May 6, 2025

200 - Façade flammability across scales and standards with Guillermo Rein and Matt Bonner

The player is loading ...

Episode 200! And for this special episode, I've travelled to London to interview Prof. Guillermo Rein and Dr Matt Bonner on a piece of research carried out at Imperial College London, with the experiments performed in our laboratory at the ITB. In this episode, we discuss the concept of flammability of the building facades and how this flammability is assessed with different testing methods available in the world. You could argue that every country has their own method, and in some case...

In this episode, we discuss the concept of flammability of the building facades and how this flammability is assessed with different testing methods available in the world. You could argue that every country has their own method, and in some cases, they use those methods even with varying criteria of acceptance. Even though the methods are as different as they can be, they all claim they test for fire safety of the external façade and are used as the basis for local regulatory regimes. Knowing that so many methods exist, we approached this with a question: Will they agree on ranking different facades? Will they show us the same results, or will each show us something else? And this question is inspired by Prof. Howard Emmons, who in 1968 went into a similar endeavour with building materials. Back then, Emmons said:

“Such profound disagreement between serious attempts to measure combustibility points out better than any argument that we really don’t know what we are talking about when we say, ‘this is more combustible than that’; ‘this is a more safe building material than that’”.

In this podcast episode, we discuss a series of 25 experiments: testing five facades, two ETICS and three rainscreen facades with a varying degree of use of combustible materials. All the material combinations were built by us in the same way, and then assessed using five test standards:

The Polish method PN-B-02867,
The international screening method ISO 13785-1 (smaller corner configuration),
The German method DIN 4102-20,
The American method NFPA 285, also used globally
and the British BS 8414, also highly influential over the world and the basis for the new harmonised EU approach.

We go into the background and rationale of the experiments, an overview of the testing methods as well as into qualitative and quantitative findings of the study.

Once the paper is published, I will update the shownotes with a link here :)

For now, you may also want to revisit previous episodes of Fire Science Show discussing the fire safety of facades –

It all started with episode 4 with Matt Bonner: https://www.firescienceshow.com/004-facade-fires-and-ai-with-matt-bonner/
An overview of current Issues with Eleni Asimakopoulou: https://www.firescienceshow.com/124-advancements-in-fire-safety-of-facades-with-eleni-asimakopoulou/
And some interesting facts about SBI method with Rudolf van Mierlo (and their development of façade testing method): https://www.firescienceshow.com/140-development-and-implementation-of-the-sbi-test-with-rudolf-van-mierlo/

This research was funded by The Berkeley Group. The experimental part was performed at the Building Research Institute ITB, with a group of tests with the Polish method performed as part of our statutory research NZP-130.

Thank you for being with the Fire Science Show for 200 episodes! Huge shoutout to the OFR for enabling this project and allowing me to share insights like this with all of you in an open-access repository!!!

----
The Fire Science Show is produced by the Fire Science Media in collaboration with OFR Consultants. Thank you to the podcast sponsor for their continuous support towards our mission.

00:00 - Welcome to Episode 200

03:27 - The Birgit Project Introduction

10:43 - Five Testing Methods Across the Globe

22:10 - Designing the Facade Test Matrix

35:19 - Qualitative Test Results and Observations

48:19 - Small vs Large Tests: Cost and Value

56:38 - Three Purposes of Testing Standards

01:01:51 - Final Thoughts and Concluding Remarks

WEBVTT

00:00:00.160 --> 00:00:03.669
Hello everybody, welcome to the Fire Science Show episode 200.

00:00:03.669 --> 00:00:13.867
That sounds kind of surreal, but I'm super rejoiced that we've reached this point and I must say the second hundred of the episodes went much faster than the first one did.

00:00:13.867 --> 00:00:18.707
I'm hugely thankful to all the guests who've participated in those 200 episodes.

00:00:18.707 --> 00:00:40.206
It's all thanks to your willingness to share excellent fire science and research and engineering with others that allows me to produce this show, and I'm also hugely thankful to how far for their contribution in the podcast is our third year of partnership in this and and they truly are the the amazing force that enables me to to do this for all of you.

00:00:40.206 --> 00:00:42.941
So huge thanks to our farm and, of course, thanks to you listener.

00:00:42.941 --> 00:00:52.728
Whether it's your first episode of the fire science show ever or you've listened to all 200 of them, I'm super happy that you're willing to embark on this journey to learn fire science with me.

00:00:53.048 --> 00:00:58.552
In podcast episodes like this, I I like to close some loops or at least revisit important topics.

00:00:58.552 --> 00:01:04.067
Actually, the last episode, 9199, we've also closed an important loop on timber.

00:01:04.067 --> 00:01:14.963
I've talked with Danny Hope and Luis Gonzalez-Avila from OFR on the Commercial Timber Guidebook and how stuff that we've discussed over the episodes of the podcast.

00:01:14.963 --> 00:01:23.331
As you know, research as growing body of knowledge that we had on timber turned into actionable document that really helps you design offices with mass timber.

00:01:23.331 --> 00:01:31.262
And in this episode episode 200, I'm trying to revisit the subject of facades and it's a subject that we've kind of started the podcast with.

00:01:31.262 --> 00:01:33.326
To do this I've went to London.

00:01:33.326 --> 00:01:37.004
I visited some good old friends at Imperial College London.

00:01:37.004 --> 00:01:43.203
I've talked to Professor Guillaume Reyn from the Imperial College London and to Dr Matt Bonner.

00:01:43.203 --> 00:01:46.611
He was a postdoctoral researcher at Imperial back then.

00:01:46.611 --> 00:01:59.400
Now he's at Trigon and we've discussed a topic that was kind of an Imperial ITB collaboration to really understand the testing methods and testing regimes on building facades.

00:01:59.801 --> 00:02:01.587
It was kind of a crazy idea.

00:02:01.587 --> 00:02:06.600
If you could rank projects on craziness, I would rank this one very, very high.

00:02:06.600 --> 00:02:21.627
The idea was simple let's bring in five testing methods coming from different parts of the world Some big, some small, some having a corner, some being flat, some having a burner, others having a crib, some having a window, some do not.

00:02:21.627 --> 00:02:33.985
Let's mix them up and test five different types of facades over those five methods and just look at the consistency like just look on how those methods help us rank those facades.

00:02:33.985 --> 00:02:36.431
Because, if you think about it, it's kind of ridiculous.

00:02:36.500 --> 00:02:39.206
Every country has their own facade testing standard.

00:02:39.206 --> 00:02:41.111
Every country seems to like them.

00:02:41.111 --> 00:02:47.631
There are attempts to unify those, but of course those are influenced by strong, let's say, political forces.

00:02:47.631 --> 00:03:02.731
As in fire research, it often happens and we don't seem to have a consensus on what's the best way to assess fire safety of a facade, and today, in a post-Grenfell world, it is a topic that bothers a lot of fire safety engineers.

00:03:02.731 --> 00:03:06.528
So in this episode we discussed the project.

00:03:06.528 --> 00:03:14.354
It was called Birgit, and Matt actually presented this at SFP Edinburgh, opening the pathway to have a podcast episode about it.

00:03:14.354 --> 00:03:19.271
Actually it won the Best Abstract Award and Matt gave it as a plenary talk.

00:03:19.271 --> 00:03:20.360
It was very well received.

00:03:20.360 --> 00:03:24.572
So I'm super happy to follow up on that with this podcast episode.

00:03:24.572 --> 00:03:28.104
There's a paper in the pipeline that explains everything that we talk in.

00:03:28.104 --> 00:03:31.612
It's not yet published but hopefully soon will be.

00:03:31.759 --> 00:03:52.890
So as soon as that comes up I will let you know and in this podcast episode you'll learn about what we've done, why we have done it, what was the general idea of why this needed to be done, and we will discuss what can be learned from the small methods, what can be learned from the big ones, and, for me, the important part is the sweet spot, where you get the most knowledge out of your money spent on testing.

00:03:52.890 --> 00:03:55.709
So I think that's more than enough.

00:03:55.709 --> 00:03:59.939
I'm super happy to invite you for the 200th time to the intro.

00:03:59.939 --> 00:04:02.649
Let's spin it up and jump into the episode.

00:04:02.649 --> 00:04:09.068
Welcome to the Firesize Show.

00:04:09.068 --> 00:04:12.590
My name is Wojciech Wigrzyński and I will be your host.

00:04:33.079 --> 00:04:42.112
The FireSense Show is into its third year of continued support from its sponsor, ofr Consultants, who are an independent, multi-award-winning fire engineering consultancy with a reputation for delivering innovative, safety-driven solutions.

00:04:42.112 --> 00:04:55.812
As the UK-leading independent fire risk consultancy, ofr's globally established team have developed a reputation for preeminent fire engineering expertise, with colleagues working across the world to help protect people, property and the planet.

00:04:55.812 --> 00:05:11.947
Established in the UK in 2016 as a start-up business by two highly experienced fire engineering consultants, the business continues to grow at a phenomenal rate, with offices across the country in eight locations, from Edinburgh to Bath, and plans for future expansions.

00:05:11.947 --> 00:05:17.163
If you're keen to find out more or join OFR Consultants during this exciting period of growth.

00:05:17.163 --> 00:05:20.531
Visit their website at ofrconsultantscom.

00:05:21.052 --> 00:05:40.033
And now back to the episode.

00:05:40.033 --> 00:05:46.966
And even though we've started this remotely and that was 200 episodes ago Can you believe that that's a long journey?

00:05:46.966 --> 00:05:50.427
And, matt, we've started with facades, so we're coming back to the facade subject today.

00:05:50.427 --> 00:05:54.151
It's also a nice, like you know, closing a loop in the podcast.

00:05:54.151 --> 00:05:57.709
Anyway, we're talking about Burgett Project, something we've recently closed.

00:05:57.709 --> 00:06:02.831
The papers are in revision process, which will take 100 years, but eventually they will come up.

00:06:02.831 --> 00:06:06.048
But let's talk what brought us here.

00:06:06.048 --> 00:06:11.908
So, guillermo, you came to me with a crazy idea some time ago and, yeah, can you pitch it once again?

00:06:12.420 --> 00:06:14.026
Yeah, this is truly a crazy idea.

00:06:14.026 --> 00:06:18.428
It's a crazy idea with a crazy founder and a crazy false start.

00:06:18.428 --> 00:06:19.685
So it's almost canceled twice.

00:06:19.685 --> 00:06:25.120
I was quite worried about this, but it didn't and we had here the dream team of why it went forward.

00:06:25.120 --> 00:06:29.711
So Matt put in their weight and Voge put in your weight.

00:06:29.711 --> 00:06:32.129
It was absolutely amazing, but it's a crazy idea.

00:06:32.129 --> 00:06:39.740
It's a crazy idea because it's in the context of what happened with the Grenfell Tower fire and everybody talking about facade fires but no one putting a solution forward.

00:06:39.740 --> 00:06:44.673
So the idea was to steer the community, to make them feel a little bit uncomfortable.

00:06:45.279 --> 00:06:53.894
When you look into how facade tests are done around the world, most countries I think probably every country in the world actually does have a facade test.

00:06:53.894 --> 00:07:04.367
They might borrow from the British influence or from the American influence or from the Germanic influence or from the other influences, but it's fascinating that they all look so different.

00:07:04.367 --> 00:07:08.627
And when I mean different, I really really really mean different.

00:07:08.627 --> 00:07:13.944
You put them together and they hardly agree on anything except that they have a facade.

00:07:13.944 --> 00:07:25.233
The rest, the ignition source, the size, the shape, the configuration, not to talk about the criteria for passing it and the question is it cannot be that everybody's wrong, right.

00:07:25.233 --> 00:07:26.682
So some people have to be right.

00:07:26.682 --> 00:07:30.055
Some people have to be making sense of how they measure the facet.

00:07:30.055 --> 00:07:31.420
How do you figure this out?

00:07:31.420 --> 00:07:43.742
How do you figure out which tests make sense, which tests agree with each other, which tests are worth to put a significant amount of money into them and which ones can be put on the side for the time being?

00:07:44.362 --> 00:07:48.672
I wonder, is a test meant to be an objective truth, or is it just a test?

00:07:48.672 --> 00:07:49.012
You know?

00:07:49.293 --> 00:07:49.653
It's a test.

00:07:49.653 --> 00:07:58.850
But if you have 26 tests depending on the nation, not based on the laws of physics, not based on the technical knowledge of the nation, and the question is do they agree?

00:07:58.850 --> 00:08:01.649
Because if they agree amazing, they agree, there's agreement.

00:08:01.649 --> 00:08:04.401
We have 26 ways to answer the same question.

00:08:04.401 --> 00:08:06.124
Is that the case?

00:08:06.124 --> 00:08:10.067
Because before we do this study, no one could answer that question.

00:08:10.067 --> 00:08:11.069
Are the 26 wrong?

00:08:11.069 --> 00:08:11.850
Are the 26 right?

00:08:11.850 --> 00:08:13.031
Are 15 right?

00:08:13.031 --> 00:08:14.112
Are 3 right?

00:08:14.112 --> 00:08:16.394
I mean right in the sense of agreeing with each other.

00:08:16.394 --> 00:08:24.906
Obviously, we don't know why it's right or why it's wrong, but the question was agreement, and this is something that had already been done, but not with facadesades.

00:08:24.906 --> 00:08:32.686
Professor emmons in the 60s did this with flammability of panels and he showed that there was no agreement and that led to a new standard yeah, so I think maybe the framing was slight.

00:08:33.969 --> 00:08:36.822
Maybe I saw it as slightly different to that as in.

00:08:37.082 --> 00:08:42.201
Obviously the premise is the same you don't like the right and wrong, but I wrote the proposal so I know why it's framing.

00:08:42.201 --> 00:08:43.725
Yeah, no, no exactly.

00:08:43.745 --> 00:08:44.307
Yeah, no, no, exactly.

00:08:44.307 --> 00:08:58.227
I'm saying how I approached it, not necessarily how we got the funding or what the purpose of the project was, but when we sat down, because basically the project was, we wanted to compare, as you say, different facade test standards.

00:08:58.227 --> 00:09:10.581
So in order to take the same facades, facades across a range of different amounts of fuel and different configurations are things that we expected to have kind of a different flammability performance, whatever that means.

00:09:10.581 --> 00:09:14.201
We can come back to that, but we expected the facades to have different performance.

00:09:14.201 --> 00:09:15.164
But we didn't know.

00:09:15.164 --> 00:09:17.149
You didn't know automatically.

00:09:17.149 --> 00:09:22.745
You might guess which ones might be better or worse, but you wouldn't be able to rank them, certainly looking at them.

00:09:22.745 --> 00:09:24.982
That would be an opinion, exactly, you'd have an opinion.

00:09:25.302 --> 00:09:28.649
A desktop study, no desktop study.

00:09:28.649 --> 00:09:32.481
And then, fascinating, he knows, Bodger.

00:09:32.481 --> 00:09:33.509
You know because you have a testing facility.

00:09:33.509 --> 00:09:48.424
So what we saw and we saw this, Matt, in real time when we, before we approached you, Bodger, before I approached ITV we were a poor project because we didn't have too much money and we went to many different testing houses and one thing that happened when we were talking to the testing houses, we realized that the engineers in the testing house knew.

00:09:48.424 --> 00:09:52.986
They knew which facade would be more flammable or less flammable according to the standard they knew.

00:09:52.986 --> 00:09:54.268
The question is, how do they know?

00:09:54.268 --> 00:10:04.094
Not because they read the book, not because they wrote the paper, not because they ran the code, not because they have a computer model, not because they are mathematical geniuses is because they have seen the test so many times.

00:10:04.094 --> 00:10:11.618
They trained themselves, they became artificial intelligence learning machines and they could actually not explain to you why, but they could actually tell you the right results.

00:10:12.100 --> 00:10:19.330
So, yeah, I think this was the kind of the topic of my PhD and sort of the previous research we had done was kind of trying to look at.

00:10:19.330 --> 00:10:22.241
And this is how we started our collaboration with you, wojciech, and with IT.

00:10:22.241 --> 00:10:37.331
Itb was trying to see whether these tests, whether we could draw out the predictive mechanisms that were happening by people performing the tests, who could tell which facades would perform well and which wouldn't.

00:10:37.331 --> 00:10:38.703
Could we draw that out?

00:10:39.025 --> 00:11:11.809
Now, obviously, because of the nature of the collaborations and the data we were able to get hold of during the PhD, this was very weighted towards one particular test method, the one done at ITB, the Polish test method, and so this kind of was building on, because when we had a few examples in our PhD data set from the British standard BS8414 and from the American standard NFPA285, and we actually found, when we tried to compare the results of these tests and sort of combine them into one data set, that was really really difficult because they were all measuring different things.

00:11:11.809 --> 00:11:22.751
They seemed to be measuring somewhat the same thing, but trying to work out the sort of causal mechanisms behind that with the kind of data they were recording was very, very challenging.

00:11:22.751 --> 00:11:27.225
Yeah, and so there was sort of this question of like, well, why are they all different?

00:11:27.225 --> 00:11:28.408
What are they measuring?

00:11:28.408 --> 00:11:33.808
And so, to kind of step back a bit, guillermo came to when he presented this project to me.

00:11:33.808 --> 00:11:35.907
He came to me with this paper.

00:11:35.927 --> 00:11:36.789
This is after your PhD.

00:11:36.940 --> 00:12:03.153
This is after my PhD yeah, this was me looking for work to do and Guillermo came forward with this project and showed me a paper from Professor Howard Emmons from 1968 called Fire Research Abroad, and in this paper there was other stuff in the paper, but part of it was he had done this kind of round-robin study where he'd asked six test houses around Europe to rank 24 materials in terms of their combustibility.

00:12:03.153 --> 00:12:17.850
He didn't define what that was, because at the time lots of people were using the phrase combustibility as if it had an inherent meaning, as if everyone agreed on this, that this material is more combustible than that material, obviously, except he was asking well, do we agree what that means?

00:12:17.850 --> 00:12:21.500
Can you tell me which of these 24 materials is more combustible?

00:12:21.500 --> 00:12:22.961
Can you rank them According to their national?

00:12:22.981 --> 00:12:24.104
standards 24 materials is more combustible?

00:12:24.104 --> 00:12:25.424
Can you rank them According to their national standards?

00:12:25.424 --> 00:12:26.186
The standards existed.

00:12:26.206 --> 00:12:26.966
Yes, exactly.

00:12:26.966 --> 00:12:29.688
So each country had a standard around combustibility.

00:12:29.688 --> 00:12:33.712
And he went to these testing labs and said just tell me what the rank of these?

00:12:33.732 --> 00:12:37.317
materials is and he got the materials and he sent the same material to each of the six labs.

00:12:37.496 --> 00:12:42.639
Exactly and they ranked them and the results were no better than random noise.

00:12:42.639 --> 00:12:46.831
They did not agree on which were more or less combustible.

00:12:47.120 --> 00:12:47.621
And the materials?

00:12:47.621 --> 00:12:48.746
Do you remember some of the examples?

00:12:48.746 --> 00:12:50.105
One of them was plasterboard.

00:12:50.105 --> 00:12:54.368
I think it's the only one that people agree on it has no combustibility.

00:12:54.368 --> 00:12:57.109
I can't remember they were polymeric.

00:12:58.144 --> 00:12:58.769
In the figure.

00:12:58.769 --> 00:13:00.500
They don't label all the materials.

00:13:00.500 --> 00:13:03.750
It's like a separate bit in the paper where you have to read other things.

00:13:03.750 --> 00:13:08.142
So I don't remember it off the top of my head.

00:13:08.142 --> 00:13:08.823
But there was a lot of plastics.

00:13:08.823 --> 00:13:10.668
There were some organic polymers um as well, yeah, like wood and things like that.

00:13:10.668 --> 00:13:17.953
So it was quite a range of of stuff and some of them were the same kind of generic plastic, but in different forms as well.

00:13:17.953 --> 00:13:25.153
So it gets quite tricky, yeah, and some of them were thermoplastic, some of them thermosetting, so they charred or melted and all of that.

00:13:25.153 --> 00:13:28.250
And that's part of the challenge with defining combustibility right.

00:13:28.250 --> 00:13:31.529
And because of that, part of the outset of this was in Europe.

00:13:31.529 --> 00:13:41.388
They sort of tried to harmonise these standards and that's what ended up eventually becoming the European reaction to fire suite of tests which kind of define combustibility.

00:13:41.388 --> 00:13:43.869
It is quite arbitrary in a lot of ways.

00:13:43.869 --> 00:13:46.160
There is some.

00:13:46.160 --> 00:13:51.663
Some it did have a physical basis, but that's not the point of this episode and I'm sure you've got another episode of the podcast.

00:13:52.163 --> 00:13:55.610
This is where voychek will like link in the podcast description.

00:13:55.610 --> 00:14:09.908
But so we saw this and we went okay, we're in a similar situation right now with facades and that if we asked, I imagine, if we did this round robin study now and we asked everyone to rank facades in terms of their flammability, would countries agree or wouldn't they agree?

00:14:09.908 --> 00:14:17.547
Now, obviously we couldn't really, or we didn't have the funding or the ability, yeah, to talk to all these to ask people.

00:14:18.200 --> 00:14:21.880
The original idea was to do exactly what MNs did with facades in Europe.

00:14:21.880 --> 00:14:23.282
Now the amount of money?

00:14:23.282 --> 00:14:24.342
No, we couldn't.

00:14:24.764 --> 00:14:29.927
Yes, that was infeasible and actually even trying to do what we did.

00:14:29.927 --> 00:14:34.572
So what we were trying to do was okay, can we just run all the experiments ourselves?

00:14:34.871 --> 00:14:39.195
Realistically, you cannot send a bunch of facades across Europe to burn them down in multiple hubs.

00:14:39.195 --> 00:14:40.616
No, you could.

00:14:40.636 --> 00:14:44.500
You could and unfortunately there's no other way to find out.

00:14:44.500 --> 00:15:35.094
I know you could, you could and unfortunately there's no other way to find out, because what we did with Matt in the Kresnik and in his PhD project, where we just took the results, purpose in mind, with the same solutions in mind, preferably by the same technical team, there's literally only one way to find out if, for the same sample, for the same type of facade and it sounds the same facade, but trust me, the Polish method facade built with a certain type of solution and be a standard facade of the same time of solutions, those are completely two different builds in terms of effort, scale, structure, amount of people involved.

00:15:35.094 --> 00:15:36.860
They just, they're just different.

00:15:36.860 --> 00:15:47.913
Right, it is even hard to say the same facade is just built from the same materials in this, with the same technical principles in mind, but you really have to put them in place and burn them down.

00:15:48.379 --> 00:16:10.394
And actually maybe to kind of highlight a particular example of this and something I've had discussions about at conferences, at places where I've presented our work so one of the things we had to decide when sort of setting up these facades because there are all these standards of very different sizes and scales was about do we include cavity barriers and, if so, where we had ventilated facades?

00:16:10.394 --> 00:16:18.549
And in some countries you would need those facades to have cavity barriers and quite often, if you're going to include them, you'd want to test them because that would give a more.

00:16:18.549 --> 00:16:21.544
You want to include them in the test because that would give a more favourable outcome.

00:16:21.544 --> 00:16:32.940
Of course, in these larger tests you are testing for in a way, they're trying to represent flame spread between multiple stories and they're considering the size of them is above one story.

00:16:32.980 --> 00:16:34.726
Some of them do multiple stories.

00:16:34.726 --> 00:16:35.692
That's a beginning already.

00:16:35.692 --> 00:16:36.179
The size right.

00:16:36.179 --> 00:16:40.652
Some of them the ones that I might prefer actually cannot.

00:16:41.403 --> 00:16:48.572
So that was kind of my question is there was lots of ways you could maybe interpret, how you would even compare across these tests.

00:16:48.572 --> 00:16:56.293
So we did include cavity barriers in the larger tests and we basically aligned their position to the British one.

00:16:56.293 --> 00:17:02.513
We said, ok, the British one puts them at these heights, and those heights are chosen because they're where the thermocouples are.

00:17:02.513 --> 00:17:09.029
And the thermocouples are there because they're vaguely supposed to represent the height of a compartment and correlate with these previous tests.

00:17:09.029 --> 00:17:19.003
That happened in Canada in the 90s, and so there's all these historic reasons about why they are where they are and why the cavity barriers are where they are and other tests aren't correlated in the same way.

00:17:19.003 --> 00:17:20.868
So already we're kind of that's the thing.

00:17:20.868 --> 00:17:26.943
Like the test, the instrumentation of the other test isn't set up in the same location, wouldn't have the cavity barriers in the same location, of course.

00:17:26.943 --> 00:17:30.840
We then chose our own instrumentation location, so that's a bit different.

00:17:30.840 --> 00:17:34.661
But then with the smaller tests the question is do you include cavity barriers at all?

00:17:34.661 --> 00:17:38.515
And we didn't, because that's not the intent of those tests.

00:17:38.555 --> 00:17:41.983
When they're screening those facades they would not put cavity barriers in.

00:17:41.983 --> 00:17:42.786
They were.

00:17:42.786 --> 00:17:50.145
They're looking at kind of the flame spread over a kind of a sample of this or the flammability over a part of this facade.

00:17:50.145 --> 00:17:55.708
And I've been challenged already about like saying, oh well, shouldn't you include?

00:17:55.708 --> 00:18:09.373
Like, if you're not including cavity barriers, it's not really representative, because the fire is now impinging into your facade in a way that it might not in reality, but then again you're getting to this mistake of, oh well, these tests are supposed to represent reality in some way.

00:18:09.373 --> 00:18:15.608
Anyway, I was more bringing that up as an example of why this sort of idea of direct comparison.

00:18:15.608 --> 00:18:19.756
It's not as simple as in the material case with combustibility.

00:18:20.280 --> 00:18:27.250
I can get the discussion a little, even a step back, because the first challenge was okay, we want to compare five facades, you want to rank them?

00:18:27.250 --> 00:18:28.864
Yeah, but how do you ensure that?

00:18:29.085 --> 00:18:33.203
we chose five because then that's when the money ran out and really we went to 25, right.

00:18:33.203 --> 00:18:34.567
So I remember with matt we put them.

00:18:34.567 --> 00:18:41.949
Actually, with matt we put all the standards, what we wish to do, obviously way more than five, all the facades we wanted to do way more than five.

00:18:41.949 --> 00:18:46.423
And then it's like we wrote a line where, with you guys, we ran out of money and it was five on five.

00:18:46.444 --> 00:18:52.336
I'm happy to do more If there's a founder listening who sees value in burning more.

00:18:52.376 --> 00:18:53.119
I'm happy to burn more.

00:18:53.119 --> 00:18:58.162
I feel the audience might need to know the final five because actually we would have loved to have 20 of them.

00:18:58.481 --> 00:19:01.826
Let's go with five tests and then let's talk to five types of wars.

00:19:02.428 --> 00:19:04.411
What is important is there were many tests.

00:19:04.411 --> 00:19:06.794
How many tests was there in that review?

00:19:06.794 --> 00:19:07.515
26, right?

00:19:07.515 --> 00:19:14.330
So we found I think there was 21 across the papers.

00:19:14.330 --> 00:19:15.439
So we wanted to choose within those 21,.

00:19:15.439 --> 00:19:17.448
Right, Because these are the ones that have been looked at already.

00:19:17.448 --> 00:19:23.076
We wanted to use countries that are relevant to the world but also relevant to us based in the UK, and we wanted to be different.

00:19:23.076 --> 00:19:24.013
We wanted to use countries that are relevant to the world but also relevant to us based in the UK, and we wanted to be different.

00:19:24.013 --> 00:19:25.395
We wanted to embrace diversity.

00:19:25.820 --> 00:19:27.989
So, we didn't want to get five that are identical.

00:19:27.989 --> 00:19:39.290
No, we wanted to actually go across boards and immediately we found the POLIS test, which first you at United TV know how to do the POLIS test, but through the PhD thesis of Matt we developed a love for the POLIS test.

00:19:39.290 --> 00:19:43.280
We actually end up.

00:19:43.280 --> 00:19:44.364
Many people might not know about the PN test.

00:19:44.364 --> 00:19:44.925
We actually really like it.

00:19:44.925 --> 00:19:47.172
There's a lot of value in it, so obviously we keep it right.

00:19:47.172 --> 00:19:52.530
Then we have the ISO, which has in size a lot of similarities with the POLYS test.

00:19:53.701 --> 00:19:55.506
Small ISO because there's also a big ISO.

00:19:55.506 --> 00:19:56.709
So we pick the two-meter tall one.

00:19:57.839 --> 00:19:59.182
It's 2.8 meters already.

00:19:59.182 --> 00:20:00.365
It's taller than a person.

00:20:00.365 --> 00:20:01.987
It doesn't fit into a fire lab already.

00:20:01.987 --> 00:20:03.450
Again, we say small.

00:20:03.450 --> 00:20:06.234
We just want people to realize what we mean by small.

00:20:06.234 --> 00:20:08.704
Then we have the NFPA.

00:20:08.704 --> 00:20:10.892
Of course we cannot go without the Americans.

00:20:10.892 --> 00:20:17.693
American is a very influential type of way of thinking and is adopted in many other countries that might not actually have a standard otherwise.

00:20:17.693 --> 00:20:21.371
So the American one is very strange because it's just two stories with two compartments.

00:20:21.371 --> 00:20:25.266
And an external burner in addition to yeah with two different English sources.

00:20:25.460 --> 00:20:42.424
Well, I'd also say, part of the choice of the Polish, the American and you're about to mention the British one as well was that we had data from my PhD and we had performance stuff, we had knowledge and there was availability to compare stuff in the literature.

00:20:42.424 --> 00:20:48.314
And similarly with the other two, we chose which Guillermo is about to explain the ISO and the German one.

00:20:48.314 --> 00:20:55.527
That was again about comparison with other data and we knew the European harmonized standard was being worked on at the same time as well.

00:20:55.527 --> 00:21:03.339
So this was about trying to choose the ones that would be most able to have comparison with other things in the scientific.

00:21:03.961 --> 00:21:07.106
yeah, okay, the bigger ones so you know I I want to recap.

00:21:07.106 --> 00:21:10.193
So we have the police, one is 2.5 meters.

00:21:10.193 --> 00:21:12.465
We have the small iso 2.8 meters tall.

00:21:12.465 --> 00:21:16.537
The nfba is five meters, more than 5.3 meters.

00:21:17.097 --> 00:21:22.069
These are already constructions, right yeah we definitely need an itv kind of testing site.

00:21:22.069 --> 00:21:22.952
It doesn't go anywhere else.

00:21:22.952 --> 00:21:24.885
Then we went to the D, the German.

00:21:24.885 --> 00:21:27.988
The German is very influential, especially because most German engineers are very proud of it.

00:21:27.988 --> 00:21:45.071
So the D is already more than 5.5 meters tall and it has a corner configuration, which is actually a configuration that has a lot of merit, because then it radiates to each other, the two panels, and then you have the British one, which is a corner test as well, and it's the tallest of them all, which is it goes to 8.9 meters.

00:21:45.071 --> 00:21:49.036
Right, this is a proper wall of a building, right.

00:21:49.056 --> 00:21:49.557
It is massive.

00:21:49.980 --> 00:21:52.328
And it's fascinating when you put them all together.

00:21:52.328 --> 00:21:59.828
There's one of the figures in the paper and in the presentation of Matt is how different they look already just by putting them together.

00:21:59.828 --> 00:22:00.471
The same scale.

00:22:00.471 --> 00:22:09.608
Not just only that the way the barriers go and the ignition sources and the criteria and the tiny details, the big details already could not be more different.

00:22:09.608 --> 00:22:21.246
And yet Polish people, international level, american people, german people and British people, they receive their protection of the environment and the safety of the environment through these standards.

00:22:21.246 --> 00:22:27.000
And we are in a context where the Grenfell Tower we know that the market and the regulation was not working.

00:22:27.000 --> 00:22:34.144
So the question is could this be one of the reasons, this disagreement in the way they want to measure and test facades?

00:22:34.144 --> 00:22:37.520
Could this be part of a disarray in the market and the regulation?

00:22:37.902 --> 00:22:44.094
Okay, and so those were the five testing methods, and we also burned down five types of solutions.

00:22:44.094 --> 00:22:54.988
My question that I asked before, but I'll reiterate how do you actually compare five different like, if you want to compare five different facades, how do you make it fair play?

00:22:54.988 --> 00:23:05.830
You know, I could build an extremely heavy let's say, timber facade and compare it to a thin aluminum foil, and the aluminum foil will probably fail, even though it's not combustible.

00:23:05.830 --> 00:23:08.368
So how did you make it for play, matt?

00:23:08.368 --> 00:23:11.230
Because I know there was a lot of engineering in that as well.

00:23:11.799 --> 00:23:23.087
Yeah, I think this was a tricky question and again we were kind of trying to base it on the previous research we'd done in the PhD and the lessons we learned there.

00:23:23.220 --> 00:23:25.647
The lessons learned from the research, not the research itself.

00:23:25.647 --> 00:23:26.631
Exactly exactly.

00:23:26.631 --> 00:23:28.546
What I mean is that I'm trying to paraphrase that.

00:23:28.546 --> 00:23:32.429
This previous knowledge, this project, didn't start with us playing with Lego.

00:23:32.429 --> 00:23:36.829
No, the team had knowledge and this was used to launch to the next step.

00:23:37.319 --> 00:24:04.164
Yes, exactly, but also the knowledge was quite a course level and so that was what we're sort of carrying over, in the sense that we've been able to identify that in the research we've done before.

00:24:04.164 --> 00:24:09.473
Across all these three different test sets we had only three types of facades were tested in 95% of cases, which are the rain screen facades, which is where you've got a ventilated cavity and some external cladding, ethics facades or EIFs, whatever.

00:24:09.493 --> 00:24:21.987
The other one is facades which is like render facade systems, where you've got some insulation and then some render which is kind of fixed to a substrate not ventilated, and then the last one is sandwich panels, which were and of those it was more common in our data set to see ethics facades and rainscreen systems in our data set to see ethics facades and rain screen systems.

00:24:21.987 --> 00:24:28.672
The other thing that gets chosen very commonly is curtain walls, but of course curtain walls aren't really flammable in the same way.

00:24:28.672 --> 00:24:36.557
In general People argue about this but they're not really tested in these tests looking for flammability facade systems.

00:24:36.557 --> 00:24:38.067
So that was the first thing.

00:24:38.067 --> 00:24:43.326
We were kind of like OK, we have these three types of facades and of the ones we see, the most common are rain screen and ethics.

00:24:43.326 --> 00:25:11.469
So that was where we started and we also realized that actually when we looked into it deeper, in general we found so only a very small proportion of the data, as the facades we looked at in our phd kind of failed these tests or performed badly in these tests, supported the spread of flame, and the ones ones that did needed it kind of seems obvious needed some sort of combustible material, some source of fuel, and it was a lot more common to fail if they also had a ventilated cavity.

00:25:11.469 --> 00:25:23.833
So we're like right, we think rain screen facades are going to perform worse than ethics facades, and we also think that the more combustible material you have, the more likely there's going to be flame spread, and so we tried to kind of cover that range.

00:25:23.833 --> 00:25:30.634
We had an order to our facades of how we had a predictive ranking for these.

00:25:30.634 --> 00:25:38.059
So we had two ethics facades, one that used mineral wool insulation and one that used EPS insulation.

00:25:38.059 --> 00:25:43.165
That's very common insulation for these kind of facades and so there's an obvious ranking there.

00:25:43.165 --> 00:25:46.238
You think the one with no fuel in it, that was kind of just a control.

00:25:46.238 --> 00:25:51.753
We assumed that all the tests would rank that the same and rank that as the best one.

00:25:52.296 --> 00:25:59.980
Then, after the ethics ones, we then had three rain screen facades, and so in the rain screen these can get very complicated.

00:25:59.980 --> 00:26:01.785
On real buildings there can be a lot in the build-up.

00:26:01.785 --> 00:26:13.743
We were kind of keeping it to this kind of essentialised rain screen facade, which was you just had some cladding, a ventilated cavity and then some insulation fixed to your testing substrate.

00:26:13.743 --> 00:26:31.580
In this case what we are varying between these three facades is we had one that had non-combustible cladding and combustible insulation, one that had combustible cladding but non-combustible insulation and one that had combustible cladding and combustible insulation.

00:26:31.580 --> 00:26:38.032
So you can see that those three cases have kind of increasing amounts of fuel.

00:26:38.032 --> 00:26:56.211
So because the cladding's thinner, it actually provides less fuel, but obviously it's also on the outside, so possibly more exposed to fire, to the sort of impinging flames, and so it wasn't really clear exactly what the ranking of those three would be.

00:26:56.211 --> 00:26:58.482
We would imagine they'd perform worse than the.

00:26:58.583 --> 00:26:59.204
It's very important.

00:26:59.204 --> 00:27:00.990
We talk a lot of people.

00:27:00.990 --> 00:27:02.152
Everybody had an opinion.

00:27:02.152 --> 00:27:02.880
I thought this was absolutely fascinating.

00:27:02.880 --> 00:27:03.555
Everybody had an opinion.

00:27:03.555 --> 00:27:03.877
I thought this was absolutely fascinating.

00:27:03.877 --> 00:27:05.405
Everybody had an opinion.

00:27:05.405 --> 00:27:09.189
Some people raised some technical knowledge, some people have been talking to their friends, some people have been reading books.

00:27:09.189 --> 00:27:10.726
Everybody has an opinion.

00:27:10.726 --> 00:27:15.969
You show this table of these five facades and the discussion would be for hours and hours and hours.

00:27:15.969 --> 00:27:16.770
I'm unprofessional.

00:27:16.770 --> 00:27:21.829
What we wanted was to provide the evidence, not the opinions, not the.

00:27:21.829 --> 00:27:22.391
I think.

00:27:22.391 --> 00:27:25.345
I feel I thought it's like the evidence.

00:27:25.345 --> 00:27:26.750
This was the objective of the project.

00:27:27.846 --> 00:27:34.282
And the amount of combustible material was kind of normalized by the U-value of the facade.

00:27:34.282 --> 00:27:37.507
So it was not based on some unrealistic case.

00:27:37.507 --> 00:27:43.166
Oh, we're going to put like 20 centimeters of EPS now because it's better when it's thicker.

00:27:43.166 --> 00:27:45.528
No, they were normalized to provide the same U-value.

00:27:46.119 --> 00:27:49.045
U-value is the resistance to heat for insulation of a building.

00:27:49.380 --> 00:27:50.163
Yeah, so this is.

00:27:51.542 --> 00:27:53.188
Which is the reason why we have facades.

00:27:53.188 --> 00:27:53.971
Exactly, yeah.

00:27:54.381 --> 00:27:59.970
Exactly that's one of the major points right now, why you would put EPS on the exterior of your building.

00:27:59.970 --> 00:28:02.509
You want it to insulate better.

00:28:03.461 --> 00:28:22.108
It's worth saying that we actually used a relatively high U-value, so not as much insulation as you'd use on a real building, and that was partially for cost-saving and partially for safety reasons, just so that we wanted to keep the fires within manageable lab conditions.

00:28:22.500 --> 00:28:24.429
Thank you, matt, I really appreciate that.

00:28:24.760 --> 00:28:26.366
It's important because we didn't go.

00:28:26.366 --> 00:28:30.925
One of the five is not the most flammable facade that everybody's worried about.

00:28:30.925 --> 00:28:38.411
It's actually so flammable that it's actually forbidden in the UK, for example, because we were afraid that it would overwhelm the testing facilities.

00:28:38.411 --> 00:28:40.848
It would I tell you this?

00:28:40.848 --> 00:28:50.229
Because when I go around the world telling people that are not fire experts and they don't visit very often our conferences they are fascinated by that detail that the testing lab already knows.

00:28:50.229 --> 00:28:51.586
Oh, I don't test that one.

00:28:52.122 --> 00:28:52.765
No, I'm not testing that one.

00:28:53.460 --> 00:28:54.884
It's like well, if they knew.

00:28:54.884 --> 00:28:57.108
Why is it that the market didn't know?

00:28:57.891 --> 00:29:00.503
No, no, guillermo.

00:29:00.503 --> 00:29:02.130
The sad part is, the market does as well.

00:29:02.799 --> 00:29:03.740
The sad part is the market does as well.

00:29:03.740 --> 00:29:04.421
The sad part is that exactly.

00:29:04.481 --> 00:29:05.886
But let's not go there.

00:29:06.381 --> 00:29:17.768
Anyway you place your order and I'm left with buying 25 facades, Sorry context, it's very important One of the two false starts of this project, the two difficulties that we had to overcome.

00:29:17.768 --> 00:29:36.071
The last one is that this happened in the middle of the lockdowns of the pandemic, of the COVID-19 pandemic, so which means that this project was literally all on Zoom and phone calls and emails and remote testing and obviously success story, massive success story, but it was really hard.

00:29:37.181 --> 00:29:41.531
Yeah, yeah, and it took us a lot of time to actually take your ideas and turn them into designs.

00:29:41.531 --> 00:29:43.386
And of course I'm a Polish lab.

00:29:43.386 --> 00:29:50.653
I can do as many Polish tests as you like, but if you ask me to do a British standard test for you, I don't really have the rig, you know.

00:29:50.653 --> 00:29:55.045
So I had to build one, and I had to build the NFPA one and I had to build the DIN one.

00:29:55.045 --> 00:30:05.676
So in the end they were not the most perfect rigs that you would have in testing houses for testing houses, who just run them as run-of-the-mill tests.

00:30:06.622 --> 00:30:12.026
And NFPA, dean and BS are very upset with you because you didn't build the perfect, of course, of course.

00:30:12.026 --> 00:30:13.586
No, it was very professional.

00:30:13.586 --> 00:30:14.502
What I mean is a joke.

00:30:14.502 --> 00:30:16.805
Is that, yeah, and we are exposed to this.

00:30:16.805 --> 00:30:17.446
We are exposed to this.

00:30:17.446 --> 00:30:18.986
Some of the listeners already have this.

00:30:18.986 --> 00:30:20.741
Everybody has opinions on this, everybody.

00:30:20.741 --> 00:30:21.623
Some of the listeners already have this.

00:30:21.623 --> 00:30:22.242
Everybody has opinions on this.

00:30:22.242 --> 00:30:22.804
Everybody has opinions.

00:30:22.804 --> 00:30:23.585
Oh, you chose the barriers wrong.

00:30:23.585 --> 00:30:24.705
Oh, you chose the wrong facade.

00:30:24.705 --> 00:30:26.567
You did wrong the thermocouple location.

00:30:26.567 --> 00:30:32.954
We can have thousands and millions of hours of discussions about the tiny details, but let's focus on the big details.

00:30:32.954 --> 00:30:34.896
One of them is almost 10 meters tall.

00:30:36.799 --> 00:30:37.622
The other one is one meter tall.

00:30:37.622 --> 00:30:45.229
I think the key thing is that we did realize that we wouldn't be exactly matching by building thing is that we did realize that there we wouldn't be exactly matching.

00:30:45.229 --> 00:30:46.776
By building our own system we wouldn't be exactly matching these standards.

00:30:46.776 --> 00:30:49.388
We didn't have the time to be like certified each one, we didn't have the budget to do that.

00:30:49.388 --> 00:30:53.182
We did actually look but you know we went out for tender with this project.

00:30:53.182 --> 00:30:59.644
We didn't automatically uh, itb was bought in as a collaborator nobody tell them what happened when we went to tender.

00:31:00.326 --> 00:31:03.452
Oh, it was extremely, extremely expensive.

00:31:03.593 --> 00:31:06.799
It would have cost, it was 10 to 100 times bigger budget than we had Right.

00:31:07.560 --> 00:31:08.402
And it was.

00:31:08.402 --> 00:31:13.712
Yeah, it definitely would have cost probably five times the amount at least.

00:31:13.712 --> 00:31:20.653
Okay, Anyway, it was a lot of money to do these by a sort of test certified rig, as it were, for these larger tests.

00:31:20.653 --> 00:31:39.444
But the key thing is we're not interested in and I think a lot of the reason I have people who have challenged me at conferences and stuff is because it's people from manufacturers or people or testing houses who are like worrying that we're trying to compare these tests or rank these tests, no we are trying to compare the tests.

00:31:39.464 --> 00:31:44.203
But the idea is we're not coming away to say this one's the best one, this one's the worst one.

00:31:44.203 --> 00:31:45.388
That wasn't the point.

00:31:45.388 --> 00:31:48.028
What we're trying to say is are they consistent?

00:31:48.028 --> 00:32:01.933
And the consistency is going to depend a lot more on these big differences, like if the consistency, if it is as important as those tiny, tiny little details that we might have got wrong.

00:32:01.933 --> 00:32:07.228
If that's what's affecting the consistency, then you know that's already a finding.

00:32:07.368 --> 00:32:12.710
Yeah, exactly, this is very, very important because we are discussing all the details, and even more than they are.

00:32:12.710 --> 00:32:27.571
If the final classification of a test of a facade or any material depends on tiny little details, I say that a standard is not useful and also look, the new European method is in development.

00:32:27.632 --> 00:32:30.829
Right, you brought some tiny details about cavity barriers.

00:32:30.829 --> 00:32:32.727
They are tiny details really.

00:32:32.727 --> 00:32:35.748
Guillermo brought the corner configuration.

00:32:35.748 --> 00:32:37.566
He said it matters a lot if there's a corner.

00:32:37.566 --> 00:32:38.489
No, no, I said it's interesting.

00:32:38.489 --> 00:32:41.308
It's interesting, but it matters because they re-radiate.

00:32:41.308 --> 00:32:49.872
There's details around windows, there are details about the connectors, there are details about what you do at the end of your building, at the top of your building.

00:32:49.872 --> 00:33:01.845
There's so many tiny little technical details and if those tiny little details can change something from very good to very bad, then we have a very big problem.

00:33:02.026 --> 00:33:02.749
We have a big problem.

00:33:02.749 --> 00:33:04.807
We should not be using that as standard.

00:33:04.906 --> 00:33:11.000
Yeah exactly that's where I found a lot of joy in this project to actually standardize my details.

00:33:11.000 --> 00:33:12.646
I did everything the same.

00:33:12.646 --> 00:33:22.030
That was the first time where we could, as a laboratory you said we build up the knowledge base, so we also build a 25-experiment knowledge base, now more.

00:33:22.030 --> 00:33:35.823
But it was the first time where we had a chance to actually do the Polish test method and build the exact same facade in the exact same way, in a much bigger way, and see what happens First time ever, because I saw smaller tests.

00:33:35.823 --> 00:33:41.905
I saw larger tests it was never the same exact type of construction that I had control over the tiny details.

00:33:41.905 --> 00:33:45.568
It was very important for us and I would like to Sorry, I want to say something.

00:33:45.819 --> 00:33:51.688
It will probably cause a stir in the academics in the audience, but it's not the first time that you saw this, boji.

00:33:51.688 --> 00:33:53.846
This is the first time that it's been done.

00:33:53.846 --> 00:33:59.490
Yeah, okay, in the wall and it's unacceptable that it has to happen in 2022.

00:33:59.490 --> 00:34:02.855
Yeah, it should have happened decades ago that it has to happen in 2022.

00:34:02.875 --> 00:34:03.700
It should have happened decades ago.

00:34:03.700 --> 00:34:06.079
Yeah, perhaps when we were starting thinking about standardizing stuff.

00:34:06.079 --> 00:34:16.195
Well, I had episodes about you brought up SBI and the Euroclass system and Van Muelen told me that facades were next to be standardized.

00:34:16.195 --> 00:34:17.179
It just never happened.

00:34:17.179 --> 00:34:19.568
So it should have happened 20 years ago.

00:34:19.568 --> 00:34:23.728
Anyway, I wanted to bring in Jakub because Jakub Bielaski from ITB.

00:34:23.728 --> 00:34:25.746
He participated in this with me.

00:34:25.746 --> 00:34:33.340
He was an immense support and he's probably the only one who saw all 25, because I think I've missed one or two days of the small elements testing.

00:34:33.340 --> 00:34:38.806
So Jakub probably has even more on-hand experience than me on testing those.

00:34:38.806 --> 00:34:42.750
From our perspective it was very interesting and I don't know.

00:34:42.750 --> 00:34:54.027
Maybe let's go through the results briefly or what we found briefly, and then I'll tell you my experience about doing them, because I think it's an interesting context to the findings themselves.

00:34:54.027 --> 00:34:58.871
Let's try and briefly cover how did these facades behave?

00:34:58.871 --> 00:35:02.108
Can you give us a five, five minute executive?

00:35:02.210 --> 00:35:02.710
update, matt.

00:35:03.000 --> 00:35:39.972
So I think, starting from a purely qualitative explanation, I think what was interesting is, if you just looked at the videos of these tests, you could see that in general, the qualitative behavior was quite consistent, in the sense of pretty much all of the rain screen facades promoted flame spread in every test, more than the ethics Well, more than the ethics and in fact the ethics facades, because of and this was a challenge with measurement it's very difficult to see actually whether spread is happening behind the render or not.

00:35:40.501 --> 00:35:57.967
Only in one of the experiments did the flaming break through the render to the outside, so it was visible, um, and we can actually we'll circle back around to that, but in general you could see this trend of like okay, they're all supporting flame spread, the insulation versus the cladding.

00:35:57.967 --> 00:36:03.706

200 - Façade flammability across scales and standards with Guillermo Rein and Matt Bonner

Listen On

Recent Episodes

Support On