Fire Science Show
Fire Science Show
231 - BESS explosion prevention and mitigation with Lorenz Boeck and Nick Bartlett
Today we cover another branch of safety of Battery Energy Storage Systems (BESS), that is explosion prevention in mitigation. I always thought you can either end with a fire or with an explosion, and boy I was wrong... but we will go back to this later. Now I bring on Dr. Lorenz Boeck (REMBE) and Nick Bartlett (Atar Fire) to unpack how gas released during thermal runaway turns a container into a deflagration hazard, and what it takes to design systems that actually manage the pressure, flame, and fallout. This is a tour through real incident learnings, rigorous lab data, and the evolving standards that now shape best practice.
We start with the fundamentals: from the overview given by NFPA855, why modern BESS enclosures—with higher energy density and less free volume—see faster pressure rise, how gas composition varies by cell and manufacturer, and why stratification matters when lighter hydrogen-rich mixtures sit above heavier electrolyte vapors. From there, we translate UL 9540A outputs—gas quantity, composition, flammability limits, burning velocity—into engineering decisions. NFPA 69’s prevention path typically relies on gas detection and mechanical ventilation designed to keep concentrations below 25% LFL, validated with CFD to capture obstructions, sensor placement, fan ramp, and louver timing. NFPA 68’s mitigation path kicks in if ignition happens, with certified vent panels sized to the actual reactivity and geometry, relieving pressure and directing flame away from exposures.
A major takeaway: the latest NFPA 855 now often pushes for both prevention and protection. Even with active ventilation, partial-volume deflagration hazards remain, especially as cell capacities rise and gas volumes scale up. We dig into venting trade-offs—roof vs sidewall, snow and hail loading, heat flux to back-to-back units—and how targeted sidewall venting can deflect flame upward while reducing weather vulnerabilities. Perhaps most critical, we talk about late deflagrations observed hours into large-scale fire tests, when changing ventilation conditions allow pockets to ignite. Active systems aren’t built to operate throughout a long fire, so passive venting becomes essential during and after ignition.
Whether you’re a fire engineer, AHJ, insurer, or developer, this conversation connects the dots between lab data, CFD, and field realities. You’ll leave with a clearer view of how to apply UL 9540A, NFPA 68, NFPA 69, and NFPA 855 in a world of stacked containers and supersized cells—plus where training can shorten your learning curve.
If you are interested by the course given by colleagues in Lund in January 2026 - here it is: https://www.atarfire.com/event-details/nfpa-855-8-hour-training-lund-university
----
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 Fire Science Show. Uh when I posted an episode with Norrider a few weeks ago on battery energy storage systems, it really blew up. It was a very popular episode, so I thought uh while I'll have access uh to some really good battery speakers, let's take them on a spin and and make uh perhaps a follow-up, more battery energy storage system content because it really seems it gets a lot of people interested in. During my stay in Lisbon when I've attended the SFPE Symposium on Lithium Ion battery safety, I've heard a lot of great talks actually, uh but uh I've heard a really good talk about explosions in battery storage systems, and I thought that I actually never covered explosions in this podcast, not just in the relationship to the battery energy storage systems, but I don't think I really had a proper explosion episode at all in the podcast, which is quite funny. 230-ish episodes in, and there are still new things that were never ever covered in in the podcast. Great to realize that we're not running out of topics in here. But on the topic that I wanted, I found a great speaker in in Lisbon. Actually, I found more than a great speaker, because I found two speakers. For this podcast episode, I've invited Dr. Lorez Boeck from Rembe and uh Nick Bartlett from Atar Fire, and they, besides uh being specialists in battery energy storage system fire safety and explosion safety, these guys also run courses on NFPA eight five five, NFP sixty eight, sixty nine, a lot of standards. And and this is actually another reason why it's time to do this podcast episode. Because we're not talking about hypothetical solutions, we're talking about quite a standardized, I would say quite a major part of fire science right now. There are standards that you can follow, there are technical guidances that you can use to design fire safety and explosion safety of your storage systems. And all of this we will discuss in this podcast episode. So I hope I got your interest quite high up. I hope that you will stay with us till the end, you'll learn where you can find our courses. And for now, let's spin the intro and jump into the episode. Welcome to the Fires Show. My name is Wojciech Wegrzynski, and I will be your host. As the UK leading independent fire consultancy, OFR's globally established team have developed a reputation for preeminent fire engineering expertise with colleagues working across the world to help protect people, property, and the planet. Established in the UK in twelve sixteen as a startup business by two highly experienced fire engineering consultants, the business continues to grow at a phenomenal rate with offices across the country in eight locations, from Edinburgh to Bath, and plans for future expansions. If you're keen to find out more or join OFR consultants during this exciting period of growth, visit their website at orconsultants.com. And now back to the episode. Hello everybody. I am joined today by two guests. Uh, first, Nicholas Bartlett from Atar fire. Hey Nicholas. Hey Warchick. Good to have in the podcast. And together with us, Dr. Lorenz Boeck from uh Rembe. Hey Lorenz. Hey Warchick, thanks for having me. Thanks for coming for to the Fire Science Show. Uh with the recent podcast episode I had with Noah Rider, which broke some yearly records on listenership. I could not miss an opportunity to do another battery energy storage uh podcast episode. And Lorenz, I saw your talk in SFP event in Lisbon uh recently. I've learned that you are doing some crazy cool things uh uh with training uh battery storage systems. So here we are to follow up on that. And I hope today I will extract from you some knowledge to my friends who are listening to this podcast about, let's say, explosion protection of battery energy storage system, but also about the landscape, the normative landscape or reference landscape in which people can design those explosion preventions. And uh first, I well, that's a general question, but why why should we care about explosion prevent what really is the hazard regarding explosion in battery energy storage systems? Lorenz, maybe you would like to start?
Lorenz Boeck:Yeah, for s ure. Um so as we all know, uh BES, you know, it's a pretty new and fast-evolving technology. It is. We've had it around for long enough that we've actually seen some events out there uh that happened, both fires and explosions. So we had an opportunity to learn uh not just you know from lab work, from theoretical work, from modeling work, but also from uh real-world uh incidents. So the explosions we've seen, most of them go back to some sorts of failures of individual cells or maybe the controls that are supposed to keep the cells happy. And as a result of these failures, we think thermal runaway. Thermal runaway is something that's it's pretty well known, even in the general population, because it can happen to your phone battery, to any of your personal devices around you. And I think the awareness is really growing. But what it does in best is it can release flammable gases that can accumulate inside the energy storage system and generate a pretty significant explosion hazard at a scale that is much greater, of course, than just your personal device.
Wojciech Wegrzynski:And uh like it's kind of funny because when you think about storing energy and the late acid batteries, the explosion hazard was always kind of there, but for different reasons, like the hydrogen production when you charge, etc. But here it's on a different level. Do we care about this explosion hazard only when you're building those mega packs on the desert or any energy storage system? It's really something that this risk is real.
Lorenz Boeck:Well, we definitely care about them at all scales, right? Going from consumer products to maybe the system you have in your house to store energy from your PV panels all the way up to commercial industrial applications and then grid scale applications. And um, those are some of the ones that I personally work on the most, is the larger applications. But yeah, we do care throughout the entire range of steels.
Nick Bartlett:Yeah, I think the hazard does change a little bit depend upon where the energy storage system is located, whether it's outdoors, whether it's indoors. But you know, when you have thermal runaway and you're producing hydrogen and CO and some other, you know, hydrocarbons anytime that's of course in an enclosed environment and it can accumulate, then you do have that explosion hazard. So, you know, it's just different different settings for the same issue.
Wojciech Wegrzynski:Let's let's try and do an explosion one-on-one in like five minutes to the listeners. I didn't have an explosion episode in the fire science show, so let's introduce perhaps some important terms. You just drop some stuff that is produced in in the thermal event uh of a battery. How does that turn to explosion hazard and like what dictates how big the explosion hazard is?
Lorenz Boeck:Sure. So you know, usually when I teach explosion protection uh in one of my classes, that's uh multiple weeks. So do this in five minutes, but um I'll try. So what we have in BES is basically a gas explosion hazard. This is not an energetic material or dust explosion hazard or any of these other great things that we might have to worry about elsewhere, but it's specifically a gas explosion hazard. And what makes them unique is the way the gas gets produced, which is from, like we said, the thermal runaway of a battery. So now we have the situation where we have a flammable gas inside of an enclosure, so typical container or enclosure of a battery energy storage system. The issue is that if that gas ignites, typically it builds so much pressure due to the combustion that this pressure overwhelms the strength of the enclosure. So typically these enclosures are built by far not strong enough to contain these deflagration events or these overpressure events. That's the issue here. So if this happens, you might rupture the container in a very, very worst case scenario. Or you might do other things before that happens. You might pop open the doors of the system, you might detach ventilation equipment and such. And all of that generates a hazard. And that hazard could affect the surroundings, could affect people in the surroundings, could affect assets or equipment in the surroundings. Those are the things we're trying to prevent, right? Really making sure that, first of all, we prevent the occurrence of an explosion, ideally, but if an explosion occurs, we manage it or mitigate it in a way that the outcome is still acceptable. So there's a lot of physics, you know, and how these explosions occur, what exactly happens, we can dive as deep as you like into that. But in a nutshell, I think that's the issue about best explosions, right? If a container with a limited strength, if there is a defigution or an explosion event inside, you build so much pressure that you might cause uh a hazard to the surroundings.
Wojciech Wegrzynski:I would like to book you for a fire fundamentals episode in my podcast. We really need to cover explosions. I did I did one good one with with Rory on on flammability limits. So the listeners are expected to understand what the flammability limits are and they're very important in here. How does the volume of the space within the battery energy storage system play a role in here?
Lorenz Boeck:Yeah, so generally the scale of the system plays a strong role because if you have an explosion in a very small space, that basically means that your pressurize will be extremely fast. Assume a certain rate of combustion or rate of flame propagation in a very small volume, you have a very fast rise in pressure, and that means we have to protect it in a certain way. If you have a large volume, your pressurize tends to be slower, but you also need to possibly relieve much more overpressure, much more gas. So the protection or ways of protecting these shifts a little bit. And specifically in BES, uh, one of the challenges is, for example, that there is not much free volume inside of these containers anymore. And I've already talked to know about this, about design changes over the past years that we see in the industry. That was shocking to me.
Wojciech Wegrzynski:Absolutely. Like I've knew very little of it, but how much has changed since I've learned a little of that? It's it's really insane. But sorry, I throw you over.
Lorenz Boeck:No, definitely. And and I had the same conversation with a good friend a few days ago where he asked me, well, are these best units still such that you can walk into them and you know look at the batteries left and right of an aisle that goes down the center of these systems? And when we talk about containerized systems, that's just not how they look anymore. They're densely packed nowadays of thousands of batteries or battery cells inside one of these units, and you're really trying to maximize energy density. So from an explosion standpoint, that reduces the free volume inside of these enclosures. So you're pushed more and more toward a scenario where you have an explosion and a relatively small effective volume that you need to deal with, that you need to either prevent or mitigate.
Wojciech Wegrzynski:Do you care about the exterior of it as well? As in is there like a vapor cloud explosion hazard outside of the enclosure? Or you consider that's already good if I got rid out of the gases from from inside?
Lorenz Boeck:Well, so from an explosion protection design perspective, the primary focus is protecting the enclosure itself. That's your first step, if you will. Making sure that you don't breach the enclosure, making sure you do pressure relief or prevention so that doesn't happen. But then of course you need to handle the secondary effects of the explosions in a safe way as well. Something that's um covered, for example, in NFPA68, but also other design standards for explosions that tell you how to predict uh the external effects of an explosion. For example, extend pressure, right? When we talk about explosion venting, you open up panels or rupture panels, then you will have external flame and external pressure propagating from that opening. And there are engineering methods and standards that tell you how to predict these effects and then how to consider them to make sure that you're doing this safely, right? Again, in the event of an explosion, making sure that you're not creating any unacceptable effects in the surroundings of the equipment.
Wojciech Wegrzynski:Nick, how big is the relationship between the type of a cell and the explosion hazard? Like, do we have like explosion safe cells or all of them have the hazard? And to what extent the state of charge also plays a role in this?
Nick Bartlett:Yeah, well, state of charge certainly does play a role in this. I mean, 30% state of charge is often cited in various studies where there is, let's just say, less flammable gas produced. But what I would say is, you know, generally when I teach NFP 855 is that all lithium-ion batteries have a pretty similar hazard. So as the cells get larger and larger, you know, typically you'll have about half a liter to two liters of gases that are produced per amp hour. So that will obviously scale. And as I know you talked about with NOAA, I mean, today an average cell for an energy storage system is 285 or 314 amp hours, but that is dramatically increasing. And we've even seen press releases for a thousand amp hour plus. So, right off the bat, you you know without even any testing that your gas volume is increasing, but you're putting it in potentially the same exact box. So it definitely changes the hazard as you uh increase the size of the cell.
Wojciech Wegrzynski:Perhaps it's a it's a good moment to move into some of the standards and and specifics because I also like want this to be very practical. We obviously are not able to teach you explosion safety, the listener within a podcast episode. I hope you don't expect that. The the gentleman here uh have a very successful um world tourney with the courses and trainings. If you'd like, you'll definitely learn later in the episode how how can you sign for that. But there are resources and standards that help you figure out your situation. And to what I understood for the cells themselves, there's a UL949540A, which which helps us quantify what I've actually asked, like how how how dangerous is the cell. Maybe let's try to briefly talk about uh about the cells, and then we'll move into standards and prevention in in the storage systems.
Nick Bartlett:So yeah, I think kind of to your point, what I what I teach when I teach NFP 855 is that there are many layers to this. It's it's almost like a cake. So you know, of course, you have the overall product, the whether it's a uh ISO container, uh, so that has its own certification. And then you you get to smaller and smaller levers. So the rack itself will have its own testing and certification, then you get down to the module, which would have like what we call a UL 1973 certification, which has, you know, I think over 30 tests just involved in the module. You know, they short, short circuit it, they over voltage it, they do whatever they can to the module to fail it. And then you get even more granular. So the cell usually has UL 1642, and then you know, there's other certifications and standards for all the protective products, like we've already been talking about explosion control. So that that's its own standard. So what results is this kind of, I don't know, it's almost like a puzzle of different layers of certifications. It's not just one standard, and um, you have to consider all of them. And I think cumulatively, that's when you get an actual product if it's been evaluated properly, that at least the standards say is safe.
Wojciech Wegrzynski:And and let's let's do the 9540A specifically.
Lorenz Boeck:Right. So that standard is critical for us in the end for designing the bright explosion prevention and protection, because that standard tells us about the hazard that the cell produces. We talked about this before a little bit, but the hazard from an explosion standpoint is all centered around that flammable gas that is produced. So the questions are how much of that gas is released and how reactive is it? From an explosion perspective, reactivity is usually quantified by burning velocity and maximum explosion pressure. Some of some of these are the critical parameters we're looking for. And we get those parameters from UL9540A tests. Like Nick mentioned already, also that standard in itself does tests at different layers. It starts with a cell level test. So it just tests the individual cell, and then it builds it up to a module level test and then to unit level test and an installation level test potentially. Um so at all these different levels, there are different tests to be conducted. But for us, from explosion design standpoint or protection design standpoint, it's especially the cell level and the module level that matter the most. Cell level, what you get is the gas composition, the amount of the gas, the reactivity of the gas, and the flammability limits. And then at a module level, you get to understand if there is a potential for propagation from cell to cell or for a cascading thermal runaway scenario that could evolve or involve multiple cells. So maybe not just one cell, but maybe there are multiple cells that can thermally propagate and all release a certain amount of gas.
Nick Bartlett:Yeah, and one of the important distinctions that you know people often get confused is that UL9540 is a certification for a product. Whereas another standard he mentioned, 9540A, it has a very similar name. So it I can see why it's confusing. That is basically a test standard that gives you data. It's not a certification. So you test your cell, and that's where you get how much gas is produced and what are the constituents, and then you test the module and you get data. So 9540A is really just a test that provides data that informs how to design around the data that's produced. Whereas 9540 without the A is a product certification. So it's really uh important distinction that's quite easy to confuse.
Wojciech Wegrzynski:All my life I've been designing smoke control and we had like NFPA 92B and A. Oh yeah. So that was fun. Uh in terms of uh you said it gives you composition flammability limits. So so those compositions are are vastly different. Can we say that there's like a single main like element in there, like hydrogen that that drives the behavior, or is it really like about the composition?
Lorenz Boeck:Yeah, there are some elements that you will typically see in most of these tests and for most of these cells. Um, those are hydrogen, carbon monoxide, but then also carbon dioxide, then a range of hydrocarbons, and then larger molecules, organic compounds. And um, in the end, those are all some sort of decomposition product of the electrolytes of the battery. The composition, how much of what you get really depends on the cell chemistry of the makeup of the cell, all the way down to um the specific manufacturer and the form factor of the cell. So it's not like you can say, well, I have a certain cell chemistry, LFP, for example, and that will have a certain gas composition, but it really varies from cell to cell. And what drives the explosion hazard you already mentioned is especially the hydrogen content. But then also the rest does matter, right? So it's not like it's not like this is all hydrogen and then a little bit of extra species, but the hydrogen content might range between something like 30% up to maybe 60%, depends what what cell you have. But then also the rest matters and contributes to the energy that is released if there is a combustion and explosion.
Wojciech Wegrzynski:I'm looking at a table from sample gas data. It's a it's a composition of species that are much lighter than air, species that are that are heavier than air. Is this mixture persistently uniform in the time scales that are interesting to you, or are you worried that at some point it's gonna separate into different layers of cake?
Lorenz Boeck:Yeah, we do see stratification in larger scale tests, which is pretty interesting. So when you uh put a silent thermal runaway in a larger enclosure in a lab, you will typically see the electrolyte vapor settling down. Toward the ground, and then the lighter gas is rising. From a UL9540A standpoint, all of this is just reported in relative concentrations, right? It's up to us as designers to make sense of that, what that really means in a realistic explosion scenario. And what is typically done from what I see is that we mostly focus on the lighter components, mostly because those are also the more reactive ones. But it's a bit of an open resource question, I would say, on whether or not we should dig a little deeper into the contribution of those heavier species. That was hard to capture, right? In tests, a lot of it is condense at ambient temperature. So you would have to run the test and sample at elevated temperatures. And then even if you wanted to, for example, use synthetic gases to replicate that in other tests, it's pretty challenging to deal with mixtures that have light and heavy compounds.
Wojciech Wegrzynski:Now it's a selfish question. Does the industry have some standard surrogate vapor or gas mixture that could be used to study the consequences of those explosions? Is there a consensus on that? Because if I wanted to like if if I'm only interested in studying the, let's say, the pressure relief or some other mechanism, I would not like to run my simulation a hundred times for 100 different combinations of gases.
Lorenz Boeck:Yeah, so generally the concept of surrogate mixtures is very important here and it's powerful. One of the necessities here is that if you're running a test with a battery, there's only so much gas that it produces. But you might need a lot more gas to then determine some of the safety parameters, for example, burning velocity or maximum pressure. So what is typically done is you look at the composition that you sampled from the actual battery test, and you produce a surrogate gas that now represents that actual battery gas. And then you run all these other tests like burning velocity and Pmax, or you could also do, for example, large-scale testing. But you need that surrogate gas. This way that I described to going from a specific battery to a representative surrogate gas and then continue on is the typical process right now that we see a lot. I would not say that there is an industry-wide standard for a surrogate gas for that reason. That is due to the variability of the composition, really. But I would love to see over the next year some consensus of how, for example, a worst-case battery vent gas could look for a certain set of chemistry, maybe some additional parameters considered harm factor size, capacity, whatever. But I would really like that because then that would give us a basis for designing in a conservative way for systems where, for example, we don't have 9540A uh data yet. That'd be a powerful starting point that you can then be more specific about in the future once you get the data.
Nick Bartlett:I think generally speaking, it it would be good to come up with some sort of general agreed-upon surrogate, but you know, then you run into the situation where like we could look at a hundred different cells and probably say that the lower flammability limit is between four percent, which is for hydrogen, and let's say nine percent. Like we could all agree on that. We could all agree the burning velocity of a hundred tests would be between fifty and a hundred. So conceivably we could just make some sort of gas that's very conservative. But then kind of what's happened in this industry is that coals and standards are actually, in some regard, limiting the development of technology because we like to benchmark things, we like to standardize things. And this is an industry where everything is very novel. So, you know, theoretically we could come up with an extremely conservative surrogate gas, but it we'd have to put that in context where you know that that might not actually always be representative. It would for certain be worst case.
Wojciech Wegrzynski:Let's maybe venture into the NFPA ecosystem of that, because like Nick, you multiple times mentioned NFPA A55. Maybe let's just clarify how the ecosystem of the standards look like, which standard that describes what, and then perhaps we can go deeper into particular uh aspects of the standards.
Nick Bartlett:Yeah, so if we start, I think at the highest level, it would be NFPA 855, which is the standard for the installation of energy storage systems. So timescale-wise, the first edition of that actually came out in 2019. So, I mean, I would say globally it probably has the most, I would say, input and refinement of probably any of the global standards out there for safety. So it's you know, today it's on its third edition and it's drastically different than what it looked like in 2019, which is, you know, pretty impressive to think it's been six years and it's a completely different document. So yeah, you start you start at at with NFP 855, and that covers commercial installations. Uh, it does cover residential energy storage systems. Actually, it has a chapter on storing and collecting lithium-ion batteries, which is a completely different approach. So then, yeah, then from there you just kind of go down with additional standards and and uh it does cover like how did how do we design the actual product, how do we install it, how do we commission it, how do we maintain it. It tries to bring together all these aspects because it's not it's not just well, how do we design the deflagration events? Of course, that's extremely important, but but there's a whole ecosystem, like you said, where you know, if you miss one one critical aspect, like you you forget to maintain your systems, then you know, of course you're gonna have a problem no matter how well it's designed.
Wojciech Wegrzynski:So it's like equivalent of NFPA 101, life safety code, like uh kind of. Okay, and and the two standards that I hear uh very often about NFPA 68, 69, what what are those? Uh and what was the role?
Lorenz Boeck:Yeah, those are the explosion prevention and protection standards that 855 points to. So instead of reiterating how this is done in 855, we just refer to 68 and 69 as existing documents. Uh 69 is the explosion prevention standard. So if you think about an accident sequence, the first thing you would want to do is to prevent, right? Before you even think about mitigating if the accident happens, but first you want to prevent. So in a logical sequence, 69 would be the first one to consider. This for BEST specifically would include mechanical ventilation systems that remove flammable gases from the enclosure, for example. And then the second standard to consider is NFPA 68, which is the protection standard or explosion mitigation standard, I would say, using defibration venting. So that talks about if there is a defibration inside of your system, so if you're not able to prevent it in the first place, then how do you protect the enclosure? How do you provide pressure relief in the form of defibration vent panels to keep your enclosure safe and prevent any rupture, any adverse external effects?
Wojciech Wegrzynski:Perfect. So let's perhaps try to first of all, like does every battery and the storage system that follows an FPA 855, that does it need both protection and uh prevention solutions on it? Is it required like universally, or there's like a scale at which it starts?
Nick Bartlett:Yeah, generally speaking, in the last edition of NFPA 855, which is 2023, you you could do either or. Because that was based on the knowledge at the time. You could do one or the other. Fast forward lots of discussions between different experts. We kind of realized that even if you have the mechanical ventilation, NFPA 69 that Lorenz was talking about, there is often still enough flammable gases that are generated even in the presence of that ventilation system to create what's called a partial volume deflagration hazard. So now it actually more or less requires both. So you're essentially mandated to do mechanical ventilation now, and you then have to analyze if there is a remaining partial volume deflagration hazard, which would then necessitate potentially installing some sort of deflagration vents.
Wojciech Wegrzynski:When you're having that ventilation system, how does how does it look like? Is it like uh, I don't know, uh a fan that you'd install in uh in a toilet on the wall, like a small device that's always on? Is it like a huge cooler, chiller that also plays the role as a you know cooling of the what what's the device? I don't think I've ever seen a new generation storage facility yet. So I am absolutely uh an idiot in terms of that. I have no idea how it looks like. You have to describe it to me, Nick.
Nick Bartlett:Well, I think I think you were a little spot on there, so it's a lot simpler than you might think.
Wojciech Wegrzynski:So thank God I mean because fire safety got a lot harder than I thought it would be.
Nick Bartlett:No, no. You're you know smoke control, so you kind of know it's this it's a similar concept, just applied differently. So typically, for most of the commercial products, there'll be some sort of gas detection. When the gas detection reaches a certain concentration, typically 10% of the LFL of the specific gas, it'll you know, it'll essentially turn an exhaust fan on, open a makeup air inlet. It's okay, it draws fresh air in, takes air out, and I mean that's really it.
Wojciech Wegrzynski:So it's not something that it's like continuously operational in the in the storage system?
Nick Bartlett:Not not typical. I have seen a few where they don't they're not really concerned about just running a fan all the time. Um, I have seen that application, but 99% of the time it's not running all the time.
Wojciech Wegrzynski:In such a storage system, do you do you actually have some sort of HVC climate control that would be running all the time anyway, or is the the gains are not that big?
Nick Bartlett:Yeah, I mean I would say thermal control of the actual container is kind of a separate topic. So these days, used to be a lot of these had actual HVAC in them. There still are some, but a lot of them have liquid cooling. So they'll have a chill chiller and they'll be circulating glycol. They'll often have some sort of dehumidification because water is not good for batteries. Um so they'll have their own separate kind of thermal and humidity control systems. So it's not like a drive to to make them all together in one mega ventilation system is like separate rolls and yeah, I haven't haven't seen that so far, not to say it doesn't exist, but I would also be concerned about control logic and sequence and how they all work together.
Wojciech Wegrzynski:And how how do you calculate the efficiency of the genit? Like how how does one dimension this? How many I don't know, air changes per hour you need per container, or that's the job of the manufacturer and you don't care as a fire safety engineer.
Lorenz Boeck:It all starts again actually with that information that you get from uh URL 9540A, because you need to start with the hazard, right? So you need to understand how much gas is produced. And critically, when you design 69 systems, also at what rate that gas is released. With a 69 system, you're looking to remove the gas at a sufficiently high rate so it doesn't accumulate. And that's actually a very interesting conversation because getting that rate of gas release is not easy, both from a testing perspective, but then also right now, just from the perspective of interpreting these uh 9548 test reports, it's not that easy to get reliable numbers. But once you have a number that you're happy with and you work with, you can design the ventilation system. There are different steps of design, I would say. You can do a preliminary design where you can use some hand calculations. There are some rule of thumbs in, but also some simple analytical models. But then what we see mostly in practice is that you use CFD simulations to actually simulate the true behavior of the specific ventilation system for that specific best geometry, considering things like how obstructed is the geometry. It's very important, right? Some of these can be hard to ventilate because they're so densely back. But then also, for example, how fast does the detection system work once there's a gas release? How fast does the fan ramp up and do the louvers open? All of these things need to be considered together to really predict how much gas is allowed to accumulate at a worst case time during this entire process. And that gets then measured against the performance criterion of 69, saying that at most during this entire process, you're allowed to accumulate 25% of LFL globally throughout the enclosure. That's the performance criterion. And the simulation will show you that. So that is the most typical thing we see in the industry right now, hopefully. And actually also as per 855, the ventilation rate needs to be verified once installed. So making sure that you're actually getting the air exchange that you have designed for, making sure that you measure that in the actual installed system.
Nick Bartlett:Yeah, and I I think to add on to that point there, and we were talking about this earlier, is you know, it's really important when you're designing one of these that you have some familiar with familiarity with all these nuances, right? Because if you're used to designing buildings and then, you know, somebody comes to you and says, Hey, can you validate this ventilation rate for an explosion control system? One of the things we see is that if you just take this 9540A data, right, it's not conservative enough, right? So a 9540A failure of a cell is some people would call it a gentle overheating of a cell. I mean, that impacts the results, right? You're gonna get a certain gas volume, you're gonna get a certain gas release rate, as opposed to, let's say, a multi-cell short circuit, which is very energetic and produces different results. So what I've seen in the past is that if you're not familiar with these nuances, like you might just take the data from 9540 and be like, oh, this is great. Let's just design the system based on this. But that's probably like designing anything, that's probably not the best approach. And we, you know, with this new NFPA standard, is actually we've been able to get some content in the annex to kind of explain, yeah, you need to take that data and build in some conservativisms and just like you would with designing any other type of system.
Wojciech Wegrzynski:What kind of time skills are we speaking about from like the failure of the cell that releases a cloud of gases to like you having your control 25% LFL? Like, do you have to do it within a minute, five minutes, ten minutes? I mean, I've seen minutes to seconds. So minutes to seconds. Okay.
Nick Bartlett:When we're talking about well, how long does it take to release all of the volume of gas in a single cell? It's usually minutes, right? So the cell has a vent on it, it overpressurizes, it'll release gas, then it goes into thermal runaway. The the gas release rate will usually change, but that's a minutes time scale. And then your time scale for detecting those gases and then turning on the exhaust fin is usually seconds, right?
Wojciech Wegrzynski:And and you're basing it on gas detection. It's not that you're monitoring some status of the battery and the BMS is instructing the uh container, oh, you need to ventilate because I'm having some odd uh measurements. You base it on measurements of gases inside the free volume.
Nick Bartlett:Yeah. Yeah. Typically you see a couple, whether they're a hydrogen detector, CL, similar products, you're monitoring the container for gases, and that's what's triggering the ventilation, not the battery management system.
Lorenz Boeck:Yeah, and that's good practice in safety systems, anyways, to keep those separate, right? That's typically a requirement if you look at critical safety systems that they operate completely separate from the process controls or the equipment controls, which in that case would be the BMS. Now, of course, the BMS has a critical part in the safety of the overall system to make sure it's kept at a healthy condition and maybe also doing things like early detection of potential upcoming failures of batteries predictive, uh yeah. Exactly. Yeah, and that's very powerful, of course, now with modern analytics to really trend these batteries and understand is there potential future failure to be expected. But um, when we talk about 68 or 69 systems, those are entirely separate from the controls of the battery system itself.
Wojciech Wegrzynski:And if if it's not just uh the cloud, if it's a fire, does this tool operate in the same way? Does it have any fire resistance properties of the fans or or it just, I don't know, dies from high temperature or shuts off?
Nick Bartlett:Yeah, I mean, typically, and this was actually addressed in the newest 855, so the operation of most of these safety systems, it's really only expected to be up until the point where fire occurs. So, you know, you might have some resilience, but it's on the minute scale. So a lot of the large-scale fire tests are hours, right? So, you know, you're not expecting these systems to operate for hours, right? Except for potentially passive deflagration events, right? I've actually seen in a quite a number of these large-scale fire tests where hours into the test of flaming fire, we've seen pretty large deflagrations. So, you know, those passive systems are actually quite important, not just prior to a fire occurring, but even uh over the duration of a fire. Because you can get underventilated conditions and things like that.
Wojciech Wegrzynski:You got my attention now because my colleague is doing uh a research grant on post-fire explosion uh resilience of steel uh members. So uh what do you mean? You had then deflagration hours into fire. That's crazy. Like why?
Nick Bartlett:Yeah. Well, uh, you know, we've seen this, so a lot of the large-scale fire tests that are being done today on these, you know, giant ISO containers, the way they'll run the test is that they'll open the deflagration vents before the test starts in order to get some, you know, to build up a fully developed fire. So one of the phenomena we've seen, and I would say at least half of these tests, is somewhere in the middle of the test, as when the fire gets large enough, I guess the you know, my my theory is that the ventilation conditions inside the container change, you get an underventilated condition, and somehow, and Lorenz could probably speak about this as for sure as well, somehow you get deflagration, right? And and sometimes, you know, the the container will deform, you might get projectiles thrown, but it's it's in the middle, could be two hours into a fire. And this observation has actually been it's it's pretty new, I would say, because we've been doing these tests this year. It's really a key observation for first responders that just because the container is on fire does not mean you cannot have uh deflagration.
Lorenz Boeck:This is like a backtraft kind of mechanism, Lawrence. I could imagine. I mean, it really depends on whatever mechanism is available to bring oxygen back into the enclosure. And I think Nick could speak from his experience seeing these tests live. I think a lot of things happen during these tests, right? You might have breach of the enclosure in some areas, you might have some deformations, you might have internal collapse potentially. It's it's hard to tell really what's going on. But the outcome is really that yeah, you see defibrations well into a fire event, which I think is a critical observation. And it teaches us something about the scenarios we need to consider for BESS fires and explosions. It teaches us that we can't just stop with this uh conventional event tree thinking where you know either you get a fire or an Fire or explosion, yeah, exactly. Right. That is unfortunately, because it creates complexity, is not sufficient for BES. But the branch, let's talk about these branches. If you have one branch where there is immediate ignition of the battery event gas, then you are theoretically speaking on the fire branch, which now could lead to a late defigration as well, right? So you're on the fire branch, but then there might be another branch off where you might just stay in a fire or you might have a late defigration. The other part of the branch is if you're releasing gas and you're not igniting that immediately, then you would go toward conventionally speaking, Explosion branch, right? So you're always seeing accumulating gas. If there is delayed ignition, then you can have defigration and explosion. But so you realize that you have two explosion scenarios here. You have one pre-fire explosion, and you have one post-fire or during the fire explosion that we need to understand and consider. And that we also need to protect against.
Wojciech Wegrzynski:Okay, well, uh, in that case, your prevention mechanism 69 definitely will not be very helpful if they die into the fire. So let's move to 68. And uh and how how what what do you do with the flagration? And what are your possibilities? Because I assume you are in now in a branch in which it is bad. Like no matter what you do, it's probably bad. Like how much not bad can you make turn the bat into?
Lorenz Boeck:Yeah, so I think the question, what is worse, a fire or an explosion, that's hard to answer. Right? They all have that individual effects and consequences. But explosions are concerning, especially for emergency responders, whoever is on site, and they are concerning for any exposures that's around the site. If you're in a densely populated area, for example, you you need to really understand the possibility of explosions and the potential consequences and what happens to your exposures. Now, thinking about those two scenarios, pre-fire, post-fire defigations, we can protect them with a good layer of protection concept. And that is, in a broader framework sense, what 855 gives us. It gives us this idea that you leverage multiple layers of protection, meaning NFP 69 and 68, to ultimately safeguard that system. 68 and 69 systems will work a little bit differently in these different scenarios, but that's a general concept to use both together and an optimized combined framework. If you have a pre-fire deflagration, so accumulation of gas, ignition, and explosion, then the flagration vent panels are a great solution here that you know are installed on the container. If there is overpressure, then the panels open, the relief overpressure, and your enclosure is kept safe. But you might also need the 69 system to help you a little bit to minimize or at least limit the amount of gas that can accumulate. Because if you're accumulating too much gas in the container, that might overwhelm 68 defigration vent panels. So that's a great branch where you leverage both together to manage that hazard.
Wojciech Wegrzynski:With the with the defigration panels, how big they are and how hard is it to open them? Like, is it just you know a flappy window that opens? Is it like a device that does respond? Like what do they look like?
Lorenz Boeck:Right. So typically what the industry uses are manufactured deflagration event panels. That means there are panels that are pre-weakened in certain locations around the circumference so that they open at a very defined overpressure. So, for example, 50 millibar overpressure or 100 millibar overpressure, which is not much, is well below the strength of these containers. So they're very purposefully engineered for that. They are tested and certified to make sure that they actually repeatedly open exactly at that pressure. And the size of the panels needs to be calculated for a specific system. So you don't just buy a panel off the shelf, you don't just put that on your enclosure, but you need to consider the explosion hazard in terms of the flammable gas. You need to consider the geometry of the system, and then you need to exercise sizing calculations or modeling to determine how big of a panel or how many panels you need to protect the system. Is there a downside to having too large panel? Like can you make your entire roof a panel? I think so. Some of the very, very severe hazard scenarios we see, meaning those where a lot of gas is generated and the gas is very reactive, would force you to a place where you almost have to rent the entire roof of the store, uh, the entire roof container. So that is one of the limitations of a 68 system if you wanted to use it standalone. And I think that was recognized by the 855 community. And that's one of the reasons why we went from the previous 855 2023 recommendation to use either 68 or 69 to now using both together to manage that. So that is a limit, literally, how much area do you have available on your enclosure where you can place the flagration vent panels?
Nick Bartlett:Yeah, and that and that issue actually plays itself out quite a bit as you get smaller and smaller enclosures, but the cell size is the same. So, you know, whereas your typical ISO container, you probably have enough vent area. We've proven that as you get to these commercial and industrial cabinets, you know, the design challenge becomes significant.
Lorenz Boeck:Yeah, that's true. And that's that really highlights that all of the things we're talking about. This is all an engineering discipline, right? You need to calculate and and predict and model how much of that protection you need. And in the case of a 68 system, how much area you need, in the case of a 69 system, how much ventilation rate you need, how you need to place your ventilation components and so on. It's not a one size fits all. It's very much system specific.
Wojciech Wegrzynski:I think I saw your colleague in Hong Kong talking about uh interesting solution from Rembe on angled product uh on angled panels. I mean, I don't do products in this podcast, but it was clever. I'll I'll low it. So he can tell about your innovation if you like.
Lorenz Boeck:Yeah, and really what my colleague Marius showed there in Hong Kong is some of the research work we're doing to try to advance solutions for this industry. So, aside from the commercial aspect, we really think that there is uh a smarter way of providing deflagration venting for these types of enclosures. So, conventionally, what the industry has been using are panels that are installed on the roof. One of the main drivers were to just needed so much area, and the roof was the only place to find that much area. And the other part was that venting through the roof means venting flames away from potential uh exposures. And so it was a safe way of doing that. Now, the downside of that is that you need to consider external effects, for example, weather effects. And many of these systems are installed in areas where you could have snow accumulating on top of these containers. You have that. If you understand the physics of defigration venting, you realize if you weigh down these panels, they will open slower. That's simple physics, nobody can change that. So they will not necessarily activate as designed under severe snow loading conditions. Now, what we did at Rembe, and we did that for different industries for many years, is think about how we can move these defigration vents to the sides of the container, away from the roof sidewalls, to protect them from snow loads, but also protect them from things like hail, for example, severe hail storms. And that's what my colleague Marius showed is a product called Targo vent. It's it stands for targeted venting. It's a defigration vent panel that still uses the same physics, so it opens under overpressure, but it has an angle limiting device. So it only opens up to a certain angle and then allows you to deflect the explosion upward and away from exposures. And again, that is an industry-specific solution here that we think makes a lot of sense and that has gained traction. You will already see this in some of the commercial systems out there. And I think it's one of the ways where the explosion protection community, and especially we as a manufacturer, can contribute and support that industry, make more application specific protection systems.
Wojciech Wegrzynski:Do we have some competing interests between preventing fires or consequences of fires and explosion prevention? Like now you say you could put this thing on the side, which obviously brings the fire nearer to the neighboring container, for example. So how big are the trade-offs and how problematic are they for you in this prevention?
Lorenz Boeck:It's definitely critical to consider the big picture, right? So you cannot really design a protection system, let's say, against explosions without considering all the other needs of the application. And what's nice now is that we are conducting large-scale fire tests. So we will understand a lot better how things like explosion protection devices interact with a fire scenario. If you had panels on the roof and they're open, then you typically see a relatively large flames exiting from these panels. And what that can do is that it can impose a pretty high heat flux to the container behind the unit that is on fire. So very often these units are installed back to back with another container. And if you have panels on the roof, you get that very high heat flux to the back-to-back container. Uh in that case, if you have a panel mounted on the front walls, then you're actually moving that flame away from that exposed container. It may expose the container across the aisle a little bit more, but there is quite a bit more separation and it's still a much less uh severe heat exposure than, for example, if you had open doors in the front of the container. So we've seen some large-scale fire tests with these types of sidewall panels where actually the panel mounting provided a benefit to the outcome of the fire test.
Wojciech Wegrzynski:We're nearing the end of the episode. I still want you to talk about the training a little bit, but I have one curveball question at the end. Would you rather see like a double height single container or two containers one on each other? Because I have a feeling like uh they were growing in height one way or another. What's the second option? Two containers, one on top of another, or just one make one ISO 40 but double height.
Nick Bartlett:Oh well, they're they're both of them are being developed, so we'll we'll see both of them within a year. Two two two double heights stacked on each other.
Wojciech Wegrzynski:Let's go. No, but but seriously, I mean, how big an issue, how how big shift of the paradigm is the attempts to optimize the energy per land uh square meter because people try to clump as much as they can nowadays.
Nick Bartlett:Well, I I I think you you hit the the nail on the head with Noah. I mean, this is this is where the industry is going. So there's one manufacturer that's trying to produce a 20 20 megawatt single container, like one giant container. It's it's too big to be shipped. Like it's just that big. And then and then the other shift is multiple manufacturers are doing the the stacked containers. So I think the the hazard is it's to be seen, like what the best protective strategies are for these new and unique uh installations.
Wojciech Wegrzynski:It's so interesting that we are in a field you said uh the NFPA was developed in 2019. That's the most developed standard because it's with us so long, already three editions, like and uh like five, six years later, we already have technology that is not just you know, outside of this, it's like a complete different planet of of safety solutions because I think it invalidates so many things on so many layers. Uh like wow, it what what a world that we're living. Uh people really need to uh to train. Let's move to that ad. So uh what what's your efforts in that area? Because I I heard so so many good things about uh the courses that you're providing.
Nick Bartlett:Yeah, so a couple years ago we decided to create a course on NFPA 855. I mean, we don't we don't represent NFPA or anything like that. And the reason we did that is because I kind of saw, like I kind of mentioned earlier, in most of these projects, there's a lot of different stakeholders, right? So there's a fire engineer, there's a risk engineer, there's the battery manufacturer, there's the regulator, HGN, there's the insurance, there's all these different people, right? And so all of them have a different understanding of what's required. And so, you know, when you're the person that's trying to shepherd an entire project to compliance and everybody has a different opinion, that's kind of a problem in general. So one of the reasons we created the course was to attempt to start to bring hopefully some sort of standardization at least, or education as to what are the requirements, what is the intention, and like why why are they requirements? We don't just talk about what they are, but like why, why, and then how to apply them. And then, you know, what was really cool in London this year is like I said, you know, in the United States, we've we've trained a lot of fire engineers and AHJs. Well, when we went to London, I mean the the crowd was it was all sorts of people. And when you start getting everybody in the room, then I think we can make you know a real difference. So um we've only done one course this year in London. We're doing a course in Sweden in January, so I hope that you know your listeners, if they're in Europe, they can attend that. It's been gosh, 14 years, 12 years since I studied in Europe. So I have very little connections there at this point. So um we hope a lot of people can come out to that course.
Wojciech Wegrzynski:If you send me a link, the listeners will find it in the show notes. I've I've only heard good the good things about the previous editions. So I guess if if you're dealing with those systems or you expect that you will in your future, whether you like it or not, it's it's probably a good idea to educate yourself, uh, not just in in the podcast episodes, but that there's an awfully lot of reading and a lot of of learning out there.
Nick Bartlett:Yeah, yeah. And one of the things you know Lorenz and I are gonna start doing is next year in May, we're gonna do like a course, it's like a day and a half, right? So instead of just teaching NFPA 855, which is which is essential and great, with Lorenz's knowledge that he's talked about today, we're gonna do a half a day just on NFPA 68 and 69 because there is just so much. Like in a one-day course on NFP 855, you can spend 30 minutes or less on that topic. So we're gonna do four hours, and we just we feel like that's an area where a lot of people could benefit from because there is just so much so many fine details that uh be great if everybody's on the same page.
Wojciech Wegrzynski:Can you recommend some open source resources that people could dig in outside of reading an FPA? I'm not sure even sure if that's open source. Like uh they used to be free to read.
Nick Bartlett:You can you can on their website, it's just you have to click page by page, it's kind of annoying. It it is free.
Wojciech Wegrzynski:A little price for for free.
Nick Bartlett:Yeah, yeah. Lorenz, I don't I don't know if you have any good recommendations.
Lorenz Boeck:You know, I'm very much on the more fundamental side of the explosion protection topic. So that can be overwhelming because there's so much work going on in the resource community. But if you are a researcher in explosion protection or also on the fire side, you know, dig in, look at your typical journals, journal of loss prevention, uh, for example, that has brought out a few studies on BES and some of the things we talked about, such as how to properly apply 68 or 69 systems. But there's a lot to be done still. Um, so there is literature, it's pretty broad, but there's so much research work to do. And I know many of your listeners are researchers. So yeah, dig in, help us identify the critical topics. We're more than happy to be part of that conversation too, because we want to drive knowledge in this field. Of course, we want to teach and and and show you what the current state of the knowledge is, but it's so important to keep driving that as grouping. Fantastic.
Wojciech Wegrzynski:Yeah, I agree on that shout out, and I'm observing this field. Uh, I think my first battery episode was episode six in the podcast many years ago. It was more about vehicles with Roland Bishop. He he's now uh involved, I believe, back then in RISO. Yeah, it's been quite a journey to observe how this field develops, matures, how those branches are most more distinct. Now you shattered my world with uh being able to jump from fire branch to the explosion branch. I was living very happily, you know, uh just worrying about one, not both at the same time. But in the end, you know what? I actually we we had an event like that in Poland. There was and it I think there was loss of life of firefighters in in that one where we had some sort it was not storage, it was more like a repair shop or perhaps some some kind of shop that that dealt with batteries in a cellar of a building, residential building, and and there was a fire followed by an explosion. So so that that was quite tragic. So we recognize already the the dark side of those events. I hope we learn more about them and how to prevent them. And in terms of those commercial applications for large energy storage, it's great that there's already sources of knowledge like NFPA standards that uh lead you to some sort of standardized technical level of those. And let's let's hope they grow and develop. And if they're wrong, I hope they fix them. If they're good, I hope they reinforce them. That that's that's my only hope. Um, guys, thank you so much for coming to the to Fire Science Show to to share this. And I hope to see you somewhere uh soon. If you ever blow up an energy storage system in Europe, call me, please.
Nick Bartlett:Thank you, thank you for inviting us. We really appreciate it.
Wojciech Wegrzynski:Thanks. And that's it, thank you for listening. I had so much fun recording battery episodes three, four years ago. And it's it's really fun to see how much more technical we can be in those episodes today. Back then, four years ago, we were talking about general things, like what's a cathode, why they burn, you know, some very initial ideas. We were dealing with with engineering problems already back then, but with really, really scarce and limited knowledge. And today we are in a world where we have really good grasp of the technology. We have a knowledge, we have solutions that have been tested, proven, solutions that have been disproven, upgrades. So much wealth of knowledge. It's it's it's such a joy to observe the whole landscape of a discipline has changed over a course of just a few years. As I'm doing this podcast, just a few years, and it's such a major shift. Anyway, I think it was an excellent summary of the problems with explosions related to battery energy storage systems via SSS. Nick and Lawrence gave a fantastic overview. Obviously there's a wealth of knowledge to be found in in their cursors, and uh if you dig deeper, I would start with an FPA sixty eight sixty nine, you really go through them and then you probably will uh pick up and learn a lot of m a lot of new things. I think it could be actually interesting to re-listen to the episode after you go through those uh standards because from my experience you always discover something new in this in those conversations when you get yourself more familiar. If you enjoyed this episode, let me know. I will try to make more episodes like this. There have been a lot of batteries recently in the Fire Science show, but uh I see that listeners resonate with this topic. I've been into two big conferences, I enjoyed them thoroughly. It has it's really a nice subject to talk, and I think uh something that uh really is interesting for a lot of you. So yeah, thank you for for being here with me. Um next week, next week's Christmas, so there's no fire signs. Spend some time with family, have some fun, take a little bit of rest. I don't think I'll be publishing the episode on the Christmas Eve this one week of the year. I would like to take free, and I really need downtime. So I hope you will allow that. But I will still see you here in the last day of the year because there's gonna be one more episode this year. I hope that gets released on the New Year's Eve, and we start uh the new year with a really solid piece of uh fire science. That's the plan, and I will love to see you there. And for now, if you celebrate Christmas, Merry Christmas, and if you don't, I just hope you have a good time and some rest along the way. Thank you very much for being here with me. Cheers, bye.