268 - Is critical velocity that critical?

Why Critical Velocity Keeps Returning

Wojciech Wegrzynski 0:00

Hello, everybody. Welcome to The Fire Science Show. Today I'm gonna take you to a journey on critical velocity, not the first one in the podcast, but somehow this topic comes back and comes back to me. I've just finished a round of design on multiple tunnels in Poland, and, uh, that was a very significant part of each of the designs. We're also going into completion of some older projects, and we still talk about critical velocity on these projects. And we have a bunch of projects in various rounds of revision, and guess what is being obsessed in every single of them? It's the critical velocity questions. And, uh, I'm sitting there, you know, tired, feeling that I am wasting my time, that every- around me is, is wasting and wasting hours of time that we could spend on making the constructions we design better and safer instead of obsessing on, on this one point that somehow has became a, a, like, a complete proxy of success or failure on tunneling projects. And I, I don't know why. Like, I... Now with, with the experience that I have, I see so much more in what could have been done, and I see less and less importance of obsessing on, on, on the precise third digit of this particular variable. and of course around us there's some things happening with the numbers as well. Uh, there's papers published on new models. There's papers published on why FDS is perhaps not suitable to calculating that. So it's also a lot of boiling in the academic part as well. And, uh, having this clash with my project experiences, I'm in a very weird position. And, and that is what I want, to talk about in this episode. I want to talk a bit about where did we get to a point where we are today, why I think this particular metric or just having a, a verifiable easy metric is important for people who double-check the projects. I would like to talk about the design situations and how do we get into a position where this number gets into equations and starts driving the design of our system. And at the same time, what are the margins of safety that we silently impose across the spectrum, making this one particular variable less and less important. and finally, I really want us to appreciate, you know, the consequences of our design choices in terms of what needs to be built in the building to support our needs, to support our numbers. Like, how does it look like to go from a drawing board, from a concept, from a number of Newtons, a number of jet fans into real world tunnel where this has to be, be built and, and what's the consequence of those choices? I hope it's gonna be an interesting episode. I'm not gonna ramble just about how bad critical velocity is, I promise. It's, it's gonna be a broader introspect-- it's gonna be a broader insight into how I carry my everyday's work. So if you would like to join me on that, you're very welcome to do. Let's spin the intro and jump into the episode. 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

The Seduction Of A Single Metric

Wojciech Wegrzynski 4:45

So here we are talking about critical velocity once again, huh? Haven't done a solo episode in ages. I hope I didn't get out of shape though. I'm well-rested, had a great summer. I hope you guys-- I hope you have had a great summer as well. You've rested a little bit, and, uh, you are now reinvigorated to do all the hard fire safety engineering projects or research out there. Uh, for me, summer was very, very interesting. Midst of the summer, I saw a little bit of critical velocity, you know, drama, let's say. Call it drama, perhaps it wasn't drama, but, uh, it was interesting. A team of Conrad Stacy again has published an interesting, uh, piece on ability or inability of FDS to, to predict critical velocity with some claims there. And, um, And there was also a response, uh, to, to, to that piece. There were some changes done to, to the code. Not sure if that's fully solved or, or fully closed. I'm gonna cover this a little bit more later on. But again, even at, at the summer, the critical velocity has somehow caught up with me. And while spending my time with, uh, family, I still had to respond to at least three different projects where, uh, verifiers had different opinions on what the critical could be on the project. Uh, so how, how did we get there? That one single number, a product of a calculation, becomes so obsessively followed, checked, I don't even know how, how to claim it in projects. And at the same time, I, I find it, to some extent, irrelevant. From my perspective, like, i- if I had to give you my opinion about critical velocity, its calculation, and its use, I would say just, like, assume it's three meters per second and be done with it. And, and just assume 3.0 meters per second for literally any project, and you'll probably be 95, 98% there, and you'll have a good design, and you'll not have to bother with all the, you know, ambiguity of, estimation of that number, and you'll be in a good position. And I'll defend that, uh, later on in the podcast. But, uh, f- first I, I, I started thinking about it, you know, from a kind of philosophical point of view. Why there is one number that carries so much impact and importance? W- w- why this particular number? Why not the total thrust of my jet fans? Why not the total number of my jet fans? Why not the power consumption? And the thing is, it, it kind of gives you the premise of safety. The, the number itself represents a state which, to some extent, could be considered a transition from an unsafe state to a safe state. I mean, that could be assumed. I know that it is not true. But I, I think in a broader context, especially when we go outside of our profession, outside of, you know, detailed, uh, understanding of the physics of the fire, all the flows, et cetera, I think it, it, it gives a lovely premise, you know. This is the velocity at which you have gained the complete control over the direction of the smoke in your tunnel fire. That's a powerful premise because it, it kind of assumes that upstream where you support fresh air, there is, like, infinite promise of safety. There, there's like the available safe evacuation time upstream by definition has to be infinite because if you have reached this velocity, the smoke can never travel upstream. That, that's kind of the promise that I think people find or people seek in the, in the critical velocity. And with that number, it kind of, you know, provides you this binary distinction of what is good, what is bad. Less than critical velocity, bad. More than critical velocity, good. And on top of that being such a sharp distinction, it also is kind of calculatable. You know, y- you don't have to have a PhD in fire safety engineering to calculate the value. It's, it's actually fairly simple equations that you put some variables in and you get a, a number. Voila. Here you are. Tell me if your number is bigger than my number. Now, and it's not the only number that carries this promise of safety in the engineering world. We have fire resistance, which is expressed in minutes and also carries a, a promise of structural safety. We have visibility in smoke, which carries the promise of being able to evacuate We have all those various simplifications of complex fire physics that in the end try to simplify a set of complex phenomena into, into something you could express with, with a single digit. And this is like, in a way it's, it's beautiful in, in fire science that we have this set of rules in which even non-being, not being a fire specialist, you can to some extent make sure that yes, this is fire safe. But at the same time, you know, when you go deep into the design of that number, into the design problem, actually arguing around that number becomes pointless because If I, if, if my velocity, critical velocity for a project is three meters per second and I somehow achieve 2.9, does it mean a failure of the systems in the building? Of course it doesn't mean a failure of the systems. Does three means unlimited, unrestricted safety in all cases? Of course it does not mean that. So o- once you get deeper into the meaning of the number, which we will in a, in a second, this promise kind of breaks. So the more you know the less optimistic you are about the number. And, now in the reality of my projects, the people I, I'm dealing with, the people who are the verifiers on my projects, uh, engineers o- on the projects, other branches on the projects, authorities, firefighters, et cetera, all the stakeholders that I need to engage and remember I try to be a great communicator. I respect my stakeholders. I respect their needs. I respect their boundaries. I respect their goals and ambitions, and I try to make sure that my project speaks to them in a way that is comfortable for them. So, so they can also calculate the number. The issue is that there's many ways to skin a cat and there's many ways to calculate, uh, critical velocity. Of course, if you have the exact same set of, uh, variables and the exact same dimensions of the tunnel and you use the exactly same equation, you'll get the exact same, uh, result which is, which is fine. But the, the issue is, uh, projects take a decade. I could have used NFPA 2017, someone can, uh, validate that with equations from a newer version and here we have a discrepancy. you have to calculate, uh, some variables like hydraulic diameter for example and, and you may get a number and someone could calculate it slightly different and get a different number. I used a different section of the tunnel or length of the tunnel to average my slope, And you have averaged over a different distance. We may end up having a slightly different local slope values You may actually use CFD to establish that value and once you use CFD to establish that value and someone else puts an equation, they're gonna get a different value and, or they can use a different CFD code and get to a different value. I find it almost inevitable that eventually on projects we have some slight variations of that number and let's say I've put a number 2.9, someone calculated, "No, no it should have been 3.1." Uh, this becomes madness to solve. It, it, it often puts us back on a drawing board adding jet fans to the project just so we can then compensate for this lack of critical velocity. And this I find a total madness. We have not only wasted time at this point, but we've also wasted a lot of resources at that point. So, uh, how do we get there? How do, how-- Like I know that not all of you are tunnel designers. I would be surprised if all of you were tunnel designers. It's a quite a specific niche in the fire safety engineering, an interesting niche. I highly recommend working in this space. It's a little bit stressful. w-what's critical velocity? Let's, let's go

How Critical Velocity Is Defined

Wojciech Wegrzynski 13:55

back to basic concepts. Critical velocity, when you have a tunnel, there is this, this assumption that, uh, for eff-efficient ventilation of most of the tunnels, you can use a concept of longitudinal ventilation, which means you blow from one side, you extract on the other side. And that's exactly it. That, that's the complete thing. You are supplying air from one end of the tunnel, you are removing smoke from the other end of the tunnel, and the smoke should go in just one direction. It should not come back upwards. That's the, that's the premise. And there is this assumption that there exists a specific velocity that's tied to the size of the fire which you choose, uh, that's tied to the dimensions of the tunnel, the, the height of the tunnel, the emitter of the tunnel, the slope angle of the tunnel, which guarantees that this flow is unidirectional, and this is what we call the critical velocity. There's multiple ways to calculate that. You can use the equations from NFPAs. Historically, um, it was introduced by Abel of Thomas, and Williamson revisited it. Uh, we, we actually had a whole podcast episode with Konrad Stacey and Michel Beyer on revisiting critical velocity, where we actually go deep into the technicalities of how this was established through experiments in Memorial Tunnel, through mathematical equations, and how a team of, uh, of, of Konrad has revisited this, this, this thing. Because there's also another model, the Konrad Stacey model, which, uh, which also, uh, I think is recognized by NFPA, which, um- takes a different approach than the previous equations to, to calculate it. It focuses more on the heat release rate per unit area and kind of the intensity of the fire rather than the total calorific, um, burning rate in the moment. And they, they put a very convincing, uh, you know, case where if you-- even if you have a hundred-megawatt fire, but it's stretched over a few hundred meters, uh, the intensity along that fire is, is negligible, whereas the total calorific, you know, power of that fire is huge. And, and in that case, the, the traditional concept would completely break, while their concept kind of recognizes that. So that, that's a, that's a thing that, uh, resonates with me. You also are able to show critical velocity, uh, through the means of numerical modeling, and that's what we usually tend to do. and this means we have to provide kind of a numerical proof that once the fire reaches its full potential, the system actually carries this smoke unidirectionally, and we do not have this thing called back layering. McLaren is just a num- name for a ceiling jet that goes against the stream of the incoming air. So You, you don't have this air coming back. It means you've reached the critical velocity. Now, with CFD, it's a little bit difficult, uh, in terms of you think of critical velocity as a single number, but I do not view it like that. if you consider tunnel as a one-dimensional object, you know, it's, it's literally a pipe, and the conditions across the pipe are uniform in, uh, in, in height, in width, and it just changes in depth. You know, it just changes in one direction. And, and everything's uniform in the cross-section of the tunnel, then perhaps, yes, this would make sense. But, uh, tunnels are complex and, There are a l- a richness of fire phenomena physics that, uh, take place in tunnels, and also the momentum exchange between two interacting streams of air, especially when one of these streams is, um, a victim of Coanda effect because the, the jet stream, the, the ceiling jet stream does, you know, kind of stick to the ceiling of the, of the tunnel enclosure. Wh- when you have this momentum exchange between those streams, it kind of, you know, depends on when you capture this exchange. In other words, if you have fire grow to a full capacity, create a ceiling jet, and then you start your ventilation to fight that ceiling jet, uh, or especially if the fire was also supported with wind that pushed it even further against your, uh, your ventilation, then you need to use a lot of force to push that jet. But if you had captured the fire when it was still small and it was growing while you had these fans blowing perhaps the sling jet has never formulated perhaps this physics was never there and and and those two velocities one to you know push the fire away from its full potential and and a velocity which i would say is the minimum velocity in which you still hold the fire in its place those can be two very different numbers and it's going to vary project to project it's going to depend on the size of a fire depend on the initial conditions of your analysis wind etc but definitely those numbers could be different those numbers could be different and keep in mind the reviewer wants one single number one single number to put in the project verify critical velocity is this much and i have calculated this different value now we fight um so with with this it's not even mistakes in the calculation process it's it's just the physics it and i had so many projects in recently where, where actually th-this was quite a problem. Like I, I can-- out of my head, I can give you three. Like one, we, we were dealing with this critical velocity, uh, for a tunnel. We designed the system, it matches what we expect, it delivers nice performance. Uh, but the, the, the tunnel can also have, you know, bi-directional traffic in emergency situations. Like there's-- it's a double tube tunnel, but there's a chance that one tube is gonna be closed for whatever reason, and then we have to do bi-directional, um, you know, flow in the other tube. And perhaps we want to reverse the direction of the, of the flow in the tunnel in that case. And in that case, in that tube, the slope is working against us, so we are dealing with completely different forces. And suddenly in my hand calculations is insufficient. Like I, I now lack my hand calculations when I put everything into the equations. My critical velocity estimated for the tunnel was, let's say, two point nine, and from my hand calculations, I see I will gonna-- I'm, I'm gonna have two point four-ish. And wow, that's an issue. Like for an emergency mode, reverse

Why CFD Refuses One Number

Wojciech Wegrzynski 20:55

mode, where the fans are obviously also less, um, efficient, now I start to not have sufficient amount of flow. I put it in my CFD because I, I understand, uh, some of that physics, and I understand that if we capture the fire early, we, we introduce some scenarios where we start accelerating the flows earlier in the case, and, and we actually demonstrate that, that it actually works because, uh, uh, we can maintain an in-directional flow also in this reverse configuration in this tunnel. But the reviewer is unhappy, and we spent l-- I would-- you would not believe the amount of hours spent on, on, on, on solving that, even though everything was fine. Another example, I'm designing, uh, a tramway network, tramway tunnels, and there's a station that connects to the tunnel. The station has a completely different cross-section than the tunnel, and now the fire is like literally at the edge between the station and the tunnel, but within the tunnel. Uh, and now I have some ceiling jet formulating and, and kind of moving towards my station, but then as soon as it reaches the station, the air captures it and, and takes it back to the tunnel. So I have this dreaded back layering W-why I do have that? Because the, the velocity, if you have a unified cross-section of a tunnel, that's great. You can have a pretty much unified velocity in the whole cross-section. But if you have a sharp change in shape of your tunnel, like transitioning from a station into a tunnel section, at this very point of your system, due to how the project was designed and other, you know, architectural features of the projects, it's literally impossible to have unified flow field in that very specific point of the tunnel. If the fire was 100 meters deep into the tunnel, no problemo. There's, like, unified velocity there, and will reach whatever you would like to reach. But here at the boundary, mm, it's, it's just impossible for me to create a completely even, flat velocity surface, if I may, in that space when there's a f- a large fire just next to it interacting and messing with everything. So yeah, I, I don't think I can ever reach critical velocity in that particular space, which doesn't mean I have failed as a designer and did not create system that protects both the station and the tunnel. Uh, yet it's a challenge. Another one, I'm dealing with a, a train station this time, quite a large train station, and the tunnel that connects to the train station. And obviously, uh, the tunnel is quite a huge beast, uh, in terms of width of the tunnel. And now my question is, like, i-if you have model that has been developed for, uh, a memo-memorial tunnel, or we have data from Roona Humber, we have data from a facility in Hihong, those are all fairly small tunnels, and I'm designing something that, that's, uh, order of magnitude wider. Does it still hold? Like, because if I have to design the amount of air that, that gonna provide me this critical velocity in that giant space, in those circumstances, like, I m- I probably have to build a power plant next to the, next to the station because there's no way I can supply this much power in there. And, and yet, at the same time, I know I-- if I don't obsess about the number, but I obsess about the safety and driving the smoke away from the, from the station into the tunnel and extracting it comfortably, I'm gonna achieve success. So yeah, those are just examples from, from my recent projects where Where this rigid thing doesn't really tell you the whole story. It-- You thought that it's a proxy of safety, it's a proxy of success, but in the end, it, it's not matching the project's reality, and yet I have to deal with that So now let's talk about what happens if I don't achieve that success. When I don't get that critical velocity, w-what exactly are the consequences of me not having this exact number in place? And, and I can miss it by, in a two ways, and one, I have underestimated the velocity, and it, um, in reality, it should be-- it should have been higher, or I have underestimated the fire, and in reality, it kind of grew bigger than my design assumptions. W-what happens then? I started with an opening that it kind of is a proxy of success, a binary proxy of success. You either have it or you don't have it. That's how it is often interpreted, but that's not the whole story. Um, it's not one single transition point between good and bad. It's a turning point at which you start have some sort of formulation of back-layering in your, in your system. And now, does back-layering immediately mean failure? And I don't think so. I, I don't think having some sort of back-layering being synonymous with the failure of the ventilation concept in a tunnel. back-layering is a ceiling jet. Tunnels are rather tall spaces, like especially if you design train tunnels, inevitably you're gonna have a very, very tall space. If you're dealing with, uh, TBM-borne tunnels, uh, you probably are dealing with quite tall spaces as well. So in those circumstances, uh, having a back-layering formulated just means that some of the energy of the fire, some of the smoke generated by the fire is moving upstream, upwards for a specific distance. This distance could be quite short, twenty, thirty meters. This distance could actually be quite severe, like few hundred meters. Of course, those are different, you know, levels of a failure. If the back-layering distance is 20, 30-ish meters, I don't think this is critically bad from the life safety perspective. Usually, you still have a stream of cold air underneath. You still have retained the visibility up to the point where the fire is located. Like, you can see there's quite a stable system where the smoke moves up and then i- is captured by, by the incoming air and, and is kind of pushed away anyway. In that case, I don't think the back clearing represents a failure of the safety strategy o-o-of the building, and it kind of is, is, is liberating in, in that-- in the fact that it's not a sharp number. If your back clearing is, like, 300 meters long, then, then of course this is usually pretty bad because the smoke will get mixed at the edge. It's gonna cool down. You, you will probably have the smoke, um, filling your entire cross-section of a tunnel eventually before, uh, the spot of the fire. This means any rescue operations upstream are difficult. This means approach to the fire is almost impossible. This is a failure of a system, and you can s-see quite the distinction between those two, you know, failures, uh, in, in-- if you do numerical modeling of the fires. Um, now what happens if a fire grows beyond the heat release rate that you have designed with? Um, first of all Vehicles are discrete packages. Like, there's a size limit to a vehicle, and also usually if you have a very huge fire, it either is an, a spill of oil or a very long truck, you know? And if, uh, if Stacy Bayer brand model is correct, it's about the intensity, not the total calorific potential. Therefore, if you have a part of the truck burning at maximum, and we perhaps have the whole truck burning at maximum, but it's at the length of the tunnel, the difference is not that tremendous. That's a very, very promising, uh, you know, uh, finding. in the classical take, even if you increase the size of a fire from, let's say, 100 megawatts to 120, 150, uh, the change is not as drastic as you would thought. So actually, the system is not really immediately losing that capacity like blink of an eye, on-off version. You've crossed the boundary of, of when the system can work, uh, and it stops working. It, it gives you a, a f- a ton of intermediate states. We actually had that one in one of the tunnels. It was actually, um, not longitudinally, but a transversely ventilated tunnel. So a different concept

Real Projects That Break The Rule

Wojciech Wegrzynski 30:05

perhaps the, um, critical velocity doesn't apply really in here. But we're also observing, you know, the smoke spreading across the, um, the parts of the, of the tunnel, how, how long the smoke layer is. And we, we-- because we had a failure of one of the fans that were extracting, and it was the major capacity loss for the system, we were doing a parametric study on what does it mean for the building. It was initially designed for 100 megawatts, and w-we have shown that up to 50, there's like literally no difference. Like the safety is the exact same, and there were a minor difference at 100 megawatts, where the lengths of the streams was a little longer, but still not at the point where the system would totally collapse and then safety would not be provided in the building. So yeah, there is a lot of, uh, capacity within the concept of, the critical velocity, but that's not our whole margin of safety. And oh boy, we do have an amazing, amazing margins of safety in our smoke control designs in tunnels, and those all come from the design situation. So, uh, when I design my tunnel, the expectation is, is that I achieve the critical velocity. Yes. To achieve the velocity, what do I have to battle in the tunnel? What, what are the forces that work against me? One, there's a resistance of the tunnel tube itself. There's gonna be obstacles, the vehicles in the tunnel. The fire itself is a huge obstacle, so to push through the fire that you need a lot of force to actually do that. Uh, fantastic work by Ingo Ries, uh, on, on that. Uh, there's gonna be some forces coming from the chimney effect in the slope of the tunnel, which can work against you. There's potentially gonna be wind that's, that, that can also act against you. Um, you also have sources of inefficiency in your ventilation, which means, if the jet fan is blowing air at 20 degrees, it's in ambient, it, it has its full potential. But if the jet fan is downstream, and some of them will be downstream, it's, it, it pushes hot air. Hot air which is less dense, which means there's less momentum transfer from the fan to the flow of air, which means those fans downstream become less and less efficient. Therefore, uh, there is a very specific design, uh, you know, scenario in which the performance of the system is at its worst, and that's when you have the fire quite early in your tunnel, which means all the jet fans are working in their worst, uh, possible conditions. Fire that's blowing down, so your tunnel must be going down, so you are fighting against the slope. Uh, you have the maximum potential of the fire, of course, and you have wind acting against your system with some uh, maximum, you know, power as well. And that's usually my, my design scenario. But if you, if you think about the probabilities of that, it, it becomes like ridiculous, and I love to show that on conferences recently. Um, first the fire. Let's say a fire in a tunnel is one every three years, and 100 megawatt fire is gonna be one out of 100 fires. That's very, very rough statistic. You can go much more detailed on that, taking, uh, real data from, uh, different tunnel networks, but that, that's roughly it. So one every 300 years, that's a 100 megawatt fire in your tunnel. Now, uh, the fire is at the beginning of a tunnel. You can assume that the probabilities may be linearly distributed along the tunnel, so every place in the tunnel has exact same probability. That's probably not always the case if you have, uh, sections of your tunnel which are wider, that's perhaps more attractive to stop at, and we know that from a Warsaw tunnel. But you could assume there's linear probability that, that the truck can stop anywhere in the tunnel. So, What are the chances that the, the vehicle is gonna stop exactly at the beginning of the tunnel? Uh, that's also a number. Um, then you have, uh, wind. What are the chances that the wind effect's gonna be considerable? And here we're usually designing it for 95th or 98th percentile wind, so it has 5 or 2% chance to, to be there present exactly when you consider that. And from the evacuation point of view, you also would like that event to happen in maximum traffic because, hey, if there's no vehicles in the tunnel, there's no one to save, so why I'm calculaking- calculating asset, asset? Of course my tunnel is filled with cars because this, uh, this dreadful fire happened in the peak traffic. Uh, so n- now you start adding those probabilities or multiplying those probabilities, and you can get to a ridiculously low number like 10 to minus 6, 10 to minus 7 probability, yearly probability of this particular fire to happen in that situation. And yet that's my design situation. I'm happy designing for 95th percentile of wind. But if I take the whole design situation, I'm designing for 99.999 percentile of, possible combinations of different scenarios. And I've, I've even done, uh, you know, um, a probabilistic calculator which, uh, just took those variables out of a spectrum and, uh, put them into the equations and calculated how much jet, how many jet fans, how much power I need to, you know, achieve critical velocity in that case for those randomized, you know, uh, input conditions, which were of course, uh, ranges that, that, that were reasonable. And I've run a, I believe, Monte Carlo of 1 million simulations on that, and I found that actually to go from 99.999 to 99.95 I could reduce the system in the tunnel by a third. Can you imagine that? I could reduce the amount of devices in the tunnel by a factor of third. One third of the system was only to bring my confidence that I can achieve the critical velocity from ninety-nine point ninety-five percent of scenarios to ninety-nine point nine nine nine. Like, we're talking about residual numbers in here, like ridiculous, and that's a third of the system. And, and, and now I approach a project with that in mind, and I argue with someone about zero point one difference in the critical velocity, which results in like fourth, fifth significant digit change in the, in the thrust of my fans. And believe me that I've lost those battles sometimes and had to redesign the system and add jet fans to compensate for that. How ridiculous is that? Well, now you could accuse me, "Oh, Wojciech, but it's all for the safety. We're all doing it for the sake of safety. Why won't we increase the number of jet fans so we are safer?" One, we're not safer. Like, that's invalid argument. Like, if you are all about safety, where are my sprinklers? If safety was truly the objective, uh, there are ways to increase the safety in this, in the system. And, uh, I very rarely am able to push the discussion to that level where we truly discuss the, the metrics of safety and, uh, ways to introduce safety of the tunnels because of the contractual requirements. And, uh, two, people severely underestimate what it means to build a ventilation system in a, in a tunnel or in any facility, if I may. You could mistakenly think that it's just, you know, the number of Jet Fans that's changing. Just add a pair of Jet Fans. How hard is that? Well, it is pretty hard. The ripple effect of, of that design choice is tremendous. One, you have to physically install a, a second part of Jet Fans. That-that's a, a cost for the Jet Fans, that's a cost for the installment. Yes, the-- you, you have to physically install the fans. You, you, you would think that it ends there, but oh boy, it does not. Second thing are the cables. Uh, the cables are ridiculously expensive if, if they are supposed to be, um, fire-rated, and especially if you have large distances, the resistance on the cables, uh, the... it, kind of becomes a design problem. You need to use really large cables, and sometimes you cannot even reach the distances that you need because I cannot put my new Jet Fans ten meters away from my old Jet Fans. No, I have to separate them in distance. I have to put them a hundred, two hundred meters away, which potentially means a lot of cables. And I've been in the projects where adding, uh, or moving the groups of Jet Fans that was required has resulted in us having to put a medium voltage station in the middle of a tunnel because we were not able to efficiently power them with low voltage, which means I have a transformer in middle of my tunnel. I've introduced an really dangerous piece of equipment in the middle of my tunnel for sake of safety, but there was no other way to solve the problem. It was not only for ventilation, it was also for pumps for water, but, uh, here we are, a medium voltage station in middle of a tunnel Um, for those jet fans, if you want to have any sort of variable control on them, they will need frequency

When Missing It Is Not Failure

Wojciech Wegrzynski 40:15

inverters. That's a huge cost. You need a room for that. And also the invent- inverters or even any startup system, they're gonna impose the, uh, limitations on the length of the cables that you're allowed to have between the inverter and the fan. So it's not possible to put an invent- inverter five kilometers away in a control room. No, no, no. You have to find a way to put them into the tunnel somewhere where they can serve the, the, the jet fans as well. That's a hell of a design problem. Um, power itself The one thing is the power consumption of the jet fans. So now you have to have an increased capacity of your electrical network to supply for that pair of fans. And my experience is that we are always lacking power on the tunnels, especially when they are deep rural tunnels away from the city. In that case, it's very difficult to, to get sufficient power, and increasing the power demand is-- that's a difficult discussion to have. But that's, you know, just the power to, to, to, to have it on the site. You also need backup. It's not one source of en-energy that you have to have. You have to have a second source of energy that also meets this increased demand. That's a hell of a cost. From the user perspective, that's another pair of jet fans you have to maintain. That's another pair of jet fans that can break. So the tr-triple effect of placing this another piece of equipment into the tunnel, it propagates through all the branches of the tunnel design, and consequentially, this can be a massive, massive change in the design. And as I said, because sometimes I have not, uh, um, reached the number that the other person has reached, uh, here we, here we are. We, we have to design for that, and that-that's quite, quite problematic. So yeah, if I had a magic wand and I wanted to live in a perfect world, how would I deal with that? Um, as I said at the beginning, I wouldn't really detail that much on calculating the velocity. I think value three for general projects, value one point five to reach confinement velocity, which is velocity at which you may have some back layering, but it's usually controlled. That's also a fair number. With those numbers, I-- as a starting point for my designs, I usually have achieved a good degree of success. Of course, I don't end my design on saying three meters and we're done. I go into detailed CFD and investigate some stuff. But, but, uh, to start with and close the discussion on the critical velocity three and then we're done. Then I would go into explicit CFD modeling or one-dimensional modeling to see the outcomes of that assumption across a range of scenarios to identify whether in likely scenarios of a fire, let's say for the 98% of design combinations, I really have sufficient margin of safety and what-- how close I am at the rest of the design scenarios. And well, usually the choice is not there. Usually, I still have to have the all my numbers match for the 99.999 design case anyway. But in the perfect world, I would probably just drop this on a risk analysis and, and, and see like how far I deviate and what are the consequences of that. And, uh, then I could be happy with that. I could do it further, and that would depend on the type of a tunnel I'm designing and some critical variables that, uh, really drive the safety of the building. One would be the slope. So if, if my tunnel is-- has severe slope, I would probably have to introduce some measures before I start my jet fans to gently move the, the, the flow in the tunnel and battle the stack effect. Um, for slope tunnels, that, that, that would be majority of my work on the tunnel. Uh, a lot would depend on the, on the traffic. So does it have high traffic volume or low traffic volume that influences the probability of the fire? Does it have significant, uh, heavy goods vehicle traffic or not? Different tunnels require different goals. I-i-if it's a railway tunnel, it's a different thing than a road tunnel. So the type of traffic is gonna drive a lot of the decisions as well, and perhaps the-- my comfort zone is gonna be defined in there, which of course I cannot say that it's within my comfort zone. I have to define that in the project outline and in fire strategy and explain why I think it's good or bad. Um, finally, the physical size of the tunnel, number of lanes, for example, that's a good measure of resilience actually. If you have a very wide tunnel, we're designing very, very wide tunnels in, in Poland right now. I found that, uh, in those tunnels, actually the consequences of fires are much less severe than if you have just one lane and the, the fire can block the entire cross-section of the tunnel. In Warsaw, to block the entire cross-section of a tunnel, you would have to have like literally a tanker fall on the side and like turn 90 degrees against the traffic, which is obviously not a typical scenario you would be dealing with. Um, and, uh, if you have that, a lot of people will escape the tunnel. And, and that's another thing, you know, the escape strategy. We're doing all of this to provide safety and, uh, the way how people escape the tunnel is not necessarily how we would like them to escape the tunnel. Uh, in Warsaw, we had a truck fire in a tunnel, and most of the people escaped the tunnel in their vehicles, you know, driving out of the tunnel, which was not what we have expected. We wanted them to leave the vehicles and escape on foot through emergency exits. Completely different thing. And, uh, yeah, one thing about that as well, I haven't mentioned that, but if vehicles move through the tunnel when your ventilation is working, oh, that's a mess because vehicles can very easily overwhelm the power of smoke control system and natural flows in the tunnel. The vehicle-induced flows in the tunnels are extremely strong. We've done an experiment in Warsaw where we had, um, flow introduced by very strong wind of the day. So we had four meters per second in the tunnel of flow by the wind. That's very, very large number. And we took 20 vehicles, and we're driving against the wind with those, uh, 20 vehicles. And we're

Hidden Safety Margins In Design

Wojciech Wegrzynski 46:55

talking about really huge two-kilometer-long tunnel. 20 vehicles is nothing. And we were able to slow the wind in the tunnel from four meters to two meters per second. That's a huge difference. And therefore, I'm absolutely sure that vehicles can overwhelm the system. And people drive in the tunnel when the fire is happening. This changes a lot of things, and again, it's something we choose to not concentrate on. We rather detail the critical velocity. That's how I would approach the projects if I have the magic wand and I could deal with them. Reality is I'm here dealing with critical velocity. I have to answer a question in five minutes after I finish recording that about critical velocity, and I will be dealing with critical velocity for end of my days as a tunnel designer because In the end, I understand that the other stakeholders have to have something that's rigid, that's verifiable, that's a number that you can run against, a performance metric that's precise, that, that, that, that's exact, that's not ambiguous. And fire safety is everything but exact, everything but precise. It's a collection of states, it's randomness of the fire, the time of a day, number of people in the building. Every fire has a population of one. So we're kind of fitting an extremely complicated system into a very rigid boundaries and on, on the, you know, the clash of those two approaches, you, you end into frustration like the one I demonstrate today, uh, inevitable. I understand that, that the reviewers need that. I helped some people write some clauses of code recently in Poland and also we needed to, to get a measurable, you know, number because they, they need means to verify that. I just wish that, uh, beyond the metric, we could also engage in conversation about the safety and that the fundamental goals of the safety of the project would be kind of higher in the rank than just one single number that is there written and that, that we can argue about. That would be, that would be my actually more realistic dream and, and I wish we eventually get there and I wish that you in your fire engineering journey will be able to, to procure fire safety engineering concepts on that level and, and with that freedom. I w- I wish you that wholeheartedly. So that-- I think that would be it in, in that, uh, podcast episode. I didn't really go into the FDS modeling, uh, thing. I, I told that I'm gonna do that. There was, uh, you know, a claim that FDS overestimates the critical velocity based on memorial tunnel simulations and, uh, apparently it overestimates it by... from, from Conrad Stacy's paper quite a lot. And that, that's quite a, a challenging claim because, uh, as I said, if you overestimate the velocity by a lot, you will need more fans into the system and those more fans translate into all the issues that I've mentioned. And it also creates a problem where if I do my CFD for tunnel in Ansys and someone cross-verifies that with simulations in FDS, we might get to completely different conclusions and that also kind of happened to me, uh, some time ago. So that, that's probably quite the challenge. I don't know how it can be resolved. I, I'm not sure if, um, big effort is taken at FDS to to change that or revalidate that. There's of course a, a whole lot of, uh, you know, uncertainty about the Memorial Tunnel and, uh, it's, it's a part of an ongoing discussion. So I'm watching this space very closely and while I'm in lucky position that I, I'm using different set of software for most of my tunneling projects, I recognize this, uh, challenge may be quite a problem for a lot of people in the industry. So I would be l- I would love to see how, how this is eventually resolved. And my dream is that we finally get a new experiment. We really need new experiment. We cannot base everything on the a little bit, uh, dual records of the Memorial Tunnel from 30 years ago. I really wish we had a new set of data that would help us def- redefine those concepts and revalidate the, the tools that we're using. That's a dream. Uh, maybe it will happen one day. We will see. Uh, anyway, I think that's everything I had on my list for this podcast episode, and I, I hope you've enjoyed that. And if you work with tunnels, I hope it was directly applicable for you. If you don't work with tunnels, I hope it was at least interesting or perhaps even entertaining. Thanks for being here with me in the Fire Science Show, and I look forward to welcome you here for another dose of Fire Science, uh, next Wednesday. See you there. Cheers. Bye.