Containing a battery pack meltdown.

[justinjja]

Any thoughts on something like this:

15kwh 14s 18650 battery pack,
Put it at the bottom of a 55 gallon steel drum
put a couple 32 packs of water bottles inside the barrel above the batteries.
Cut a hole in the lid and install a cooling fan.

If all the other safety measures failed for some reason,
Battery catches on fire, melts the plastic water bottles and gets doused in water.

In theory all the energy goes toward making steam instead of burning down my garage.

Does burning an 18650 release more energy than fully discharging it?

15kwh is 54 mega joules, or enough energy to boil only about 5 liters of water.

And if you are wondering about just the steel barrel without the water,
It would only take about 2.5 mega joules to melt the lid on a steel drum.

 

[Mudd]

I've heard of some people on here talking about putting sandbags above groups of cells on like a mesh grate. I don't think lithium and water mixed too well.

 

[justinjja]

My understanding is that there is actually almost no lithium in a lithium ion battery.
Google suggests 0.6g in an 18650.

 

[Texas-Mark]

I would use the sandbags instead of water. What if the jugs break for some reason? Water and electronics rarely go well together. I have used the water jugs in other applications though.

 

[time2roll]

My primary concern would be to contain and vent the fumes outside. Some type of metal box with a sealed lid would be good enough for me.

 

[Rider]

I'm thinking it's a potential disaster. Exo-thermic reaction, confined space, add water...

The power of the steam might be enough to blow the drum apart.

 

[Texas-Mark]

My primary concern would be to contain and vent the fumes outside. Some type of metal box with a sealed lid would be good enough for me

Rapidly expanding gasses in a sealed metal container rarely will end well.

 

[justinjja]

Google says 0.6g of lithium per 18650.
so there is something like 720g of lithium.
Burning lithium is 43.1MJ/KG so thats an extra 31 Mega Joules of heat.

I mentioned there would be a hole in the lid with a fan, so no pressure.

 

[spambake]

It won't work. The bottles won't melt. You can actually boil water in paper or plastic containers. Once the water is gone the plastic will release toxic gases.

Maybe sand will work better.

 

[justinjja]

It won't work. The bottles won't melt. You can actually boil water in paper or plastic containers. Once the water is gone the plastic will release toxic gases.

Maybe sand will work better.

Even if they didn't melt, boiling water would expand until the water bottles exploded, same result.
And ya If I have 1200 lithium batteries on fire, I'm not going to be worried about melting a little extra plastic lol.

Leaking is definitely a concern, I know those 1 gallon "milk jug" bottles are really flimsy,
but something like a water bottle or 2 liter soda bottle seem much more durable.

 

[Rider]

I mentioned there would be a hole in the lid with a fan, so no pressure.

Doesn't need to be covered. When the water is instantly vaporized buy the heat, tremendous volume increase and pressure is exerted. Ever see a manhole cover get blown off from an electrical short? And that's in a very open space below ground.

 

[Short_Shot]

Water will work fine if you have a few thousand gallons to down them in.

Use sand. Or better yet don't use lipo cells.

 

[justinjja]

What is so great about sand?
Sand would just get hotter and hotter until the wood in my garage reaches 300c and ignites.

The beauty of water is the phase change, it takes a huge amount of energy to boil water.
And until that water is boiled away, the whole barrel is going to be sitting at 100c.

I think mineral oil would burn if you got it hot enough.

 

[Short_Shot]

What is so great about sand?
Sand would just get hotter and hotter until the wood in my garage reaches 300c and ignites.

The beauty of water is the phase change, it takes a huge amount of energy to boil water.
And until that water is boiled away, the whole barrel is going to be sitting at 100c.

I think mineral oil would burn if you got it hot enough.

Go ahead and Google a lithium pack fire and see how the firefighters put it out.

 

[justinjja]

Go ahead and Google a lithium pack fire and see how the firefighters put it out.

The reason they use a ton of water when an EV catches on fire is because they don't have a giant steel drum they can fill up with water and throw the car into :D

The firefighters are actually trying to achieve what I'm describing here. keeping the batteries in constant contact with water so they stay cool enough to not ignite the rest of the batteries in the car.

 

[time2roll]

Rapidly expanding gasses in a sealed metal container rarely will end well.

Yes sealed so the gas goes out the vent. And don't use a plastic vent.

 

[time2roll]

I could see adding a water inlet and outlet to connect hoses and flood the box.

 

[justinjja]

Math is fun :D

So 32 liters of water can handle something like 300 Mega joules of energy.

I want to stay under 300c, starting point is 25c, specific heat of sand is 830J/kg*c
So solving for kg I would need 1300kg of sand to match the water,

But realistically only need absorb about 100 MJ,
So that drops it down to 400kg of sand.

And then, as long as the barrel isn't touching a wall it could get a lot hotter than 300c before igniting anything the garage.
So maybe as little as 200kg would work.

200kg of sand is 0.122 cubic meters, or a little over half of the 55 gallon drum.
So it is just barely doable

 

[Substrate]

18650 is merely a size, not an indication of chemistry.

Theoretical questions such as these show that you are unsafe and need MUCH more real-world knowledge before you hurt yourself or others.

At the very least, stop thinking of 3.7v nominal 18650's, and stick with 3.2v nominal LFP / LiFeP04 chemistry.

 

[justinjja]

Figured it was pretty self explanatory, but these are 3.6v ncr cells.