273 - ASET/RSET concept - flawed or great?

Why Revisit Old Fundamentals

Wojciech 0:00

Hello, everybody. Welcome to the Fire Science Show. one of the challenges with making this podcast is that I try to be attractive to both the senior tenured members of our society, but also very open and approachable to the new up-and-coming fire safety engineers who are may perhaps just beginning or perhaps people in between those two. And, you know, it's sometimes challenging to find topics that can serve both of these ends of the audience at the same time. I think that, uh, one group of such topics that actually is that actually could be interesting for both tenured fire safety engineers and those who are just on the beginning of the journey is to discuss some very well-embedded old fundamental concepts of our discipline. I've done some of those in the past, and today I will do one more. Today, we will be talking about ASSET/RSET, the available and required safe evacuation, safe egress times, which is the fundamental concept of fire safety engineering, absolutely fundament... I'm a smoke control engineer. For me, this is everything. This, this is my entire life. I'm dealing with both sides of the equation because that's the way I have to prove that my solutions on the building will work and will provide adequate means of egress to anyone inside the building. The concept is with us for 50 years and, uh, I think it's, it's a great idea but with flaws, especially flaws in implementation. There are also voices like, uh, in 2010 paper by Vyto Babrauskas which are very, very critical towards the, the concept of ASSET/RSET which we'll discuss in here and I don't fully agree with, but, uh, it's an interesting perspective. It's something that's very simple. ASSET/RSET is a very simple concept. It's something that is very useful and also if you don't use it properly or perhaps if layman uses it in a very literal way just as a comparison of the number of seconds on one side, on the other side and is blind to what safety means in the building, you're not really moving safety forward. It can go both ways and, uh, both ways will be discussed in this podcast episode. I hope you stay here with me. Uh, let's spin the intro and jump into the episode.

Sponsor Message From OFR Consultants

squadcaster-e4ag_1_05-27-2026_144803 2:37

The Fire Science Show podcast is brought to you in collaboration with OFR Consultants, a multi-award-winning independent consultancy dedicated to addressing fire safety challenges. OFR is the UK's leading fire risk consultancy that this year celebrates its 10th anniversary. As experts in fire engineering, they are fully committed to delivering preeminent expertise to protect people, property, and the environment. With over 30 chartered engineers and a team of fire researchers at their core, they continually explore the challenges that fire creates for their clients and society so that the best research, experience, and diligence can be applied for effective tailored solutions. In 2026, OFR will grow its team once again and is keen to hear from industry professionals who want to collaborate on fire safety features this year. Get in touch at ofrconsultants.com. And now back to the episode

Wojciech 3:37

I have

ASET RSET As A Simple Promise

Wojciech 3:38

to start with I don't understand why all the evacuation words start with E. Egress, escape, evacuation, and the, the shortcut is ASET, ASET. Uh, I've, uh, had some challenges with which is the correct one to substitute for the E. Is it safe egress time, safe evacuation time? I guess I, I'm at the point where I've learned to not, uh, care about that, that much anymore. I think, uh, following the standardized methodology, escape is probably the most correct one. But it, it's a funny tidbit that, uh, we have three words for the same thing and, uh, they all start with the same letter leading to this. I, I wonder how many legal battles were held in courts about, uh, the meaning of word escape, egress, and, and evacuation. But anyway, I hope you don't end in your-- in a court by-- for your engineering. I hope that listening to this podcast, uh, takes you further away from h-having done engineering, uh, that ends up in the court. So let's talk about engineering. ASET, RSET concept, I, I think, uh, it is one of the most beautiful, easy, and well-crafted messages that fire safety engineers can give to the, to a public or to a layman. Not as the ASET, uh, has to be higher than RSET, w-which is, let's say, a mathematical formulation of the, of the philosophical idea, but the philosophical idea in, on its own. You have to be able to escape the building before the conditions within will cause you harm. That's how I understand this in a broader sense. ASET, RSET. RSET is my time which I need to take to leave the building, which is my baseline. This is, this is how much time the building has to give me because this is the time I need to, to be able to safely escape it. And the ASET part gives me how much in a plausible fire scenario I have to do that. If I have more time than I need, that means that I can escape the building before any harm is done to me by the fire. Not 100% guaranteed, of course, and, uh, local caveats and, and then some other things come into the equation. But in principle, this describes a very sound core engineering principle on how do we provide adequate safety in a building. Um, I always say that, uh, you're not supposed to escape a building, return to a building and escape again. That's not the point of fire safety engineering. If this is possible and we keep the available safe egress time, uh, for a very long time, so the RSET is met with giant safety margin, it just means the building's very safe. But if you are core to the engineering analysis, if your ASET is just barely larger than RSET and your ASET is huge compared to RSET, both of them led to the same outcome. The people have escaped. I'm not saying this is, uh, the reason to optimize for the shortest ASET, uh, just to meet the RSET. It, it's not about that. But, But I, I just want to say that there's more to safety than just escape time. Now, where this principle can go wrong is, when people start treating this as the sole only source of confirmation that the building is safe, and they treat it literal, like up to a second, introduce, uh, safety margins, and, suddenly we start optimizing for seconds.

When Literal Compliance Breaks Design

Wojciech 7:08

This happens to me quite often. There are few types of buildings in which this is perhaps challenging. Um, all the types of, of, of venues where actually it kind of makes sense if you have a venue with huge capacity. I think this, this equation is extremely important and has to be very, very strictly followed. But it happens to me also in tunnels, where I have, uh, cases where I have, uh, the biggest train that could ever stop in the tunnel parked next to a evacuation escape door, which is blocked by the fire now, with the largest amount of people inside the train, of course, and them having to travel the highest distance. And now we're looking into when they escape the tunnel. And with my ventilation system, it's not like once they move past the train, they're already safe. So It's just movement afterwards. It's not escaping fire anymore. It's just, you know, moving away from the train that's on fire. There's no harm possible to them, thanks to the systems that operate in the tunnel right now. But still, I have to show when they exit the tunnel because of the legal description of what fire compartment is and when the evacuation ends in the fire compartment. And now I compare conditions from two ends of the tunnel. In one end, my tenability criteria go down. On the other end, people are still escaping. And someone, a lawyer, can tell me, "Oh, your A-set, R-set is not met because the R-set continues and A-set is lost in this location." Fair, it is, but no harm is ever done. So this is where I see it being very challenging in implementation, and the literal understanding of the concept actually leads to non-optimum safety design. And actually, with, with the classical approaches to fire safety engineering in tunnels, for example, I-- at this boundary conditions, I find it almost impossible to fulfill this, uh, definition, to the letter of the law in the definitions of my fire compartments, et cetera. So I'll, I'll come back to that later on because this is something where we really need to discuss our scenarios.

Where ASET RSET Came From

Wojciech 9:15

But before we move forward, uh, let's try to have a dive into the history of the concept. Uh, in the paper by Vyto, uh, from 2010, the link is in the show notes, um, Vyto describes that the method was first introduced in NIST at 1975 in studies for smoke, uh, sensors, smoke detectors, which kind of makes sense. Uh, we'll come back to that later on. I've always credited, uh, Dr. Leonard Cooper, uh, for coming up with the, with the concept as he kind of formalized it in a paper from 1983, I believe. That's-- That was my reference, uh, of the origin of the formal concept of the ACLR set. And, uh, since then it spread, uh, very widely. It spread into, uh, British, uh, standards, uh, 7974s. It spread into the ISO ecosystem. It become introduced in the SFP, uh, handbooks, in the SFP, uh, evacuation, uh, guides. So it kind of became the staple core concept of, uh, fire protection engineering. I think if you think about it, the principle that you have to have more time to escape the building than it takes for the building to cause you harm, I don't think it's, it's very radical or very innovative one, you, you may say. I mean, it seems obvious in hindsight that, that you have to have that. But to formulate that as a core engineering principle and a target to meet by the engineering design actually takes some cleverness to, to formalize that. So I would like to give, uh, due all the credit to the fantastic people who have brought this, uh, concept to the world of fire safety engineering. They were largely at NIST, so thank you NIST for introducing yet another great thing, uh, to, to the fire safety engineering, uh, society. Now, I'm not sure how this is implemented in your local, uh, law requirements or your local guidances, Poland ASET/RASET was introduced to the law in, I believe, 2009 or 2010, which incidentally is the same year when, uh, Vyto wrote a paper criticizing it. But regardless, uh, it, it got introduced into law requirements for smoke control, where it literally says that smoke control has to keep the conditions on the evacuation routes, uh, tenable for the duration of the evacuation time. Which is an interesting way to put it, I would say, because it allows us to do some things that perhaps if you are core to the idea of ASET/RASET, you cannot do, because here we, uh, have the direct introduction of, of tenability, but the tenability at the evacuation routes or at escape routes. So I'm not checking for it in the remote parts of my building. I'm not checking it in random parts of my building. I'm tracking whether the evacuation conditions are available locally at the parts of the building that people use to escape. And to an extent, it also allows me to understand that if there is a part of the building, if the building is very large, there's an origin of fire, uh, place where people would escape immediately, and the tenability could drop down. But I n-understand that it takes them, let's say, one minute to escape this particular region of the building, and then may take him 10 minutes to leave the rest of the building. The numbers are absolutely made up. Um, I can stop tracking, you know, the, the ASET, uh, in, in the room of origin because I know there should be no wonder they already escaped this, uh, this space. We will, um, come back to that because this is kind of the foundation of a local ASET, local RASET approach, where you track it more locally rather than globally, which I'm a huge fan of. And as again, within this, this definition I have in my system, I can kind of do that. I also understand that in many systems, you would just have a global value for your whole building, and, uh, and that's it. And that's also fair, uh, especially depending on what is, you know, the use of this, uh, ASET/RASET comparison. Is it for smoke control? Is it for life safety, or is this like structural collapse, for example? Those would be also quite different, uh, problems, uh, to deal with and perhaps in, in, in s- especially in the structural end, h- using the global, uh, ASET/RASET would be, um, much better Now move, let's move to the concept itself. It's beautiful for fire safety engineering because it unifies two ends which are very difficult to unify. That is the, the fire domain, the available safe egress time, and the human domain, the evacuation domain, the required safe egress time. So let's, let's discuss some aspects of, calculating those times, um, on, on its own because they, they also come with caveats and with some challenges, and then there's even more challenges to integrate them properly into one ASSET RSET, uh, relationship.

RSET Starts With Choosing Time Zero

Wojciech 14:04

Um, let's start with RSET. Uh, I, I think it's, it's better to start with the, with the human, with the evacuation part. So in principle, required safe, uh, escape time is the time that it takes you to escape the building in case of a fire. First question: What is your time zero? What, what does zero seconds-- What's the-- If you have a timeline of required safe escape time, what is the point zero on that timeline exactly? Because philosophically, I think, uh, the most reasonable point is the start of the fire. I think that that's reasonable, right? The fire has started, and We are beginning our analysis here, but it's also sometimes quite reasonable and useful to declare this time as your detection time, the moment where the fire is detected by the building or even alarm time when the fire is announced by the building. Why, why those three options? So if you only consider something like an alpha T squared, uh, fire and engineering design fire concept in, in your analysis, then probably starting both timelines at zero when the fire is initiated could be reasonable. The issue is with the, the, um, incipation phases of fires. And I've seen those analysis in practical engineering. I unfortunately seen a lot of them, especially in car parks. It happens sometimes that people would take a, a curve from a fire experiment on a, a vehicle, and this curve has like an incipation phase, a phase in which the fire is very, very little. It's not yet detectable by the building. It just lives there and very slowly spreads until an event happens that makes it accelerate and grow to a large fire. And this incipation phase can last minutes, tens of minutes, really. I've, I've seen CFDs for car park where the evacuation has ended before the fire even started. So I, I think there's a massive abuse because these are not the, the same timelines in that case. When you account for detection, it is a little bit better. I, I think detection is the m- the best point to start ASET, RSET. Why? Because the fire has to reach a certain threshold to be detected, um, which means you're not gonna go away with a very, very tiny fire unless you engineer for it, unless you do like smoke aspiration sensors or some-something very clever to make sure that you detect fire very, very early. In such case, yes, that, that makes sense. But this is an outcome of engineering. And on the human part, the fact that you have detected the fire does not yet mean that the fire is announced in the building. Um, you could have two-stage alarms. You could have some additional false alarm, procedures that delay the release of the alarm in the building. You could have automation that just takes time to release the alarm in the building. It's not necessarily the instance the fire is detected. It's not necessarily the instance the information is released and reaches everyone in the building. So this also allows you to investigate this part of the fire safety engineering spectrum in your building. I think this is the best part. And also you could have a time where the alarm is announced in the building. So in this case, you can say, okay, everyone is now informed about the fire and we start calculating RSET. In the fire safety analysis that I do, I would usually tend to alpha T squared fires. I would not have long incipation phases, the zero in my RSET calculation is not necessarily the zero in my fire simulation because the fire has to reach a special threshold at which it's detectable. And then I calculate, establish that and give that as my zero. So the timelines are kind of split between each other, but I believe this is correct. Now what

Pre Movement Assumptions And Bottlenecks

Wojciech 18:12

happens when the, uh, we are past the zero on the RSET curve? Um I will not go into details of how we do evacuation modeling because there have been a, a ton of episodes on that already in the podcast. Uh, but, uh, I would just like to comment on the engineering assumptions that you as an engineer have to do. Uh, the evacuation timeline is not just one time, it's multiple things that are added to each other. This is something sometimes grouped into groups of pre-movement and movement time. Sometimes we go into more details where we could have, time of alarm, time of detection, uh, decision-making, uh, analysis, uh, times, wh-whatever you would like. You, you can complicate this, this a lot more. But the, the-- fundamentally, it's two times. One is the time that it takes you to start moving, and the other time is how long does it take you to move through the space to escape the building pretty much. You-- I think it's, it's fair to simplify it into those two stages of, uh, preparing to escape and escaping split by the moment when the decision was made that I have to escape. Now, as engineers, we have to make assumptions for both ends of this. So for pre-evacuation time distributions, we would go with some assumptions from the published document 7974 Part 6. Or I personally very highly value work of, uh, Rino Lovreglio, which was published, some years ago during his, uh, short research stay at NIST, a very productive research stay. And he, he's published a great paper, uh, reviewing the pre-evacuation time distributions for very specific occupations, which I find extremely useful, uh, for, for my needs. So thanks, thanks, Rino. I'll link, uh, I'll link it in the show notes because it's that good. those pre-evacuation time distributions kind of represent the time it takes for a person to do all the stuff they have to do before they start escaping, which is interpreting the information, gaining the information, doing actions with others. Although it's, it's questionable whether this includes for assistance and, uh, seeking, uh, others. Like if you can imagine a situation where you have, uh, children in a kids play area of a shopping mall and the, the parents currently on shopping and there's a fire alarm in the shopping mall, it's very unlikely that the parents will not try to come back for their children, which definitely is not included in the pre-evacuation time distribution for, uh, occupants of a mall. So, so there are things that would go way beyond, uh, the distributions, but that-that's the point of the distribution. It's a statistical attempt to capture a complex phenomenon into a range of values. We usually have, um, them presented in, in multiple ways. We have the time distribution it takes for the first occupants to start evacuating, and we have a time that takes 99% occupants or all of the occupants start evacuating, and those are usually different times, and we usually assign them as a range with a distribution in between those times. Now it's-- for me, it's important to understand, like, how does evacuation fail? Like, w-what happens that y-y-- leads to situation where you don't have enough evacuation time? So in the evacuation analysis, you, you may have two kind of bottlenecks. One is, uh, congestion at the, the escape routes where everyone starts evacuating and they queue at the doors, at the staircases, at corridors, and basically the queue formation is what dictates the evacuation time. Or you could have, uh, a staircase or, or, or e- evacuation pathway capacity and the length of them, which means that, uh, people have to get through a long, distance, and it's the time they are moving that dictates the, the, the escape. This is very highly linked to the pre-evacuation time distribution as well because if you assume that everyone, evacuates at the last time and then they all queue at the exit, it's not a very plausible scenario to be honest because once the process starts, people will start following each other. So I may be wrong, but I-- there is a concept that, uh, you take 30% of the people in the room and, uh, assign them the whole distribution of the evacuation time or you may have a maximum amount of, of people in the room but you assign them the initial, uh, delay of the first people to evacuate because we we, uh, accept that very soon they will-- the rest will follow. I, I think it's, it's quite reasonable to, to have this distinction because otherwise if you have the room full of people and no one starts-- they, they start evacuating at random and people just stand there. That, that's very, very unrealistic scenario. I would, though I, I, I would avoid, um, suggesting there's a huge realism in those, uh, uh, simulations. They don't capture real human behavior in fire. They are just useful engineering concepts just like ASIT, RSIT is. There can be also a case where, um, the pre-evacuation time is just vastly higher than the time it takes for people to escape, and most of the time, you know, would be actually taken in the decision-making phase. So a case that comes to my mind are-- is hotels or residential buildings where occupants would not be very likely to escape immediately. They could be sleeping, they could be unaware, they could be wearing headphones. And in that-- those cases, the pre-evacuation time distributions can be extremely huge. And, uh, in that-- those cases, like, the movement time is, is sometimes even irrelevant to the escape. In other cases, like, uh, office skyscrapers, the lot depends on the strategy of how you evacuate the building. So whether you let all the floors at the same time, one floor at a time, because here the capacity of the staircases is largely gonna be expired, uh, very quickly when, when you start evacuating the entire building at once, which means people will be just stuck in the giant queue. And in that case, uh, how long it took them to react to the alarm is largely irrelevant. It's, it's relevant on the floor of origin, but for the rest of the building it's, it's not really im- that important because they're gonna be stuck on the staircase for the most of the evacuation time. So different ends of the spectra. However, it's, it's very important that, uh, we make sure that, uh, people start evacuating as soon as they can. In terms of engineering, in some cases, we have to help that this is achieved. Imagine, uh, having evacuation of a sports arena, evacuation of a concert, evacuation of a cinema, I believe that, uh, we can actually help people take the decision to evacuate by inevitably shutting down the, uh, reason that people are in the room. So if it's a cinema, turn the lines off- lights on, shut down the movie. No movie anymore. We just escape. if you give people premise that this is a temporary situation or not really sure whether they can continue their activity or not, they're gonna hesitate in moving, and this can cause significant, significant delays in, in the escape time. Actually, there's, there's a whole science about the human response to alarm signals and, um, evacuation messages, which is studied by multiple fantastic people in the world. I perhaps should bring more of them into the podcast, um, as well. Um, the next thing is the population distribution density. What, what do I take for my RSAT calculations is, is just like I said, it's the either smaller amount of people but with the long distributions or the biggest amount of people that I could have in my room. In my law, it's actually defined by law how many people per, compartment I have to consider, and the numbers are really huge, and it's very challenging. Sometimes we're getting definitions based on the architecture of the building. So if it's a venue for 20,000 people, I'm gonna have 20,000 people inside. Sometimes it's dictated by, for example, a train. If a train has 1,500 seats, um, I'm gonna have 1,500 people inside the train, no matter how popular or congested the line that train operates at is. It's just assumptions, and I think the society is expecting us to, to take the most, uh, challenging yet plausible scenario for this. I'll come to-- back to that later because it's also something that's worth contemplating on. Um, another thing is the scenario. So scenario-wise, we tend to do sensitivity analysis where we try to, uh, block some of the evacuation exits to identify whether closing one of the doors can dramatically change the outcome of the whole evacuation process, and I think it's, it's extremely valuable core engineering exercise to be done. Of course, if you have just a single escape route and a single door at the end of it, if you block it, it's not gonna work, so... But that also gives you information how critical the door is and how much care has to be done to make sure that it is, uh, open for the evacuation process. So broadly speaking, that would go to the RSIT end of the spectrum. I'm not going very deep on the concept. I just wanted to highlight some challenges with the alarm and, and pre-evacuation time distributions because they may have this decisive role on the comparison between the ASIT and RSIT criteria. Uh, how to calculate RSIT, that's another story for another podcast one day. now

ASET From CFD Tenability And Design Fires

Wojciech 27:37

let's go to the ASIT, available safe escape time, and ASIT we usually would do with, uh, computational fluid dynamics, CFD simulations, where we would consider some kind of a design fire for the compartment in which the fire occurs, and then we would track tenability criteria. Tenability criteria are thresholds of different parameters that allow us to define whether evacuation is still possible or not. They're a little bit arbitrary, but let's say based in the fundamentals of fire engineering, the most important of them is visibility in smoke, and I have a lot to say about visibility in smoke, but that's another podcast episode again. Um, important thing, losing visibility does not yet mean you, you perished in the building. There's no dose of visibility that's gonna create harm to you. There are other reasons why people perish in the room, in the room or in the building. Uh, loss of visibility on its own is not yet one, but can lead to others that can cause fatality, which is why we consider it. And due to the exponential nature of the Lambert-Beer Law, the visibility is usually the first one to reach its tenability ratio. Therefore, it kind of became the most important tenability criterion. But we have more. We have temperatures, we have, um, heat fluxes, oxygen concentration, toxic species concentrations. You can track, uh, effective doses, uh, FEDs, uh, on the occupants. So th- there are multiple ways we can track whether the tenability is maintained. We usually track it, as, as I mentioned in Poland, on the evacuation routes, and we usually track it at the height of one point eight meter, which is linked to the height where the, the, the head of the participant would be, which also is very arbitrary and very challenging. Actually, we, we had a whole survey, uh, on tenability criteria done, and there's a paper published in the Fire Safety Journal that, uh, I will drop a link to in the show notes because I think it's a great, resource, especially for, uh, younger fire engineers to summarize all the concepts of tenability and how they are used in the fire safety engineering Now, the second part of ASET, which is very arbitrary, is the design fire. And basically, the performance of systems in your building is a response to the, to the made-up design fire that you put inside. And everything depends on the design. If, if you make it too small, you're gonna have endless ASET. Make it too high, you may lose tenability very quickly. I think the art of fire safety engineering is the ability to pick a correct design fire for a specific occupancy in specific case, in specific time, and make it plausible, and make the assumptions kind of foolproof, kind of, you know, safe. We, we-- for example, we've recommended with my friend Gabriele Vigna, we've recommended sooth hills of 0.1. Not because it's like we have a great physical explanation of why this value is, uh, applicable for broadest group of materials. We likely would not be able to provide such a proof. But for the reason that it's fail-safe, it's due to exponential nature of the law. If you make a mistake, but you have put this value, you, you're on the safe side. So it's kind of an implicit margin of safety, in there. Uh, there, there's a huge responsibility on the design fire, fires in this ASET, RSET, and, uh, I would say we don't perhaps do enough due diligence on verifying those, not just in terms of the size of the fire, but also in terms of the location of the design fire. For example, in a shopping mall, it will have completely different outcomes. If you put the same fire on the highest floor of your mall, on the lowest floor on your mall, is it underneath a projecting balcony? Is it underneath a void opening? Is it within a, a shop, shopping unit that spills the plume into the mall? Those are four completely different fires, and you could have the exact same alpha T square design fire describing all of them, and, and yet they will lead to completely different outcomes. So this scenario definition is, is perhaps the biggest, uh, challenge and biggest responsibility of fire safety engineer out there. There's also the system response, how our systems take care of the smoke. And here I'm mostly thinking about smoke control systems, but th- this also includes sprinkler systems which act on the fire itself to prevent the growth of the fire. It could be detection systems which are necessary to trigger all the, the, all of the smoke control in your building. So th- there's multiple layers of interaction there. In smoke control, it's not just how much you extract the smoke and how quickly you clean the building from the smoke, uh, providing tenability at the evacuation routes, but also how do you input the makeup air into your building because this mixing can actually be a significant cause of interference between the smoke control and the, the fire and the smoke layers. You mix the layer, you lose the buoyancy, the smoke is everywhere, you start losing your tenability. So the system response is also critical in determining the, the ACET time Okay, so now we have, uh, ASET, we have RSET, we can do the comparison.

Babrauskas Critique And Residential Limits

Wojciech 32:52

So let's perhaps, uh, open up, uh, the space for some criticism on, on why the concept, was called a flawed concept by Babrowsky, Fleming, and Russell, uh, some years ago The-- I've, I've read the paper, I've refreshed the paper yesterday, and what the paper is about is really mostly residential scenarios. And I would say that, yes, the concept is flawed in case of those because of the uncertainty in estimating the pre-evacuation time for the RSET analysis. I don't think you can ever do a correct RSET analysis for a residential compartment because there's just too many things to come into that, and it's too specific to the limited amount of occupants that are within that compartment. Unfortunately, in this case, I am with Vyto, the concept breaks very much and is not very useful. In this case, the concept is useful in terms of designing new solutions for alarming people in new, new types of smoke detectors, new types of alarm systems, new types of reaching people with information that they have to escape in-- from the building. Perhaps in the future, we will have some digital twinning capacity. I know the work is already ongoing in multiple places in the world, where actually the conditions within the building could be tracked as the fire grows to provide people with more precise information about the nature of the hazard that they are, uh, found themselves in. And, uh, if we have that type of information, perhaps we ha- could have better kind of impact on the decision-making phase. Because if you tell people, "Okay, you know, the fire in the building is growing so fast, your life will be threatened, uh, and you may have five minutes to escape," that's a more powerful message than just a beeping sensor which you have to investigate what it's doing. But this is not something for today, this is something for, for the future. And at the same time, I think when we engineer large public spaces, offices, venues, tunnels, et cetera, the concept is less flawed. It's not perfect because it misses some of the interactions, but it's not, it's not the worst one. I, I think it's a very useful engineering assumption. So while I, I agree with, uh, Vyto and his colleagues on the specific, uh, set of cases and circumstances they have brought in their paper, I would not yet call ASET/RSET as a flawed concept overall, because overall it does, for me, uh, do its job in places where I want it to do its job. But it doesn't mean that I don't see, uh, ASET/RSET failing. Um, first question is like, what does it mean that the ASET failed? What does it mean that the RSET failed? Uh, you can be very to the line of the standards. ASET fails when tenability criteria is breached. RSET fails, uh, if you-- RSET ends, not fails. RSET ends when people escape the building. Um, but is, is, is this really the ultimate case? Like, if the tenability fails, but in a space where I don't have humans anymore, is this a failure of, of ASET or not? And, uh, it's, it's a philosophical argument. Unfortunately, we're in engineering science and, uh, we're not having space to do philosophy in here. But, uh, it's, it's a very important question. Like, uh, is, is it one global estimation of where the conditions have failed or is it more like temporarily linked with the evacuation process to see whether they failed at the location where the agents are in the simulation? It's, it's very difficult. Uh, I, I would lean towards the second because it's, for me, more natural I don't think there's a building value of ASET. There's a human value of ASET. Uh, every occupant has their own. but yeah, it, it brings some challenges, especially when you start discussing the concept with laymen who, who are not, um, advanced enough to, to go deep into the concept and, and think about it like this. In terms of RSET, imagine I'm escaping the tunnel. It's, it's very close to, to, to me, the, the examples. You have a fire in a train, then you move upwind to the portal of the tunnel. You have passed the train, which is literally the only combustible source in the tunnel that is in there, and you have longitudinal ventilation that is pushing fresh air towards your face in such a way that the smoke is not able to back layer. It's not able to move into your direction anymore because that's the concept of critical velocity and that this is the basis, the fundamental of my engineering design. So basically, as soon as you move beyond the, the train, there is literally nothing in the tunnel that can cause harm to you fire-wise Yet, uh, I'm not sure if I can say the RSET has ended when people move past the, the train. For me, the A- the, the RSET ends when they exit the tunnel completely, and they may have, uh, half a kilometer to move. That could be additional few hundred seconds added to the RSET in-- and, and the fire in the tunnel grows. The conditions next to the train may worsen eventually. And I-I've been in those challenging discussions where I would be said, "Okay, look at the seventh minute of your simulation. The conditions next to the train are definitely not within the ASET range." But I'm like, "Yeah, but everyone in the evacuation layer is already way beyond the train." It's a very challenging discussion, and, uh, I don't have a universal answer to that. That, that's the, the fire safety engineering part in, in, in this analysis where you have to go deep into that case by case and, and figure it out as you go. Another question

Worst Case Misalignment And Safety Margins

Wojciech 38:34

is, so we calculate the RSET. You may cal-calculate a lot of scenarios in your building and find the RSET for all of them, and then you likely take the worst one. You have the ASET. You may do the same. You may do a lot of CFDs in your building, with different design fires, and you may take the worst ASET, uh, you have obtained in those analysis, and then you compare the two. But is the scenario for the worst RSET the same scenario as you had for the worst ASET? And, this again, in tunnels, uh, it's often that, that the worst scenario for evacuation is perhaps a very good scenario for the smoke control and vice versa. The most challenging scenario for smoke control could be very easy for evacuation because how the obstacles are placed against the evacuation routes, what's available, what's not available. So in this case, I find it very difficult to, to have this, considered at a global, uh, level of the entire building at once. And, and something that we need to have scenario by scenario analysis and, and compared to of each other and have perhaps different R- RSATs for different design fires and different scenarios that you're considering in your CFD layer. a beautiful way out of that is also doing multiparametric analysis and consider those times as probability function distributions. As probability functions, then you can compare the probability of ASAT being, this high, RSAT being this high. It does not exclude the possibility that, uh, those scenarios are misaligned, but it gives you a reasonable approximation of the margin of safety you get in the building. Um, margin of safety is actually another concept that's very interesting because in Poland, in some cases, we have explicitly mentioned that there is a margin of safety, but where does it really live? For me, the margin of safety lives in the pre-evacuation time distributions, in the movement speed that I choose. It lives in the design fire that I choose, in the soot yields. Uh, you know, there's where I have my m- design, uh, margins of safety, but sometimes people tell me, "Okay, and now you have to multiply the RSAT time by one point three to be sure that, um, it's safe because we now introduce explicit margin of safety of 30%." When we are forced to do that, we sometimes do this, uh, sometimes we battle very long, but, but usually we do it. Sometimes such a margin of safety can cover you for, additional challenges that are hard to account for, like, uh, people that need assistance or pe- disabled people who have, uh, very different, reaction or movement properties than the rest of general population. Sometimes it's, it's, it's decent to do that, but sometimes it just leads to very, very weird, uh, design. And

Maps Agent Based ASET And Digital Twins

Wojciech 41:16

now finally, how to track ASET, RSET because there are interesting concepts, two, two very interesting concepts that I would like to introduce. One is the ASET, RSET maps, which basically means that, uh, once you do your calculation for smoke, you do your calculation for evacuation, instead of providing a number for the whole building, you present it as, uh, a map. For RSET, it means, for example, from the location where every single occupant has started, you know how long it took them to reach the escape. Of course, it cannot be just a single scenario. It has to be distribution because there will be some randomness in pre-evacuation, et cetera. But with enough, uh, simulations, you can map it, like which parts of the building take longest to escape. And for ASET the same time, the same thing, you can track tenability criteria in time, note the time where in every location of your building the tenability criteria fail, and now present it as a form of ASET map rather than one value of the building. And if you have two of those, it actually provides you with capability to investigate the most onerous locations in your building that you really have to focus in your safety engineering, uh, to make sure that people can safely escape those. It just highlights, you know, the spaces in the building where perhaps the ASET falls the shortest, the RSET is the longest, and, uh, if, if they both happen to be in the same exact location of your building, it's a great target for your fire safety engineering in that space. Um, another concept that I find very interesting and perhaps useful is to track ASET on the occupant. So now we're talking about coupled fire evacuation simulations in which the fire simulation solves the flow field in the building and sol-solves the conditions in the building that people are exposed to. But you don't assess this in relationship to the building or evacuation routes. You assess it in relationship to the agents that are now moving through that flow field trying to escape. What it allows you to do is to track the individual tenability criteria of every single occupant at their own and based on that, have an ASET value that is not the building, value, but rather the occupant value. And if every occupant could leave the building, the exit of the last one is your RSET, and if they never reached inability, it means that the criteria, the ASET, RSET was explicitly met. I'm not sure how streamlined it is to implement. I would, um, guess you could do it with Pathfinder. I think you could do it with, uh, JUPIT-Sim that we've talked about in the podcast a few weeks ago. So there are ways to account for that. I think it's a very attractive way to doing fire safety engineering. It's not a new concept because original fire models were meant to do something like that. It's just very challenging in implementation. And of course, there is the whole, whole, actions, decision-making, uh, layers which we absolutely do not account in this approach. So now for some final words, I think ASET-ARSET concept is very useful for fire safety engineering analysis, and it's here to stay with us for a very long time. It captures a fundamentally important question whether the building provides people with the means of escape sufficiently in a way that they will not sustain harm from a fire. Means of escape is also an important distinction here. It doesn't guarantee you that everyone will escape, but it tells you that everyone had ability to escape, and it was within their own, you know, judgment and decision-making and actions whether they used this ability or not. I-- Okay, it, it, it's perhaps difficult because we perhaps don't tell people explicitly what ability they have to escape. Uh, but this is up to the community engagement and population education where we make sure that people are more aware. Such campaigns are, for example, happening in tunnels in Poland. After some failed evacuations w- in which people have not done what they were supposed to, we had some media campaigns that tell people how to act in the tunnels. This is interesting and perhaps a very valuable, um, you know, thing for fire safety engineering that brings society closer to our ASET-ARSET concept rather than having us to adjust it for the m- misinformation in the society. Uh, another interesting thing is, you know, whenever you bring ASET-ARSET nowadays, people start contesting pre-evacuation time distributions by saying, "Ah, but the first thing people do is to take the phone out and shoot a video of the fire," which I don't have an answer to. That's something you have to ask Reno or, or other colleagues in the human evacuation space. Does it really change the pre-evacuation time distributions that much? Or perhaps when the fire is large enough, they will evacuate anyway. Um, when I'm doing tunnels, road tunnels, my design fire reaches ten megawatts, like, very rapidly, like, I would say within first two minutes of the simulation. Ten-megawatt fire is a fire that I am horrified with, and I am someone who's doing this for a living. I cannot imagine a layman standing next to that fire shooting video of it. Well, technically, you're not even able to stand next to it or hold a phone next to it. So While I agree that there could be a pre-evacuation time distributions of whatever minutes you have, when you're faced with catastrophical fire that is obviously huge, I don't think you're gonna stand there and shoot a video. You're gonna, you're gonna take steps and they're gonna be a large ones, I tell you, my friend. Uh, this even happened to, to us, uh, you know, when we were doing some large experiments and they surprised us. We were taking fast steps even though we were perfectly knowledgeable of, of the nature of stuff that we're dealing with. So that's a, that's a question for the future. And finally, the concepts of mapping, the concepts of agent-based ASET. I think those are extremely attractive. The concept of coupling the building system to a digital twin to evaluate the current conditions and the current likelihood that the ASET for specific locations of the building is this and this to guide the evacuation. Fantastic. Uh, dynamic systems that guide people, especially in skyscrapers, so you make sure that the people who are threatened by the immediate action of the fire can escape immediately to the staircase, which is empty, and then start introduce others into the staircase to empty the whole building if a total evacuation is necessary. and fantastic optimization can be done there. Um, and lastly, you know, making sure that, where, where we have ability to act as fire safety engineers, as in prescribing the scenario of closing a movies- movie theater, shutting down the movie, telling people there will be no ability for them to continue their activity, making sure that they actually escape. I think those little things are critical where we can really make sure that the world, real world scenario will be as close to, to our engineering approximation that we have made before.

Practical Takeaways And Closing Thoughts

Wojciech 48:30

So yeah, ASET, RSET, a flawed concept, a great concept, uh, something in between. I personally highly admire the, the, the, the sense of, of the idea a- and the fact that it has been turned into a useful engineering principle. And I'm very curious, uh, what do you think about the concept? If you would like to comment on it, uh, send me an email, and if you see me somewhere, we are happy-- I'm happy to engage in a discussion about this, uh, with you. So yeah, uh, apologies for the little worse, uh, audio. I'm quite, uh, my throat is sick, and it's really... That's a difficult, uh, situation for a podcaster to not be able to speak properly. But I, I thought I have to deliver Fire Science to you, especially that last week, uh, because of a combination of stuff, I've missed an episode. So here we are today. I hope you have enjoyed this episode, and next week I'm sure to deliver my promise that, uh, new Fire Science is gonna head your way. So, uh, let's meet up there. Thank you for being here with me in the Fire Science Show. Cheers. Bye.