smoke that is greater than the temperature of ambient

00:35:58 in other words because the hot smoke is hotter than ambient there's going to be a force that is going to push the smoke up so in this direction

00:36:08 direction i will have a pressure of the smoke and this is plus it goes in that direction but the size of the duct determines how much friction i have which is going to

00:36:21 much friction i have which is going to create a fraction that is going to go down and that is going to be the pressure of the friction okay but that's negative okay

00:36:31 okay now if i have a door that is open that is going to create a pressure

00:36:37 pressure if i have two doors open that's going to change the game now what gives me the velocity the surrogate it gives me the balance of all these pressures because at the end what

00:36:48 these pressures because at the end what i want to see is what is the net the pluses minus the minuses so i want to make sure that i understand what is the net to be able to understand what is the flow

00:36:59 what is the flow so the flow when i say that is two meters per second the one thing that i don't realize is that that's those two meters per second assume a typical pressure of a fire

00:37:13 assume a typical pressure of a fire of one compartment of the size of a typical fire in a unit

00:37:19 okay so if i'm making that assumption and i'm designing for two meters per second then i need to be clear about the fact that that is the

00:37:30 about the fact that that is the assumption because what happens now if i have a fire

00:37:35 fire that engulfs the entire building that fire is going to create

00:37:43 a huge delta p

00:37:45 delta p a huge pressure difference that is infinitely bigger than this one and therefore what is going to create is an imbalance on my system

00:37:56 on my system that now is dominated by what is the controlling mechanism so the system inevitably is designed for a set of assumptions and those

00:38:09 a set of assumptions and those assumptions need to be perfectly clear and if those assumptions are not perfectly clear arguing about whether the system will work or will not work

00:38:21 work or will not work becomes a futile exercise because it is the assumptions that determines the performance and i might design a perfect system to work for a certain set of assumptions

00:38:32 to work for a certain set of assumptions but i need to be perfectly clear that that system is going to work in a different way if those assumptions fail

00:38:43 those assumptions fail okay so in principle what i have is a system

00:38:48 system that was designed for a set of assumptions and among the set of assumptions is that the delta p that is created by the fire is the fire of one unit

00:38:59 the fire is the fire of one unit and if i fail to guarantee that then i have no guarantee how the system is going to work now not only is designed for that but it's also designed for to enable

00:39:11 it's also designed for to enable firefighter intervention because if you actually have the fire in the top floor and you have the firefighters coming in from the bottom floor you're basically pulling fresh air into

00:39:22 you're basically pulling fresh air into the stair now if you have people opening doors in multiple floors with smoke you might end up in a situation where your hydrostatic pressure or the pressure

00:39:33 hydrostatic pressure or the pressure that is created by the temperature increase might end up actually bringing the smoke into the stair or into the lobby and all these variants are a very fine equilibrium where all the pluses have to be compensated for

00:39:45 the pluses have to be compensated for the negatives to give you a net and in which direction that net goes is a function of what are the pluses and the minuses and if you introduce this type of dominating factor then obviously

00:39:56 type of dominating factor then obviously what is going to happen is that the system is going to perform in a way that is different to what it was intended okay so as you can see assumptions

00:40:07 assumptions are very important and and certain assumptions are going to take a more important role than others and we have to be extremely careful about this because for example we're

00:40:19 about this because for example we're going to make assumptions about compartmentation and uh and those assumptions might not be correct nevertheless we have four levels of redundancy

00:40:30 redundancy okay but we might be making assumptions about a smoke management system well we know we have no redundancy because if the fire exceeds one unit the delta p changes and your smoke management system is going to have

00:40:43 management system is going to have a different performance okay so

00:40:46 okay so these are the kinds of nuances that are necessary if we really want to understand a fire safety strategy so at the end the objective

00:40:58 at the end the objective of introducing compartmentation and smoke management is to deliver a safe area now the safe area could potentially be the stairs and if

00:41:09 could potentially be the stairs and if you have an eager strategy that requires to evacuate people you will have to maintain that safe area for a very long period of time until people fully evacuate

00:41:18 evacuate the building okay

00:41:21 okay but if you're going to implement a stapled strategy and this is something that becomes to some extent contentious you know if you are going to have a staple strategy where your safe area is

00:41:34 staple strategy where your safe area is all the spaces occupied by all the other occupants of the building then

00:41:40 then you just simply cannot allow for fire spread okay now people are going to argue and this is going to be explicitly said many times

00:41:51 times that

00:41:51 that an absolute statement of no external fire spread is unachievable you know this is going to be the argument on an endless

00:42:02 argument on an endless basis now i'm not necessarily disagreeing with that you know it it might

00:42:07 might making absolute statements might not be necessarily appropriate but

00:42:14 but you can modulate those statements you can say that the objective is no fire spread

00:42:19 spread and you can modulate it by saying to a certain probability you know you can modulate it by making a statement that no fire spread but if fire spread happens here is my plan b my

00:42:32 fire spread happens here is my plan b my contingency this is how we pick it up the redundancy so we can put all this but we can still set the objective as being no fire spread and we can qualify according to what we need in

00:42:44 qualify according to what we need in very different ways because in essence by principle if people are going to be remaining in other areas and there is no

00:42:55 in other areas and there is no contingency plan then in essence by principle no means no and there is no way about it okay because keep in mind that it is our

00:43:06 okay because keep in mind that it is our acceptance that occupants are not responsible for understanding fire safety so

00:43:14 so the decision-making that will prompt them you know to leave a unit if they come to the conclusion that the risk has reached an acceptable level it is a

00:43:26 reached an acceptable level it is a decision that we cannot impose on them so we have accepted that we are delivering the safety and we are not expecting the occupants to make those decisions that are informed decisions

00:43:39 decisions that are informed decisions so

00:43:40 so in essence this is the principle behind no external spread

00:43:45 spread now

00:43:46 now obviously if we have a staple strategy even if we are evacuating people you know we would like uh the structure to remain in place collapse of buildings is never an acceptable outcome so the way

00:43:59 never an acceptable outcome so the way in which we set the objectives for structural performance is by saying that the structure will have to withstand the burnout of the fire so even if there is no intervention when all the fuel is

00:44:11 no intervention when all the fuel is consumed and the fire dies on its own you know the structure should still be standing you know which is effectively the case of grenfell tower so

00:44:20 so in principle that enables us also to have a level of redundancy to our egress because we have plenty of time from a structural perspective and and

00:44:31 perspective and and it allows

00:44:33 it allows for fire brigade intervention without the risk of the building falling on them and it enables also the possibility of extending

00:44:42 extending you know our evacuation time to as long as it is necessary so finally we have the fire brigade and the fire brigade obviously plays a role

00:44:53 fire brigade obviously plays a role and and again uh there are the actions of the fire brigade that we need to take into consideration you know but also we need to understand what is the objective that we set for them

00:45:05 objective that we set for them so when we look at the protocols of the fire brigades the objective that we set for them was to suppress a one unit file as they did in the case of grenfell tower

00:45:17 in the case of grenfell tower okay so in principle if we don't set those objectives correctly then we have the potential of

00:45:26 of misleading the fire service in a direction in which they will not be effective in their activities so there's an element of explicit information that needs to be transferred you know which

00:45:37 needs to be transferred you know which is what we are requiring from them now documentation obviously is essential because the training of the fire service is very different to the education of a structural

00:45:48 to the education of a structural engineer or an aerospace engineer a mechanical engineer so we're going to have different languages we're going to have different levels of competency in different areas and therefore the communication becomes fundamental

00:46:01 communication becomes fundamental so

00:46:02 so in essence that gives you a sense of how you put together a fire safety strategy and how dependencies and assumptions become extremely relevant now

00:46:16 once we have the fire safety strategy there has to be an element of transparency in its assessment so somebody has to be able to come and make an appraisal

00:46:28 and make an appraisal and assess if that strategy is actually going to work in time things might deteriorate things might change modifications might be made and we have the need to conduct the fire risk

00:46:41 the need to conduct the fire risk appraisal or assessment so that's an integral component of the process to make sure that the system as design continues to perform in time so i'm going to use as an example uh the the

00:46:53 going to use as an example uh the the newly

00:46:54 newly produced uh past 9980 you know to try to um explain uh how we have dealt with this concept of appraisal and and how

00:47:06 of appraisal and and how we are leaving to the competent professionals an enormous space for interpretation that we need to understand

00:47:17 understand how that enormous space of interpretation correlates you know with competency so if i just take one statement you know from the document

00:47:28 from the document you can immediately realize uh that that statement in itself carries a very significant weight in numerous forms

00:47:39 in numerous forms so basically the statement says that the only way external fire spread can be prevented in its entirety is to build a building out of entirely non-combustible materials

00:47:50 materials and not have any windows or other openings in the external envelope now

00:47:57 now let me try to explain that this statement

00:48:01 statement carries an enormous set of assumptions okay it is based mostly on on a document from 1966 by langdon and thomas and law

00:48:12 from 1966 by langdon and thomas and law and uh where they did a series of tests where they actually showed that the expectation is that when you have a building and this will be different floors of the

00:48:24 and this will be different floors of the building and you have a fire here the expectation is that the fire will end up here

00:48:31 here but they demonstrated that the reality is that actually under the circumstances of their tests the reality is that the fire actually went in

00:48:42 went in so they conclude that expecting that a fire cannot ignite the floor above is an unrealistic expectation given the circumstances that

00:48:55 expectation given the circumstances that they

00:48:56 they run the tests now i am not going to disagree with this i actually think it's a perfectly valid argument particularly if

00:49:05 if the window is open there is a window is obviously a barrier that will slow the penetration but eventually the window might fail but if the window is open you have a direct battery

00:49:16 battery okay now there's the first that's the first assumption is that effectively we're working with the configuration that they used now

00:49:27 the configuration that they used now there were further studies in canada that actually complemented this so the configurations are more than one but the reality is that they are the assumption this it works for these configurations now the next assumption is the

00:49:39 now the next assumption is the difference between smoke migrating into a space and actually having ignition to have ignition you need a sufficient amount of heat to reach the materials in here which is

00:49:51 to reach the materials in here which is a much much more onerous condition than having smoke penetrating now

00:49:58 now in principle this is an appropriate statement to make but you have to disclose all the assumptions that go in there because the reality is that we also all know

00:50:10 reality is that we also all know that i can change this problem and if i start extending

00:50:17 this balcony with non-combustible materials to the point that this flame never makes it to the end

00:50:29 to the end okay

00:50:30 okay then

00:50:31 then i can actually design a building with windows

00:50:35 windows that actually works because the flame will never get into here and the smoke will go up so by the time it's blown by the wind it's so diluted that it doesn't produce any harm so in principle i can design a

00:50:48 any harm so in principle i can design a building with windows okay so that's not the limitation now even the use of combustible materials well i can allow myself to put combustible materials here

00:51:02 because the flame will never have enough heat to ignite it now obviously in the tests that langdon thomas and thomas and law did their

00:51:14 protrusion is two feet and they found two feet as being insufficient and i understand that because it's not enough to divert the flame completely away

00:51:25 away but

00:51:27 but when you translate this into a statement that says

00:51:32 that says that

00:51:34 that to be able to design a building for the no spread condition you have to design it made out of totally non-combustible materials and

00:51:45 totally non-combustible materials and also that it has to have no windows there's two things that you're doing one is you're not presenting your assumptions and two

00:51:54 and two you're implying that we have to allow for fire spread so that's the implication that given that you cannot avoid it you have to

00:52:05 that you cannot avoid it you have to allow for fire spread now

00:52:08 now that in combination with the use of other types of language

00:52:16 very rapidly brings you to a position where you realize that interpretation knowledge of the assumptions understanding of the dependencies

00:52:27 understanding of the dependencies knowledge of the conflicts becomes absolutely essential to be able to answer the questions because what you're saying is that the building for the building to be

00:52:39 building for the building to be considered acceptably safe you have to define what acceptable means then

00:52:47 then we have now switched from a condition of no spread to a condition that implies time and they say the time required for fire to spread

00:52:58 the time required for fire to spread externally from one compartment to another or from one story to another needs to be sufficiently long in other words i need to be able to quantify the spread rate

00:53:10 quantify the spread rate or the failure time of the system and somebody has to do that in an explicit way now that is going to require not only understanding combustibility but it's going to

00:53:22 combustibility but it's going to understand need to understand the thermal mechanical behavior of the system because these systems are going to be heated and as they're heated they're going to deform you know and all this needs to be understood all that now again is being

00:53:35 understood all that now again is being piled again and again and again and then it is combined so as to allow for safe

00:53:42 for safe escape

00:53:43 escape and the intervention of the fire and rescue services how do we communicate to the fire and rescue services what is the acceptable spread rate if we

00:53:56 what is the acceptable spread rate if we don't write it down

00:53:59 how are they going to understand that they should expect a four-story fire within 10 minutes of ignition if we don't show them how we calculated things how we

00:54:10 how we calculated things how we estimated things what assumptions did we make and what are the dependencies and conflicts that are potentially there so

00:54:18 so i can continue to read every single one of this but i think you know there is uh truly no point because they all go more or less in the same direction you know not unduly harmed

00:54:29 not unduly harmed what is the definition of unduly harmed you know prevent it from escaping what does it mean to be prevented from escaping what is visibility is it temperature is carbon monoxide you know

00:54:40 temperature is carbon monoxide you know what is what we're talking about how do we quantify all these things at the end land into statements such as do not compromise the communal means of escape

00:54:54 compromise the communal means of escape and what is worst of all this is before those needing to use the escape routes so

00:55:03 so if you think about it this terminology from my perspective requires more and more to emphasize the competency of those individuals

00:55:15 competency of those individuals delivering this analysis and furthermore their capability to communicate in a way such that information gets transferred in a manner such that everybody can do their job

00:55:25 their job so

00:55:27 so at the end you know we need to understand you know what provisions are going to be put in place for fire spread because as we recognize if we have a state put

00:55:38 recognize if we have a state put strategy we have to find compensation for not fulfilling the requirement of no so without stating clearly what are those provisions

00:55:49 those provisions then we have no guarantee that things are going to work correctly at the moment

00:55:55 moment of the event what is tolerable what is acceptable how our time and firefighter capabilities quantified all the things need to be explicitly addressed

00:56:07 need to be explicitly addressed what is unduly harmed needs to be explicitly addressed you know and how do you guarantee that escape is viable you have to have a plan there is no point in

00:56:18 there is no point in having a situation unfold in front of you if you don't have a plan on how to deal with that situation now a further problem that tends to to

00:56:29 now a further problem that tends to to happen is that without the benefit of hindsight and the knowledge you know of what happened in grenfell tower i wonder how a designer will answer the

00:56:40 i wonder how a designer will answer the question if four levels of redundancy were sufficient because in principle i could use every single one of these words

00:56:47 words and argue

00:56:49 and argue with perfectly uh valid arguments that for this particular building four levels of redundancy was sufficient to allow for the flame spread that we

00:57:00 to allow for the flame spread that we observed

00:57:02 observed and i could actually even argue that the fire service should have had a plan and it was their fault for not having a plan to intervene in that case and i could make many many different arguments

00:57:13 make many many different arguments to ask a designer to justify designing this building exactly the same way as i did on the first place and and i think this needs to be very

00:57:24 and and i think this needs to be very very very clear because in many ways you will say that is professional responsibility moral and ethical behavior

00:57:32 behavior but at the same time it is competency because if you don't understand the consequences of your actions many times you might think that a problem is much much simpler than what it is and i do

00:57:45 much simpler than what it is and i do think that this is not a simple building many people will claim in a simple problem like this we do not need the appropriate documentation i'm sorry but this is not a simple building this is a

00:57:57 this is not a simple building this is a very complicated building by the mere fact that we allowed external fire spread so

00:58:04 so as you can see even when you're doing an appraisal an assessment you need to have not only a very clear idea of what is what you're aiming for but also you have to have the

00:58:15 aiming for but also you have to have the level of communication and competency that is appropriate to the language that you are introducing in your documentation so

00:58:24 so which brings me to the concept of the fire safety strategy and what are the basic components that a fire safety strategy should have to be able to allow me to to start thinking about the

00:58:36 me to to start thinking about the performance of a system because as you can see i haven't even started talking about the performance of an external building envelope you know all i have done is set up the stage and

00:58:48 all i have done is set up the stage and i still haven't got to the point where i can actually define the performance so the first assumption that we make is that there will be a single event a single fire which inevitably the moment

00:59:00 single fire which inevitably the moment that you have multiple fires in multiple units that main assumption is going to break down now

00:59:07 now i'm going to implement a detection system that is going to enable me to try to establish when uh people are going to begin the process of egress or they're going

00:59:18 the process of egress or they're going to be made aware that there is an event and then i'm going to put the four levels of redundancy associated to compartmentation to try to achieve the objective of guaranteeing a safe area

00:59:30 objective of guaranteeing a safe area that safe area could be the full building

00:59:32 building if if it's a stapled strategy or it could be uh the stairs if i'm going to enable an eco strategy and i need to resolve through the smoke management process the conflict

00:59:43 conflict between the lobby being a preamble to the stair okay so basically i've more or less handled all this and

00:59:53 and then i have the no external spread and i know that that no external spread is going to underpin a number of different things and therefore there are many dependencies associated to the no external spread

01:00:06 associated to the no external spread and then i have the structure and the structure has to withstand the fire and eventually that gives me the rest of the building as a safe area and that also enables a staple strategy so

01:00:18 so now

01:00:19 now if you think about it that is delivering building safety is the building doing its job

01:00:24 its job okay now that can fail and we recognize that and we have an extraordinary redundancy you know which is the fire brigade so the fire brigade will come in

01:00:35 so the fire brigade will come in and try to address any issue that actually escapes the design premise by which this building was to deliver the safety that we're aiming at so

01:00:47 so on top of that we have to add proper building management because many times provisions for that you know we have to add maintenance we have to add inspections and all these things guarantee us that through time

01:00:59 guarantee us that through time the building is going to continue to perform

01:01:01 perform okay

01:01:02 okay so

01:01:03 so if we think about it detections they have redundancies you know in most units we'll put multiple detectors you know and those redundancies are there to try to make sure that if one fails others will pick

01:01:15 sure that if one fails others will pick it up

01:01:16 it up obviously with compartmentalization as we we've seen compartments have multiple levels of redundancy because we recognize the weaknesses and we recognize the dependencies and the potential conflicts

01:01:28 potential conflicts now the fire brigade represents a redundancy to the whole building operation and we've put that in place now in the case of the structure the way by which we achieve the necessary

01:01:40 by which we achieve the necessary robustness is by using an assessment criteria that is effectively very conservative so we basically design structures to perform under the worst possible fire

01:01:53 perform under the worst possible fire for the longest possible period of duration which is burnout and by making the test a bounding test where nothing can get worse than that we've added a massive set of safety factors into

01:02:05 massive set of safety factors into structural behavior now we can criticize the method a lot but the reality is that in practice we have implemented a very robust approach which is increasing the sa the factor of safety to a magnitude

01:02:17 sa the factor of safety to a magnitude that is extremely large because we recognize the dependencies on the structure so we've done that appropriately now when it comes to the no external spread

01:02:28 spread and we start looking into the dependencies we realize that no external press bypasses compartmentation and has the potential as i explained to

01:02:40 has the potential as i explained to disable smoke management so in principle that's a very strong dependency that has already killed two levels you know of my problem now there's a dependency on the

01:02:51 now there's a dependency on the structure because the assessment method that we use that is so robust for structural analysis is an assessment method that is predicated on the rest of the building being called so the loads can be redistributed so in principle

01:03:04 can be redistributed so in principle there is a dependency on structural behavior that we need to recognize there will be a dependency on what are the safe areas because if we are allowing the fire to move away from the

01:03:16 allowing the fire to move away from the space we are effectively changing what is a safe area and so we need to take that into account and of course there's a dependency with the fire brigade who in principle has a

01:03:27 the fire brigade who in principle has a protocol to fight a fire of a certain nature that could be completely changed and they will be presented with a fire that is completely different so as you can see we have created

01:03:38 we have created a massive amount of dependencies on the building envelope and its capacity to prevent the spread of the fire now

01:03:47 now we have no redundancies they're only dependencies and the problem that we have is that on top of that we have a conflict because the moment that we give priority

01:03:58 because the moment that we give priority to the fire brigade to fight the fire we recognize that if

01:04:05 that if the smoke management system is not working as intended then the potential for having smoke in the means of egress and propagating itself to a number of different places

01:04:16 itself to a number of different places is very high so just by the mere operations of the firefighters opening stair doors to put the hoses and fight the fire are you're creating a potential conflict because

01:04:27 creating a potential conflict because when we had the one unit fire we had a state put strategy that created a sequence

01:04:33 sequence the people in the unit evacuate one once that's terminated firefighters fight the fire that's two and then people can migrate out of the building so in principle we

01:04:45 out of the building so in principle we created purposely a sequence and we assume that sequence will be respected but if this fails then we are created potentially a conflict because we have to evacuate people while intervening in fighting the

01:04:57 people while intervening in fighting the fire so we need to be able to recognize the things now a fire safety strategy is not only listing all the things that you did it's primarily listing what your assumptions are what your dependencies

01:05:08 assumptions are what your dependencies are and what your conflicts are and if you're not explicit about all these things and make them extremely clear you will become the victim of words such as adequate you know useful and all

01:05:21 as adequate you know useful and all these things that that in principle are part of the understanding of the problem now as i say buildings last in time

01:05:32 buildings last in time and buildings are complex systems and they require people from multiple disciplines to intervene you know and uh and therefore communication becomes essential and

01:05:43 communication becomes essential and that's why you know having the appropriate explicitness in all the things is absolutely fundamental

01:05:52 so if if that is okay mr sherman i have two more slides yes and that i would like to go through uh before we take a break of course now you you indicate when you

01:06:04 course now you you indicate when you think it's a convenient time okay um i think if we get close to half fast 11 the sonographer might like a break anyway but um no no it will take me probably 10 more minutes that's fine of

01:06:15 probably 10 more minutes that's fine of course

01:06:16 course okay so so then we have to move into the concept of performance objectives so we need to now be quantitative now now is when i start talking about

01:06:27 now now is when i start talking about how we define performance for the building envelope in as much as for everything else and in some cases it's actually quite easy but in other cases it's very very complex so

01:06:38 so the detection in the case of the detection establishing performance of a detection system is actually very simple because an ideal detection system will

01:06:50 because an ideal detection system will give you immediate warning instantaneously so if the time to detection is zero you have the perfect detection system so the technologies will aim to zero that's the performance target the only

01:07:03 that's the performance target the only thing that slows them down is that you have a lot of nuisance alarms the more sensitive you make your detector the more nuisance alarms you get and therefore the less people are

01:07:14 get and therefore the less people are going to be prompt to react appropriately to the smoke detector so that's pretty straightforward so technology as it evolves what the technology is trying to do is target zero

01:07:25 target zero very simple target and technology is dedicated to try to minimize nuisance alarms so that they can bring the detection as close as possible to zero because if you make detections zero then

01:07:38 because if you make detections zero then you can ignore that part of the problem and people what people do is what really matters and that's the only times that actually matter now other aspects that you need to take into

01:07:49 other aspects that you need to take into account you have to put sufficient detectors

01:07:51 detectors because the smoke has to get to the detector so you're going to place the detectors strategically to make sure that that time gets as close as possible to zero but as you can see i have a very

01:08:02 to zero but as you can see i have a very simple and straightforward performance target

01:08:04 target now

01:08:05 now with compartmentation again i have a very simple performance target do not allow fire or smoke passage until the burn out of the fire

01:08:16 the fire now i have i can devise a method to establish that and i can devise failure criteria that enables me to address that particular target so

01:08:27 so these are the easy ones so defining compartmentation and defining detection as performance targets are actually quite straightforward with external fire spread the problem is

01:08:38 with external fire spread the problem is a lot more complicated because you have this

01:08:43 this apparently impossible target of can not allow a flame to spread okay that of course is a difficult target to achieve particularly in modern systems are so complex

01:08:55 systems are so complex at the same time the system should not allow smoke to migrate

01:09:00 migrate so the formations and things that could allow the passage of smoke could potentially be a problem even if a fire doesn't ignite in a different place and

01:09:10 and and of course the system cannot ignite and spread a flame when solicited by a fire external to the building which is the one that we understand better which is a fire that is in another building and produces some heat that is going to

01:09:22 and produces some heat that is going to ignite the facade but the performance criteria then start becoming a lot more complex and much more difficult to quantify

01:09:31 quantify you know with the structure again is simple you know we are basically going to require to sustain no failure according to a predefined failure criteria and according to a predefined bounding test

01:09:43 bounding test and with the fire brigade is extremely complex because uh

01:09:51 uh well we recognize that

01:09:54 that the emergency response is a response to building failure which is the task okay

01:10:01 okay then

01:10:03 then we start introducing things like foreseeable events because protocols for firefighting need to be defined and therefore foreseeability needs to be assessed

01:10:15 foreseeability needs to be assessed and

01:10:16 and you know

01:10:18 you know we have to achieve an acceptable outcome and once again we run into the same problem of what is an acceptable outcome we can always take the position you know that that fire fighting is such

01:10:31 you know that that fire fighting is such an impossible task in nature that any outcome is acceptable but we recognize that that is potentially not an acceptable approach or we can establish clear performance

01:10:42 or we can establish clear performance criteria of this is what you need to achieve

01:10:44 achieve and again that becomes quite difficult so we need to be quite careful on how we use this redundancy to the system in a way such in which we set performance criteria that are appropriate to the

01:10:56 criteria that are appropriate to the capabilities of the fire brigade but they don't exempt them from their duties so

01:11:02 so in essence once i have those performance criterias i find ways to target them and uh and again in the case of the detection i have to quantify the activation time so i have a

01:11:14 quantify the activation time so i have a standardized test that i use to characterize and calibrate all smoke detectors that enter the market very very straightforward so in the case of compartmentation this is what brings the concept of fire

01:11:26 is what brings the concept of fire resistance so i have a test and that test delivers a fire resistance rating which basically tells me for how long according to the test the system will fulfill its functions and i can

01:11:37 will fulfill its functions and i can classify you know these buildings in a fairly simple way in the case of external uh fire spread uh

01:11:47 uh things again get very complicated not because there's issues of flammability but there's issues of thermo-mechanical performance of the system in the case of the specific building

01:11:58 in the case of the specific building because the interactions between the components are going to be critical to the way in which the system behaves so it becomes a system performance and the system performance pertains the building

01:12:09 system performance pertains the building you're trying to design and therefore it is really fundamental that this is looked upon in the context of the design that we are producing

01:12:19 producing and in essence the objective is to support the fire safety strategy so whatever built external building envelope we design it has to support the fire safety strategy that we have

01:12:30 fire safety strategy that we have implemented this is why for example when we design facade systems for offices where we have sprinklers we have an evacuation strategy we have complex smoke management systems we might

01:12:41 smoke management systems we might tolerate

01:12:42 tolerate certain things that we will not tolerate for example for a residential building where we have none of those features so

01:12:50 so in many ways the concept is that we have to support the fire safety strategy but now we don't have a test we don't have a target it's not an activation time

01:13:01 it's not an activation time it is something that is much more complex than that and it becomes a system performance and again with the structure is fire resistance again so no need to talk about that anymore

01:13:12 about that anymore and uh and with the fire brigade you know we tend to resort back to the statement that they have to perform according to what is being requested from them by a fire and rescue services

01:13:23 from them by a fire and rescue services act of 2004. whether that is clearly defined or not that's a different story but that's what we are charging them with so i think that this would be a good moment for me to

01:13:34 this would be a good moment for me to stop if that's okay yes certainly well thank you very much all right well we'll stop there uh we'll resume at half past 11 if that suits you

01:13:43 suits you and i think i should say to you as i've said to every other witness please don't discuss your evidence while you're out of the room with anyone perfect thank you thank you very much could you give the option please

01:14:01 thank you very much mrs grange hoffman thank you

01:34:05 would you ask professor torreira to come back in please

01:34:15 all right professor well we're ready to carry on when you are okay thank you very much

01:34:19 very much so um

01:34:21 so um in the second part of the presentation i am going to focus on actually the quantification of performance for um external facade systems and um and

01:34:32 and before uh we look at the object in itself uh we need to try to understand uh to some extent or at least to a sufficient extent uh

01:34:43 sufficient extent uh how the different physical processes are going to uh interact you know with the system in itself

01:34:50 itself so uh

01:34:51 so uh the

01:34:52 the example i'm going to be using to try to provide this explanation is the simplest possible example which is just simply a combustible material and how a fire will spread you know

01:35:04 and how a fire will spread you know within a combustible material i will move on then to look into more system behavior aspects of it but initially to try to describe the different phenomena

01:35:15 try to describe the different phenomena i want to use a more idealized and simple scenario so uh in in in simple terms a facade system is going to be subject

01:35:26 a facade system is going to be subject to two different types of solicitation uh if you look at the at the image in the left

01:35:33 the left uh you will see the growth of uh an internal fire so this is an internal fire that starts burning um in in the interior of the unit and is going to

01:35:46 interior of the unit and is going to deliver a heat flux an amount of heat per unit area to uh the structure now the structure obviously comprises of the ceiling the

01:35:57 obviously comprises of the ceiling the floor the internal walls but also comprises of the external system so the external system is also going to receive this solicitation now if you look at the

01:36:09 this solicitation now if you look at the the diagram above the photograph uh you will see more or less the magnitude of the heat that you receive so uh in the highest point which is in

01:36:20 so uh in the highest point which is in the center of where this fire is you will find that you get numbers of the order of 500 kilowatts uh per meter squared

01:36:28 squared now uh you could see 500 400 200 300 as you move away from the source of the fire now what this shows you is that the fire in the interior of the compartment

01:36:40 fire in the interior of the compartment tends to be much more onerous than the fire in the exterior of the compartment now normally for the wall the internal walls the ceiling and the floor we will

01:36:52 walls the ceiling and the floor we will address this through the process of fire resistance now from the perspective of the external components of the wall we will

01:37:03 we will not address this we will make the assumption that the external components will not require any fire resistance and therefore they are not addressing the solicitation from the interior now that

01:37:15 solicitation from the interior now that in itself is a limitation and a very very strong assumption that we're making because effectively it assumes that in the intersection between

01:37:26 between the external part and the internal part there will not be a potential gap that gets formed now this is this is a i think in many ways a legacy of the past because we used to have

01:37:38 of the past because we used to have systems

01:37:40 systems where

01:37:41 where you had a floor slab and we used to sit and hang components out of the floor slab

01:37:49 slab so inevitably by assigning a fire resistance to the slab the interior fire was protected and uh any solicitation that came in this

01:38:02 any solicitation that came in this direction any heat flux will have to breach a fire resistant floor before it could get to the compartment now that is going to change dramatically when you have a

01:38:14 to change dramatically when you have a system

01:38:16 system that is pinned by a connecting element that you're going to fill

01:38:22 with something in here because you're going to have the same solicitation but this time is going the performance is going to depend not only on the properties of the

01:38:35 depend not only on the properties of the infill

01:38:36 infill but how all these things move together so all of a sudden you have moved from component performance which is quite easy to establish to a system

01:38:47 quite easy to establish to a system performance where is the relationship between the building the external envelope and the connecting elements that will determine what will happen this by definition is a much more

01:38:58 this by definition is a much more complex system and that currently remains fully unaddressed so

01:39:05 so that will be the internal aspect now the external aspect as we can see here results from the acceptance that the building envelope will fail to the outside and will project a flame

01:39:18 the outside and will project a flame that because of buoyancy will bring flames and hot gases up and that's the second photograph that you see in the middle now

01:39:27 now the

01:39:29 the this particular projection of the flame can be characterized by a region where the flame is going to exist which is the flame length that i call lf and it's

01:39:40 flame length that i call lf and it's going to be the heat coming from this flame but it's going to come from the external solicitation now

01:39:48 now temperatures on the outside of a compartment tend to be much lower than the temperatures on the inside of the compartment because air tends to dilute the heat

01:39:58 the heat buoyancy brings a lot of air that cools down the

01:40:02 down the the smoke resulting in a much lower temperature and you can see from the plot that i have in here now the bottom part of the plot

01:40:13 now the bottom part of the plot represents the edge of the window and as we move up we're moving up away from the edge of the window towards the top of the flame

01:40:24 of the flame and we can see if we follow the red curve

01:40:27 curve we can see how we start at about 60 kilowatts per meter squared and we descend as we move away to values of the order of 10.

01:40:38 order of 10. so you can see that this solicitation is very different so now we have a much lower heat flux and that much lower heat flux is a decaying one that is going to basically

01:40:49 decaying one that is going to basically shrink as you move away so if you get far enough and this is a concept of protecting by means of a spandrel if you get far enough you will get to such low heat

01:41:00 enough you will get to such low heat fluxes that you might have the capacity of preventing ignition okay this is i can achieve that by putting a spandrel i can achieve that by putting a projection so basically i can

01:41:12 putting a projection so basically i can move my end point far enough so that i'm low enough now obviously as i depicted in this diagram that can be very far so it might be impractical and we need

01:41:24 so it might be impractical and we need to take that into consideration but nevertheless is something that we can play with

01:41:28 play with now i plotted two lines i plotted the red lines and i plotted the blue lines the dotted lines represent the range of values that were measured and the blue

01:41:39 values that were measured and the blue lines and the red lines represent different scenarios people have done many studies on external plumes and uh and the the heat fluxes and what you can see is that in

01:41:50 fluxes and what you can see is that in the worst case scenario you might reach a peak value that is not 50 or 60 as you have here but it might reach up to 100 or 120 so those are the greatest values

01:42:01 or 120 so those are the greatest values that people have measured so it starts at about 100 120 and then it will decay back into about 10. now

01:42:09 now depending on what materials you're using depending on the geometry of the compartment depending on the openings of the compartment you know the things will change the decay will be more pronounced or less pronounced the peak value will

01:42:20 or less pronounced the peak value will be higher will be lower but the physics is more or less the same now the additional component of the process is the external heat flux which is what i call qe

01:42:32 is what i call qe in in in the bottom now that's any heat that can come up from either another burning building as we can see here or he come up from another surface that is burning so once

01:42:44 another surface that is burning so once you get multiple ignitions you can have heat from one flame feeding onto the other which is what tends to happen once a fire develops that you start getting several areas that are burning and

01:42:55 several areas that are burning and they're cross feeding each other so in essence if you look at the diagram on on on the right you will see that you have two sources of energy the external heat flux that

01:43:07 of energy the external heat flux that comes from the outside and the heat flux that comes to the flame from the flame okay now if this sources of heat flux are capable of igniting the material

01:43:19 are capable of igniting the material then the flame will spread okay

01:43:22 okay now this is this is a very important aspect of the problem because uh the capability of a heat flux to ignite a material depends on two

01:43:35 to ignite a material depends on two fundamental variables the first is that here is the material here is your flame you have to have sufficient heat so this has to be big enough so that is

01:43:48 so this has to be big enough so that is capable of igniting a material just to give you an example uh if you put a desk in the sun it will not ignite because the heat flux from the sun is too low so it can warm

01:43:59 from the sun is too low so it can warm up but it will never reach a point where it starts degrading and igniting okay nevertheless if i make it big enough then

01:44:07 then it could potentially ignite now the flame is a physical space

01:44:14 space so that flame is going to provide heat only to a certain region so the flame cannot heat very far away it will only heat up a certain region

01:44:26 heat up a certain region and this is what i call lh so this is the heating length okay now this here is the burning length

01:44:38 so it's the area of the material that is already burning so this is already burning this is what is providing fuel so that the flame can continue to exist

01:44:50 so that the flame can continue to exist the flame now is combusting with the air that is feeding it and is providing a heat

01:44:56 heat that is heating up this heating length now what happens if the flame starts shrinking

01:45:05 shrinking to the point where the burning length is now

01:45:09 now the same size as the flame then the flame has nothing to heat and the system will stop spreading so the condition that tends to lead to

01:45:20 so the condition that tends to lead to no spread is a condition where lh is approximately equal to zero in other words my amount of fuel starts shrinking

01:45:31 shrinking so the flame becomes shorter to the point that it cannot heat anything beyond what the burning length is eventually this material that was

01:45:43 eventually this material that was burning will consume itself and once it consumes itself then it will stop burning now if while i have a heating length while lh

01:45:54 while lh is greater than zero in other words i have an area that is being heated there is a propensity for the flame to spread

01:46:03 spread now this condition of lh equals to zero okay is is is really a very simple representation of much more complicated problem but the in essence gives you the

01:46:14 problem but the in essence gives you the physical basis of the concept of no spread now you can achieve lh equals to zero in many different ways so for example you know you could potentially insulate

01:46:27 know you could potentially insulate to try to prevent the heat from getting there so you're effectively by other means achieving the same thing you can increase the ignition temperature to an infinite value by encapsulating

01:46:39 to an infinite value by encapsulating the material with a non-combustible material

01:46:43 material and you are achieving the same objective so you can play with different measures you know aluminum foil ceramic renders aluminum plates all these different ways

01:46:55 aluminum plates all these different ways to try to create mechanisms by which you are bringing effectively this value to zero the problem is that this is my very simple case of a single material where

01:47:07 simple case of a single material where lh only depends on the relationship between the amount of fuel and the heat flux that comes from the flame if i start creating concepts such as encapsulation i start introducing other

01:47:20 encapsulation i start introducing other variables like mechanical behavior and the moment i introduce mechanical behavior then it is not the fact that lh equals zero is that i am protecting the material by means of a mechanical system

01:47:33 material by means of a mechanical system in such a way that i have created an equivalent system that renders lh equals to zero and this is how i start building this complex form of behavior

01:47:44 this complex form of behavior so in principle the idea is that the way in which a fire spreads is that i produce fuel the fuel produces a flame the flame brings heat flux

01:47:55 flux through an area that is given by lh and once the temperature of this reaches its ignition temperature the flame jumps okay and it jumps a certain distance and

01:48:07 okay and it jumps a certain distance and that distance divided by the time to ignition will give me the velocity okay and i have a flame spread velocity now

01:48:16 now in essence the concept in itself is a concept that we've understood since the 1970s and uh there's a famous paper called controlling mechanisms of flame spread that explains all this and it's a

01:48:29 spread that explains all this and it's a paper that dates you know from from the late from the early 1980s that summarizes work since the 1970s now the application of this to the more complex systems is something that is a different

01:48:41 systems is something that is a different story that has again

01:48:44 again brought all sorts of other complexities that have made things a bit more difficult

01:48:49 difficult so

01:48:51 so in essence the objective if what we are aiming to is no spread is to find a way by which we do not

01:49:03 is to find a way by which we do not allow

01:49:04 allow a preheating length that can enable the ignition and spread of the material whatever mechanism we use whether it is natural when you have a

01:49:15 whether it is natural when you have a single material or whether you introduce barriers to try to prevent this from happening you know that is the objective for no spread now

01:49:24 now if the objective is to quantify spread then the problem is different we allow spread to happen and the only thing that we do is we measure

01:49:33 measure how fast it takes to jump this length so if it's going to jump what i call a delta lb so a piece of lb if it's going to jump this amount

01:49:44 if it's going to jump this amount then i need to quantify the time that it takes for it to jump so that i can look into how rapidly it's spreading so these are basically the performance objectives and it depends

01:49:57 objectives and it depends on your fire safety strategy which one is the one that you're going to pick if you have a stapled strategy and you're aiming at not spread you will be looking into

01:50:08 looking into target targeting your design to meet the

01:50:12 the no

01:50:13 no heating length condition while if you are going to allow for flame spread your objective is to know how fast

01:50:21 how fast and so that you can create the entire fire safety strategy around a quantifiable and acceptable level of flame spread so

01:50:31 so basically

01:50:34 the next step is to try to go back and look into what are the parameters that control this process okay and the diagram to the left shows

01:50:45 okay and the diagram to the left shows you a sequence of parameters that are going to influence

01:50:51 influence lh and the flame spread velocity so

01:50:55 so i am not going to go into the details of the different parameters but i do want to explain a little bit the sequence because they are related to the way in which we test

01:51:06 which we test so uh

01:51:09 so uh what we were discussing was that the material would produce fuel and that's the mf on the left hand side of the diagram uh the

01:51:21 uh the amount of fuel is going to depend on how big is the area that is burning and that is going to be the lp in other words the term on the top is how much fuel is

01:51:33 the term on the top is how much fuel is being produced and the the one on the bottom is how big is the area that is producing fuel the bigger the area the more fuel you're producing

01:51:43 producing the more combustible the material is the faster it's going to produce fuel so if you have for example a liquid fuel like kerosene the burning rate or this mf will be very large well if you have a fire heavy fire

01:51:56 well if you have a fire heavy fire retardant material the mf will be very little

01:51:59 little okay and uh but

01:52:02 but the second term uh the lp is going to grow in time because as the fire spreads you're gonna have a bigger and bigger and bigger and bigger and bigger area that uh that that burns so it is part of

01:52:15 that uh that that burns so it is part of what needs to be calculated so that is going to progress in time and i need to be able to calculate that to be able to understand how rapidly the fire is going to spread so

01:52:27 so the product of the area burning times the amount of fuel that is being produced per unit area will be the total muscle fuel and that is the next term mf without the double prime

01:52:39 mf without the double prime now you have now the fuel and the fuel will combust and it will have a heat of combustion and that's the energy that is released per unit mass of fuel

01:52:51 per unit mass of fuel so you have the two terms the heat of combustion and the amount of fuel and the two of them multiplied will give you the total heat release rate so that's the amount of energy that the

01:53:02 that's the amount of energy that the system is producing so on the top i have the amount of fuel that i produce per second and on the bottom i have the amount of energy that i produce per kilogram of fuel that i

01:53:13 i produce per kilogram of fuel that i generate

01:53:14 generate if i multiply the two of them i get the heat release rate and that's the value of q dot

01:53:20 of q dot so q dot is a typical representation that we use for the heat release rate now

01:53:28 the heat release rate in essence is one of the most important variables that we address in fire because in many ways

01:53:37 ways it is the first clear manifestation of what the fire is which is an amount of energy that is being released every unit time

01:53:45 unit time now that heat release rate is going to control the temperature of the flame and it's also going to control the temperature of the smoke so the two next terms that you have to the right will be

01:53:56 terms that you have to the right will be the temperature of the flame and the temperature of the smoke tf and tg and it's also going to control the length of the flame now this is very important because if

01:54:08 now this is very important because if you have a material that does not have a heat of combustion in other words it's non-combustible then your heat of combustion is zero

01:54:19 then your heat of combustion is zero there is no heat release rate and therefore you have no flame temperature and you have no flame length and the problem stops so the easiest way to prevent spread is by

01:54:31 is by eliminating combustion okay

01:54:34 okay now if you are going to allow combustion to happen

01:54:37 to happen then you have to continue now the process

01:54:40 process and now you have to quantify how that temperature of the flame and that temperature of the smoke is producing a heat flux and that heat flux is heating the material

01:54:52 flux is heating the material you also have to be able to quantify what external sources you might have and that's the term on the top the blue term and you also have to be able to quantify

01:55:03 and you also have to be able to quantify how the air by flowing because it is fresh air is cooling the whole process but also is driving the geometry of the fire

01:55:13 fire so this is why for example many times we choose certain geometries in the test you know why do we have the room corner test

01:55:21 test we have the room corner test because that allows the air to come in a certain way that produces the longest possible flame

01:55:29 flame and that is considered to be the most onerous case you know in bs8414 we have a corner why do we have a corner for exactly the same reason the air comes in a certain way that produces a flame that is the

01:55:41 way that produces a flame that is the most elongated and therefore is considered the most onerous case so i cannot ignore how the air comes in because the air is going to determine the geometry of the flame and therefore

01:55:54 the geometry of the flame and therefore is going to be critical in me being able to determine if i have an lh or not okay

01:56:02 okay now

01:56:04 now the next row are the fundamental parameters that will control these things these are properties of the material and uh and there's no need in going into detail other than just understanding

01:56:15 detail other than just understanding that these properties are going to be controlling all these things

01:56:20 things so the the fee the parameter fee is what determines uh the potential of the flame to radiate heat

01:56:29 heat the parameter t infinity is the ambient temperature that's the simple one t ignition will be the ignition temperature of the material the higher the ignition temperature is the more difficult it is to ignite the material

01:56:41 difficult it is to ignite the material the lower the ignition temperature the easier it is to ignite the material and k multiplied by rho multiplied by c is the thermal properties and that's what we call the thermal inertia

01:56:52 we call the thermal inertia a material that has a very high thermal inertia is very difficult to ignite and to spread a material that has very low

01:57:00 very low thermal inertia is very easy to ignite so if i combine all the parameters up to lh

01:57:07 up to lh with all the material properties to the right i can eventually estimate the flame spread so as you can see

01:57:18 then that flame spread and that's the dotted line will deliver how much area of the material is burning and the faster your flame spread the faster it grows and the

01:57:30 faster it grows and the more

01:57:31 more exponential growth you get for the whole process okay so they're linked so there's a feedback loop in this whole thing so

01:57:40 as you can see this is a complicated problem

01:57:43 problem and uh and we need to be able to try to address this problem in ways that are tractable

01:57:50 tractable but nevertheless to be able to address these problems in ways that are attractable the first thing that we need to do is to understand it once we understand the problem then we can try to find a simpler

01:58:02 then we can try to find a simpler formulation okay that's the objective when we go from

01:58:08 from the knowledge base into something that is practical or viable for people to be able to design and to be able to work you know that has to be based on the fundamental

01:58:21 to be based on the fundamental understanding so we need to take this problem in its full complexity and try to understand it and once we understand it we can come up with simpler methods of assessment okay

01:58:31 okay now

01:58:35 as i said i just touched on the complexity of the problem and

01:58:41 and as you can see from the photograph on the right

01:58:44 the right there is a photograph of the windsor tower in madrid and the reason why i picked that photograph is because if you look at the columns you will see the deformations induced by buckling

01:58:58 the deformations induced by buckling so you can see the magnitude of the deformations that you can have because this materials these railings are being heated

01:59:08 heated so we cannot trivialize to what extent mechanical deformations are fundamental in the behavior of the systems

01:59:17 systems and and they need to be considered at every stage in the assessment of a building envelope these are not simple deformations these are massive deformations that can create enormous

01:59:28 deformations that can create enormous gaps and they can create the possibility for penetration now

01:59:34 now needless to say on the left hand side i extracted this from a phase one report you will see when we start adding all these components and we have insulation you know we have

01:59:45 you know we have a composite system of aluminum and combustible polyethylene material you introduce a cavity in the middle you start creating all sorts of different

01:59:56 start creating all sorts of different mechanisms that complicate the problem much much much further so our need to understand gets enhanced now this is a problem that

02:00:09 gets enhanced now this is a problem that i feel is very common that the moment that we address a complex system in the area of fire our tendency is to naturally resort

02:00:21 tendency is to naturally resort to

02:00:22 to let's do a big burn so if we don't understand we burn

02:00:28 we burn and uh and to me this is really really problematic because uh

02:00:35 extracting proper information from a large-scale test requires not only an enormous amount of proper measurements requires a very clear understanding of

02:00:47 requires a very clear understanding of the objectives and the way in which the test has been set up but it also requires an enormous amount of competency to be able to interpret the results

02:00:59 the results and

02:01:01 and doing a large scale test because i don't understand cannot be

02:01:07 cannot be the process to be followed you understand you formulate the appropriate test and if you need a large-scale test to fill the gaps of system behavior so you get the

02:01:18 system behavior so you get the information that you need then you implement the appropriate measurements and you deliver the test with the appropriate results that fill the gaps that you need but we cannot

02:01:30 the gaps that you need but we cannot use the large-scale test as a mechanism to compensate for our ignorance or incompetence because all we're doing is creating a bigger problem that we are going to understand even

02:01:41 that we are going to understand even less

02:01:42 less so

02:01:45 the forms of encapsulation will vary you have fully encapsulated systems as you will see in in this case of a you have partially encapsulated systems as you can see in b

02:01:56 encapsulated systems as you can see in b where the top and bottom are left open or you can have one side encapsulation as you can see in c and all these systems will behave in a different uh manner so

02:02:08 different uh manner so all these mechanisms are happening all these mechanisms need to be taken into account and all these mechanisms are part of understanding the problem now

02:02:19 now let me if we can run this video that will give you an example of the one of the most important aspects of understanding is when we create a complex system and this is an aluminum

02:02:32 complex system and this is an aluminum composite panel you can see it clamped on this is the aluminum composite panel is clamped on one end and this is a heater that is providing a

02:02:43 and this is a heater that is providing a heat flux

02:02:44 heat flux against the aluminum composite panel um is it running

02:02:52 so the moment that we remove the the cover it starts heating up and what you need to observe is the failure mode the failure mode of this will be at some

02:03:03 the failure mode of this will be at some point

02:03:04 point the front

02:03:05 the front aluminum panel will actually split and we will stop the video at the moment it splits so you can see what happens the moment it splits then the aluminum is no longer attached

02:03:18 then the aluminum is no longer attached to the polyethylene is allowing the polyethylene to be exposed which is effectively enabling the existence of lh while the aluminum is attached

02:03:29 while the aluminum is attached to the plate the aluminum is basically taking the heat

02:03:35 heat and therefore it's going to happen um in here so as you can see if you look carefully at this area here you will see how the aluminum plate has

02:03:47 you will see how the aluminum plate has split

02:03:48 split and the moment the aluminum plate has split if you let it run again

02:03:53 very rapidly you see how the smoke starts being produced because now is the polyethylene that is being heated and and and quite rapidly you will see that ignition follows

02:04:22 and once the ignition follows it will continue to burn and restraint but as you can see this is the type of understanding that is fundamental you start by looking what

02:04:33 fundamental you start by looking what are the failure modes you try to understand what those failure modes are you try to see if there's ways by which you can prevent them and only then if you realize that there's some system elements that you cannot describe

02:04:47 system elements that you cannot describe with these tests you progress to the next level of the problem so if you um if you look at the details you know basically what is described in this

02:04:58 basically what is described in this diagram is exactly that the aluminum composite panel has the two aluminium plates attached to the polyethylene and while they're attached they're protecting the polyethylene

02:05:10 they're protecting the polyethylene infill

02:05:11 infill as soon as the gap gets created and it opens enough to allow for the surface to heat

02:05:18 heat then immediately what you're going to get is the propensity for ignition you are effectively going from a condition of lh equals to zero because the aluminum has eliminated to a condition where lh is the entire surface

02:05:31 condition where lh is the entire surface of the polyethylene so

02:05:34 so as i say if you look at figure b this is a different form of the same thing so this is a ceramic rendered combustible insulation and you can see how you can achieve fundamentally the

02:05:45 how you can achieve fundamentally the same thing while the ceramic render uh is intact the the insulation will not burn that's the image to the left but as soon as it cracks and opens then the

02:05:57 soon as it cracks and opens then the insulation ignites and effectively what you get is complete combustion of the insulation so there's many ways in which we can achieve this and all this need to be understood but if you look at the

02:06:08 be understood but if you look at the forms of mechanical behavior and mechanical failure the figure to the left with aluminum composite panels is the split

02:06:16 the split that is the failure mode well in the other case is the actual failure of the mechanics of the ceramic rendering these are two completely different

02:06:26 different failure modes that are induced by the same heat

02:06:30 same heat but they're behaving in completely different ways so we need to understand these things before we can look into the performance of these materials so to me this is what is fundamental and

02:06:43 so to me this is what is fundamental and i think this is one of those things that in fire safety we have completely forgotten in any other area of engineering before we can produce a product whether it is a

02:06:54 can produce a product whether it is a mobile phone you know a satellite or a car we need to demonstrate that the product delivers the functionality it was designed for so if if a car doesn't run nobody is

02:07:06 so if if a car doesn't run nobody is going to buy it you know or if a phone doesn't connect you nobody is going to buy the phone so we have a very clear relationship between product development and functionality and the same way when

02:07:18 functionality and the same way when we're talking about copied vaccines you know it is a whole demonstration of performance that enables us to actually put the vaccine in the market and we spend millions

02:07:30 spend millions in product development now when it comes to fire we forget that fire safety is a functionality and that functionality needs to be

02:07:41 and that functionality needs to be assessed

02:07:43 assessed and that requires product development and it requires an investment from the manufacturers in delivering the appropriate functionality for fire

02:07:54 appropriate functionality for fire this is completely different that delivering all the other functionalities that you need to deliver acoustics thermal aesthetic everything and then see

02:08:04 see if i can pass the test it's a completely different process and i think we need to look at products that are supposed to deliver

02:08:16 at products that are supposed to deliver a fire safety functionality as products that need to be studied developed and assessed before they're even introduced into the market

02:08:26 market and and to me that is a really really important thing because in a way moves us away from this idea that

02:08:34 that fire safety performance assessment should always be done the fastest the cheapest and the simplest way why is this functionality that is so critically important for society

02:08:46 critically important for society so ignored what we're spending millions in making sure that the color of the render is actually the right color i mean these are the kinds of things that are difficult to understand but we

02:08:59 that are difficult to understand but we really need to be clear

02:09:02 clear that fire safety is a performance that needs

02:09:06 needs you know to be assessed as a functionality it's one of the many functionalities that these multi-purpose materials or multi-purpose systems have so

02:09:17 so if you think in those terms then you start thinking there in

02:09:23 in first understanding the performance of your components then

02:09:27 then bringing you to system behavior and start seeing how the different systems interact what you see in this diagram is basically a concrete slab with a connection element with the

02:09:39 connection element with the fixation railings the glazing all these components are going to deform and move and interact in a completely uh different way and they need to be addressed in a

02:09:50 and they need to be addressed in a correct and comprehensive manner now obviously

02:09:55 when we address this complex system behavior

02:09:58 behavior we get to a position in which we might require large-scale tests but nevertheless a large-scale test is never going to be

02:10:09 a large-scale test is never going to be reality

02:10:10 reality a large scale test will always remain a test and that test is going to provide us with the information that we need to be able to make an assessment so we can criticize enormously the

02:10:22 so we can criticize enormously the concept of a desktop study but what a desktop study is is inherently engineering we take all the information that we have and we make an extrapolation that

02:10:33 and we make an extrapolation that delivers the required performance we demonstrate it in a competent way we explicitly state our assumptions our dependencies our conflicts you know and

02:10:45 dependencies our conflicts you know and we explicitly justify everything we did once we've done that we have done a rigorous extrapolation that gives me confidence that the system that i tested to gain

02:10:57 that the system that i tested to gain information

02:11:00 analyzed through a desktop study enables me to confidently establish that i have met the performance criteria and again i want to insist that that has

02:11:11 and again i want to insist that that has to be explicit and it has to include not only the thermal flammability combustibility behavior but also the thermo mechanical behavior of the system because we cannot ignore the

02:11:22 because we cannot ignore the deformations so you can have a perfectly non-combustible material that is an extreme hazard because it deforms and breaks apart very rapidly okay so

02:11:34 okay so in essence i'm not going to go into great detail of this because i know professor bisbee is going to talk in in great detail but in essence uh the tests that were conducted

02:11:45 essence uh the tests that were conducted at university of edinburgh are an example

02:11:49 example of how you look into the different failure modes to try to give answers to very specific questions that the chairman had asked from professor bisby so

02:12:02 bisby so a test was very carefully developed

02:12:06 developed on the basis of an understanding of the system but also on the basis of a set of preliminary tests that informed professor bisbee of the different

02:12:18 professor bisbee of the different failure modes of the characteristic behaviors

02:12:21 behaviors that can be observed and those set of very small scale tests with all that information allow him to progress to the definition and the instrumentation of a

02:12:33 definition and the instrumentation of a test of a higher level of complexity you can call this an intermediate sort of scale test but it was sufficient to answer the questions that were being

02:12:45 to answer the questions that were being asked

02:12:46 asked you know the intention of the tests were never to develop a

02:12:51 a design solution the intention of the test was to provide answers to very specific questions that had been asked now that particular test provided the

02:13:02 now that particular test provided the necessary informations you know that enabled us to be able to with great certainty answer those questions in a very precise manner

02:13:13 manner so

02:13:14 so the focus of the tests it was in trying to understand the the mechanism by which the different components of the system enabled the progression

02:13:27 of the system enabled the progression from

02:13:28 from a system that had an lh equals to zero to a system that had an lh that was greater than zero and therefore enabled the uncontrolled vertical fire spread

02:13:40 the uncontrolled vertical fire spread and what you see from the the figure on on the bottom left in um

02:13:47 in um in the small box you will see that the it's it's just a the the full set of data and you can see how eventually it takes off and it burns at a very large

02:13:58 takes off and it burns at a very large heat release rate so we are

02:14:02 developing a series of tests that enables us to see the failure so you can see it has failed now the question is what was the process or the sequence that followed and what

02:14:14 or the sequence that followed and what were the different elements that influenced that sequence so

02:14:20 so uh you're going to have the basically separation as i showed that first manifest itself by a piece of the system falling off and that's the first

02:14:31 the first little peak that you see in here you have before that you have melting and dripping that produces a flame that is sustaining the process independent of

02:14:43 is sustaining the process independent of the presence of an igniter and then eventually you're going to get the full

02:14:48 the full failure or the full separation of the panel that happens at this point here that is going to lead to the takeoff of the fire

02:14:59 to the takeoff of the fire is once the whole system splits that you have the fire taking off now the mechanisms that lead to that are a combination of a number of different things and as i say professor brisby is

02:15:11 things and as i say professor brisby is going to touch on this in detail so i will not get into the in into the details

02:15:16 details but in essence what you see in the images to the right is how a system where effectively the fire is not spreading on the first two images that you have

02:15:28 on the first two images that you have here is stationary and it's only burning because dripping polyethylene is sustaining a little flame all of a sudden takes off and the moment it takes off it becomes

02:15:39 and the moment it takes off it becomes uncontrollable so

02:15:44 understanding failure modes many times is achieved by conducting small scale experimentation that allows us to understand the failure modes and enables us to design better systems

02:15:56 enables us to design better systems so if we understand that it is that separation that actually matters then we can introduce mechanisms to prevent it and if we still recognize that probably the melting and dripping and burning is

02:16:09 the melting and dripping and burning is still too much of a hazard even though this thing might not separate we can actually change the infill for something that doesn't melt so we can start playing with different things to improve the product until we

02:16:21 things to improve the product until we achieve the functionality now obviously the small scale experiments will support larger scale experimentation because that information then can be used to

02:16:32 that information then can be used to design and implement a larger scale test that eventually brings us to understand you know the full system behavior so i am not

02:16:42 i am not saying

02:16:43 saying that we have to rely on small-scale tests

02:16:47 tests i'm saying that small-scale tests are a fundamental and essential part of gaining the necessary understanding so that we can then progress to identify

02:16:58 so that we can then progress to identify what variables are missing that need a larger scale to be able to assess them

02:17:08 for a fire safety strategy that supports a staple strategy we need to now think very carefully what is the target so obviously there is the condition of no fire spread

02:17:20 no fire spread but we can also allow a condition of no fire spread until the brigade can enact any an eager strategy so that is a perfectly possible option but it requires creating that

02:17:33 option but it requires creating that option and creating it explicitly now we recognize that the no fire spread condition is a very difficult condition to attain

02:17:41 to attain and it requires a significant number of constraints that might result in many of the current systems that are very effective for other functionalities having to be simplified

02:17:53 having to be simplified either because the performance cannot be achieved or because we don't have the knowledge base to be able to predict all the different behaviors so either way you know we have to understand that this

02:18:05 you know we have to understand that this is a complicated problem but nevertheless it is a potential performance target but if we know that we cannot meet it then we have to find ways like for example

02:18:16 we have to find ways like for example what i stated there to try to be able to establish

02:18:22 another mechanism by which we support the fire safety strategy to make sure that we allow for some level of flame spread in which case

02:18:33 spread in which case you know we have to be able to quantify that flame spread so

02:18:39 so as i said the lh equals to zero condition is very difficult to you know to characterize and uh and the other thing that is very difficult to characterize is uh when you are standing outside a building knowing

02:18:52 are standing outside a building knowing when that lh equals to zero condition has been breached so that expectation from the fire brigade is one that needs to be looked upon in a very careful way so if we are going to expect the fire

02:19:03 so if we are going to expect the fire brigades to identify when a departure is necessary then we need to make sure that we understand when it happens how it manifests itself and that we communicate that in an

02:19:14 and that we communicate that in an explicit and effective way so

02:19:20 in the case of a fire safety strategy that supports a controlled evacuation strategy basically what we're targeting is an allowable rate of spread and we should be able to quantify the

02:19:32 and we should be able to quantify the the rate of spread and and basically if we can quantify that then we can design all the alternative provisions to make sure that our egress times still remain much much smaller than our

02:19:45 still remain much much smaller than our available time okay that's the objective so

02:19:53 now we know what the targets are and and we need to also understand the difference between conformity and compliance

02:20:04 compliance so

02:20:05 so a building fulfills a fire safety function

02:20:08 function and as such the building has to be compliant

02:20:12 compliant the whole process to the left is a process by which the manufacturers the people that assemble systems people that assemble systems into the

02:20:24 people that assemble systems into the building the construction all the different

02:20:28 different uh

02:20:29 uh stakeholders in the system from manufacturers to designers to builders

02:20:35 builders at every stage we need to be able to demonstrate that and provide the necessary information that shows that

02:20:46 that the parameters that we're using to quantify our performance are being attained so if i state that to be able to design

02:20:57 that to be able to design a building with a particular eager strategy that requires a particular flame spread velocity i should be able to provide all the information so that somebody can verify

02:21:10 information so that somebody can verify and assess the conformity of that information and make sure that that flame spread velocity is what the material delivers so in essence

02:21:21 so in essence conformity for me is providing in a sort of in a competent and

02:21:27 and honest manner is providing all the information that you need to be able to establish that what you are receiving is exactly

02:21:38 is exactly what you need so you have calculations those calculations depend on certain parameters depend on certain quantity quantified information and that

02:21:49 quantified information and that information needs to be verified and if that information is properly verified then you're confident that your calculation is appropriate now obviously the verification goes all the way through to the delivery of the project

02:22:01 through to the delivery of the project because obviously it is not the same thing to design something than to actually build it you know and implement it because things can change you know things can be modified or things can not

02:22:13 things can be modified or things can not be done at a level of rigor that is not appropriate and somebody has to be able to establish that it still conforms with the expectations that you have of the performance of that system

02:22:25 the performance of that system so to me

02:22:26 so to me it is extremely important that in the process of understanding the behavior of of an external building envelope it is essential to understand

02:22:38 envelope it is essential to understand that there's multiple levels of conformity that needs to be met before

02:22:43 before we get to the building and we establish that the building is compliant and uh i touched on this in in in great detail in in my report because i do

02:22:54 detail in in my report because i do believe that there is some element of confusion and many times we talk about products being compliant but the reality it is that the building is the one that is compliant because

02:23:05 is the one that is compliant because it's the building that delivers the fire safety

02:23:07 safety um strategy this fire safety strategy is the one that is compliant and the products are just components that fit there and there has to be a professional

02:23:18 there and there has to be a professional with sufficient competency to amalgamate all that information and deliver a compliant building and from my perspective that is the role of the fire safety engineer okay so

02:23:30 okay so in essence we have the performance objectives and we have a clear separation between conformity and compliance

02:23:42 conformity and compliance and and again that brings us back you know to to the original question of what is a safe building but now what i'm going to do is quite rapidly look at this through the perspective of approved document b and

02:23:54 perspective of approved document b and our current building regulations so that that finally leaves me with uh the tests and why they answer the questions or not so

02:24:04 so in a proof document b we always talk about the five functional requirements and in the case of external fire spread what we say is that the external walls of the building should adequately resist

02:24:15 of the building should adequately resist the spread of a fire over the walls and from one building to another having regard to the height use and position of the building in other words the external fire spread or the external

02:24:26 the external fire spread or the external walls of a building have to support the fire safety strategy okay and

02:24:32 okay and but approved document b has no explicit objectives instead what they do is they have five functional requirements now those five functional requirements

02:24:43 now those five functional requirements you can consider them as the pillars of the fire safety strategy but the reality is that they're just components that need to be stitched by a competent professional that has to be

02:24:55 competent professional that has to be able to marry them in a way such that they actually work together and this is why when you have a state put strategy it is the staple strategy that defines the use and position of the

02:25:07 that defines the use and position of the building and therefore it requires you know for b4 to be no spread unless you put alternative provisions

02:25:18 unless you put alternative provisions so all these things are the

02:25:22 the space in which a competent professional should be able to operate now and and as you can see the space is quite significant because the reality is that b1 to b5 are only the tip of a very

02:25:35 that b1 to b5 are only the tip of a very very large iceberg you know of assumptions dependencies and and conflicts so we need to be able to engage with these things in an appropriate way now

02:25:46 now the compliance is compliance with the fire safety strategy

02:25:51 strategy so it is the fire safety strategy that is compliant with a set of objectives that unfortunately remain undefined and what is a safe building

02:26:03 and what is a safe building is the question that still remains undefined nevertheless we have to show compliance against that performance criteria so

02:26:13 so in many ways you're almost saying that you have to assess compliance to the comfort you know of the competent professional which is a very tricky proposition

02:26:25 which is a very tricky proposition because it puts an enormous onus on competency and therefore an enormous need for properly regulating competency so

02:26:34 so then it sits the fire safety strategy sits on all these pillars and these pillars then

02:26:42 then have behind them a series of potential approaches okay and we're all very familiar with those potential approaches you've seen those diagrams again and again and again

02:26:55 those diagrams again and again and again but they're just potential approaches nevertheless despite the fact that there are potential approaches that are based on major assumptions that have major limitations that hide complexities that

02:27:08 limitations that hide complexities that high dependencies that hide conflicts

02:27:14 we have to extrapolate those

02:27:19 those guidance

02:27:20 guidance into something that enables us to define a comparative level of safety and

02:27:28 and interpretations as you can imagine can be very broad and therefore very very problematic so in essence without an overall objective for the fire safety strategy

02:27:40 for the fire safety strategy there's very few mechanisms other than professional competency that enable us to be able to establish what is a safe building

02:27:51 what is a safe building so

02:27:57 if we cannot establish what a safe building is then how can we link this to the physics of the problem you know here we have a very complex physics with a lot of different

02:28:09 physics with a lot of different variables

02:28:11 variables and we have to be able to establish that this building envelope provides me a mechanism to support the fire safety strategy so that it delivers a set of

02:28:24 strategy so that it delivers a set of unknown objectives so for a test to be useful we have to clearly establish the phenomena that needs to be quantified and the tolerable error and uncertainty

02:28:36 and the tolerable error and uncertainty we can do that so the physics is there it might be very complicated and we might not be able to do it for the most complex systems but we could actually do it for a number of different systems and

02:28:47 it for a number of different systems and we can actually do it right we can establish errors and we can establish that we have confidence that we've put enough safety factors to be able to be in a safe place so that we can do

02:28:59 in a safe place so that we can do you know we we have to understand the relationships that govern the phenomena we do that we understand the relationships you know we know what is a flow in a cavity and we can understand what a flow in a cavity is you know we

02:29:11 what a flow in a cavity is you know we know how to stop it you know we have tools to do these things and we can even set targets like lh equals to zero now

02:29:24 the next step is that we have to extract from these relationships that the parameters that influence the phenomena and here's a list you have in this box all the parameters that influence the phenomena

02:29:35 phenomena you have to determine the measurements necessary to define and infer the parameters professor bisbee did that with the tests that he conducted so you actually can do it you know and you can you have to

02:29:46 you know and you can you have to demonstrate that all inferences are unique and you understand all your errors and uncertainties we know how to do this so the physics is there it has been studied it has been recognized and

02:29:58 been studied it has been recognized and we know how to do it which is a very different proposition is are we competent enough to apply this information nevertheless

02:30:08 the problem is that if we don't have a clear target even then we can be as competent as we want

02:30:15 want and we really don't know exactly how to use this information now

02:30:20 now instead

02:30:21 instead we resort to surrogates and i talked about maximum increased distances i talked about velocities these are all surrogates tests are surrogates performance in a test becomes a

02:30:33 performance in a test becomes a surrogate and if we don't understand all the assumptions behind them it is very difficult to relate the surrogate to the real performance variable that we're interested in now our tests

02:30:44 now our tests are currently the existing tests can be input tests which are basically a test that i conduct to extract a parameter maybe two three multiple parameters but

02:30:55 maybe two three multiple parameters but i want the test purely to extract information that then i can input into a model

02:31:02 model and it is the model that gives me the performance that's what we call first principles design so i can construct a mathematical formulation that describes what is the flame spread velocity

02:31:14 what is the flame spread velocity but that mathematical formulation requires the t ignition it requires the thermal inertia it requires the parameter fee and i'm going to use tests to assess

02:31:26 to assess what is the value of phi what is the value of the thermal inertia and what is the value of the ignition temperature then i input them in the model and i can calculate the flame spread okay so that's an input test

02:31:38 that's an input test a bounding test is the fire resistance test where you find what is the worst worst worst possible scenario the hottest fire for the longest duration and if my system survives that i can

02:31:50 and if my system survives that i can have certainty and confidence that it will survive any other fire that's a bounding test so i've created a worst case condition that allows me to bound the problem

02:32:02 bound the problem and then there's the scenario tests the scenario tests are the most complex set of tests because you need to be able to first

02:32:11 to first establish that the scenario is representative of the condition that you want to study then you need to be able to establish that that test is going to provide you information that is valuable for you to

02:32:22 information that is valuable for you to be able to extrapolate performance and then finally you need to be able to make sure that you are making the measurements and the analysis that is appropriate for that extrapolation

02:32:33 appropriate for that extrapolation so scenario tests are are focused on the idea of reproducing reality or some form of reality that can be extrapolated to reality to the real reality

02:32:47 now a typical example of an input test is the one followed by the linear route or the difference between non-combustible and combustible so if i

02:32:58 so if i conduct a test and that test determines that the heat of combustion is so small that it can be neglected and therefore the material will not burn

02:33:11 and therefore the material will not burn then

02:33:12 then the rest of the problem as you can see from this diagram disappears okay that particular problem now is limited to the fact of establishing this

02:33:25 limited to the fact of establishing this and the problem that i have solved is the problem of combustibility so in in other words it is not going to burn that doesn't mean that the system is going to perform

02:33:36 mean that the system is going to perform correctly all that that means is that that material has a heat of combustion that can be neglected okay so that's an input so if i have a model

02:33:47 the model becomes in many ways useless because the key parameter is zero and you multiply anything by zero and it can give you zero so in principle i don't need a model it's very simple if i want to assess compostability i

02:33:59 if i want to assess compostability i have a simple test that is an input test that gives me the heat of combustion so it's a perfect example of how you can take some information and

02:34:09 and extract something that is of value now

02:34:13 now this is where the confusion comes in because then you use terminology like limited combustibility limited combustibility requires to create a threshold

02:34:24 create a threshold okay but the moment that the heat of combustion is not zero you cannot cross anything

02:34:33 anything what happens is that the problem now exists and unless you analyze the entire problem and come into a position whether where you either establish that the flame spread velocity is adequate

02:34:45 the flame spread velocity is adequate or lh is equal to zero

02:34:49 you cannot claim that because it's limited combustibility it will deliver the performance that you need you still have to do the full analysis because at the end your performance is

02:35:00 because at the end your performance is defined by lh equals to zero and the flames travelocity so if you don't do all the analysis limited combustibility by itself has no meaning

02:35:10 meaning now if somebody has done the analysis and establish a clear threshold and that clear threshold has been demonstrated then you can simplify the problem

02:35:22 then you can simplify the problem and then classify it as limited combustibility and deliver a simpler version of an analysis of something that we already understand

02:35:35 something that we already understand now in the absence of that analysis that cannot be done and in essence that analysis unfortunately when you have complex

02:35:47 unfortunately when you have complex systems

02:35:48 systems becomes part of the design process so this is a perfect example of an input test that can deliver you something that

02:35:59 test that can deliver you something that can be tweaked in a manner that gives something that is potentially in the absence of an analysis not satisfactory so

02:36:09 so class zero is is a very similar example

02:36:14 example because it is really not an appropriate approach to characterize what the test aims to characterize so this is a test that aims to characterize surface fire spread or surface flame spread

02:36:27 surface flame spread nevertheless the whole configuration of the test does not reproduce the system behavior does not reproduce the failure modes does not have any of the characteristics that enables me to

02:36:39 characteristics that enables me to understand when lh stops being zero if what i'm testing is an aluminum plate that aluminum plate will always give me lh equals to zero

02:36:51 lh equals to zero but if i put polyethylene behind an aluminum plate then what i'm interested is when it splits that's the failure mode that matters you know not the result

02:37:03 you know not the result of

02:37:04 of a test

02:37:05 a test that effectively is testing the aluminum and is not testing the system behavior so

02:37:13 again the combination of the two standards uh set an arbitrary threshold for the heat of combustion that has no relationship with what follows after

02:37:25 with what follows after because once i have heat of combustion i have to calculate everything and the flame spread test doesn't bear no uh reflection of the failure mode that i need to understand so in

02:37:36 that i need to understand so in principle that test provides me no information it's meant you know to provide me information and i could instrument a test that is very similar i could change the way in which

02:37:48 similar i could change the way in which i arranged the frame i could do a lot of different things to improve that test and obtain valuable information but used in a standardized way

02:37:59 in a standardized way effectively for this type of products doesn't give me the information that i need at all so it doesn't give me any uh information on how to establish lh and it gives me no information on how to

02:38:11 and it gives me no information on how to establish a flame spread velocity

02:38:16 now finally is bs 8414 and br135 this is a scenario test you know and we can criticize the scenario all we want you know we can

02:38:28 scenario all we want you know we can criticize the way in which the system is you know uh assembled all we want nevertheless is a test that attempts to reproduce a

02:38:39 is a test that attempts to reproduce a scenario

02:38:40 scenario and as br 135 establishes it leaves it to the interpretation of the competent professional so br 135 does not establish a passing

02:38:52 so br 135 does not establish a passing criteria

02:38:53 criteria b r135 establish failure criteria beyond which you know there's no hope so if that happens then you don't want to use this material and we have the right

02:39:05 right you know to cut the tail okay that's the right that we have but beyond that the test is just the test now

02:39:15 now there is no

02:39:18 assessment that comes with the test that is up to the competent professional so

02:39:25 so the competent professional has to determine what the objectives of the test are because without really having a clear objective it's just a test for the sake of running a test

02:39:36 sake of running a test so what is the information that they want to have you know if they want to understand how cavity barriers work then they need to set the test in a way such that they gather information of how the cavity

02:39:47 gather information of how the cavity barriers work if they want to understand how the insulation contributes to the burning you know of the external cladding then they have to provide the appropriate information to be able to understand that maybe it's an analysis

02:39:59 understand that maybe it's an analysis of the the system you know but there has to be an explicit analysis and there has to be explicit data and information and instrumentation

02:40:10 data and information and instrumentation so

02:40:12 so before conducting a bsa414 it has to be very very clear what the objectives of the test are and uh and how to embed that concept in what is the

02:40:23 how to embed that concept in what is the risk that is represented by flame spread you know it could be surface it could be cavities you know all the statements are there

02:40:30 there in br135 nevertheless the statement that is missing is the role of the competent professional in determining all these

02:40:41 all these details

02:40:43 details the test as it is it is purely just a test

02:40:47 test now as i said the test doesn't have passing criteria it just has failure conditions and that's perfectly fine you can create failure conditions as a mechanism to

02:40:58 failure conditions as a mechanism to eliminate a certain tail that you deem from the onset no matter what the objectives are as unacceptable but the fact that you don't fail according to the failure criteria doesn't mean that you have met the

02:41:11 doesn't mean that you have met the performance criteria associated to your particular fire safety strategy for your particular design that is left to the competent professional to establish so we have this failure criteria and i

02:41:22 so we have this failure criteria and i have no objections you know to the failure criteria other than the fact that mechanical performance is treated in such a summary way that in principle leaves a lot of space for potential

02:41:34 leaves a lot of space for potential failure modes that are not captured at all in a test that is deemed to be a system test so if it's a system test then you need to incorporate all the potential failure modes and this is where the whole

02:41:46 modes and this is where the whole sequence comes in if i have done a proper

02:41:50 proper product development trying to understand the functionality and how it works and i have progressed that from the small scale through an intermediate scale until i have confidence that my system

02:42:01 until i have confidence that my system is going to operate i'm going to know exactly what i want to get out from the large scale test if i have not done any of this background work there's absolutely no way that a

02:42:14 work there's absolutely no way that a large-scale test that is very ill-defined because you don't know exactly what you're looking for and you don't know exactly what you're measuring you know it's going to provide you the information that that you need

02:42:27 information that that you need so

02:42:28 so at the end it is about incompleteness of the information you know the facade system might not reach the br 135 failure criteria and yet still be entirely incompatible with

02:42:40 yet still be entirely incompatible with the fire safety strategy of of a residential building with a stay put policy in place and and that i think is something that i would like to make absolutely clear

02:42:51 would like to make absolutely clear that that uh that the fact that it doesn't reach the failure criteria doesn't mean that the system in its application meets the performance criteria you know the information provided by the

02:43:03 you know the information provided by the thermocouple arrays is very very limited and uh and i i i don't need to go through the details all this is in in in my report and uh but at the end this

02:43:15 my report and uh but at the end this test

02:43:16 test that we need to have that enables me to establish that the flame is not going to spread and it's going to support the staple strategy or is going to spread at a certain rate

02:43:28 spread at a certain rate that is going to support an evacuation strategy is not information that is collected in the test you can argue that by looking at two data points you can establish some sort

02:43:41 data points you can establish some sort of flame spread velocity but the reality is that if you are to do that

02:43:47 that then you are accepting that if i put 10 more thermocouples and i actually measure the proper flame spread rate and i can fit it to an appropriate model that can allow me to

02:43:58 appropriate model that can allow me to predict what happens after

02:44:01 after you exceed the height of the test wouldn't that be much better so in essence we have a test in this minimum expression with this minimum

02:44:12 minimum expression with this minimum amount of data and we are arguing that we can extract information that is so much more refined because it depends on so many variables that requires a much more

02:44:23 variables that requires a much more extensive set of measurements then why not just do them you know why do we need to stick to the minimum expression when the risk is so significant so to me this is a really

02:44:35 significant so to me this is a really important aspect and in both the case of the staple strategy and the case of the evacuation strategy

02:44:45 strategy the information that you extract from br135 is very very limited

02:44:53 at the end there is another aspect that needs to be addressed

02:44:57 addressed and and you will find out and uh i put a citation in there that is great controversy on whether this scenario is an appropriate scenario or not

02:45:08 not here we have three examples of very similar tests uh the left-hand side is bsa414

02:45:14 bsa414 sorry

02:45:15 sorry the number is is wrong there's

02:45:19 there's nfpa285 and the fm global test and you will see in the figure on the left that each test has a very different description of what is the heat flux

02:45:32 description of what is the heat flux that you need to apply so whether we are applying an appropriate scenario is still a matter for debate and i'm not even arguing that this scenario as it is cannot be

02:45:45 that this scenario as it is cannot be used

02:45:46 used but already in itself understanding to what extent it describes the appropriateness of the scenario is something that needs to be done

02:45:55 done because

02:45:56 because you know we have the data we know what this decay is and we know what the peak values are so why we cannot characters and say well this test applies for this type of

02:46:07 this test applies for this type of conditions that leads to a value that is more or less this value in this position but that information is still unavailable so in essence

02:46:19 in essence uh the the test provides very limited information the final test that i'm going to look into is the european reaction to fire classification test and and this is a test that was developed

02:46:30 and this is a test that was developed for a very specific purpose which is to try to understand the propensity for flashover

02:46:35 flashover so that was the objective that was the scenario and uh and the test that that was used to define that scenario was the room corner test which was a large scale test and a very large amount of data was

02:46:48 test and a very large amount of data was produced to establish a very good correlation between a classification which is a simplification of a well understood problem

02:46:57 problem that enabled them to establish a classification for propensity to flush over

02:47:04 over now

02:47:05 now the test through a benchmarking study was shrunk to a level where it produced better data it was more consistent and it was at a scale that was more practical and it was demonstrated through this

02:47:17 and it was demonstrated through this benchmarking that actually the classification matched and you could actually get the same use the same classification for the sbi and uh and

02:47:28 and uh and be able to predict the propensity to flash over okay the phenomenon of flashover is very different to the phenomenal vertical flame spread is controlled by very very different parameters

02:47:39 different parameters now therefore that classification that is meant to be for the purposes of of

02:47:48 of of addressing flashover cannot be used to try to address vertical flame spread so it's just simply the use of that classification is what is inappropriate now the test has a

02:48:00 what is inappropriate now the test has a lot of good measurements there's a lot of good data and potentially there's information that can be extracted from the test

02:48:06 the test and it could be perfectly appropriate for people to use that information in such a way that they can conduct an appropriate analysis to establish the the performance criteria that they need

02:48:17 the performance criteria that they need but it is not uh to be used as a means to classify extrapolating a classification that was intended for a completely different scenario

02:48:28 scenario the variables are different and and the classification therefore provides no meaningful information

02:48:36 as i said several times none of the current performance assessment testing procedures addresses the formations thermo mechanical behavior

02:48:44 behavior or failure of the encapsulation and given the strong dependency of mechanical performance and encapsulation the absence of this performance criteria represents a serious limitation to

02:48:55 represents a serious limitation to everything that we're doing today so whether we understand combustibility and whether we understand how to design combustible systems

02:49:06 systems in many ways is very limited by the fact that we still don't even understand how to design non-combustible systems and unless we put focus on how to design non-combustible systems we are

02:49:18 non-combustible systems we are constantly introducing failure modes without paying any attention uh whatsoever so the the real question is how did we get there and uh and and this is a

02:49:30 get there and uh and and this is a diagram that is in in my report and the only information that this diagram shows you is how through the years knowledge has lagged need

02:49:41 knowledge has lagged need you know we've been building buildings way before faraday understood a candle and uh and we've been creating tests way before we understood ignition combustion

02:49:52 before we understood ignition combustion or flame spread beyond that you know way before uh you know we even unders you know we needed all this information before we

02:50:04 we needed all this information before we even professionalize the discipline of fire safety engineering so ife doesn't come in until 1918 the american version sfp until 1950 and uh so there's a long long history

02:50:18 and uh so there's a long long history of us doing always the same thing if you don't understand anything burn the entire building that's the process that we've been followed that's the advice that we have been following you know for decades and

02:50:31 been following you know for decades and decades and decades if you don't understand burn it you know that is the way we've been operating now the one thing that we need to recognize is that if we look at the end of the timeline we've caught up

02:50:43 end of the timeline we've caught up so now we have most of the knowledge but what we haven't changed is the mindset and we're still trusting those who tell us if you don't understand

02:50:54 understand burn it

02:50:55 burn it and that's not the case you know we can understand we can set the performance we can set the functionalities and we can actually come up with a proper calculation of proper desktop study that

02:51:08 calculation of proper desktop study that actually uses all the information that is available to actually deliver a

02:51:14 a conformity to the performance that we expect

02:51:17 expect so

02:51:18 so this is how we got there and i think we need to change that mindset because until we change that mindset

02:51:24 mindset fire safety is never going to be seen as a functionality it's going to be seen always as just the last stage of the whole process the one that we want to do very rapidly where the

02:51:35 we want to do very rapidly where the only good answer is don't change anything you know where effectively it's all about cheap fast and simple and we have a very complicated problem that cannot be addressed in those terms

02:51:47 that cannot be addressed in those terms so

02:51:48 so explicit performance objective again i go back you know we need to know what a fire safe building is you know how fire safety performance is expressed in building regulations now we

02:51:59 expressed in building regulations now we understand it you know it is expressed in a way such that leaves an enormous space for the professional to exercise their competency nevertheless we have no idea who is the person that

02:52:11 we have no idea who is the person that we expect to deliver this knowledge and if we look at what we do you can see in the column on the left on all the yellow things the

02:52:21 the performance objective remains a matter to be defined by competent professionals the overall safety objective and fire safety strategy have to be defined by the competent professionals by blending

02:52:32 competent professionals by blending these pillars in a way such that it is coherent you know the requirements how requirements be one to before support the fire safety strategy therefore is defined by the competent professional

02:52:45 is defined by the competent professional interpretation of guidance and supporting testing you know to be able to

02:52:51 to extrapolate or in one way or another when compare performance again is the determined by the competent professional and the applicability and validity of

02:53:02 and the applicability and validity of guidance and supporting testing to be defined by the competent professional in other words the onus on the competent professional is enormous at every level all the gaps need to be

02:53:13 at every level all the gaps need to be filled

02:53:14 filled and

02:53:16 and and all those gaps require a very very high level of competency and as i said in an area that is multi-disciplinary and that it requires a very significant understanding of multiple different problems that go from

02:53:28 multiple different problems that go from human behavior all the way to legislation so so we have to be quite careful so in summary i would just like to end by saying that knowledge in support of fire

02:53:40 saying that knowledge in support of fire safe buildings is currently available and the traditional practice of regulating before understanding is no longer necessary so

02:53:49 documents like past 9980 unfortunately still follow that approach and i think we need to be very very careful in the way in which we use them performance testing for systems

02:54:00 performance testing for systems delivering fire safety functions should be done as part of product development and then assess for conformity it's not about compliance it's about making sure that whoever is developing a product is

02:54:12 that whoever is developing a product is providing you with all the information that you need and that information is verified

02:54:17 verified and materials products subsystems and systems have to be assessed for conformity and it's the fire safety strategy that has to be compliant because that's the functionality that we want at the end that is the performance

02:54:28 want at the end that is the performance requirement so current building regulations put an extraordinary weight and competent professional practices you know this is inconsistent with the weak professional competency framework that we have currently in use

02:54:40 that we have currently in use and uh and this will end by telling you not i am a mechanical engineer i have three degrees in mechanical engineering i can stand in front of you and argue endlessly how competent i am you know but the reality is that it's

02:54:52 you know but the reality is that it's going to be up to you if you believe me and uh and the problem is that other people will stand in exactly the same place coming from different backgrounds and arguing exactly the same things and if you don't understand what competency

02:55:04 if you don't understand what competency is and we don't have a clear definition of competency we create a confusion of competency that ends up devaluing you know the capacity of people because a competent individual recognizes

02:55:16 a competent individual recognizes complexity recognizes the implications of their decisions and therefore proposes solutions that are complex in nature that require deep analysis and deep thought an incompetent individual

02:55:28 deep thought an incompetent individual is unaware of his incompetency they provide simplistic solutions that are cheap and fast

02:55:35 fast and the problem is that if we don't define competency appropriately at the end you know people like me or other people are going to be standing in there and we will not be able to tell the difference so i hope that i have clarified my

02:55:49 so i hope that i have clarified my report and i'll be more than happy to answer all the questions when the time comes

02:55:54 comes professor thank you very much it's been a very interesting morning thank you thank you

02:55:59 thank you that's great do you uh won't say anything else at this stage no um we can break for lunch now and then we have a presentation from professor bisbee this afternoon right well thank you much again thank you frustra and we'll see

02:56:11 again thank you frustra and we'll see you again next week when you are asked to answer some questions in relation to your report thank you thank you very much

02:56:23 yes thank you very much well we'll break at that point and we'll resume at two o'clock please yes thank you very much