Sodium ion safety test ends in flames - Bluetti power station

[thm_w]

Bluetti pioneer sodium portable power station 900Wh: https://www.bluettipower.com/products/sodium-ion-battery-pioneer-na
"All-Around Safety: Sodium-ion chemistry, AI-BMS, structural safeguards, and electrical security keep every charge worry-free."

Apparently there are many options for sodium ion, some are explosive, some are poisonous: https://www.sciencedirect.com/science/article/pii/S2352152X25013271

You can find plenty of other videos where sodium ion pack is punctured and nothing happens. Is that what we are seeing here, a difference in chemistry?

Yes I get its dumb and unrealistic that he shot it full of lead. But it still seems excessively volatile.

[5U4GB]

Looks like a stupid clickbait video.  Next thing he'll post one showing you can set a car petrol tank on fire with APIT rounds, which is about as likely to happen as the event he's created.

[uer166]

Looks like a stupid clickbait video.  Next thing he'll post one showing you can set a car petrol tank on fire with APIT rounds, which is about as likely to happen as the event he's created.

There's no obvious clickbait to me.

I think it's quite a good video demonstrating that marketing bullshit is just that: bullshit. Someone should do it with a LiFePO4 pack demonstrating the same.

Bottom line is that there's currently no intrinsically safe high density battery technology, and thinking that there is would put a integrator into a bad situation.

[Simmed]

my summary of what i saw
WTF
a battery impale test?
no no no
impale test is for kids
its battery shooting test !
what did he use to shoot it ?
i dont know, i only saw the shooting part
shooting it 12x ?
it starts burning
shoot it another 6x ?
there ! kaboom !


 
his target grouping is not very nice
come now
the title should read
"this battery is not bullet proof"
or
"i shot this 18x"
i bet the 2nd title gets more views

[Marco]

If you vaporize a block of iron it can burn violently too. Intrinsically safe still means under realistic circumstances.

[nctnico]

What you need to keep in mind is that a battery contains energy and that this energy can do damage when released in an uncontrolled way. And this is true for every device which can store energy. Even a simple coil spring is potentially dangerous if it can store enough energy. For example a coil spring like the ones mounted on a car, can snap your arm or leg off.

[Siwastaja]

Shooting at it is a decent way to do a penetration test. Result would have been same on a classic nail penetration, which simulates completely realistic threat models. Physical damage happens. The video demonstrates that the thermal runaway feeds itself after the initial damage, and the amount of energy released, and the projectile action, all real problems with Li-ion and similarly Na-ion batteries.

All the safety talk is just marketing. As uer166 says there is no high energy density battery chemistry available at this point, and sodium ion as a replacement for li-ion is only for cost reduction, nothing else; energy density is worse, safety is not any better.

But apparently the novelty of the technology adds some wow factor, when it's really just a worse product in all respects, except price. Specifically I would not want sodium ion in mobile applications (like this power bank, let alone an EV), because it's not that much cheaper, not any safer, and compromise in energy density is significant.

We need smaller batteries, not larger. We are going backwards: we are in a situation where a normal family sedan weight has gone up from 1300kg to 1700kg when made an EV, which is still barely acceptable as there are significant advantages for an EV which offset that extra weight; fair enough, using state-of-the-art battery tech it's a decent car with decent range and the increased weight must be accepted as a compromise. But then the Chinese pushed LFP into EV mainstream and so it went to 2000kg and consumption to 20kWh/100km due to sheer weight of the car, the car eats up tires faster than ever before, but car journalists praise them because max quick charge power went up a little bit (range went down, but they don't care about that) and price went down by a few thousand.

But what is the total cost of ownership?

And now if you look at the media, car journalists are waiting for sodium ion batteries, the Chinese promise they are here soon, so then the weight goes up to 2200, maybe 2300kg and consume even more energy to move that mass, with even less range. But they are a little bit cheaper to initially buy, and they charge fast, and that's all that seems to matter.

Really, the only application I see for sodium ion batteries could be stationary (home or grid-level) energy storage. It's like margarine, it's a cheap substitute for the real thing, not some kind of health product / safety improvement.

[Siwastaja]

What you need to keep in mind is that a battery contains energy and that this energy can do damage when released in an uncontrolled way. And this is true for every device which can store energy. Even a simple coil spring is potentially dangerous if it can store enough energy. For example a coil spring like the ones mounted on a car, can snap your arm or leg off.

Yep, it's impossible to make a completely safe battery, but that doesn't mean they have to be this dangerous. Specifically, the mechanism for thermal runaway which feeds itself, shooting out flammable electrolyte which is not part of the actual energy storage but extra load on top of that, and acting as projectiles are all negative features which fundamental physics do not require. So it is entirely possible to see significantly safer (not perfect though) high energy density battery technologies in the future. But obviously this isn't one of them.

[Marco]

Shooting at it is a decent way to do a penetration test.

Once.

[Siwastaja]

Shooting at it is a decent way to do a penetration test.

Once.

Result would have been likely the same if shot once, if the exact spot where to shoot was known. At least if repeated 10 times with different units.

The idea of these sort of tests is not to show how probable it is that the same triggering event happens in real life. The point is to prove the presence of the chain reaction, and the severity of that chain reaction, and using a strong trigger is the best way to reliably achieve that.

For safety, you want failure rates near 0%. For testing the safety, you want failure rates near 100%, so that testing is quicker and cheaper. But this confuses some people.

It's the same idea as with accelerated aging test: you create unrealistic conditions exactly to amplify the phenomenon you are interested in. In this case, you amplify the odds of it happening; but the actual consequence is real. Just that it would be harder to artificially test if you tried something benign.

Yet, if a product is claimed "safe", then it means it should not go randomly in flames and shoot burning projectiles out. Not on any single customer, even if 1000000 is manufactured and sold. And if a mechanism, triggered with shooting it multiple times with 100% success, exists, then it means 0.001% of them will exhibit the same behavior from some very usual usage pattern, like just using it as intended (internal fault), maybe dropping it on a concrete floor, or something similar.

Shooting test does not of course show how often that will happen, but given enough of these things in the wild I can assure you it will happen; this video demonstrates what will happen, not when.

And don't get me wrong, I'm all fine accepting that risk, really. Just, I don't think there is anything wrong with the video, really. But I don't tend to panic by seeing things I already know.

[Marco]

Depends. If it it's round cells, it's likely about getting lucky. If it's pouch cells, it could just as easily be about overcoming the thermal balance. More holes, more heat, cooling stays about the same.

[mtwieg]

The video is pretty straightforward about what it is. It's a reality check against people making unrealistic projections about new technology.

I wish they would have explained specifically how the energy is being released (I don't think its sodium metal combusting like lithium does, I guess the electrolyte is a very volatile solvent?).

[helius]

The reactivity series for alkali metals increases with heavier elements. Sodium is more reactive than lithium, and the most reactive is cesium. (Francium is so rare that for all practical purposes we do not need to consider it.)
But the cells under discussion do not contain significant quantities of either sodium or lithium as metals. They are called lithium-ion cells for a reason. The energy released is from the combustion of the electrolyte.

[Marco]

a battery contains energy

Any piece of metal contains potential energy in a reducing environment, a lot of it. The old scifi trope of "what if you could discharge all the stored energy in a battery at once" is not relevant, because under non scifi situations you can't.

Electric current limited reactions are limited by the fact that the current simply blows away shorts when it gets too large. It's plain old oxidation by atmospheric oxygen which gets you, a question of reaction speed and how fast material can be exposed to atmospheric oxygen once things heat up. Can the electrolyte flow? Do the materials swell enough once things start oxidising to increase the rate of exposure to oxygen to problematic levels? Those are more important than the theoretical potential energy of oxidation, even for highly reactive metals.

The battery can no more oxidise instantly as a block of iron.

[Mr. Scram]

You can find plenty of other videos where sodium ion pack is punctured and nothing happens. Is that what we are seeing here, a difference in chemistry?

Yes I get its dumb and unrealistic that he shot it full of lead. But it still seems excessively volatile.

It may be as simple as a bullet ending up in just the right spot to cause a short, or dislodging something causing a short, and the resulting heat and fire doing the rest. When stuff in a container heats up and builds pressure you will get pops, crackles and ejections. Something like burning wood crackles and pops and ejects material as well.

The reaction doesn't seem very volatile to me. It's a large battery pack with a bunch of energy and it's mostly a smallish fire. I've seen smaller lithium ion batteries causing a lot more fire, smoke and mayhem.

[thm_w]

What you need to keep in mind is that a battery contains energy and that this energy can do damage when released in an uncontrolled way. And this is true for every device which can store energy. Even a simple coil spring is potentially dangerous if it can store enough energy. For example a coil spring like the ones mounted on a car, can snap your arm or leg off.

I agree but this is only 900Wh, and there are plenty of videos of people impaling similar power density LFP or sodium ion pouch cells and all that happens is a bit of smoke. So there is some clear difference here beyond energy.

The reactivity series for alkali metals increases with heavier elements. Sodium is more reactive than lithium, and the most reactive is cesium. (Francium is so rare that for all practical purposes we do not need to consider it.)
But the cells under discussion do not contain significant quantities of either sodium or lithium as metals. They are called lithium-ion cells for a reason. The energy released is from the combustion of the electrolyte.

OK so is there some way to figure out what electrolyte Bluetti cells are using, and if safer alternatives exist.
It looks like there are a wide variety of options: https://advanced.onlinelibrary.wiley.com/doi/10.1002/adma.202313572

The details I can find are "NaFeMnO₂ (layered sodium manganese oxide) cathode with a hard carbon anode." I don't see any electrolyte mentioned.

It may be as simple as a bullet ending up in just the right spot to cause a short, or dislodging something causing a short, and the resulting heat and fire doing the rest. When stuff in a container heats up and builds pressure you will get pops, crackles and ejections. Something like burning wood crackles and pops and ejects material as well.

The reaction doesn't seem very volatile to me. It's a large battery pack with a bunch of energy and it's mostly a smallish fire. I've seen smaller lithium ion batteries causing a lot more fire, smoke and mayhem.

I would understand that if it was one bang, but they just keep coming, and the flames are quite intense.
Yes li-ion is worse, but I feel like we didn't watch the same video. LFP would probably do nothing in a proper enclosure, this thing would easily burn your house down.

[uer166]

LFP would probably do nothing in a proper enclosure

LFP can runaway as well.. Even if it doesn't cause a giant flamethrower effect, the gas from internal reactions will easily kill you in an enclosed-ish space like an apartment. There's FAR too many absolutely terrible dangerous LFP packs that are made by people that think that it doesn't matter because they're inherently safe (they're not).

[Mr. Scram]

I would understand that if it was one bang, but they just keep coming, and the flames are quite intense.

While I'm not familiar with the construction of this specific device, other power banks of this size seem to typically contain a sizeable number of cells or pouches.

Yes li-ion is worse, but I feel like we didn't watch the same video. LFP would probably do nothing in a proper enclosure, this thing would easily burn your house down.

It's a large chunk of matter on fire. It doesn't seem to escalate or spread to anywhere near the same degree lithium ion regularly does. It mostly seems to be a thing on fire.

If you burn a chunk of the same size of some wood types, it's likely to produce a larger, more violent fire. We're perfectly comfortable stuffing our surroundings full of wood, dry as a bone, and even building our houses out of it.

I don't think it needs to be completely inert in the case of a mishap. It just needs to be less susceptible to catch fire, burn less violently, be easier to put out, or produces less dangerous gasses while on fire than lithium ion, or any combination thereof, and it will be a meaningful improvement.

[nctnico]

We need smaller batteries, not larger. We are going backwards: we are in a situation where a normal family sedan weight has gone up from 1300kg to 1700kg when made an EV, which is still barely acceptable as there are significant advantages for an EV which offset that extra weight; fair enough, using state-of-the-art battery tech it's a decent car with decent range and the increased weight must be accepted as a compromise. But then the Chinese pushed LFP into EV mainstream and so it went to 2000kg and consumption to 20kWh/100km due to sheer weight of the car, the car eats up tires faster than ever before, but car journalists praise them because max quick charge power went up a little bit (range went down, but they don't care about that) and price went down by a few thousand.

But what is the total cost of ownership?

And now if you look at the media, car journalists are waiting for sodium ion batteries, the Chinese promise they are here soon, so then the weight goes up to 2200, maybe 2300kg and consume even more energy to move that mass, with even less range. But they are a little bit cheaper to initially buy, and they charge fast, and that's all that seems to matter.

Later this year Changan is planning to put an EV (Nevo A06) on the market with a sodium-ion battery. Range is 400km (probably 300-ish km in real world driving conditions). It will also be available as a series hybrid. You are right that EV batteries are moving backwards into lower energy densities.

[mtwieg]

It may be as simple as a bullet ending up in just the right spot to cause a short, or dislodging something causing a short, and the resulting heat and fire doing the rest.

LFP can runaway as well.. Even if it doesn't cause a giant flamethrower effect, the gas from internal reactions will easily kill you in an enclosed-ish space like an apartment. There's FAR too many absolutely terrible dangerous LFP packs that are made by people that think that it doesn't matter because they're inherently safe (they're not).

I'm not super experienced with lithium batteries, but I have worked with large supercap arrays, and when talking vendors about safety the message was roughly the same as LFP batteries: the electrolyte (all of these cells used ACN) may vent under overvoltage/overtemperature conditions. In that form, the electrolyte is extremely inflammable and quite toxic. However, the cells alone will not cause the electrolyte to ignite (I'm not sure if that includes puncture conditions, that wasn't a concern in our application). That requires a secondary ignition source, making the question of safety much more complicated.

I wonder if in this test the bullet is also acting as the ignition source (i.e. it's very hot or creates sparks when impacting the battery), which would not happen with a typical nail penetration test.

Aren't there standardized penetration tests? Can't imagine there aren't...

[5U4GB]

Aren't there standardized penetration tests? Can't imagine there aren't...

Yup, the standard is the nail penetration test:



Trust a US person to "test" a battery by shooting it... (yes, I know this is stereotyping :-).

[Siwastaja]

Trust a US person to "test" a battery by shooting it... (yes, I know this is stereotyping :-).

It's a very... American way 

And yet - no one claimed it's a scientific way, or a standard test. But that doesn't make it totally useless. It demonstrates well how such a battery behaves once in thermal runaway.

[Simmed]

there is that (i think is Finnish guy?)
hydrualic press channel
and a curious samsung lithium that did not explode
 

[Mr. Scram]

You are right that EV batteries are moving backwards into lower energy densities.

Lower energy densities aren't necessarily an issue when that's offset by longevity or cost, preferably both, as long as the difference is within reason of course. Considering the lifespan of an electric car is often dependent on the battery pack, having a 10% less dense pack that will last twice as long isn't a bad tradeoff. At some point there will be a crossover point where the decreased wear yields a better range, even if the power density was lower at the start.

Stuffing massive batteries in every car because people want range is a separate problem. It makes sense for some cars, but many will be lugging around excessively heavy battery packs. I don't think the solution to that is technological, though, it seems to be mostly a cultural issue, possibly combined with the state of the support network. If fast chargers are reliably available around every corner, people will worry less about range.

I feel a real portion of the range anxiety I see is caused by too few chargers, broken chargers, excessive charging rates at some chargers, and having to sit at a charger for extended periods of time. That messes with people, leading them to wanting big battery packs to be less dependent on factors outside of their control. I suspect that more hard wearing chemistries would actually help, as people have more trust in the range they get both now, and a few years down the line.

[nctnico]

You are right that EV batteries are moving backwards into lower energy densities.

Lower energy densities aren't necessarily an issue when that's offset by longevity or cost, preferably both, as long as the difference is within reason of course. Considering the lifespan of an electric car is often dependent on the battery pack, having a 10% less dense pack that will last twice as long isn't a bad tradeoff. At some point there will be a crossover point where the decreased wear yields a better range, even if the power density was lower at the start.

Stuffing massive batteries in every car because people want range is a separate problem. It makes sense for some cars, but many will be lugging around excessively heavy battery packs. I don't think the solution to that is technological, though, it seems to be mostly a cultural issue, possibly combined with the state of the support network. If fast chargers are reliably available around every corner, people will worry less about range.

I feel a real portion of the range anxiety I see is caused by too few chargers, broken chargers, excessive charging rates at some chargers, and having to sit at a charger for extended periods of time. That messes with people, leading them to wanting big battery packs to be less dependent on factors outside of their control. I suspect that more hard wearing chemistries would actually help, as people have more trust in the range they get both now, and a few years down the line.

You really need to think hard about why ICE cars have a typical range of 700 to 900 km. Why would that be? Why isn't half sufficient? Why not double? I'm sure it is not some random design choice!