Science Friday - Supreme Court Case On Regulation Of Vapes | Predicting Lithium-Ion Battery Explosions

Episode Date: December 6, 2024

The Supreme Court hears a case centering on the FDA’s denial of applications for nicotine-based vapes containing flavored liquid seen as addictive for young people. And, damaged lithium-ion batteries can cause deadly explosions. An algorithm could help detect when they’re about to happen.Supreme Court Hears Case On FDA Regulation Of Flavored VapesOn Monday, the US Supreme Court heard arguments in a case that could change the Food and Drug Administration’s power to regulate nicotine-based vapes. It revolves around the FDA’s denial of applications from two vape companies that sell flavored liquids in their products, citing that the liquids presented a danger in addicting young people to nicotine. But a lower court rejected that denial, saying the agency was inconsistent in its approval process. The decision could impact the FDA’s regulatory power to ban the sale of some of these vapes in the US.Ira Flatow is joined by Rachel Feltman, host of the Popular Science podcast “The Weirdest Thing I Learned This Week,” to talk about the health implications of the case. They also discuss other top science stories of the week, including why there’s been a large decrease in cervical cancer rates over the past decade, why Venus probably wasn’t able to support ancient life (contradicting some hopeful theories), and new information about when humans might’ve started to wear clothing, thanks to some clues from lice.A Warning For When A Lithium-Ion Battery Is About To ExplodeLithium-ion batteries are used in all sorts of electronics, like smart phones, laptops, and e-bikes. That’s because they can store a lot of energy in a small package. Unfortunately, this also means that when a battery is damaged, there’s a lot of energy to expel, which can lead to explosions and fires.While there are no national statistics available, in New York City alone, there have been 733 fires started by lithium-ion batteries since 2019, which killed 29 people and injured 442. Most of these fires are a result of e-bike or scooter batteries catching fire.Now, researchers have trained AI algorithms to be able to better predict when a lithium-ion battery is about to explode. The battery makes a hissing sound two minutes before an explosion.Ira talks with two researchers at the National Institute of Standards and Technology about their work on the subject: Dr. Andy Tam, mechanical engineer in the Fire Research Division, and Dr. Anthony Putorti, fire protection engineer and leader of the Firefighting Technology Group.Transcripts for each segment will be available after the show airs on sciencefriday.com. Subscribe to this podcast. Plus, to stay updated on all things science, sign up for Science Friday's newsletters.

Researchers are working on tech that could give you a two-minute heads-up if a lithium-ion battery is about to blow. You could potentially build this technology into smoke detectors that are on the ceiling, but you could also put them potentially in a product to also provide detection. It's Friday, December 6th, and you're listening to Science Friday. I'm sci-fri producer Deep Biedershmetz. Lithium ion batteries power so many devices in our lives, and companies like them because they can store a lot of energy in a small package. Unfortunately, that means when those batteries are damaged, all that energy can lead to explosions and fires. But researchers have trained AI

algorithms to predict when a lithium ion battery is about to explode. Thanks to a key sound, it gives off two minutes before going off. We'll get to that story in a bit, but first, here's Ira Flato with the top news and science this week. Earlier this week, the U.S. Supreme Court heard arguments that could change the food and drug administration's power to regulate nicotine-based vapes. It revolves around a case where the FDA denied applications, of two vape companies that sell flavored liquids in their products, citing that the liquids presented a danger in addicting youth to nicotine. But a lower court rejected that denial,

saying the agency was inconsistent in its approval process. Well, here to read the tea leaves of this case and enlighten us with the Science News of the Week is Rachel Feldman, host of the popular science broadcast, the weirdest thing I learned this week, and Scientific Americans podcast Science Quickly. Welcome back to Science Friday, Rachel. Thanks for having me, Ira.

All right. Can you tell us what's going on with this case in the court? Yeah. So back in 2020, a couple of companies filed FDA applications for vape products. And they included a bunch of flavors that the FDA flagged as being, you know, potentially really appealing to kids. And as you said, a lower court then said that, you know, sort of proper legal procedures hadn't been followed. And what we're seeing now is basically the result of an appeal that the FDA filed under Joe Biden's administration.

So the question isn't should all flavored vapes be allowed. You know, the FDA has definitely cracked down on flavored vapes and no one is against that in principle. but what the Supreme Court is going to decide is like how much power the FDA actually has to enforce that. So in other words, we'll see whether the courts think the FDA should retain its enforcing kind of powers. Exactly, yeah. Or are there going to be a ton of loopholes for companies to say, no, the FDA unfairly targeted us for these flavors, which just to be clear, in this case, includes stuff like pink lemonade and a flavor called Jimmy the Juice Man, peachy strawberry. which, you know, it is really hard to imagine that not targeting sort of the TikTok generation. But, you know, we will see how the Supreme Court rules, which should happen sometime in the summer. And at least the preliminary sense is that it looks like some members of the Supreme Court

were not impressed with the argument that the FDA overstepped here or misstepped, that it seems like the Supreme Court might be leaning toward siding with the FDA. And if the Trump administration keeps coming to the Supreme Court with more challenges to other regulations, this might tell us how it might rule in those cases. Yeah, well, and, you know, vapes are an interesting thing in particular when it comes to the Trump administration because he recently posted on social media that he was going to once again save flavored vapes, which is interesting because his administration, you know, did come out against vapes targeting kids and being available to kids. But he now seems to be taking issue with, you know, the anti-flavored vape cases. It's hard to actually know where they stand on this issue in particular. But, yeah, certainly we're seeing a lot of challenges to agencies' power. So it'll be interesting to see how this case plays out.

Yeah. All right. Let's stick with health for a second because we have some really great news this time about a huge decrease in cervical cancer death rates. What's happening there? Yeah. So researchers were basically saying, you know, it's been almost 16 years. since the HPV vaccine was introduced. So we should start to see the initial impact of that vaccine on cervical cancer deaths

because, you know, the people who were first vaccinated are getting to be in the age groups where, you know, we would expect cervical cancer diagnoses and deaths. And they found something really promising. You know, they were seeing in the data like a steady drop in the number of deaths from cervical cancer in U.S. women specifically under the age of 25. But then from 2016 to 2021, instead of that study drop, they saw like an absolute plummet. And while they weren't looking specifically at which of these women were vaccinated, there's really no other factor at that time that could explain that sudden drop in deaths. And it's especially compelling given that, you know,

we saw things trending downward, which we can attribute to better screening, you know, more public awareness. So it does seem quite likely that the vaccine is to thank for that big drop in deaths. Wow, that is good news. Well, speaking of what may not be such good news, I want to move over to the Arctic ocean where scientists are projecting that we could see the first ice-free day there before 2030. Wow, that's pretty early, isn't it? Yeah, it's definitely troubling. This is a study looking at the idea of an ice-free Arctic, which, just to clarify, because I think some folks have been a little confused by the headlines on this, we're not talking about literally no ice. I mean, it's the Arctic for there to be literally no ice.

That would be like our planet would be uninhabitable if it was that warm up there. But it's less than a million square kilometers of ice is considered functionally ice-free. because that is so below the range of what is normal. And also with that threshold, basically we would functionally have what you call a blue ocean event, meaning the ice that's supposed to help reflect solar energy and like insulate that part of the world would be gone. Wow. Even in the wintertime.

Right. Yeah. So when we talk about functionally ice-free, you know, we're saying there's so little ice. that for climate purposes, we're dealing with water instead of ice on the whole. And yeah, what this study found is that in the worst case scenario, we could be seeing an ice-free day in just a few years. Previous studies have focused on the first ice-free month or the first ice-free summer, which would take a little bit longer to happen.

But researchers are pointing out, you know, even just having one day where we met this ice threshold would show that we had reached a really bad turning point. The good news is that if we stick to 1.5, you know, for our climate target or even stay at 1.6, we will be very unlikely to see ice-free days in the near future. Like, there is a lot we can do. This is a worst-case scenario. It's not an implausible worst-case. scenario, but it is one that we have some power to prevent. Yeah, well, we reached 1.5 a lot earlier than we thought already. Yeah. So let's let's not keep doing this because we could be with this kind of news all day. I want to move on to Venus.

Let's get us off the planet for a second. Yeah, let's talk about a place that has apparently always been a hellish. Ice free. Yeah. And there's some news there. Tell us what the news on Venus is. Yeah. So the news on Venus is that, you know, there's been this idea that because Venus is about the same size as Earth and about the same distance from the sun, that it might be sort of our evil twin is a phrase that gets used a lot, that maybe it used to have water on the surface and some kind of runaway greenhouse gas effect, you know, messed with its atmosphere, left it totally hellish. It's got these sulfuric acid clouds. It's like 900 degrees Fahrenheit on the surface. Not a fun place to be.

But this new study looked at the content of basically what's coming out of the volcanoes on Venus and found such a low percentage of water steam that they think the interior of the planet must be super dry. And in fact, they say it's so dry that it's really implausible for there to have ever been liquid water on Venus. Really? Yeah. So instead of this evil twin idea, I mean, maybe it really is. is our evil twin because even though it has these superficial similarities to us, it seems like it was never a planet like Earth. It was probably always pretty hot and dry. But they say that it's still possible that it once hosted life. It would just be something much weirder than we've

ever seen before, like maybe microbes that evolved to live in those sulfuric acid clouds. We really don't know. We don't know. Let's go back to some health news because there's a study that has important findings for middle-aged women, right? Yeah. So I really love this study. It looked at what the researchers called Vigorous Intermittent Lifestyle, Physical Activity, or Vilpa, for short. And those are the little bursts of increase in heart rate that don't happen during, like, a high-intensity interval workout.

But it's like just everyday stuff. You're climbing stairs. You're carrying something heavy. You're chasing a pet, whatever. And, you know, there's. there's evidence to suggest that any kind of movement, even if it's not structured exercise, does have a cardiovascular benefit. And they wanted to see how much. And this study was pretty massive. They looked at data from, I want to say, 22,000 middle-aged folks in the UK.

And they did find that just a few minutes of this sort of incidental activity every day had a really strong cardiovascular effect. It was about 16 percent lower risk of things like strokes and heart attack in men, but it was much stronger in women. We saw like a third drop in risk if they had just a couple of minutes of this kind of activity a day. And once we get out to around five minutes, we're talking about like having the risk of a heart attack. So just a great reminder that every little bit of movement really does count. And yeah, just do what you can. And scientists have designed a new implantable device that could monitor in inflammation in the body. That's a first, right? How does that work? Yeah. Yeah, it's really cool.

They've compared it to, you know, continuous glucose monitors that, you know, you can wear and that, you know, measure your blood sugar levels continuously so that you can get, you can see trends. You can understand what's impacting your glucose levels. And we don't have anything like that for inflammation. And basically, it would sit under the skin and it has. has these strands of DNA that compared it to like a tree shaking fruit off of its branches. The strands of the DNA can stick to proteins, but then sort of shake them off so that they can grab more to get a continuous reading. And there's so many inflammatory diseases that, you know, cause chronic illness that cause terminal illnesses. So this could be a really cool tool to

help doctors and patients understand what's going on. It's always good to have. It's always good to have you, Rachel. Always great to chat, Ira. Thank you very much for taking time to be with us today. Rachel Feldman, host of the popular science podcast, the weirdest thing I learned this week, and Scientific Americans podcast, Science Quickly. After the break, how researchers trained an AI algorithm to detect lithium ion battery explosions. Stay with us. Hi there, Ira here, letting you know that we have a dollar-for-dollar match right now. So any donation you make will be doubled. And this week we're celebrating Giving Tuesday, which means now is a great time to double your impact and show your support for Science Friday, a nonprofit dedicated to making science accessible to the

public. So please go to science friday.com slash support to make your donation and invest in the future of science journalism. Again, that's science friday.com slash support and thanks. You know about the dangers and widespread use of lithium ion batteries. They're used in all sorts of electronics like smartphones, laptops, e-bikes, and this is because they can store a lot of energy in a small package. And unfortunately, this also means that when a battery is damaged, there's a lot of energy to expel, leading to explosions and fires. You may have read the headlines about e-bike batteries catching fire. In New York City alone, there have been 733 fires started by lithium ion batteries since 2019. 29 people have been killed. 442 have been injured, and there are no currently available

national statistics. But here's the possible good news. Researchers have trained AI algorithms to be able to predict when a lithium ion battery is about to exploit. But, here's the possible. and they have documented how the battery gives off a sound some two minutes before the explosion. Joining me now to talk more about their work on the subject are my guests. Dr. Andy Tam, mechanical engineer in the fire research division at NIST, that's the National Institute of Standards and Technology, based in Gatorsburg, Maryland. And Dr. Anthony Ptoorty, fire protection engineer and leader of the firefighting technology group at the NIST in Gatorsburg, Maryland.

Welcome both of you to Science Friday. Thanks for having us. Thank you, Ira. You're welcome. Dr. Ptoldi, let me start with you. Some of the basics. Why are lithium ion battery fires more dangerous than other types of residential fires?

Well, there's a couple things that contribute to that. One is rapid growth. So when the batteries are damaged, overheated, there's an internal fault, they can catch fire and eject toxic combustion products very rapidly. And then that serves as a very good ignition source to ignite other materials that surround it. Right. The other is that the combustion products that are emitted from the battery are a little bit different than your normal smoke from a fire.

They contain a lot of fluorinated compounds such as hydrogen fluoride. They also result in toxic debris that contains. heavy metals and metals such as nickel, coal, and cobalt, manganese, and of course, lithium. And so those are the kinds of things that make them more of a challenge

as far as a fire hazard is concerned. And when they get going, I mean, they're really hot burning, aren't they? Yes, and typically in a product, you don't just have one battery. You have a number of batteries that make up a module, and so if you have one go into thermal runaway,

then it tends to overheat the batteries next to it. So you can have multiple of these battery fires. In addition, fires related to lithium ion batteries are much more difficult to extinguish. So that should be kept in mind as well. And there's research going on as far as extinguishment and how to deal with these fires. Right. And Dr. Tam, let's talk about this sound I mentioned in the introduction. What kind of sound does it make? And what's going inside the battery that makes it make this sound? So the entire thing is called by thermal runaway. So basically, it is a self-sustained, destruct process that will

occur when the temperature of the battery increases. And when the temperature increases, there will be a internal chemical reaction, and that sort of caused the temperature to further increase. And during this process, there are three important events, when the battery undergoing a thermal runaway due to, let's say, thermal abuse. First, there will be a safety virus. break gauge moment with a very little amount of gas coming out, there will be a hushing sound. So there's a safety valve that breaks and then there's a hissing sound? Yep, yep, yep. And then secondly, it will enter to what we call our venting stage. And at that stage, there will be more and more hot toxic gases, just like what Tony mentioned. And then finally, there will be an ignition

stage in which you will see flame jets and the surface temperature of the battery can be as high as 1,000 degrees C. Wow. And we were only talking about like one single cell if we have a pack of batteries. This means that the ignited battery will trigger what we call the thermal runaway propagation in which the surrounding batteries will get ignited and at this point the fire will become very difficult to put out. Is there some way for a detector to hear the hissing sound in that two minutes beforehand,

the signature that it's going to explode? Oh, yeah, definitely. And this is exactly what we were doing. And then we want to make use of sound and also machine learning to develop a fast-responding, easy-to-use an accurate model to detect the safety valve-breakage moment, which we consider as one of the earliest stage of the thermal runaway. So to solve this problem, you developed an algorithm to be able to detect this safety valve explosion sound?

Tell me how this works. Oh, we actually do this as a team, and we have a collaboration with other institutions. Basically, we conduct a series of experiments with different state of charge, different battery orientation relative to the microphone, collect different sound signature. And then each of the test lasts about 10 minutes. And then basically during the experiment, you will hear some hushing sound,

and then you will see some kind of small jets when the safety valve breaks open. And in my opinion, this sound signature is very unique. And we extract this sound signature together with audio data from various human activities. to develop an effective earlier stage thermal runaway detection model. You mentioned explosion. What we're really seeing here is, you know, we have this relief valve on these cells. And so what it's doing is there's a jet of hot gases and fire and particles that are coming out. So the whole cell isn't, you know, coming apart in a big explosion. Really, what you're seeing is this jet of flame once this phenomenon occurs.

Wow. And Dr. Patorre, is this something you can take advantage of? of in a product or safety somehow? Typically, when you think about buildings, your home, your office, you're going to have smoke detectors that are on the ceiling. You could also have sprinklers. And so you have some distance between the battery or the product and the detector. So you could potentially build this technology into detectors that are located there, but you could also put them potentially in a product to also provide detection.

And that might be able to be done faster than the more difficult and time-consuming process, at least in the near term, for getting these types of capabilities in other kinds of fire detectors. Right. Could you build it into the battery itself to detect its own, you know, its own hissing sound and put out an alarm somehow? It can be integrated. Certainly there will be a possibility to do that, but we want to make it in a way that it can be portable. Let's say now you already have your lithium iron battery. You don't want to buy a new one. So we want to build this technology such that it can be used for any old devices. And how far or how close are we to be able to do that? Dr. Tam, I know you've applied for a patent for this technology.

How long would it take for this to be a safety feature that's available for consumers? First of all, we want to be safe. So we anticipate to conduct more experiments to collect data with different types of batteries, because right now we're only looking at one battery. And then we will also have to look at different microphone and also different environmental settings. And at the same time, we will have to deploy our model while we are doing the experiment and see whether or not our model would need further improvement. And then this is something that we'll be working on in 2025.

Why do we see these e-bikes? I was talking about this in the introduction about how many explosions there have been. Why does that happen? I'm not familiar with all the statistics as to how prevalent each cause is, but you do have manufacturing defects in the batteries, and the quality of the batteries can vary depending on the manufacturer and the origin of the battery. You can also have issues with repairs that can be made to the devices,

replacement batteries that were not specifically meant for that device. You can have chargers that are used that are not the correct charger, especially if you needed a replacement. You can have issues with the device getting wet, with it getting too hot. You know, if you leave it in a place where there's high temperatures. So there's a variety of different types of things that can cause the malfunction that leads to the fire. Well, until we have your product, Dr. Tarr,

Until then, give me a suggestion about how folks can stay as safe as possible if they have e-bikes or other devices with lithium ion batteries. So Tony sort of mentioned it a little bit. So basically, we would encourage using proper charging practices. For example, using right charger, don't overcharge the battery, and avoid extreme environment. And then while we are charging the battery, make sure that we are charging it in a safe place. And then just to anticipate, maybe it's good practice to think ahead of time. Let's say if the battery really go wrong, what you would want to do to try to mitigate the process or plan for the exit route.

And we believe that this will be helpful just in case there will be an accident. Also, when you're buying your product, you can make sure it's listed by a qualified test. and certification lab to make sure that the product is a good product. I'd also like to mention the National Fire Protection Association. They actually, on their website, they actually have some good one-page summaries of what you can do to use these products more safely. And also, as Dr. Tam indicated, especially if you have an e-scooter or something like that, you definitely don't want to store it or charge it in an exit route.

That's a really bad place to do it because if something goes wrong, wrong, that can block your exit. Those are good tips for everybody who's got them and who will be getting them for the holidays. Thank you both for taking time to be with us today. Thank you. Thank you very much, Ira. You're welcome. Dr. Andy Tam, mechanical engineer in the fire research division at the NIST, and Dr. Anthony Petorty, fire protection engineer, leader of the firefighting technology group at the NIST in Gatorsburg, Maryland. And that's all the time we have for Today, lots of folks to help make the show happen, including Kathleen Davis, Robin Casmer, Melissa Mayers, Jordan Smudjick. On our next episode, we're rounding up our favorite science

books this year for kids. I'm CyFRIre producer Dee Peter Schmidt. Thanks for listening.