Fire Science Show
Fire Science Show
190 - Car park fires review with Zahir
With the emergence of electric vehicles, fire safety and dynamics have entered a new domain, raising crucial questions about existing protocols, design fires and data gaps. Today, our Wojciech Wegrzyński welcomes Zahir, Associate Prof. at University Putra Malaysia, to discuss the findings from their latest papers, compare methodologies, and highlight the differences between traditional combustion engines and electric vehicles.
The conversation covers various topics, from the nuances of fire dynamics to the importance of context in risk assessment. Zahir shares his extensive experience studying vehicle fires, including the evolution of electric vehicle dynamics that users should never ignore. With thought-provoking insights, this episode emphasises the increasing need for robust, comprehensive data regarding car fires and the unique challenges posed by electric vehicles.
Join us in this engaging exploration of fire safety science, and don't forget to subscribe, share, and leave a review!
Papers! PAPERS:
Podcast episode 135 - Contemplating a car park design fire
(and the paper by J. Hodges from last year is here)
---
Badania przedstawione w odcinku podcastu omawiane przez dr Wojciecha Węgrzyńskiego przeprowadzono w projekcie realizowanym an podstawie umowy UMO-2020/37/B/ST8/03839 do projektu badawczego nr 2020/37/B/ST8/03839 pt. Skutki oddziaływania wiatru na pożary budynków w wieloparametrycznej ocenie ryzyka z wykorzystaniem metod numerycznych.
Thank you to the SFPE for recognizing me with the 2025 SFPE Fire Safety Engineering Award! Huge thanks to YOU for being a part of this, and big thanks to the OFR for supporting me over the years.
----
The Fire Science Show is produced by the Fire Science Media in collaboration with OFR Consultants. Thank you to the podcast sponsor for their continuous support towards our mission.
Hello everybody, welcome to the FireScience show. I was looking forward to this episode for quite a long time. A year ago I've published an episode when I've talked about crafting a design fire for car parks and we were about to submit a paper back then. And finally the paper's out. It was published this month in FireTechnology with my student, bartosz Miepówka, who did most of the work. Great job, bartosz. Anyway, the next day, the Fire Technology published another paper on a very similar subject, also compiling the data sets from the heat release rate measurements of car fires.
Wojciech Węgrzyński:Taking a little different angle, I welcome that addition a lot because it comes from Zahir. Zahir is a professor at University Putra Malaysia and his PhD done on car park fires in Canterbury was a strong inspiration for all work we've done in this space. So I highly welcome an addition from someone who I look up to in this space as an authority. Anyway, of course, I have a podcast, and papers don't tell you the full story of the research done. So I thought let's invite Zahir and let's talk over through our papers, let's talk over through what we have found, because a lot of findings are very similar and, as you will see, we share a lot of ideas together. However, there are some things that differ in our approaches. In my approach, we went into more like steady state approach, identifying the key events that happened during the car fire. This is what the previous podcast episode on car park design fires was all about.
Wojciech Węgrzyński:While Zahir in his PhD, he developed a method to establish design fires and he applied the same methodology to the new data set that includes electric vehicles. So he actually presents some fire curves and we get into that discussion Discussing how do you make one? Is there a point of making one? Why did he choose to do it? Why we have chosen to not give it to the public and, overall, we just talk about cars and fires and, of course, 2025, we're talking about electric vehicles. That is what makes both of our research important. In 2015, when Zaheer published his PhD, the dataset was quite complete for internal combustion vehicles. However, since then, a lot has happened. A lot of EV fires were conducted in laboratories, some new data was obtained and yet we still have some raging gaps in data, which will also be discussed. I guess that's enough teasing. I've already told you what the episode is going to be about. It's a good one. You probably don't want to miss this one Car park fires, electric vehicle fires that's an important part of the modern fire science, so let's cover that in the fire science show. Let's spin the intro and jump into the episode. Welcome to the fireiresize Show. My name is Wojciech Wigrzyński and I will be your host.
Wojciech Węgrzyński:This podcast is brought to you in collaboration with Ofar Consultants, a multi-award winning independent consultancy dedicated to addressing fire safety challenges. Established in the UK in 2016 as a startup business of two highly experienced fire engineering consultants, the business has grown phenomenally to eight offices across the country, from Edinburgh to Bath. Colleagues are on a mission to continually explore the challenges that FHIR creates for clients and society, applying the best research experience and diligence for effective, tailored solution. In 2025, there will be new opportunities to work with OFR. Ofr will grow its team once more and is keen to hear from industry professionals who would like to collaborate on FHIR safety features this year. Get in touch at ofrconsultantscom. Hello everybody. I am joined today by zahir, an associate professor at university putra, malaysia. Hello, zahir, good to have you in the podcast hello, good morning yeah, I'm having my fanboy moment.
Wojciech Węgrzyński:I'm your biggest fan and you know that and for the exact reason why I brought you into the podcast. Actually, I remember many, many years ago when we were working on the car park problem in Poland and it was a challenge let's say, 2014, 2015. 2015, we've released our book in Polish on car park fires. I remember around that time you were doing your PhD and you've done amazing stuff on car park fires and amazing stuff on summarizing data, and for me, it always have been a reference point. And what's uh, even more, we have revisited that at literally the same week of april last year, where we've both our groups, both submitted the papers independently on reviewing the car park data. So super, super interesting that we've ended up in this point together. Now, um, anyway, I would like to ask you about your journey in the world of car park fires. So if you could briefly tell me how did it started and what's the reason you've entered this world and what's the reason you've continued this research during your stay in spain recently.
Zahir:Okay, uh, thank you, wok. Firstly, I would like to thank you for inviting me to be in this podcast. I'm honored to be one of your guests. I've been listening to your podcast for quite a number of times and I think it's a an unbelievable and great podcast. Uh, as fire engineer, fire scientist, okay, so, as you all know, my name is Zahir and currently I'm permanently employed as an associate professor in University Putra, malaysia, but currently I'm having my sort of like postdoctoral leave in Universidad de Navarra, spain.
Zahir:So about your question about my journey doing this research of vehicle fires in car parks, actually it started in 2011 when I began my PhD in University of Canterbury, new Zealand, under the tutelage of Dr Michael Spearpoint and Professor Charles Fleischman. During that time, when I was enrolled in University of Canterbury, I didn't know much about fire science, fire engineering and so on, because I came from the background of chemical engineering. So I began and began to to learn about fire engineering. I attended classes and so on, and it happens that later on, I was doing research in vehicle fires in car parks. So I've been working in that area for four years until I finished my PhD.
Zahir:The PhD was we were looking to find the worst case scenario of a car park fire. So that's sort of like the whole objective of the PhD. So what is the potential worst case scenario in a car park? So we were looking into the literature, we were looking into various resources that we can access openly, because there are so many documents around the world some of them are like their native languages and so on but we were trying our best to find all the information that was there during that time. So, using the statistics that we can get, using all of the information of experiments or test data that we can get, in the end, what we did was that we were trying to, sort of like, do a risk analysis in which, as we all know, risk analysis consists of probability and also consequences. We were looking into the probability of fires to happen in car park and also we were looking into, if it happens, what are the consequences. We are looking into all of those data, we do some analysis and so on.
Zahir:In the end, it turns out at the end of the thesis we concluded that the highest risk was a single vehicle fire. So it was based on the statistics and so on. So it is what it is, and then I finished my PhD in 2015. I went back to Malaysia and since then I've been permanently employed by a university from Malaysia and then for a number of years I've been doing something else other than vehicle fires. I've been doing like even to the extent that I'm getting into peak fires forest fires because it's sort of like a national problem at that time. So I've been doing many things and around 2020 or 2021, I think during the COVID times, or 2021, I think during the COVID times I was applying for the Mary True Reaction Postdoctoral Fellowship, in which at that time, I was still eligible to apply for it. And then it turns out that I think that my proposal was good because I was awarded the grant.
Wojciech Węgrzyński:You got a Marie Curie action on that. I didn't know that really I did.
Zahir:Hence the stay in Spain.
Wojciech Węgrzyński:Congratulations, Zahir. That's a big one. I was not aware of that. That's a considerable achievement on its own, and this one was on electric vehicle fires. So let's say round two of your research. Was there any specific reason why you chose that it needs to revisit? Was the electric vehicle hazard like so different from you what you've seen during your PhD?
Zahir:I think around 2020, 2021, I think it's sort of like five or six years after I finished my PhD there was some group of engineers in Malaysia who came to us and they asked sorts of questions and so on, and during that time there were this emergence of new electric vehicles and emerging technologies and so on. And then one of the questions that came from the engineers was they were to design car parks. Do they need to use different design fires to design car parks? So that's sort of the questions that came to us at that moment and I didn't have the answer during that time because I didn't know. So that results into the proposal writing of the Marie Curie Action Postdoc Fellowship, in which we are trying to look into these problems of electric vehicles and to find the answer. Is it any different with combustion vehicles? So that's sort of like the simple question we had during that time. So, yeah, that leads to what I'm doing now in Spain.
Wojciech Węgrzyński:For us the journey into this problem and then how we ended up writing our review. It was like twofold story. So first we were working a lot with the car parks corporate design. We have settled down on traditional design curves from tno in poland and and I'm still quite happy with them I've been involved in the cn actions where we write the european standard for jetfan systems, which has completely different, like the megawatt curve. It's a different story.
Wojciech Węgrzyński:Anyway, that was quite okay-ish in the 2010s. And then I remember a particular day, like I vividly remember how it unraveled. So prior to that, everyone was talking oh, electric vehicles, they burn. Completely different, there's going to be a hazard. And we were like, ah, no, it's not, like it didn't look that bad, we didn't have that bad, we didn't have data, we didn't have knowledge. There were just like individual fires, not not like you did not have statistics, you had maybe five, five cars that burned down all around the world. At that point, 2019, and I remember that evening, there was this fire in shanghai. There is like Google Tesla fire in Shanghai, or whatever. You everyone's seen the video like one white vehicle and it goes like there's a flare gone, like in seconds, there's a flare coming out of that vehicle, you know, and I saw that video and I'm like holy shit, this is so different from anything else that I have worked on with so far with car fires and based on the size of that flare, I immediately realized it must be around a megawatt, maybe one and a half, maybe 700, I don't know megawatt-ish, you know.
Wojciech Węgrzyński:And on that particular day I was working on a car park, so I was doing a CFD for a car park, for a commercial project, so I immediately stopped that commercial project, you know, and I restart the simulation, but starting assumption that it doesn't grow linearly to, because in TNO curve you reach 1.4 megawatts, but you do that in four minutes. Instead of that, I just start the simulation immediately with one megawatt flat, like, like in the video. It just started and I run the simulation, I wake up in the morning, I watch the results and, holy shit, they're like, they're much worse. Like in this particular car park that I was working on, where I had tenability for my design fire, I suddenly lost it when I did this new setting and I remember I even texted Guillermo, like we do the comparison, like we're not ready for this. This needs research, and at that point we're. Also. It was quite the lucky thing because I just had awarded an internal grant for multi-parametric vehicle fire analysis. Evs were not part of that grant but that morning, oh hell, they became a part of that grant and we've done like hundreds of CFfds on on different scenarios involving very quick fire growth in vehicles where we've shown that if your car park is tall enough it's not a problem. It only became an issue when the car park was not high enough. And that for me at that point I've settled down on EV problem because I understood that the difference in the early phase you can mitigate it with height, you're good. Then some years later we start the grant on multi-parametric wind and fire and our subject is a car park and we come back to the problem. Okay, so what kind of a design fire do I put in this academic research which is not a commercial project? If it was a commercial project, hey, I would just go TNO curve and I'm done. Perhaps I would modify we sometimes modify the TNO curve to include for this sharp, quick growth at the early phase of the FHIR. To illustrate the EV problem. Like you know, runaway problem For commercial. I'm set For research. I need to justify right. So I hire a student for that, bartosz, and I give him a simple task that I hope that it will take him a week to finish, bartosz, just go through the literature. Take Zahir's PhD. Look what has been published afterwards. Let's just summarize this and we're done.
Wojciech Węgrzyński:Three years later, we published this paper because it was a loophole. Like the amount the scatter of data is insane. But when you try to understand what is causing this scatter of data, wow, that's a whole world and a whole story that unraveled for us. We really wanted to have a good justification for our design fire and what we found is such vastness of different things that happen in those experiments and, hey, it's explainable. Like you can explain everything you see in those curves almost everything. But if you want to compare a car park in which a tank ruptured versus a fire in which someone broke a window, like it's so difficult to average this. You need to go through that and that's the reason why we went into three years of writing this paper and summarizing this in a way that we try to show the reader, like what leads to the data and why we eventually gave up on finding a curve.
Wojciech Węgrzyński:We'll get back to the curves in this episode if anyone wants to now turn off, because I said we will do the curves later. But anyway, that's our story. It's fascinating how cars burn. Anyway, your PhD, main source of inspiration for us. You've done the job that we did on your own, reading through all the reports. So perhaps let's clarify how do we obtain data on vehicles, like when you have a data on a vehicle fire? What do you mean by that, what do you actually have and how we got it?
Zahir:Okay, so during my PhD. So I was looking into the severity of vehicle fires, and what I mean by severity is that we were looking into the hit release rates of a single passenger vehicle fire. And then we were trying to look into the literature and then any accessible test data or experimental data that we can get. So we were looking into data from the I think the earliest was around the end of 80s and then the 90s and then up until 2011 and 2012-ish. So we were looking into the test data in terms of single vehicle fire tests or experiments that we can find In my PhD.
Zahir:We were looking into the problem as we assume that vehicle fires in car park is sort of like a traveling fire, in which that, for example, if a vehicle starts to burn, then it releases energy and then on that particular vehicle there will be like heat fluxes that reaches to other combustibles.
Zahir:And then we were doing another sort of like sub-research that we can estimate or predict when is the next vehicle to ignite and so on. So in the end we will have a profile of the vehicle burning in car parks, because we think that when we are talking about vehicle fires in car parks, it's not a simultaneous vehicles are going to burn at the same time. So it's one vehicle. It will not only start, but the highest possible scenario is for one vehicle to burn and then it spreads to another vehicle, and so on, and so on, and so on. So that's the basis of why we assume it is sort of like a traveling fire. So that's why we are looking into the severity in terms of heat release rates, because in heat release rates we are able to sort of like predict what are the heat fluxes that reaches the combustibles and so on. So yeah, that's what we were looking at when we are looking for car fire data.
Wojciech Węgrzyński:So for your research it was mostly the heat release rate, right.
Zahir:Yes.
Wojciech Węgrzyński:But you've said it was from the single vehicle experiments. No-transcript have the ability to capture the heat release rates. So really it's not that you can have calorimetry results of multiple cars going off, the second vehicle igniting and you capture those curves and you would also not capture them independently. You would see one lump curve of all of them. You have to grow that from the single vehicle data. In the era of EVs do you see the same trends following In the era of EVs? Do you see the same trends following? I also don't know of multiple EV vehicle experiments and I also don't know of experiments where a firewood traveled into an electric vehicle. Does this observation still hold?
Zahir:Okay, I think one different observation about electric vehicles in this regard, when we are talking about fire spreads in vehicles, is that there is the potential of electric vehicles to spread much quicker. This is one of our observations. Usually, when we are talking about electric vehicles, they have their batteries installed underneath the passenger cockpit, the passenger area, in which usually the manufacturers they have a good separation between the batteries and the passenger area. So one of our worries is that when fire happens, or once the battery is burning, it can get deflected to the side, in which it might potentially ignite the vehicle next to the burning vehicle much quicker. That's what worries us and that is sort of like our next step in our research in which we are planning to look into this problem, because when we are talking about battery fires, they might potentially get into jet fires of very high velocity. So there's this potential of fire to be spreading much quicker, but at the moment it's not supported by any data. But this is just our worry and our observation of what can be.
Wojciech Węgrzyński:It's one of those things that sounds plausible. Like you have a battery underneath the vehicle, it may be exposed quicker. But then again, the battery is, like in many vehicles, it's already a part of structural design of the vehicle. So it's behind many millimeters of solid steel. It's not like many batteries dangling under the floor of the car. It's really like behind quite a sound case.
Wojciech Węgrzyński:And when I talk with people who are actually burning electric vehicles, a common thing people say is they were surprised how difficult it was to make the battery go off, like if you had to attack it from outside. So again, I'm looking forward to that data, because it's something we definitely would like to have to understand how easy it is to transfer the fire. The other thing is also that, taking this jet fire away, the battery is technically in the best place for the vehicle, because in a normal car park fire, the thermal radiation would be from the top, from the smoke smoke layer, from the flames, so so the battery is literally in the most shielded place of the vehicle out of all of them. So perhaps in the end, like minus the jet fires, maybe that's the best place. I I don't know. Yeah, again, something to look for. Uh, for me one observation that that comes to my mind. If any fire engineer very easily understands the concept that if you have a fire like a one megawatt fire, you put it in a small room, it's a big fire. You put it in an aircraft hangar, it disappears, it's irrelevant. So the place where the fire happens, the surroundings of the fire, are as big part of the design fire as the megawatts themselves, like the context in which the fire is placed.
Wojciech Węgrzyński:And here, working with those design fires, we see calorimetry being done in different ways. Like you have cars burned under very tall hoods, like literally a car lying in an empty space and five meters above it there's a hood that extracts smokes. You had TNO, which changed a car park into a calorimeter. You had BRE, which built a small room that looked like a car park and burned like a small passenger garage. You had Joyer, who did in a hood but with a ceiling and a configuration that perhaps imitated a car park. So it's very interesting in how many different conditions those experiments are Like. From your insight into this research, how big variety of the conditions in which the EVs were tested Did you see? You went very deep into that in your paper on the experiments.
Zahir:Yeah, I think in terms of that it is something that we are completely aware, because obviously the test conditions, when we are talking about the test conditions, even how the labs were set up obviously is going to be different and all I mean what you have already explained. But the main idea of our work is to look into this variability of tests and so on, and that's why the idea of probabilistic input came into the work, because looking at all of these variabilities for example, some of the tests, they started the burning from the world. Because looking at all of these variabilities, for example, some of the tests, they started the burning from the batteries, and then even I think there was one or two which tried to ignite the vehicle from the tires or something. So obviously when we are talking about where's the location of the ignition itself will obviously affect the burning of the object.
Zahir:So the idea of our paper is that when we are talking about a probabilistic input, we are trying to sort of like mirror the real-world scenarios in which the randomness of the fire itself, it could start from anywhere and with different growth and so on, and with different growth and so on. So the idea is that if we have more data, so the more reliable the analysis is going to be. But we have to work with what we have, and then even the data that we have now is actually have been filtered from more other data that I don't think can be used in the world in itself. So, yeah, that's the idea of the paper itself to include all of these potential scenarios that might happen in the real world. So we have this probability or this distribution or collection of all of these tests to produce and to find the tendencies of these fires and the growth and even the peaks and so on. So that is sort of like the idea.
Wojciech Węgrzyński:You bring me to a very interesting case, because it's also something that was a rationale for us. So we've seen a lot. I mean, this is a practical problem. You said you started it as a practical problem. Someone approached you hey, zahir, how do we do it? You need to have a good answer for that and this practical problem. You would like to say with some certainty that the average hit-release rate is this or the 95th percentile. Is this right? But to get that, we found that you really need to have a complete data. Your PhD was a very, very complete data set. After we went through like 2,000 papers, there was barely very little that we could find that was not already covered by you in 2015. Since then, of course, you have not covered the future. Perhaps you should work on that, zahir, but it's challenging to give those average values when you're working with such a completely varied data set.
Wojciech Węgrzyński:Some vehicles were burned with open windows, some with closed, some where the tank was full, some with the tank empty, some with the tank in which there was a hole.
Zahir:Yes, they spread off close on the metering.
Wojciech Węgrzyński:yeah, it's ridiculous. It's ridiculous, but would you consider this how to say, this collection of evidence we have from experiments as perhaps a collection of potential outcomes and then, within those outcomes, find some probable scenarios and move on? Or you are just looking for this one worst fire, you take it as the worst possible and then you work up probabilities from I don't know some failure trees or other means. How do you turn this data set into probabilities?
Zahir:Okay, that is an interesting question. I mean, looking back into my paper, is that we are trying, first we are trying to collect all the available data that there is in the wild from the resources, and then we try to get as much information from all of those resources and we present them in our paper, and then we we try to do some analysis here and there and so on, and at the end it is up to the engineers to to make a decision. So sort of we are sort of like collecting all the data, all the basic data, and then the engineers can make decisions out of it. So that's one part. Another part is that in our paper we are also focusing on something that we have introduced before. For example, in my PhD.
Zahir:We have introduced a method that we can form or create or construct a design file in which I think, if you, you notice, and and in this paper, there's a small section of what kind of values you can put into the growth part and then the decay part, so that you can construct a design file.
Zahir:So I mean it's up to the engineers if they want to use their own method, if they want to interpret the data in their own way, or else they can use our method of constructing a design file that we introduced before. So yeah, and in the bib, in short, we explained that these data sets can be used to construct a design file of electric vehicles or combustion vehicles that you want, to use it for any kind of purposes you want. So we are actually presenting options for the users, for the engineers, to use whichever they like. And then, like I mentioned earlier, we presented all the information that we can get, even to the extent that what happened during the test. So we we put sort of like a timeline of each test. So yeah, from from the resources, from the papers, from the literature we can get.
Wojciech Węgrzyński:So we explain everything I think we're reaching the point where we will talk about the curves. And this is interesting because here our approach is kind of split, where you try to find the curve and we'll talk about the curves. And this is interesting because here our approach is kind of split, where you try to find the curve and we'll talk in a second. We, what we did, we, we gave up on the curve. I thought that the introducing a time component on such incompatible data, like simply averaging the curves, or or or drawing an envelope to cover all of them, like I found both ways to be a lie. If you average, like, how can you average a fire where the tank was ruptured, with a giant nail releasing 40 liters of gasoline on the floor of the car, fire with a fire where a little burner was set next to a tire? It's impossible to average that. And when you try to draw an envelope around all curves I know such approaches were also done you have multiple tests which peak at different times and you start to add those peaks together, like, you lose the connection to the total heat release rate of a single vehicle, which is a product of its mass right. There's only this amount of energy that the vehicle can release and that's it right. It's not going to magically like grow to 20 megawatts and stay at that for 20 minutes. So we really saw no good way to do the curve. But instead, when we combine those data points and we looked at the location of the ignition, the size of the vehicle and so on, we started to see some perhaps you can call them patterns or groups of fires that you could say the fire the vehicle is ignited like this, it will burn with this rate. The vehicle is ignited like this will burn at this rate.
Wojciech Węgrzyński:You could also do a simple thing like percentile of fires, and that's what we used in the grant in the end, because I needed this for my wind and fire grants, so I needed an answer to that grant. So what we did was we used a 50 percentile, 95th percentile, 99th percentile. We used them as steady state fires and just did the wind part. I gave up on the transient phenomenon. On the transient part, I understand from your research that you went into a path of looking into transient, which is interesting, and the concept was that you've noticed that there will be a growth and decay phase with the significant peak and if you find where the peak is and what are the coefficients for the growth and decay? You'll end up with some sort of a curve. Can you walk us through this concept?
Zahir:Okay.
Zahir:So first is that when I was doing my PhD, like I mentioned earlier about the traveling fire, so we assume that vehicle fires in car park is a traveling fire problem, so that we need to look into the heat risk rate profile of the design fire vehicles. So not only the growth part is important but the decay part as well. So when we were looking into the statistics, when we were looking into the data of heat risk rates profiles for experiments for single vehicle fires, we are not doing the analysis only on the growth part, but we are also doing analysis on the decay part, in which we think is important because when we are looking into the decay part, it still has energy being released, even though it's much lower, but it contributes to the prediction of vehicles fires or the next combustibles around the burning vehicle, so it is as important. So the idea of creating this design fire is is. It's not from our own. Actually we were looking into the work by ma in which they have this kind of way of how to sort of like profile design fires as well, but it's not for the purpose of vehicles. So we were looking into their work and discussed and why not we try this kind of method and apply it for clinical fires. So, and then we ended up looking for more older work during that time by other researchers and so on, and we sort of compared, I think, five or four works which has been done before to profile a design fire. We did some comparisons, did some statistical analysis and so on and we end up with what we have now.
Zahir:So I mean we we have lots of critics towards our way in terms of doing this. I think, uh, there was some, there were some researchers who were asking us about when we are doing statistical analysis. Obviously there will be biases and so on. But in our defense, I mean in all datasets there will be bias, but the idea of explaining everything, or the awareness, the awareness of the biases itself will reduce the harm, or the awareness, the awareness of the biases itself will reduce the harm.
Zahir:So there's the idea in our paper in which we are aware about all of these biases and I mean the different tests have different way of conducting the test and with different configurations, different everything, but we present it to everyone that these are the things that they have done. But we presented to everyone that these are the things that they have done so and we made people aware that it's not going to be a I don't know homogeneous problem. And, yeah, making the readers aware about the biases on its own will reduce the the harm. So that's sort of like you know with effects. But going back to the design fire, so we had done these comparisons and so on and it turns out that the method that we have now is the best fit of the data vehicle files that we have, and then we use it and we apply it even for this work. So in the end the engineers have decided to use this method. They will be able to construct a design file based on the probability distribution.
Wojciech Węgrzyński:And because it's a growing phase, there's a decay phase, so there's a very distinct peak in that curve. How did you decide, like, at what time the peak will be? Because I also see different proposals in your paper about where the peak is and that quite drastically changes also the growth rate of the fire.
Zahir:Yeah, this was one of the discussions as well. I mean, some of the experiment data they have like not only one peak, but some even have like three or four peaks and so on. So we have discussed about this and in the end we decided to choose the highest peak out of the data. Aha, okay, so assuming that it reaches the highest and we use the highest peak from the experimental data rather than having like two or three peaks, which complicates more the probabilistic analysis, because in the end we understand that this is a probability and we just decided to take the worst case. So, yeah, that's the idea of how we in the end decide to select or to choose the peak of the heat risk for the experiment data.
Wojciech Węgrzyński:And for an engineer, what you're giving is more like a methodology rather than a worked out answer, and the engineer should find their own values. Is that what you postulate?
Zahir:Yes, yes.
Wojciech Węgrzyński:Yes, that's the the idea, because it's going to be different whether you design for the 80th percentile, 95th percentile, for a median fire, that this is gonna give you a completely different courses of fires, especially that the peaks will move left, right, like you. Yes, you just said um and okay, but this started as an EV research. From this perspective of running this study, what difference do you see in this design proposal, design fire proposal, between the EVs and internal combustion engine vehicles? Because we also see a difference in the proposed shape and size of the curves.
Zahir:Okay, so this is also an important discussion that we have in our paper. So during my PhD, I've collected experiment data for internal combustion vehicles. Obviously, the work was done in 2013 and 2014. So in this paper, now that when I want to try to compare the results for electric vehicles and combustion vehicles, I have to get sort of like an update of experiment data for the combustion vehicles as well, because, yeah, I mean over time, the components inside these vehicles are changing and so on.
Zahir:So that's why, in our paper, not only we are collecting experiment data for EVs, but we try to collect experiment data for combustion vehicles from around 2014 up until now, in which that is what we have done, and then we try to sort of like compare in terms of the, the growth rate and the peak and so on. It turns out, with the data that we have, it seems that for the electric vehicle fires, it seems that it burns quicker than combustion vehicles. But this is something that we need to use this information with care because all of the biases that we have talked much earlier so whether this is true or not, I think in my conclusion I've mentioned about there's a need of more experiments to be done. So, yeah, in the end I can present where is it, but this has to be used in care Very cautious.
Wojciech Węgrzyński:And I find one thing that is challenging with the electric vehicle data is that I do not really see data for fires that do not start in the battery. Like I really lack this transition from a vehicle fire into the battery fire. I understand why If I'm a fire researcher and I get my 50,000 euro electric vehicle to burn down and I have one of them where I'm going to start the fire, of course, in the battery, because that's the exciting part, right, but we really do not have scientifically plausible explanation on how fires propagate from outside or inside the vehicle into the battery.
Wojciech Węgrzyński:If the battery is not a subject of a direct attack, I can only hypothesize that a fire until the battery ignites will be the same as an internal combustion, because, hey, it's the interior, it's the plastics, it's just the vehicle itself, which is not that much different, which are very well aware as consumers and users of vehicles. It's the battery that's different than engine and no tank. But I do not have that data point and also it's something that I can only give as an anecdotal proof. But there has been a fire I think it was in Australia where on some open air parking lot there were multiple Teslas that burned down, multiple like six, eight like multiple, and as far as I understood, none of them had battery participating in the fire. So you had six vehicles burned down. None included battery.
Wojciech Węgrzyński:There was this Cybertruck explosion in the US a month or two months ago. Allegedly everyone started oh, it's an electric vehicle, explosion, right. Allegedly the battery survived the explosion. It was not. Definitely the explosion was not the battery and allegedly it survived. So you have this anecdotal proof that it's not so easy to transition into the battery when you have an external fire. But we don't have good data and if you would care to have this traveling, you know leapfrogging fire model in your car park. It's very important for that question. Will that battery go easily or not, because it's a considerable part of the overall heat release rate, right?
Zahir:Yes, yes, true, I mean, this is the whole point of my research stay in Spain anyway, to look into this problem. I mean, when I began my stay here in Spain, I was not only looking into this specific paper. I was sort of like exploring what is really the problem with electric vehicle fires. It's not only about the design fires that is one question that I was trying to answer but actually we found out that one of the main problems of electric vehicle fires was that the risks are understudied.
Zahir:It's a new technology. It's the same for, like, photovoltaic, pv and so on, because it is something new to be, to be society and when we were starting to look into each of the potential consequences that can happen and so on, and we found out even the problem that you have mentioned. I mean, if the fire began outside of the battery, what will happen, and so on. So there is no experiment that has shown that, or I've studied that, so we don't really know the answer. But it's not only that.
Zahir:I think there's a lot of other things in terms of the risk of electric vehicles. I understudied that and I think personally it is worth to look into this and do some experiments on this. It's only that when we are talking about burning vehicles and so on, it's it's a large or very huge magnitude of research. And also, I think from Michael Spearpoint's paper on electric vehicles I think he published it last year he mentioned about the technology evolved much quicker than the research. So the safety research is not keeping up to the technology being introduced into the society. So I think this is a problem as well. I mean, once we are starting to do the research on this, new problems arise, but it doesn't stop us from working on it. So, yeah, I think personally we have to investigate more and find ways to make the world safer Fantastic.
Wojciech Węgrzyński:Anyway, dear listener, if you want to really dig into the data and the knowledge, do you have a hell of a reading in front of you Because we did not hold with Zahir on our papers. Mine is 38 pages, zahir's is 42. So I have an 80 pages of reading. But the good news, they are both open access and available thanks to the fantastic VAR technology journal. So I will just refer people to the papers, really, and the links are in the show notes.
Wojciech Węgrzyński:If you would like to look into data or look into the polished findings, you will find them there and I hope with this podcast uh interview, we've cleaned a little bit how we got there with zahir and what the concepts were, because that's the thing they don't allow you to publish in papers. Journals should do better. If I write a paragraph on my thoughts and emotions, they're gonna say it's unscientific and should have not been in the, in the paper. Therefore, we need podcast episodes to clarify that. Anyway, zaheer, so going back to Malaysia, you're going to continue research on electric vehicles. Any plans already? Did you already purchase the vehicles to burn down?
Zahir:I think the main problem not the main problem, but I think it's widely known in the academia that it's a bit hard to get funding. So I think, when I'm going back to Malaysia, it's about engaging with the right people, and then engaging with the maybe industry and so on, and then look for opportunities for doing further experiments and so on. But we have so many ideas, but the first thing is to find some funding. I think that is the most important thing. Yeah, but the first thing is to find some funding. I think that is the most important thing. But, like you said, yes, I will be going back to Malaysia and I will be focusing on this research and any other emerging technologies problem that will arise in the future. This is something, the area that we are planning to focus on.
Wojciech Węgrzyński:Zaheer, thank you so much for coming to the Fire Science Show. It was a pleasure to have you here. I hope you've enjoyed your stay at Nevada. Congratulations on having the Marie Curie action. That's a big thing. I didn't know about that and I'm very happy that you got it. Even though it's about to end, I'm happy that we've shared the news with the world perhaps a little late, and all the best for your comeback to home. Yeah, thank you very much. Thank you for inviting me and that's it. Thank you for listening.
Wojciech Węgrzyński:Perhaps the episode could have had more juicy parts of the data in it, but I thought that you know what the papers are. There Papers are open access. They're extremely long because of the amount of data that's covered in both of them and if you really want to make your mind on the details, if you really want to look at what we know about vehicle fires, you will find that in the papers. So the links are in the show notes. You can just download them and look for yourself what they're all about. I think the key lessons for me from this episode is that I understood why Zaheer went for the curve part. I mean, you know my state is I don't like curves. For me, vehicle fires are event-based fires and you have to consider this on the key event basis and you cannot just plot it on the key event basis and you cannot just plot it on a timeline, because you don't know when the events will happen and those events will significantly change the fire.
Wojciech Węgrzyński:Zahir, however, was inspired by Maurer. He was inspired by the classical approach to design fires. He has distinguished a very distinct growth period. He has identified very distinct decay period. Between them there's a very distinct peak and you can simplify most of the fires to some course like that, because most of them would go like this, and by defining how high the peak will be and when it will happen, you will get your growth and decay coefficients, which makes it fairly easy to mathematically describe the fire with some T-squared relationships, and I think it's not bad.
Wojciech Węgrzyński:Some things Like if I was doing a structural design or if I was really going into the probabilistic methodologies that Zahir mentioned, I probably would not have any other option but use a design curve and in that case, those curves which can be fed with probabilistic input data. They would fit very well my model. If I'm, however, doing my simple ACID-ACID calculation, I'm not sure because those curves will miss a lot of key events like battery thermal runaway or quick transition into the fire of the passenger compartment. They don't follow quadratic growth in my opinion. So I'm not completely sure if these curves do the justice for acid arson analysis. I know some people would say they do. I know some very small and influential people who say it but do. But in my opinion, dealing with those things for more than 15 years, I'm not so sure. And if I was doing structural I probably would go there, because having those curves instead of steady states allows you to have multiple peaks at the same time or split the peaks, which is more interesting to see how those different fires interact with each other.
Wojciech Węgrzyński:Anyway, I'm still worried that the question of the car fire in a car park, the design fire for a car park, is unsolved and I wonder if it ever will be solved. We have some good ideas on how to go with the event thing further. Perhaps in some years I'm going to make another podcast episode telling you how we solved it and if you have a solution in your mind, if you have a good idea on how could we approach that, that's very welcome. Oh, and one more thing In the podcast. We've mentioned that, but there was a third paper from Jonathan Hodges, and Jonathan was also a guest in the podcast and he's published his paper a year ago Also a very interesting take on the car fire problems.
Wojciech Węgrzyński:So there's a lot of resources for people who want to study car fires out there, and I'll probably link to Hodge's paper in the show notes as well. So you have complete resources for your needs and that's what we do here in the Fire Science Show we provide resources to our fellow fire engineers so you are more confident in doing your fire safety engineering and more open-minded. I hope that that's a goal we achieve in here as well. So thanks for being here with me and next wednesday another interesting episode, interesting take on fire science from our side. Thanks for being here with me, cheers, bye.