Loading summary
Transcript62 lines
- [00:00]
A
I'm your host, Ed Porter. Welcome back to Transmission. Here's something most people get wrong about battery fires. They picture a cell fault triggering the issue. But only 11% of incidents actually start that way. The more common cause, how the site was built, how the site is operated, the design around it. The good news, failure rates have dropped by 99% since 2018. The harder truth is that fewer than a third of incidents ever get a root cause attached. The question isn't really are batteries safe? It's how much better could we get? Today I'm joined by Dan Charlotte Burke, Director of Asset Management at Gore Street Capital, to walk us through all of it before we dive in. If you want to go deeper on battery availability, ask CO Moto Ng's AI analyst. And if you're enjoying the show, please drop a comment, give us a rating. This helps us reach others who might find the show useful too. Hello, Dan, welcome to Transmission.
- [01:06]
B
Hi, Ed, It's a pleasure to be here. I feel like I've walked into kind of like a famous area for me.
- [01:12]
A
Yeah, well, it's our pleasure to have you here. I wonder if people do know this sort of like Houston backdrop, having seen a few episodes, maybe it's like it just, it looks so specifically London anyway.
- [01:26]
B
Yeah, yeah. And actually this is for me also like it feels almost like a rite of passage as well. Now in the industry, like if you've got like a voice and lots of opinions and worthwhile things to say to get onto this podcast is kind of a big thing.
- [01:40]
A
So our pleasure, and we love to make sure that people with views are platforms so that anyone can listen to what's going on in the space because we think it's super interesting. And let's get into it. So what's one thing that everyone gets wrong about fires in the battery industry?
- [01:56]
B
So this is an interesting question. I mean, I think there's not one thing that everybody gets wrong, obviously, but if you come from like the perspective of what does the public generally think is going to happen as well? I mean, we typically, I think if you ask the random person in the street, you know, if that battery catches fire, what's going to happen? Their vision of it is the whole site's going to go up in flames and you're going to have this massive catastrophic event which could cause sort of wide reaching harm to the area. And if it's not actively dealt with, that whole site's going to be up in flames and it's a full write off. In reality, that's a vanishingly small fraction of energy storage sites that we've got operating today. It's really not something you're likely to see in a modern site whatsoever because design has improved so much that propagation from one container to another is massively reduced as well. And part of it's as well thinking that you need to actually come in and actively put out the battery fire to stop this propagation as well. There's a lot of requirement for say, water on site to we use it for sort of boundary protection where you actually sort of spray one container next to the one that's on fire and you try and prevent that propagation. And that does help to mitigate propagation across the site. But if I was to actually change my answer, I think the use of firewater and the requirement for boundary protection is one of the big common myths. It's not to say it's not required in some instances, but genuinely nowadays the let it burn approach, where you don't bring water in at all, like that site's very unlikely to have the fire propagate across the site anyway.
- [03:38]
A
Okay, so it's two answers in one. Right. So you've said people might assume that a fire starts in the battery site and that means all of the battery site goes up. And you're saying that's not true.
- [03:47]
B
Certainly in newer systems, very unlikely. Yeah, it has happened. And unfortunately last year in California, the Moss Landing fire is obviously going to stick out in a lot of people's minds because it was a huge, huge event. The largest battery fire that we've really ever seen.
- [04:03]
A
Okay, well, for people not familiar with Moss Landing, let's go through it straight away. So what is Moss Landing? Why, why do people know it?
- [04:12]
B
So Moss Landing is a facility, it was several facilities actually operating in California, unfortunately, next to sort of protective wildlife areas as well, and very, very much in the vicinity of local population. So in terms of an area to put a giant battery storage system, pretty suboptimal. And what happened was after actually a history of fire events at that site in sort of neighbour or sister facilities that are right next to it, what happened at this site is a thermal runaway occurred in one of the modules and propagated across the site. You know, you've got, I think it was something like a gigawatt hour plus of, of actual energy storage within one building.
- [04:53]
A
And if I'm right, that was split across. So the Moss Landing isn't just one big block of batteries. There was sort of one block of batteries in one like warehouse and then there were two sort of other Blocks, maybe one other block sort of somewhere else on site.
- [05:05]
B
That's absolutely right. And the one that I'm talking about here is a site that was basically the battery was built into an old turbine room. So it's actually quite different from a typical battery storage system today because everything's built within one room. And, and the way it was set up is quite an old design where you'd have the battery racks one after another without any sort of segregation between them. So the, I mean it was set up that if one module and rack does catch fire and propagates, it can propagate throughout that entire facility. As I understand it, they were somewhat reliant on an active sort of fire suppression system using water. And you know, we can come to it later, but realistically that's unlikely to have worked and obviously in practice proven it didn't work.
- [05:49]
A
So how has that changed to today's system? So in the sort of that old fashioned system, you've got lots of racks all lined up, one next to each other inside a building. If I go to a battery site today, how does that look different?
- [05:59]
B
So most systems today, I mean the overwhelming majority are actually sort of products put together where those products are. Battery storage containers, they might be just batteries. They might be batteries with some equipment that can convert the DC power into AC and then get that onto the grid more easily.
- [06:16]
A
And that just looks like, like 20 foot shipping containers on concrete blocks, standard
- [06:19]
B
ISO containers that I so like the international standard. And you would, you know, you would barely be able to differentiate it from the sort of shipping container you see on a lorry going down the motorway. It's really a standard looking shipping container, often with some kind of, you know, it's going to have its own, like fire suppression within it. It's going to have some kind of an insulation to attempt to prevent the propagation from container to another. But really when you put these on site, they're in plain air as well. So they've got ventilation around them as well. They've got H Vac units, which is the air conditioning that can control the internal environment. And it's a very, very, very different story. And nowadays as well, the chemistry has changed somewhat as well.
- [06:59]
A
Let's do that. Let's talk about the chemistry change. So Moss landing would have been old chemistry, that would have been nmc, nickel, manganese, cobalt.
- [07:07]
B
Yeah, so that was an older chemistry that's actually come out of the EV industry. So electric vehicles initially were the absolute driving force between the R and D that went into production of Battery, battery cells. And that was basically turned into the energy storage systems that we used to have and made to work nowadays. Lfp, which is lithium ion phosphate, is the typical cell type used. There's quite a few benefits to it, but from a production perspective you can basically make the same amount of energy storage for less than cost of nmc. Also doesn't use materials like cobalt, which is a ESG benefit too. But it is widely considered to be quite a lot safer than nmc. But I want to be really cautious when talking about it and just say it's not inherently safe. It's still a lithium ion battery. It still carries the same risks that you'd have in the older nmc, which is nickel, manganese, cobalt chemistry. And if not well managed, you're still going to be looking at a similar kind of outcome with a large scale fire on site and potentially even higher explosion rates.
- [08:20]
A
Okay, so I was doing a little bit of research on this between NMC and LFP and I said, so it very much matches with what you're saying in terms of the risk. And NFP is much lower. The sort of point at which it starts to combust is higher. So that means there's more leeway to try and get that far under control or to try and sense something going wrong on the site first. And actually maybe let's come to the data point in a second. But what I was kind of really fascinated by is sort of like, why is that happening? And so if you forgive me for the slight detail, the NMC systems, the bond to the oxygen atoms is slightly weaker in them than in lfp. And so nmc, when it burns, is sort of able to release oxygen, which is obviously bad for a fire, whereas LFP holds it quite strongly. And you think that's a bit counterintuitive because you kind of think LFP like phosphate, we've all seen phosphorus in various forms and you kind of think, oh, that's going to be quite dangerous. But no, phosphate, when it's sort of in its oxidized state is very stable. And yeah, there are several kind of examples of phosphate based fire retardants and things. So it's a fascinating chemistry episode that I'm sure we could get into. If there's no, if there's an aspiring battery chemist who wants to come on and explain all this, then we have a slot for you.
- [09:29]
B
Yeah, I mean, I will say there's a couple of elements to that and I mean, stepping slightly maybe out of my comfort zone because I'm not a Chemist. But if you've got these cells that are going through some kind of electrolyte breakdown, the reaction is exothermic. And that's a really important point when it comes to fire safety of batteries, because as these cells break down, they are generating their own heat. And actually heat can induce that reaction in the first place, which is the very reason that you can get to a point where you have thermal runaway, because the heat that you're generating through the. The reactions that are happening is enough to trigger more reactions to happen. And you get this sort of cascading event where the fire can propagate across the cell module, through the rack, and potentially across a site, but pretty unlikely. Now, the reason LFP is generally considered so much safer is its thermal stability. So the two elements to that which are most important is when it does burn, it burns at a lower temperature generally than NMC chemistry. And also it takes a higher temperature for that combustion to initiate as well. So you end up in this sort of double whammy kind of benefit that it's going to be harder for it to catch fire from the heat. And if it does catch fire from the heat, it's going to take a lot more of that for it to trigger more cells to do the same. The second element, which you're absolutely right to point out, is the presence of oxygen. So oxygen, you know, it's an excellent fuel used in sort of rocket science and such because it's just so good at burning and producing energy. And in the exothermic reaction going on within those NMC cells, oxygen is produced, so it creates its own fuel source to propagate that sort of combustion.
- [11:07]
A
Okay. I think people will be fascinated by this. I hope they will. But we've. And so we've done a couple of parts of this. So we've done the first part which was, okay, look, we've learned about the design, we've kind of moved on from the sort of let it burn concept to actually we need to manage this and we need to put things in their own containers. And we need to make sure those containers are insulated and spaced appropriately, appropriate distance from each other, and make sure they're in open air so they can ventilate. We've also moved the chemistry on, so from NMC to lfp. Is there like a really interesting data part to this? Like the amount of data that you now have as an asset owner and an operator? Is that just worlds apart from kind of where it was 10 years ago?
- [11:44]
B
I think that that's it's going to depend on the site. But in principle, yes, your standard site today, if you can. Let's, let's imagine that you're coming in and you're new to the industry. You've never built an energy storage system before and you're just going to buy off the shelf products and hope that someone puts it together. You don't have a project manager, you don't have a design engineer, you don't have any of that going on. You, you're just gonna trust the supplier to give you a good site. Well, the default nowadays versus what it was is so much better from a data perspective. The data is well integrated throughout the system. You've got a control system you can interface with and you can typically get data points for things like cell voltage, current and temperature, where in the past that wasn't necessarily available. In fact, some of the older systems don't even have plant controllers. So even if the data is there, it's kind of stuck within the container
- [12:34]
A
and you never find out.
- [12:35]
B
Yeah, it might go to the original equipment manufact manufacturer like the cell supplier and they might be monitoring it to some degree. Okay, but you're trusting it.
- [12:43]
A
Yeah, but the thing that's sort of fascinating about this is that you get that 20 foot container, you're not getting data on temperature, as you say, for each of these cells. As soon as you see one cell kind of going over a limit that you set, you could kind of say, well actually no, let's shut down that block and start to look at it.
- [12:59]
B
So that is a really. I've got a bit of a story on this without putting anybody's names out there obviously, but you would think in real time, can you do something with the data to actually stop the thermal runaway happening?
- [13:12]
A
Yeah.
- [13:13]
B
What you're going to see in practice, if that cell starts to go under thermal runaway, the temperature spike is going to be so fast that really you could come to the conclusion when you look at the data there was nothing you could do. I have spoken to large suppliers in the industry that have looked at fire events that have actually come to that same conclusion. And this is years ago. Now we do know better industry. The data itself though, this is something that I've been a huge proponent of for years actually and we've used it even to drive down our insurance policies. So it's like from a commercial perspective, an investor perspective, it's a win win scenario. That same data, if you can get it from your system and put it up into the cloud and put sort of advanced algorithms against it. What you're doing is basically having a physics based model which you apply machine learning to, to say how is the cell going to behave? And you can issues potentially, you know, weeks, potentially even months ahead of time with that same data. So in that scenario where I was told by a supplier that, you know, they couldn't have done anything with the data because the temperature spikes so quickly, you can't stop the system quickly enough, what you get there actually is, you know, weeks or months of data ahead of it saying that cell is actually exposed to a much higher risk. It's not saying it's going to go into thermal Runway either, but it's giving you the tools as an operator to choose to take it out of service. And I can say we put this across pretty much all of our fleet and it's sort of third party software that goes on top of all of the data we're getting from the site. And we have reacted to those signals telling us to replace modules.
- [14:54]
A
I think this is so fascinating. It's easy from the outside to see the battery industry and think, oh, it's just the same battery that was sort of plonked on a concrete plinth 10 years ago is the same one we see today. And I think these stories really helped to sort of dispel that myth. But let's, let's put me outside of the battery industry and let's say that I, I'm going to take the opposite position to you. I want to say, Dan, I've heard everything you say. I think it's all rubbish. I don't believe any of it. Of course you're going to say all of this because you work in the battery industry. So like what does the data actually say? That that's because that can't be debated. Right? That's the.
- [15:30]
B
Well, probably not really. It depends how accurate you think the data is. Firstly, there is unfortunately quite a lot of sort of unreported incidents that are out. You know, I'm going to, I'm going to turn to the, the, some of the data from an industry group called epri. So they're the Electric Power Research Institute and they're a sort of nonprofit organization from the States. But they have built this failure incident database with the goal of when a large scale battery system catches fire, reports on it or just, just record that it's happened and attempt to fill in some of the gaps as to what caused it. And what's the cell type, who's the supplier, all that stuff and build up this database of what's gone wrong and how. So looking at that data, there's actually quite a positive trend in there. If you go back to the sort of first Data points In 2018, what we've seen is the failure rate's gone from about four incidents per gigawatt hour of deployed energy storage to nowadays quite a lot less than 0.1 incidents in the same volume, which is a. There's been actually a 99% decrease in failure rate on an energy capacity level. So that's actually a really positive trend in terms of the data that said volume is obviously increasing as well, and sites are getting larger. And one could argue that if you very, very poorly designed a site, you know, instead of going from the days where we used to have, say, a 10 to 50 megawatt hour site, we've got gigawatt hour sites now being built, so the exposure of it going wrong could be higher. But then it comes to a design choice as well. So we've talked about the difference between LFP and NMC batteries, but we also have much better design as standard practice in terms of the equipment that's put in there, not just to monitor through the data, but to actually actively manage ventilation within the system, or if there is an explosion, to channel that explosion too.
- [17:30]
A
Is there a bit of a narrative, Sorry, I will come back to what you said about the reporting and the reporting of fires, because I think that's really super important. Let's come back to that. But just first, there's a narrative question here, which is maybe in 2017 we had 1 gigawatt hour of batteries, let's say, and so four fires was kind of the total number on the system. Now, when I look at sort of the system we have globally, we are probably somewhere between 50 and 100 gigawatt hours of batteries globally, and that number is not getting smaller. So we are going to have way more batteries online. So the gigawatt hour number is going to be much bigger. So does that mean that even though we're getting much better at managing it, the sort of absolute number of incidents you'll see will be higher?
- [18:13]
B
I am going to do a bit of a politician's answer and say it depends. So in the business as usual, if nothing changes today, right, and we continue as we are, there's no reason that failure rate should drop. So if the volume is going to increase, the absolute number of incidents would increase. There's a really interesting data point, I think, which is about where does the fire initiate? And a lot of People would fairly have the preconception that a battery storage fire is going to be triggered by a battery cell just catching fire for whatever reason. It's not to say that doesn't happen, but I think there was a joint study done with the team at EPRI that I already mentioned, but the Pacific Northwest Laboratory also in the States, and they were looking at the actual root cause to what caused these battery installations to catch fire. And it was only about 11% of those incidents were caused as a root cause of the cell having an issue and actually triggering the thermal runaway. The really good takeaway, I think, or if I was to be optimistic for a minute, there's the takeaway in there, which is 65% of the incidents were caused by a combination of both operations. So how the system was being used in the run up to the incident and integration, so how it was actually put together in the first place. So if we continue to improve design as well as the diligent elements of how we commission a system and test it and do checks before we enter operations, which is always up against a commercial element of trying to make revenue as soon as possible, you probably face delays. There's going to be pressure on that commissioning program. Do not let it be squeezed. You need to do the commissioning. Right. But the positive takeaway for me there, in a roundabout way, what I'm saying is those two elements that drive the vast majority of incidents and battery installations, they're manageable. We've improved it through design. We can improve our operations, we can learn more from what goes wrong to stop operating sites into those conditions, and we can continually improve our commissioning efforts. So that stats of 99% reduction in the rate from 2018 to 2025. Well, if we're going to say that the volume is even going to increase 100 fold, then why should the number of absolute incidents increase? Why not target in another seven years, another 99% reduction? Maybe we can't do that, but you can at least get rid of that 65%.
- [20:35]
A
And I think it's really interesting to talk about when those incidents happening. So Will Murray, who I know listens to this podcast, was telling me the other day about the bathtub theory of incidence, which is that if you take the X axis and you make it the years in operation and your Y axis is incidence, then the problems start at the beginning. So, like the beginning of the site, and at the end of the site, you also get this. So the shape of that curve looks like a bathtub. And I think what you're Saying kind of backs that up, which is that yes, we've kind of, we've got out a lot of the risk from that middle period. But if we can get better practices around our installations and our decommissioning and decommissioning is probably one that we haven't learned so much on so far, then if we get to that point, we could be, as you say, getting that failure rate down even further.
- [21:18]
B
That is a really interesting, that's such an important takeaway as well. There's unfortunately a large scale incident. It was either the 30th of April or 1st of May this year. So a site actually did catch fire. Unfortunately that was actually going through the augmentation process. So the owner had basically, you know, said, you know, this has reached end of life on the NMC cells. They're going through a process where they're going to turn it into a two hour system, decommission the existing cells. And that's when the fire happened. So we know there's obviously a risk there at end of life. I talked about operations being a key part of, of operating batteries into failure. Well, as you operate these batteries, degradation. So basically if you had 100 megawatt hours of energy storage and day one, you might actually say that 70. When you've got 70 megawatt hours usable, you know, your, your battery is probably starting to reach end of life. And you can think of it kind of like a phone battery. If you've got an iPhone and you ever use your sort of battery health indicator, you might find that, you know, it starts to get really temperamental or the battery life feels really, really short or unreliable and you're seeing sort of a health of maybe like 75% still on the battery. That's actually starting to get to the point where it's pretty hard to operate that battery well and reliably well. That's also going to cause potential issues when it comes to sort of the scenarios that would trigger thermal runaway in a cell as well. You've potentially got sort of more internal resistance in that cell creating more heat. It's been subject to more mechanical stress, which is ultimately what's actually driving degradation. All these factors over time may create the conditions that could lead to sort of more risk in the battery itself catching fire. And that's an end of life problem which we've not seen all that much yet because most systems are still relatively new and there's only a sort of small fraction of them that are reaching that end of life stage. So the Fact that we've even got one large scale event that's happened kind of at that end of life stage is definitely something we should be alarmed by.
- [23:21]
A
I think it's a really frank assessment of like where the, where the sector is and I think that it's really good to have that view. Now we put a pin in the topic of reporting earlier. I want to come back to it. So you mentioned that sort of not every battery fire gets reported. What do we need to do in terms of reporting? Like what should the industry, how does the industry get reporting?
- [23:43]
B
Right, so yeah, I've done quite a lot of this actually over the years through the various sort of groups that I've either chaired or we set up like an owners forum where we could have sort of frank discussions kind of purely as owners who had concerns. This was years ago but it's become quite an established thing now where we could have these sort of Grantham Institute or sorry, Chatham House Rules. I'm going back to my imperial days there. Chatham House Rules apply. So you don't use any names. Right. And you actually just share knowledge with each other as best as you, you can. And the idea is, you know, we can all support ourselves by sharing our lessons learned, getting those out into the industry and helping others avoid the same mistakes. That's absolutely a necessary thing to do. When I'm saying that you know, a huge chunk of the incidents are caused by operations in the first place. So that does require people to put the time in though. You need to have responsible operators that are willing to put that effort in. And I still think there's a bit of a concern that not, not everybody is at that level. Personally I'm pretty taken aback by this study that twice did on the TWA ce. So the battery analytics firm, they were looking at the, the incidents that were reported in the Net Pre database and trying to link them to sort of public information on the root cause. And this, this was probably like a couple of years ago now. I'm not sure how much the stats would have changed but what they found was less than a third of those incidents and those are the reported incidents. And I'm saying many of them probably aren't getting reported or quite a few only not even a third of them had a root cause associated with them. I'd be shocked if the vast majority didn't have a root cause that was known by someone that had done the study after the incident took place. Because imagine you're an operator of a battery storage system and it catches fire. I'VE already said that most of the time that's not going to propagate across the site. In the vast majority of incidents, really, that is the case. So you're likely to be trying to restart the site at some point afterwards. Well, if you're an asset manager, you're not going to want to restart the site until you know what's gone wrong and that you're going to be avoiding it in the rest of the equipment. So it just doesn't align that we've got less than a third of those sites reporting the root cause. And I think that's a bit of a shameful statistic, to be honest. That is a global statistic. It's not something we can necessarily deal with here in Great Britain alone, but we can set the example. And one thing that I really keen on seeing is, you know, more of a sort of mandated requirement to. If there is an incident, it's kind of a matter of public interest that you share your root cause assessment of what went wrong and you get that knowledge out into the industry. There are some operators that are really good at that. And I've seen examples of battery fires where, you know, those teams that have come in either doing the decommissioning and rebuild or just manage the emergency response have actually come to either sort of the Electricity Storage Networks Working Group or even the one with the Health and Safety Governance Group, which is with desnes, and they've come in and presented on what could have been done better and their lessons learned. And that is really good to see. But unfortunately I've also seen exactly the opposite where sort of a leader of a company has said that they've done enough in this really interesting example of a battery fire where there was actual propagation between containers, one of the only ones I can think of, they basically said, we've, we've done enough, we're not going to talk about it.
- [27:09]
A
Okay.
- [27:10]
B
And that, that shouldn't be a possibility. You shouldn't be, it shouldn't be allowed. And if nothing else, a council should be able to say it's your prerogative not to report it and share the information. But we, for the interests of our sort of constituents, like, we're not going to let you operate again.
- [27:25]
A
We want to see good practice put in place. Yeah, it feels, it feels like on one hand, it feels just very good that we've been able to go so far without even having this kind of best practice sharing in place. Yeah, it's your, like, how do we do another 99% reduction this feels like an obvious place that we could go to better information sharing and maybe to kind of move on to sort of the next part of this. And I'm sorry you said desness for people listening who don't know, it's essentially the UK government body that is in charge of energy, but just to kind of go on to the piece around when a fire does happen. You talked about sort of good operational process. I'd love for you to talk me through that process of like from minute zero the fire starts through to sort of the end. Like what does that. Because I imagine that's all planned out at length. What does that look like for you? How does that work?
- [28:16]
B
I'll do my best. I might generalize a little bit. I'm not going to say that I'm an absolute expert on this, but I will say somewhat it does depend on the site. So we'll go through sort of a generic approach and then there will be places where we branch off. So we touched on it earlier, weeks ahead of time. If you've got data and you're processing it in an intelligent manner, you may see indicators of a cell that's in some kind of stress position and could go into thermal runaway. In an ideal world, that's where you stop everything. You take that module out of service, you replace it and you never see the issue occur. Obviously in this scenario we're not going to do that. We're going to the point where something is causing this event within the cell and that could be overuse. It could have been extreme cycling. Cycling is our term for sort of like importing and exporting energy through that battery and putting energy through the battery. It could be that you are sort of. You've overused the system and it's experienced too much mechanical stress. There could be internal conditions that enable sort of a short circuit between the electrodes. Whatever it is, that cell is going into a sort of breakdown event. The electrolyte is starting to break down under heat. So you're going to see quite a lot of heat given off that cell and it will probably appear if you're monitoring it as a very abrupt sharp temperature spike inside that cell. Now you're creating heat to potentially create further reaction and kind of cause at this point it's probably not really runaway, but you're causing more and more of a breakdown within the electrolyte. There's an interesting and kind of concerning point hit like 130 Celsius or so generalizing. But you then kind of break down the sort of separator between the two electrodes, which are called the cathode and the anode. And if you've removed that separator and you've got a potential difference across them, you can have a short circuit. That short circuit can also generate loads of heat and all sorts of problems there. So now we're starting to get really extreme temperatures within the cell. Depending on the cell type, those temperatures could be enough to produce the oxygen, say, and combustion is happening. You've got a fire going on or it could just be enough heat to continue the reaction without actually causing any combustion at all. During all this, gases are being given off. We've already mentioned oxygen. One of the big ones that people are rightly concerned about is sort of hydrogen fluoride. So this is a very toxic gas that's given off, but you've also got hydrocarbons and things like ethene and such. You can detect it. It's called electrolyte vapor. And that's one thing you can actually be looking for in practice to stop operating the cell into a runaway condition.
- [30:59]
A
And whilst this is happening, is the container sort of turning itself off electrically or the fire suppression systems, are they turning on in the container?
- [31:08]
B
So that might depend on how you set it up. So one thing that does exist out there, it's not a fail safe at all. But if you've got sensors that are detecting that electrolyte vapor, those G specifically only going to be present if this breakdown's happening, you can tie them to like an emergency stop which just shuts down operations of the container. We've talked about how this is a reaction that sort of might create its own fuel though, and also gives off heat which can trigger more of a reaction. So it might be that it kicks in at the right time to prevent that propagating and maybe it can level off a little bit, but it might be that it kicks in. And unfortunately the event's already taking hold. So stopping operations of the container is not necessarily enough. Okay, the fire suppression system might kick in based on your sort of temperature sensors. They might actually be looking for sort of just smoke like carbon monoxide. So if there's no combustion, it's not necessarily going to be kicking in. So it could kick in or it could not. It depends how you set it up. The battery management system should be recording all these temperatures and it should have already stopped operations of those batteries as well. But it might not. You know, it could be kept within sort of a module level and it could be that the other modules in the rack are still saying, you know, we're fine, and the rack battery management system is continuing to operate. It's really important to get that integration right.
- [32:28]
A
Okay. And at this point, we're sort of seconds or minutes into this process with
- [32:32]
B
seconds going in two minutes, I would say. I think it probably also does depend on the cell setup. I'm thinking like more of a typical kind of like NMC chemistry going into LFP today. There are some cells that might actually there's modules, you know, that have systems that are designed to just, you know, effectively drown the cells in water to take away that heat and stop operations. That's a real best practice approach. For what it's worth. Like the problem is the excess heat that's being generated by the reaction.
- [33:00]
A
The quicker you can take that heat away.
- [33:01]
B
If you can take enough heat away, which means doing it early on before, there's just so much it's hard to deal with. If you can take it away then and stop it propagating and then stop operating the battery, brilliant. You've dealt with it. So that could happen. The vast majority of of, you know, systems don't have that kind of technology. What's going on now though? So it's continuing into this propagation. We've got gases being given off. The module is going to be starting to swell. That's a real concern, I think, from a fire safety perspective, because if that module is going to swell and more and more and more with this explosive gas, you might not have combustion. This is where potentially there's a downside to LFP chemistry. And it's just important to be aware of like the risks on, on both sides, like so that LFP chemistry versus nmc. It's more likely that the NMC chemistry is, you know, that those systems are kind of already on this combustion going on. There's already a fire. Those explosive gases are being used as fuel to continue the fire. In the LFP scenario, that fire might not exist. So the gases build up and up and up and eventually what could happen? One thing that could happen, I'm not saying the only thing that would happen, the module swells to the extent that eventually you have this sudden dissipation of these explosive gases. And you know, you have terminals on the module. They've got a positive and a negative. There's a potential difference across them. That potential difference isn't normally enough. Well, of course it's not normally enough to ionize air and have current travel through it, but the resist the sort of conductivity of the Electrolyte vapor is sparsely different to air. And when that, that crosses through those terminals, it can be enough for a spark. And what you've got there is an environment where you've not had combustion. You've had a lot of, a lot of explosive gases created already. And that spark causes an explosion. That's something that I'm, I would say is, you know, a way worse scenario than a fire. So the explosive risk is, is a, is a huge problem. So that explosion might now happen, that container is going to explode and then we have the continuation of the combustion and all that. But let's say the fire suppression system does kick in. This is another problem that I've potentially. This may be one of my contrary positions. Like the fire suppression system. Typically we use kind of like a couple of types of fire suppression. Either you deoxygenate the room to remove the combustion, so you take away the fuel, so you stop the sort of propagation of fire, or you just sort of smother it with like a aerosol, like kind of of salt compound sort of thing. And it just like smothers all the equipment and it prevents that interface with the air so it kind of avoids oxygen like entering the equation that way. But those sort of fire suppression agents, they don't remove heat and, and heat, the creation of that heat is thermal runner is, is driving the thermal runaway. If you're not dealing with the actual cause of what's going on, all you're doing is delaying it it. And if you take away the fire and you're producing explosive gases, then you're effectively turning a fire what would be just a fire. Just a fire. And you can create an explosion risk that wouldn't otherwise necessarily have to be there.
- [36:13]
A
I think it's great that you are really, you've clearly put a lot of work into this as a business as well. You've thought about all the potential outcomes. Like the. I think the really bad answer to this question would be yeah, there's a fire, but we kind of let it burn and it's fine. That would be, I think the really bad answer. I think what's coming through in this is that you're clearly thinking about all of the potential eventualities of all of this. So that when things do happen and this is how you prepare for sort of worst case scenarios is that you are ready for them and you understand them. And I think that's super critical. Let's move on from sort of where we get to with the site in terms of like a response now. So People sort of like, as you get to the fire burning, the fire brigade is cooled. You have sort of the responders. How, how are they instructed to deal with a battery sign?
- [37:06]
B
So actually, yeah, this is where good practice comes in. You should bring the firefighters to site. You know, throughout the projects, probably like, you know, during even the construction, there's a fire. As soon as the batteries are on site, there's a risk.
- [37:16]
A
Yeah.
- [37:17]
B
So you should be bringing them in, taking them into kind of the, the picture of how you're operating the site and giving them the sort of awareness of what's going to go, what could go wrong, how it go wrong, and help them with their own sort of procedures. When they get to site, they'll be taking over at the end of the day, like the firefighters are going to do what they believe is the right thing for that site. You can inform them with getting those sort of emergency response plans made very, very visible on site. We install, they're called girder boxes, but really they're just these sort of like red boxes that go on the outside of site and they include really important information about what the site is and how to deal with it if there is a fire and sort of key risks associated with it. And that's absolutely, really helpful for the firefighters if, you know, imagine in the absence of anything else, like, at least they've got kind of a how to sort of summary of like what to do in the event. So say they get to site at this point, this is where they will likely be looking at doing boundary protection of some sort of. For a typical site. So hopefully the site is a modern site with active ventilation and deflagration. Deflagration is kind of like dealing with an explosion and how it travels. So you, you want your equipment to be designed so that if there is an explosion, it's not going to go outwards towards where people might be. Older equipment, unfortunately, isn't necessarily like that. And this is where I start to get really worried. Right. Because you're bringing people onto a site to do something, and most sites you don't want to just haphazardly say, let it burn, and that's fine. But if you do let it burn, the chance of it spreading across the whole site on most sites is very, very slim. Why bring firefighters in, especially if you haven't. They're not necessarily trained on what the site is. They don't necessarily know what the risks are. There's a really good example of this where it could have gone a lot worse than it did, which is The Liverpool Battery fire, I think it was in, in September 2020 that did explode. And the only reason firefighters weren't on site when that happened is the monitoring team. They were dismissing alarms remotely. And actually it was only sort of the local people that had called the firefighters when they heard the explosion after the thermal runaway had kicked in. And so the firefighters got to site late. But the 247 monitoring team, and there was one really should have called them. They had the tools and they had the information to call them before and if they had done those firefighters would have been on site. They actually thought at first this is where the Gerder box recommendation came from. Actually their own sort of Merseyside Fire and Rescue services produced a report on it afterwards and they recommended these sort of boxes which tells them what the site is. They got to site and it had already exploded. But they initially thought it was like a large refrigeration unit. They didn't know what it was and if they'd been trying to put it out, you know, there was a 500 pound, so, you know, massively heavy door that was blown off its hinges and had the force to travel 70ft. Now if that, if a person had been on site when that happened, you know that's a horrendous event and you know, that should have been mitigated out. But this is an older design kind of a system.
- [40:28]
A
It's kind of, it's pretty important, right? It's both to be really frank with where the designs have been historically, but it's also really important to kind of show that, that the designs have changed in terms of where we are today. Where we are today is not perfect but because we will keep on improving designs and how we respond to this. But I think it's a fascinating look into kind of the history. In the interest of time, I'm going to move us on to a final question. A contrarian view. So Dan, if you've got a contrarian view about the energy space that not many others hold.
- [41:00]
B
Well, on the fire safety topic, I'm just going to come back to the fire suppression systems themselves. I've been trying to sort of of point out that they are helpful. They don't remove the risk, they don't take away the heat necessarily. Be very careful with the design. But I wouldn't just start from the position that a fire suppression system is going to reduce the risk of a battery fire. It might actually sometimes do the opposite. And then the firewater side as well, we didn't really, actually get into it. But the large scale fire testing and such that's going on. In terms of best practice of design, we're seeing there's systems that have been tested 10cm apart from each other and best practice right now is like three meters because a guy, a standard is saying it. But 10 centimeter separation, 5 megawatt hours worth of batteries, which is a lot of energy at full state of charge. So the worst possible kind of scenario with the fire suppression system turned off if that's put into a failure event. You know, there's. We're seeing modern project, modern products like a huge amount of them now coming out of China in particular, they are not propagating.
- [42:06]
A
Okay, so even 10, you said 10 centimeters.
- [42:08]
B
As low as 10 centimeters is what I've seen so far, 15 centimeters. But it's to the point where it's actually much of a muchness as well because you put these things too close together, you can't actually operate them because if you had a component fail or something, you wouldn't be able to replace it. So it really gets to the point where it's just academic actually how close you can put them together because the operations and maintenance personnel need them further apart. Yeah, so. So you don't necessarily need firewater and actually I would just very much support do the design. Right. So it's a bit more passive.
- [42:39]
A
Yes.
- [42:40]
B
I'm not saying in every scenario you're not going to need it. And maybe it is worth having on hand because you can have transformer failures, there's more than just batteries on site. But wherever you can reduce people going to site to deal with that incident, you should, because you can't put it out.
- [42:55]
A
It sounds like far suppression. Got it. I mean that's obviously what you want. It does what it says in the tin. Right. Far suppression. But what you're saying is that actually when you get deeper into the picture, perhaps fast suppression is not exactly what you need and modern standards are kind
- [43:07]
B
of moving us through it. It's not termination, it doesn't stop entirely the fire, it suppresses it and it does it for some time, but then it doesn't stop it overall.
- [43:17]
A
Yeah, it's a fascinating insight, Dan. This has been such a great view into the battery world and how we're managing safety. It's been hugely frank and candid, which I have loved. So thank you very much for coming on transmission.
- [43:28]
B
It's been a pleasure.
- [43:28]
A
Yeah, we look forward to hearing more soon.
- [43:31]
B
Great, thank you very much.