049 - What burns inside a battery with Francesco Restuccia

Wojciech Wegrzynski: 2:14

Hello and welcome to the Fire Science Show episode 49. Great to have you here again. Over the time. I have noticed that there are some subjects in the podcast that generate. above the usual. Uh, interest in my audience. And among these topics, there's definitely battery fires. I think lithium-ion batteries, something that touches all of us. as engineers, because it's something new and it's something that everyone expects us to handle in a way. Even if in your scientific or engineering career, you don't necessarily touch the topic. Eventually you will be asked about the questions of burning electrical vehicles or is energy storage at your house safe or not. And I think we should be ready to meet. Answer these questions, hopefully. Based on science and evidence. And this is what I'm trying to get you in the podcast and also get for myself in the podcast. By inviting world-class experts. We had Roeland Bischop from RISE who discussed electric vehicles. We had DK Ezekoye who discussed the battery systems. And today and other great scientists this time, a young talent from Kings College. Dr. Francesco who also happens to be my good friend and whom I admire a lot because his scientific scrutiny and the way how he performs research is on another level. All the effort he puts into quantifying uncertain. It is understanding the first principles. In experiments he's doing. The way, how he designs. This is so good. And you will hear that. In the public is that he's really good. And he knows what he's talking about. Even though every second sentence he claims he's not an expert. He certainly is. And with Francesco, we go deep inside the cell, the battery. To understand what makes battery. Dangerous. Why batteries burn. What evens leads to a cascading failure of battery and. Little more about if self heating can be an issue in batteries. And that's very interesting. Francesco is a. Scientists who've made a career out of investigating self-heating and I find it's really great that he applied all this knowledge. He gained on self heating of various materials to batteries now. And that, that looks very, very promising. So. It's a tough episode, but I think if we want to be prepared to deal with batteries, to talk about batteries, to provide safety. The facilities that deal with batteries, we need to understand what happens inside the battery. And this episode is all about it. I hope you will enjoy it. And one more question to ask if you do enjoy, and if you'd like the podcast. Maybe there's a chance you could leave a five star review in your podcast app. That would help me a lot with my search engine optimization. The algorithms love five stars. So if possible, and. You can just pause for a second, jump into the review part of your podcast app and leave something nicer. I would be very grateful for that. And once you've done with that, please press play and enjoy the episode. Let's go. Hello everybody. I'm today here with Dr. Francesca Restuccia of Kings College in London. Hey, Fracnesco

Francesco Restuccia: 6:00

Thank you for having.

Wojciech Wegrzynski: 6:01

Welcome. Welcome, man. happy to have you here. And another take on batteries today. the previous episodes about batteries were really well received. So I assume the audience in the podcast is very interested in this technology and who's not in the world of fire where, being scared about the dangers of lithium-ion batteries. Every. Every day and last week had a, Mike Spearpoint on the show and he also brought, uh, numerous, things to consider, not to worry about what to consider when doing a car park, uh, with this, electric vehicles. Actually, my Mike said that the first vehicles were electric that, that, that killed me. That's first car parks were meant for electric cars now to think about that. Well, you have run the full circle, but Francesco. I didn't ask you to talk about electric cars. We can discuss, the vehicles, on some other day. I want some hardcore chemistry, man. I want to finally know what the hell sits in the battery that, um, makes it, so scary to everyone around. And I've learned that I should not, poke one with the nail to see what's inside.

Francesco Restuccia: 7:13

Accurate?

Wojciech Wegrzynski: 7:14

Yeah.

Francesco Restuccia: 7:15

The active material inside might not like it.

Wojciech Wegrzynski: 7:18

Yeah. So yeah, please tell me what kinds of devils technology lies inside the lithium-ion, uh, cell and, uh, yeah, just so let's start with the basic, like, how is god damn battery built? I know it's from outside,

Francesco Restuccia: 7:32

so I'm not a chemist, so any chemist might be horrified at my explanation cause I will simplify it. Um, but effectively. You have lithium-ion batteries. Um, and they're named after the active material that's inside. So for example, when you read lithium cobalt oxide, and you read L I C O O two or LCO, you know, that just means that the active material is lithium cobalt. so the material. The cathode is made of is lithium cobalt or the other popular one is our Alamo. So lithium manganese oxide. Um, and again, it just means that the active material, so the Capto is made up of lithium manganese, in layers. So they have an oxide layer. and the third common one is NMC. So the three that you usually read most about since you mentioned vehicles is NMC, which is, basically a cathode combination of nickel, manganese, cobalt. and so they're usually used for power cells. So that's why they're used a lot in electric vehicles, because for example, when you read the 18 6, 5, 0, so. Research papers these days I read, that tried to do a sample cell start with the 18 65 0 cell. and that's just a cell that can deliver, I don't know, about two to 3000 mili ampere hours. So four to five, um,

Wojciech Wegrzynski: 8:47

I have a flashlight with just one cell and it, and it lights for 12 hours. So for me, that's 12 hours of continuous light capacity in, such a cell.

Francesco Restuccia: 8:58

Uh, yeah. And then they can charge, so they have this specific energy. Right. it could be for example, NMC is. between 150 to 200, what our per kilo. So quite high energy dense, right? It's, it's, it's very, very big. Um, or then lithium manganese oxide one, which is less powerful one, right? There's a hundred still, right? It's a hundred watt hours per kilogram. So these. Quite a lot of energy that can be stored into them.

Wojciech Wegrzynski: 9:26

Okay, but that that's the cathode what's what's else in there.

Francesco Restuccia: 9:29

So you have the anode and the electrolyte, and in fact that's where the fire comes in. so these batteries are made up of and then electrolyte substance inside and the electrolyte substance, varies. and in fact, it's usually quite hard to find out what the electrolyte is, when writing papers, uh, again, I am not a chemist. So when I started looking into this. I was a bit confused at what the electrolyte was. So I asked the chemist, I asked the chemical engineer, what is the electrolyte? And they said, it's full of additives. It's really complicated. It varies battery from battery. Um, and, but you always know what the cathode is because that's what the battery is named after. So, and that's what really is what provides the that's the energy density

Wojciech Wegrzynski: 10:09

So, so let's, let's agree on calling, uh, electrolyte, the battery juice

Francesco Restuccia: 10:13

Exactly. Yeah. So you have the juice, you have the anode and you have the cathode, In, uh, for a fire problems, you know, why are we interested in these materials? So these are the materials within a battery when you cycle a battery, so you connect it up, you make it active, then electrodes flow, right? So you have your electrical cycle. and part of the energy that's released is your electrical energy. And part of it is what's known as Joule heating, right? So it was just the electrical, the cell impedance, your cell has some resistance and it releases some heat. Enjoy

Wojciech Wegrzynski: 10:41

One one second, once the back, uh, but, the way how it is built, I was actually quite confused when I watched the video of batteries being produced. But because when I was a young kid, I had a passion for chemistry and blowing stuff up and actually setting things on fire, which has not changed significantly.

Francesco Restuccia: 10:59

I was about to say now you'd basically just do that for a

Wojciech Wegrzynski: 11:02

Yeah, I made it into living suits, actually. That's quite cool. Um, and I don't go to jail for that. That that's another benefit. , but, uh, I I've played with batteries, uh, by playing, I mean, I've disassembled them and they, there was like a graphite core. Uh, there was a metal encasing and there was a black, all the smelling substance in it. and, I think. When I then, so 30 years fast-forward, uh, how batteries made that they were like, uh, sheets of, materials spins into a cylinder. So it's like a continuous sandwich that's just, rolled,

Francesco Restuccia: 11:39

Yeah. So, so there also, so for the. Cylindrical batteries. You're effectively you start with the powder, right? So you start with, some material. Um, you Dan make really, really, really thin sheets. and then you rolled them up into jelly rolls. So you rolled them up into really, really all these really thin sheets into a drill. Then you compress it into account right. Of metal, so that it's protected. and then you. Basically put a lid on the top, which has the plus bit so that you can connect the wire to it. and on the bottom you put it on node, right? So you put the cap on the node,

Wojciech Wegrzynski: 12:16

the only when I saw this, how it is made, and only when I finally understood what is inside, because I, I had this image in my head from the chatter that it's like a road of something in the middle, a substance, and, the encasing that surrounds it only when they realize it's, layers of cathodes and anodes. all like jelly rolls together in one continuous roll. I've realized that this. Why these, um, mechanical failure of batteries, like why, why crushing it, or why nailing it immediately creates such an effect because essentially you're suddenly connecting the things that should be separated. you, you create the short circuit between the cathodes and anodes. And because they are so close together because you probably want to have the biggest surface reacting all the time. So there's like, those ones are hundreds of these layers together. We Pierce to them completely and suddenly you've created hundreds of short circuits

. Francesco Restuccia: 13:15

You, you just pointed out the item that I may stop, honestly, in the rolling, you put a really, really thin membrane, which is called the separator that separates the two. Right? So it almost looks like a sheet of paper. Right. And then you fill it with the liquid electrolyte or the juice as we called it. And then you put the cap.

Wojciech Wegrzynski: 13:29

Yeah. So then when these short circuits, it's just all the energy that's stored in the battery and we try to put as much energy in it because of obvious reasons is it's just released in a split of second, actually. and th that that's really. Interesting, but I didn't want to talk that much about mechanical failure of batteries, because, uh, you also happen to be, quite a knowledgeable person about, self-feeding processes. And, this is something I look forward. So you you've started mentioning, um, how, why the battery produces heat. Maybe you can, uh, finish that thought.

Francesco Restuccia: 14:06

Can I add to, can I add one thing just out of curiosity? Cause you, you actually sparked my curiosity on, you know, that the battery manufacturing, from what I hear from the manufacturing, people that matter the battery manufacturing process is done a certain way because it was adopted from existing industries. Right? So when we used to do Kodak films, right, we already had the industry for that, which is why the. Making, from what I understand from the material scientists, the process was just on the adoption of the one from the film industry that Kodak it. So, and that's why it was done that way, because from an engineering point of view, I always thought there must be a more optimal way to do it, but they already had the industry. So, so they just adopted an existing industry. But again, I'm not a material scientist. So this is from my manufacturing colleagues. Cause I asked why I like, from an engineering perspective, there must be a more efficient way to make this. Um, and that's also why now they're always looking at future batteries and you know, there are. Materials that they're trying to think about, but yeah.

Wojciech Wegrzynski: 14:57

But there were also prismatic cells, poach cells. And I guess these are done in a slightly different way. I, I would guess. Okay. So you started mentioning why the battery heats up. it it's meant to produce electricity, but it also, because it has a resistance, it's not a superconductor. It must produce a heat at the same time. So first, what are the rates like how much a battery heats and how much it gives back electricity?

Francesco Restuccia: 15:23

Oh, so the vast majority is electricity. So the extreme cases, the one you pointed out earlier, which is when you start having short circuits, right? Cause if you start having short circuit, then you increase that resistance of your battery massively. Or if your battery starts degrading, then that impedance internal impedance increases the higher, the internal impedance, the more Joule heating you produce. So, you know, a fresh battery, new. With no degradation with no, short-circuits probably, I would guess 99% electricity, 1% teach, right. Or maybe 2%, very, very small amount of it is heat. Um, but the more your battery degrades, the more internal impedance you create. And so the higher the heat. And so actually it's not a constant, it depends on your batteries, aging, how much it produces. And it depends. the most extreme cases when you start creating a short circuit, , because then that short circuit creates a peak of impedance, right? And so then a large portion of your energy stored becomes released in heat rather than electricity. and that's a time issue as well because. we do heat transfer for fire, right? So transfer is a time dependent process. And so if your heat is released very slowly, over a long period of time, you go into my other area of expertise, self heating. If your heat is released very, very quickly, then you have a local peak of temperature increase. And then the rates chemistry rates are exponential. And so they go up much, much faster.

Wojciech Wegrzynski: 16:46

But, but even if it's 1%, , and let's assume you're, , car driving takes, I don't know, 30 kilowatts, 50 kilowatts, a hundred kilowatts. If, if, uh, if you enjoy your life as good cars, but, it takes like, let's say 50 kilowatts, a kilo per hour. And even if it's just a 1%, that's 500 Watts released into a metal tank.

Francesco Restuccia: 17:08

And even more important than the charging, right? So, you know, we always talk about fast charging. If you use all of your battery in two hours driving, but then you want to charge it in 10 minutes, you're doing the exact same electrical cycle backwards. Right. So instead of. The energy going one way, it's going the other, but you're still producing heat. The heat generation, the Joule heating is still based on that impedance. But if you want to, that's the biggest problem they have with fast charging, right? Is if you want to charge your electric vehicle in five minutes, which is impossible at the moment. But imagine you did the amount of heat that you will be producing is massive because. Kilowatts that you're putting in are huge. and so the charging rate and the discharging rate, so you're using up of your vehicle that electricity higher discharge rate means higher C number is what they call it in the battery world would mean more heat produced per unit time.

Wojciech Wegrzynski: 18:00

yeah, I was , always surprised when I saw this information about. Quick chargers and people trying to connect, like I don't half megawatt charger into a truck and I'm like, I run the fire laboratory. We do run a, high capacity fans in high temperatures. When I have to test a hundred kilowatt fan, it adds so much to my logistics because it's like huge, huge cords. It's dangerous. It's a lot of electricity, man. Then you just go. Yeah. And I will just block this half megawatts socket into my current, go for a coffee. , that scares the hell out of me. But I guess the, I hope they've worked, it worked it over, but again, if it's just, if it's just 1% that goes into heat, that's a considerable amount of heat

Francesco Restuccia: 18:50

And I've had this issue when designing my lab. So when I was designing my lab, I wanted to have scalability for my battery testing. So I have a nice big battery tester, but the first question I got from the health and safety team, as you know, the crazy academic who wants to put something in is how much heat will you be generating? Because this, as you've asked us to put in. A current limiter of 60 amps, 60 amps is very high. So how much heat are you going to be producing? So I had to do some back of the envelope calculations, not with 1% heat, but you know, I, I went with 10%, , just to, maximize, but yeah, it is, it is a huge concern. Um, from a research side in the lab, I can imagine from a design side when you're building the cars or building the charging stations just as big, right. Because you have more people using it, you have more chance of error. You have more risk of a failure. , Wojciech Wegrzynski: Okay. So before we go into the battery, finally igniting, we've discussed, the, the heating processes due to short circuits, uh, mechanical failure. We've briefly discussed the, charging and, and just energy consumption. And let's go into self heating, I guess if I can take it a step back in between the two, I can explain the stages maybe of the heating that might help. So in a battery, , you have reactions happening at different temperatures and you're going to get to self eating, um, with. Happy to talk about, but I guess maybe it's when your battery, it doesn't matter how your battery is heating up, right. Self-feeding or not selfie thing. , when your battery reaches a certain level of temperature and it cannot dissipate the heat, usually that's about a hundred degrees. Uh, when you get to a hundred degrees, your solid electrolyte interface. Um, so that's the thin layer, that is on your cathode. It starts to decompose. Right. Um, and so if that starts to decompose, that causes exothermic reactions and that adds its own heat. So, you know, we say it's an active material. and as an active material, as I, when I think of chemistry, the reaction rates from a combustion side always happen. Exothermic reactions always happen at different temperature ranges. Right. And so in the battery, usually you first have. Heat being created from the solid electrolyte interface decomposing, then from the electrolyte electrode decomposition at a higher temperature, and then from the electrolyte decomposition at the higher even temperature. And so you have the inner materials, those active materials they were asking about all start reacting at different temperatures.

Wojciech Wegrzynski: 21:16

Okay. So can we just remove the one that reacts to the lowest temperature and

Francesco Restuccia: 21:19

Bye. You need to, you, you need you that that's what causes the battery to work.

Wojciech Wegrzynski: 21:24

Can we make a magic material that will go off at 200? Not another 100.

Francesco Restuccia: 21:28

the SEI layer, so to say the solid electrolyte, the composition layer is just like a passivating layer that forms on the electrodes. And so as your battery operates, you're forming this layer. And so your. it's kind of covering the intercalated lithium, lithium in the negative electrode. Um, and so when it starts decomposing, then the intercalated lithium is exposed. Right. And so it's that exposed to the electrolyte. And so it causes even more reactions.

Wojciech Wegrzynski: 22:00

So basically you enter a path of no return and, this additional heat generated by this exothermic reaction will hit the thing more up, up to a next point where another thing will start contributing. And another thing, another thing, and you end up with, this thermal runaway, I guess there's this.

Francesco Restuccia: 22:18

Yes. And actually the last material ironically, which is maybe slightly cutter, we name all the batteries from the active materials. Right. So I said, I never, but that's usually the last material that Dan starts. Cause that is the one at the most high temperature. So that's actually the last bit that starts creating heat. And do you usually, um, because normally at the higher, higher temperatures than the positive electrode can start becoming unstable and decomposed.

Wojciech Wegrzynski: 22:43

let's say, uh, the, the, all the heat released in a fire of a single cell is a hundred percent. Do we know how much each of these steps contributes to this? Uh,

Francesco Restuccia: 22:53

Oh uh, complicated. So it depends on the state of charge of the battery. Um, so it depends also how much energy is in the battery. Um, and it depends on what electrolyte is being used. So there are some papers that have looked at the different, contributions. And from a self-heating side, we have looked at the contribution in temperature rates of the different rates have increased, but not the individual materials. So we looked at the temperature

Wojciech Wegrzynski: 23:19

Okay. So, so from this temperature to this is this

Francesco Restuccia: 23:22

yeah. Yeah. So at the macro level, rather than at the chemistry level, at the chemistry level, Quite a challenge. So because you start having, so the materials that react to start forming other materials. And in fact, if yourself, then over pressures, for example, you were talking about prismatic cells. So if we earlier, so if your personal ethics cell bursts, the safety event, and then some of this gases that are being produced inside, start reacting with the oxygen outside, then that causes even more reactions. And the gases that are being produced are. Very complicated. So there are a lot of carbohydrates, but there's also hydrochloric. Um, Gus is a hydrofluoric. Um, and so it can get, yeah, um, very complicated, very quickly.

Wojciech Wegrzynski: 24:04

well, I guess this is fundamental to understand this, Macro fires of batteries because they start like the, the ones that are from the individual cells, because obviously if the, if the car park is on fire and your battery catches fire from that heat, I guess the cell chemistry hoods not mattered that much.

Francesco Restuccia: 24:28

It just acts as a heat

Wojciech Wegrzynski: 24:29

it's as

Francesco Restuccia: 24:29

once you have a heat source. yes. It's the same with our pool fires. Right? If you have a pool fire, once your pool fire has a size, it's the heat release rate from that size, doesn't matter if you were burning, , C H C7 or C eight, right? , it's the heat that's being released from those flames. Uh, so he really straight, but you asked me earlier about modifying. Items inside. You're not far off. I actually think that is one of the solutions is, , if you can change the additives and the electrolytes to add things that maybe passively act for the duration of your battery life, but then if your temperature is above a certain temperature, they act as a suppressant or the heat remover or whatever it is that could be a solution. Um, or, you know, if you can have maybe a ceramic coated separator in between right then the. It's harder to, create short circuits from, right. Um, okay. Yeah, that's a cost. Yes, exactly. It, a cost that's a weight and weight is a big problem. Uh, but yeah, that could be what way. Um, so it's, it can start from the chemistry modification so you can change the catheter. You can change the, I know you can change the attitudes and the electrolytes, or as we are not chemist, neither myself nor you. we, we can look at the bigger size, right? So we can look at, a bigger size. Well, what can we do with a battery management system or what can we do with, separation, right? Uh, separation between batteries within a pack. Uh, what can we do with detection? What can we do with different suppression

Wojciech Wegrzynski: 25:59

I think it's a brilliant starting point. And despite the, like, I'm, I'm on the edge of understanding, I understand literally up to 60% of what you said. So I think, and I think I'm quite successful, but I'm going to take away. The take away is, , I think there's this thought that, uh, the things happen at certain temperatures and, uh, essentially you, you, you reach a track. Point which initiates a set of events that lead to failure. And I think this, uh, this is a great takeaway because, it doesn't happen by, by pure magic or unicorn farts. It's, chemistry that leads to that. And you've already mentioned many things that influence this, thing. The energy capacity of the battery, the sorry, the charging state of the battery, the processes is a charging is a discharging, the, chemistry of the electrolyte, because I think this is, this is where the confidential magic is happening. if everyone is using the same cathode materials, I guess they're playing with electrolytes to differentiate and pattern some things to earn more coin on, on, and finally, maybe not exactly, but you have a rough idea when this thing becomes dangerous, because if you don't enter this cascading, level of discuss this cascade of events leading to a major failure, you don't get a major failure.

Francesco Restuccia: 27:23

Yes. And we've talked to very generally, right? So it can also be, so I've talked about the system in charter discharge or in use, but also passive, right. That's where self-heating comes in. When you mentioned self-heating is, you know, what's a active circuit. So I said into charging or discharging rates, but batteries still react. You know, your battery is always reacting and when it's not connected, right. It's still, it's still producing heat, even in open circuit conditions. Um,

Wojciech Wegrzynski: 27:46

okay. I, there was a, , Guillermo broad on Twitter, , some instruction for a laptop, which mentioned fire in so many places. And I've responded showing a picture of my old laptop that two years ago, my battery has suddenly swollen. Three times its size. And I recall it happened literally overnight. I didn't use the laptop. So, when the, the device is idle, I guess there are still things, , , happening in

Francesco Restuccia: 28:11

Yeah, because mass diffusion, right. You know, you, you're always going to have some electrodes travel through. you're always going to have diffusion through your, from one side to the other. So your battery will, oh. Even when it's not connected, your internal resistance is in itself a connection. Right. so you're going to have passive resistance, not just active

Wojciech Wegrzynski: 28:33

Okay, but will it happen when the battery's discharged? Like when it's a charged level?

Francesco Restuccia: 28:38

When charged double zero, that's the dust. You don't get to that condition cause that's really bad. So when everything is on one end, um, your. You're basically I've done have exposed the material. So remember I said that you're creating this SEI layer, if you've moved everything to one end, so the battery's completely at zero you're, you will have exposed that and it's not designed to be exposed. And so, in fact, I think it starts reacting even further. Uh,

Wojciech Wegrzynski: 29:05

so, so you're telling me my laptop is lying to me when it says the charge zero.

Francesco Restuccia: 29:10

Absolutely. Yeah, there is no such, yeah, you, you will never allow a device to go down to

Wojciech Wegrzynski: 29:16

Okay. But what does that mean? That's cool. I have a laptop with one battery, but, on the factory where they produce laptop batteries, they probably have a 1 million of them stuck together. Like what, what does this mean for storage of batteries? Like when you produce them in, in large volume in quantities, You, you, you have a tons of devices that, and you just said that the reactions are ongoing, whether you like it or not. So, so they are produced and often they are shipped as charged, batteries.

Francesco Restuccia: 29:49

and that's why they put a limit. so you have an ideal charge rate to just to think 30% from the standards now. but also colder temperatures. So that's the other thing you do is to reduce this self discharged, right? You lower the temperature. So what do I do when I store batteries in my lab is they have to be in a fridge below 10 degrees, , ideally 10 degrees. and then that is the optimal. range for them not to reduce the self discharge. So self discharge is also temperature dependent. Um,

Wojciech Wegrzynski: 30:18

Okay about the stupid question. If I, myself only in winter works, uh, shorter. And if I expose it to minus 20, it's going to die very soon. So it loses this capability of react, which also from your perspective makes it safer.

Francesco Restuccia: 30:32

Well, uh,

Wojciech Wegrzynski: 30:33

I'm sorry. I'm

Francesco Restuccia: 30:34

arise when you go to too low other problems where I So I'm not a material scientist or a chemist, but you have a sweet spot of temperatures for which the materials are designed. If you go too low in temperature, I believe you might start degrading the materials in other ways. but I'm not an expert, uh, on this. so there is a guidance on the range. You want to keep it in, for the. Battery to be healthy. and I don't know what that lower limit is.

Wojciech Wegrzynski: 31:01

matter, but but it exists that that's what matters the guidance. So, , based on experience, tests, knowledge, uh, the scientists have figured out the optimal charge level where the let's say the risk would be minimum.

Francesco Restuccia: 31:16

Yeah, charge and temperature. charge and temperature. Right? So I think charge, they S they shipped them at 30%, , state of charge and temperature. I don't know, but there is an optimal temperature, which is usually your battery. The optimal temperature for a battery normally is between 10 and 25 degrees above 25 degrees. It's not good for the battery below 10 degrees might start having other issues. But for example, when I store them in the lab, it's between six and 10

Wojciech Wegrzynski: 31:40

Okay. So, batteries produce heat. There's an optimal, uh, temperature at which the let's call it risk being lowest. Uh, maybe it's, , incorrect, but, more understandable by the fire audience. So, uh, do, are we managing this heat? that's something that interests me. So if I have a bunch of batteries in my car, Which I don't because I don't have electric vehicle yet, but if in my hypothetical car, if I had it, would the battery management system be monitoring the heat as well? Because I know it is monitoring electricity.

Francesco Restuccia: 32:14

Good question. So your battery and electric vehicle is made up of many, many cells. Thousands of cells, you for a question of cost and space, you don't monitor every single cell temperature voltage you can do by line because you can have a row. And if it's in series, you have the voltage there,

Wojciech Wegrzynski: 32:34

And the voltage you probably would like to monitor to know the charge level anyway.

Francesco Restuccia: 32:37

Correct. And you'd usually do that by row. So every six or 12 or whatever that row is. , but temperature you cannot do for each one. Um, just because it adds massive costs. And so temperature sometimes is monitored, but it would be monitored at probably a pack level, not at a cell level, but everything I've been telling you now is about a single cell. So imagine you have a cell that starts deteriorating or cell that. Goes into some kind of initial failure conditions. Usually in an electric vehicle, you're not monitoring that cells temperature. and some people say that, okay, by you're monitoring the voltage. So you will see a voltage drop, but sometimes, and we shown this, in literature, including a paper I am on, we've shown that sometimes the failure happens before the voltage drops. So in fact, I have a case in one of the papers we've written that the battery was basically on fire and the voltage was. Constant. And then it dropped 15 seconds

Wojciech Wegrzynski: 33:30

a very, very brave battery. I assume it was Ukrainian fighting till the end. Uh, okay. and any other things you can monitor, like maybe you can measure in impendance of the cells.

Francesco Restuccia: 33:41

Yeah. So impedance is another thing they do, and they measure impedance, especially when they're trying to figure out how your battery has degraded. So imagine you want to have a second life of a battery. So the elephant in the room for batteries, my opinion is recycling, right? Just like we had for plastics is when your car or whatever system you're using reaches its end of life. You either recycled the battery and that has a lot of costs involved, or you try to redeploy it in the second life in something that requires maybe. high spec, right? Because your battery has degraded. So how do you then figure out how much your battery has degraded? So normally what they do is they then measure the impedance of the battery. So you can use impedance systems that measure what the new like internal impedance is, so that you know how much your battery has degraded. What it doesn't tell you is what conditions it was exposed to. So it doesn't tell you if it was kept in a warehouse at 40 degrees, and maybe so it doesn't tell you what has degraded, um, but yeah, battery recycling entire topic, but for me, that's the apart from the fire safety, which I find really interesting on a global scale there is what's the plan. And of life for the batteries. We have it for nuclear, right? If you have a nuclear power plant, you want to design a nuclear power plant, you are not allowed until you have presented the plan of what happens to all the spent fuel. What happens to the land, what happens to everything that goes with that, right? Um, even if you want to put a new oil rig out into the sea, you have to have an end of life plan for that platform and what happens. But with batteries, there's not much.

Wojciech Wegrzynski: 35:13

people who are listening to the podcast are, are open-minded fire engineers who wants to understand what the future brings. And I think what you mentioned now, if we think we have an issue with electric vehicles or lithium-ion batteries today, which. Thing is that the big issue, but let's say the, the public, would say it's a big issue today. Like when you bring to the question, the, uh, out of life batteries or repurpose batteries or 30 year, all the batteries. That's going to be an interesting issue we are going to have, and we better start looking for solutions, how to manage this. Should we incrementally increase the fire safety regulations for the batteries? If even though we don't have really good regulations at the moment, but it seems that we're struggling with something that that's not even the problem yet.

Francesco Restuccia: 36:11

Yes. And it's, I think it's going to become a problem in a couple of years, right? Once, batteries are scaling up, it's getting up and we're using more and more higher high energy density batteries once people. And there are companies that have started doing second life of batteries, you know, what are the regulations around that from a safety side, if I'm a fire. Scientists. And I care about when the system will fail and potentially cause a fire. And I have no clue of what's changed inside. I just have this battery pack, um, and I've been told the impedance has changed and that's it. I have no clue what I've exposed it to you. the history. So the history, the history of it, what's cycling, it's been exposed to what charging has been exposed to, you know, that degrades the battery. So it's from a safety side, I find it concerning,

Wojciech Wegrzynski: 36:58

And how the graded battery is more dangerous because of this increasing beadings this increased,

Francesco Restuccia: 37:04

yeah, potential short circuits. Yeah. Potential short circuits. Right. So for example, self-heating right. So

Wojciech Wegrzynski: 37:11

Yeah, go on. Go,

Francesco Restuccia: 37:12

Step into self-feeding ignition. uh, so Xuanze He and Zhenwen , Hu and, uh, Guillermo Rein. , and I have worked on self-heating at Imperial. Um, and, , a lot of what I'm going to say is the Zhenwen the Xuanze his work and their PhDs. , and you know, they studied a lot on what happens when you have XL. These X's are chemical reactions are always happening, right. So they increase your temperature. Okay. I care about self-ehating ignition. I've done self-heating ignition for non batteries, , for, you know, other reactive media in my PhD. And the problem is very easily post to me is, well, the behavior is the same. So what happens if in an open circuit cell, instead of abusing it, instead of adding an internal short circuit, I just considered a battery itself and the heating that it's caused, if the, if the heat cannot be dissipated, , And the heat starts building up. Then I might reach a critical temperature for which self-heating ignition can happen and thermal runaway can happen. Right? So I can reach a condition where these exothermic chemical reactions are causing enough heat to be produced, that the environment around it is not providing enough cooling. And so eventually the temperature will go up and all be. If that's a single cell, you do the calculations and Xuanze done experiment to, and I, you know, you get a ridiculous high temperature that you will not reach in our warehouse, but if you increase the size. So with self-heating theory, if you go back to seminar or if you go back to, the work in the eighties, or even some of my work in my PhD, you, you see that for most reactive media. If you increase the size, the temperature threshold for which you will reach a critical. temperature that will eventually reach the thermal runaway will decrease. And so by how much it decreases, it changes. And so what they did in their PhD was they looked at this phenomenon for battery. So, they said, okay, A stack of batteries and those batteries are all heating up. The center core battery might be heating up more, right. Cause it's not releasing that heat to the environment. So you're leasing it to batteries next to it, which are also causing heat. Right. and eventually you might reach the condition where you have temperatures that cause ignition temperatures that you can reach and warehouses,

Wojciech Wegrzynski: 39:25

the state of batteries is idle. They they're not being

Francesco Restuccia: 39:28

Yeah, they're open circuits. So imagine you have a warehouse, right? So, uh, Zhenwen did this calculation, , in his PhD, which was that he took the size of a warehouse and he said, okay, imagine I'm packaging batteries, , in this warehouse and it's old batteries at what size. Will I need of number of batteries for them to ignite that room temperature in the warehouse. And he found the number to be very, very high, you know, racks and racks of them. But that's, if there is no installation, there is no packaging. And so he said, okay, now what happens if I start adding packaging? Right. So packaging causes insulation, in fact, so it lets you retain the heat and he saw that that temperature started dropping massively. Um, and one of his papers found effectively. Depending on what insulation you have between the batteries. It actually massively lowers the temperature for which self-heating ignition happens. Um, and Xuanze then did this experimentally. So he measured self-feeding ignition for a stack of batteries. Um, and he saw that as you increase the stock size, the temperature for which they reached ignition reduced.

Wojciech Wegrzynski: 40:32

and then this was , passive self-feeding that you would encounter in, when you store the batteries, uh, where there, like you said, that there was a certain charging point at which they should be stopped. So did, did you guys try, like different charging points

Francesco Restuccia: 40:49

Yes. Yeah, absolutely. Very good question. So, so that was one of our first questions. Okay. We did everything at a single star charging point initially. So Xuanze his first experiments, that he did, back in China, he had state of charge of 30%. So he had the low one that you use. Transport by then he said, okay, but what happens if I increase it? Right. So what happens if I go to higher states of charge? And in fact, he found that you have more electrical energy stored inside, but he also found that your temperature threshold for, self-heating ignition, um, massively reduces. And so the more state of charge, the lower temperature, it will get lower critical. The temperature for which ignition is triggered,

Wojciech Wegrzynski: 41:31

And by by ignition, you mean the beginning of this cascading processes or something

Francesco Restuccia: 41:35

So the beginning, so basically where the process becomes unstoppable. Right. and at some point, yeah. What, what is that trigger temperature? So if I go back to something that we've looked at in fire from the 1980s is coal, right? So this was done a lot for coal fires is, you know, what is that? Temperature for storage. So Cole has the same problem, right? Call it a reactive material. It oxidizes with oxygen. So if there's oxygen in the room, your coal is going to be producing its own heat. It doesn't matter that you're not burning it. It's gotta be producing heat. and so the. Mine's that use coal or power plants that use coal always had that policy of first in, first out, right? The coal that gets into storage first is the first one that's going to be used. You don't leave it in storage for a long time. You don't start using the one that comes later. Uh, not because of, you know, aging of Cola and, you know, like whiskey the age, the better it's the opposite. It's, you know, the more aging it has, the more heat is had a chance to produce. And so for coal, they found obviously that the bigger, the storage. The lower the temperature for which you would have that trigger. And so that's what we were looking at from a battery site is can we find a similar behavior, to what we have seen in many, many fuels, over the history and fire science and which maybe the battery community was not really used to thinking about.

Wojciech Wegrzynski: 42:55

And okay. And when you did this experiments, I know there is a fancy technique that you have, ruins for everyone else. And that that's Accelerating Rate Callorimetry.

Francesco Restuccia: 43:04

yeah, so, so that was a, the last chapter of Xuanze's PhD thesis with Guillermo Rein at Imperial. Um, You know, the way we get kinetics that we use for a lot of our stuff is using accelerating rate calorimetry. We often investigate thermal runaway promise there. So those, you know, probably there's the bring you to runaway by assuming adiabatic conditions. So you assume that your conditions exactly. And you put into this machine, that's called accelerating rate calorimetry and out pop the kinetics effectively. That's a really simple way of putting it, but what Xuanze and Guillermo and I, and a couple of others, thoughts as well. We do heat transfer. Adiabatic seem surprising when you have a system that has its own temperature generation, right. And it's not microscopic, right. A battery is not a millimeter

Wojciech Wegrzynski: 43:55

Let us assume the batteries appointing spaced.

Francesco Restuccia: 43:58

and so exactly. So, so Xuanze did measurements, um, to figure out if you do not ignore the heat transfer. So basically what you're assuming is that you're ignoring internal heat transfer, right? When you use accelerating rate calorimetry and he said, well, what happens? Can we ignore, internal heat transfer? So within the cell and external heat transfer at the cell surface, because in a accelerator colorimetry you have a constant temperature wall. and he found that actually. For, I think he used, I'm on the paper. So I'd be, I need to be careful here with not getting things wrong. Um, so you use lithium cobalt batteries, lithium cobalt oxide. So for his lithium cobalt oxide, prismatic cell, he found that actually for. Temperatures are much lower than thermal runaway. So, you know, very low temperature then the temperature variation was between one, 1.5 degrees. so between zero and one and a half degrees, so relatively small, right? The, the heat, uh, that you are missing the negligence that you're getting from the heat transfer effects by that he found that when thermal runaway does occur. So when you reach those higher temperatures than the. Temperature that you've missed the change that you've missed by ignoring the heat transfer effect can go from 10 degrees to 130 degrees. And so he found that actually, if you ignore heat transfer, external heat transfer effects, then you are underestimating the heat of reaction of the cell by about 12% for the cell that he was using. That's a lot, you know, if the heater reaction. Underestimated by 12% means that your kinetics, the estimate that kinetics will have a bigger error and that error will grow as the number of cells grows. Right. So I said that that temperature for feeding will decrease as you increase the size. So imagine you have more and more cells, then that error will compound.

Wojciech Wegrzynski: 45:47

Yeah. And did they go back to the warehouse calculations knowing that.

Francesco Restuccia: 45:52

so, so the paper, the paper came out. Last week, in fact. So it's a very new paper from his thesis. and, basically our conclusion was that, um, if you ignore heat transfer effects, then the thermal runaway parameters that you're quantifying using this technique will have. Errors that can propagate in your battery safety design. And so if you go back to the warehouse, you know, make sure that if you're using that data for your calculation, you are aware of the errors so that you can use it in your prediction, because I mean, you do way more CFD than I do, which I can, you know, that's, it's fine. As long as you know what the error is, you can account for it and you can do error propagation. But if you don't know what it is, and you don't have an error propagation, then you're going to have a lot of issues with

Wojciech Wegrzynski: 46:36

Yeah, I had a very nice talk, about, I both life the engineer with Mike Spearpoint and there was this consistent level of crudeness being mentioned that, when you do, um, very fine simulations and you have a. Under pinning assumption that has a major error, no matter how fancy your technique was, this, uh, initial error will propagate to whatever you do and will increase the uncertainty of whatever you've done by a lot. And if here. The first thing, you would assume that the reaction rate is this, and you've missed it by 12%. And that's the underpinning point of every single analysis you do for, for the next of your research. that's a huge miss and not, uh, not the biggest one in the industry, but

Francesco Restuccia: 47:27

But I learned this with Guillermo. So I did my PhD with Guillermo Rein and when I started doing so I started looking at self-heating from modeling side and I realized that I couldn't find the data I wanted and literature, because I wanted the errors. I wanted to know what the uncertainty was and I couldn't find the uncertainty. I said, well, if I model need, I don't know. I don't mind how big the uncertainty is. I need to know what it is. So then I ended up doing experiments to get set data, and then. Put the errors and Guillermo and I were having discussions in my first paper, in my PhD because my aerobars are quite large. And so then when I upscaled to real life scenarios, I was having these really big error bars. And in fact, if you look even now with Xuanze, Zhenwen in our battery papers, when we upscale our error bars are quite big. I went back to Guillermo and we were looking at my figures and I was like, I'm really worried and give him, I said, but why, you know, if that's the uncertainty that we're getting, that's the important thing. And that's, when you then use the data, you need to know what the uncertainty was. So we sent the paper for review. I got demolished by one of the reviewers. I got a little bit demoralized. We resubmitted the paper. To combustion and flame in fact, got great reviews back, no problem with the errors. Uh, in fact, he knows very good that you specified all the data. And so for me, it was actually a beginning. It was a bit worrying, but in fact, I learned, and I've kept this throughout my career. And now that I'm an academic and I have my own PhD students, I am very careful always with, um, or even with my postdoc on error. Uncertainty analysis is, you know, I don't mind if there is big, but it needs to be accurate so that if you then use it for modeling. You have as accurate as you can get data.

Wojciech Wegrzynski: 48:56

It's something that my, , my mentor professor Czernecki said, do you want to be, uh, roughly correct or precisely wrong? And, uh,

Francesco Restuccia: 49:05

I like

Wojciech Wegrzynski: 49:06

roughly correct is sometimes, , that's all you can get and, to understand the limitations of, of your errors means you can. I mean, The data does not need to be super high fidelity to be useful models do not have to replicate reality in a hundred percent to be useful. These are tools for engineers. And I think here another lesson that comes for the general fire safety engineering audience, who may not be that big fans of cathodes and anodes. But, here we are talking about quiet Difficult engineering topic where so many fields of science interlap. And then again, if this battery's burned down, it's us fire engineers was going to be blamed it. We didn't solve this issue and metabolically going through it, understanding the limitations, understanding the impact of the small factors gives you, uh, in the end, the holistic view on. On the system, in this case, on the battery cell that, gives you space to work with. And it's not that we're going to magically solve the problems of, batteries if one exists, exists, but, , we need to learn how to live with them and how to manage them in a way that is acceptable. And I think this is an achievable goal. What do you think.

Francesco Restuccia: 50:28

Yeah, I agree. And in fact, you know, I, if I go back even to Xuanze's first experiments, right? He's one single cell in self-heating feeding requires a room temperature and the cell at low 30% state of charge. So, you know, really safe. The one for travel is single cell, super healthy, no issues requires something like 150 degrees or 160 degrees room temperature, which is never going to be achievable. If then two cells requires 154 cells acquire a hundred and thirty, forty five, you know, and then you scale and you get down figuring out that from what you can gather experimentally or numerically.

Wojciech Wegrzynski: 51:05

Um,

Francesco Restuccia: 51:06

If you can figure out what the errors that you are compounding are. For example, when we do experiments, if I am doing battery experiment and I have a pack, I I'm ignoring the material. That's inside the ball. Cause I'm ignoring the separation. I'm ignoring a lot of things. Right. But it should be good enough that we can then use it. If we try to do everything perfectly. Look at CFD. If we were to do a direct numerical simulation of Navier Stokes or a box that's bigger than a meter cube, we'd be spending 150 years waiting for the solution. And so often. You know, we do combustion, right? So in combustion we need to optimize our parameters. So if I'm looking at turbulence, um, and I care about the really precise chemistry and chemical kinetics, then I will set up my simulation one way and I will make some assumptions if I care about a really fast solution, but that gives me a temperature range for which my behavior will be in. And. Can afford to have a 50 degree error, then I will run it with much less chemistry, detailed chemistry, or we'll run it with fixed chemistry parameters and get my quick solution out. And so as an engineer, I mean, we have to wear two hats, right? So I'm a scientist, so I want things to be accurate and I want things to be precise, but as an engineer, I need it to be useful. Right. And so you need to find a balance in between the two and having to wear both hats. The fun part about research, I guess is. but yeah, we need to remember that it has to be useful at some point

Wojciech Wegrzynski: 52:31

Yeah. And the third hat of managing a laboratory in health and safety, but that's not the topic for the podcasts

Francesco Restuccia: 52:36

and teaching and teaching hats and admin hats, but let's ignore those for

Wojciech Wegrzynski: 52:41

I also have a podcast hat it's great. If you want to fit. It's fantastic. Thank you. Thank you so much. I think this is a difficult topic, but the more, you know, your enemy the less you are a fearful of it. And, I really want, my, my audience, my people to understand what is the issue we are dealing with? It's in a way, a mystical, magical technology. There's the fact that you can. Capture a lightning inside of a small metallic pouch, and then use it to power your electric cigarettes. And someone once said, I wanted to, to charge my book, uh, but my wife was charging her cigarettes. The future is stupid. Uh, it's a, thing that enables us. And, if you look at the world, it's is actually one of the technologies that. That is necessary to change the planet, the fossil fuel to everything, to a more sustainable future and, uh, fire engineers.

Francesco Restuccia: 53:38

I like Guillermos approach to this. So I'm going to steal Guillermo to this, which he uses in other

Wojciech Wegrzynski: 53:43

you should. Yes.

Francesco Restuccia: 53:44

he says, you need to think about when you think of a fire problem. It does. I mean, technology is useful, right. But you need to think about how to prevent it, how to compartmentalize it. If a fire happens, how do you compartmentalize it? How do you detect it and how do you suppress it? And all four things. Equally important because we always want to have prevention, right. We never even want to have a fire, but our job is also to figure out what happens if we do have a fire. So how do we compartmentalize it? How do we detect it? So we know it's happening and that's sort of what we were discussing earlier about voltage and temperatures. and how do you suppress it? battery fire suppression is really, really complicated because the chemistry can get really complicated. The toxicity of some of the gases can be, there's

Wojciech Wegrzynski: 54:23

I think even, even access to the cells that are reacting is, is the major

Francesco Restuccia: 54:28

Yeah. Um, and so, and so it's, it's putting all of those together. That's important. And so our technology is our future technological future is always bright and it's always improving, but from a fire site, we're never going to prevent every single fire. And so we always need to think about the other aspects as well.

Wojciech Wegrzynski: 54:44

Fantastic. Thank you so much for sharing this with us. And, yeah, I, I will, link to all the works that have been mentioned in the talk in the show notes. And, , please send me, uh, if you have something to add to that, because I think you've mentioned that quite a lot of interesting research being done at your lab and that, at Imperial College, And, it is absolutely worth sharing. I I'm sure some people would like to dig in more

Francesco Restuccia: 55:12

Absolutely. We'll be very happy to, and it's not limited to me. I'll send you the works we have done, but there's works from the U S works from Asia. You know, battery field is really growing. So I've shared our experiences, but thankfully there's a lot of researchers

Wojciech Wegrzynski: 55:25

Thank you. Thank you so much for doing that. And thanks for coming to the Fire Science Show. I hope you've enjoyed and

Francesco Restuccia: 55:30

you so much. Yes. Thank you for having me.

Wojciech Wegrzynski: 55:32

Cheers

Francesco Restuccia: 55:33

Cheers.

Wojciech Wegrzynski: 55:34

And that's it. Thank you for listening. I hope this was a five-star worthy episode from my perspective. It was because I finally learned what makes the batteries burn. And it's an interesting lesson. In the end. I would like to bring your attention again to the elephant in the room that Francesco pointed that is the second life of batteries. If we want a sustainable future in a circular economy. There is no way we can just let them rot on the dump wastes. And that will probably lead to catastrophical. Fires as well. So I don't think it's even the choice. We need to find a way how to reuse batteries. We need to find a way how to give them second life. Because of how they're manufactured. It's probably going to be difficult to disassemble them to first materials and then assemble again. So we probably need to find a usage for used lithium-ion batteries. We probably need to find a way. How we can benefit from them again, once there. Ended their life cycle In the primary device, they were in like car or laptop or something. And I think it's going to be a huge challenge. I know the focus today is on the safety of cars. Buses. Carparks. Energy storage at buildings, mass energy storage facilities. Maybe warehouses, but trust me in a decade or two, when there are billions of use cells that needs to be refurbished or used. A lot of us will be dealing with the issues of how safely give battery second life. And I think. We should look forward to that. There is no sustainable future without reuse. So we need to be a significant part of that. And, through the podcast, I will try to find a great guests who can talk about. The second life of batteries, even though there's not that much knowledge yet. But, uh, hopefully we can start working towards solutions before the problems start arising. And so, yeah. That would be pretty good. And for now. I think that's it for the today's episode. Thank you very much for staying till the end and for listening to the show. And if you enjoyed it, there's another episode for you waiting every Wednesday. So see you there. Cheers.