4680 batteries, what is the advantage?

[tangible1]

Why are you assuming Battery Day was not grounded in reality? Do you think Tesla was making up the numbers they presented in Battery Day? If so, what is your evidence for this? Because such an assumption would not be in keeping historically with Tesla's presentations of the past.

Dude, I said wishful thinking.
Most postings take a tack specific to their own interests and expounds.
Suddenly every xxx model is getting 500mi range - because Battery Day said so, by their 'math'.
Enough already.

We all find out when Tesla talks. And they're not as yet, but will be soon.
In the meantime, shut out the whining of the blogs. Not of value at this point.

 

[MontyFloyd]

Good. At that time, the model 3 LR had about 325 miles EPA range. A 54% increase would yield 500 miles. Can't wait!

May not happen.

We have a Corolla and Civic, both have just over 300 miles per tank (330 typically), lasts almost 4 weeks before need fill up.
If people drive more, then fill up every 2 weeks.

In rough numbers, longer range = more KWh needed = more battery mass.
4680 is more about cheaper production than more energy storage (although it does have a little more), and there is only so much energy any battery can store.
This the crux, the bigger the battery, the more energy to move, the less efficient. More range is huge increase in mass. There needs to be a sweet spot of range / convenience / efficiency.
Conversely in fuel car, the added fuel weight (5 lb / Gal) is trivial for the added miles. Someone could easily double the tankage with a small (trivial?) loss of efficiency from added mass.

Bottom line, a 4680 Tesla may have same range as comparable 2170, but be thousands cheaper.
It is all about what the cars market is.

 

[tangible1]

Why are you assuming Battery Day was not grounded in reality? Do you think Tesla was making up the numbers they presented in Battery Day? If so, what is your evidence for this? Because such an assumption would not be in keeping historically with Tesla's presentations of the past.

It's less that Battery Day wasn't grounded, it's that people do not understand the math or the dynamics of power application, and run off the reservation with their ideas.
They feel it's justified because.... "Battery Day".
I encountered one guy who claimed to have reviewed the presentation 10 times.
And so was 'expert' more than anyone else, absouluely MYLR was gonna be 500 miles, etc etc whatever wish fulfillment was required.

 

[Wikoogle]

Sandy Munro stated that the 4680 will last a gazillion years, and he would know.

Plus, in the video showing off the 4680, tesla played a lyric in the background “i will go a million miles” Seems pretty reasonable to assume that 4680 batteries will last many many decades and for a million miles due to improved heat discharge and the far shorter travel distance for electrons due to the tabless design.

Can anyone tell me a reason why I am wrong to think this?

 

[gt2690b]

Sandy Munro stated that the 4680 will last a gazillion years, and he would know.

Plus, in the video showing off the 4680, tesla played a lyric in the background “i will go a million miles” Seems pretty reasonable to assume that 4680 batteries will last many many decades and for a million miles due to improved heat discharge and the far shorter travel distance for electrons due to the tabless design.

Can anyone tell me a reason why I am wrong to think this?

It's also supposedly using dry battery electrode but I'm surprised there's not more announcements about it

 

[nativewolf]

Still no one has a 4680 battery pack disassembled and tested as far as I know. Love for someone to point me there.

 

[Mengy]

Still no one has a 4680 battery pack disassembled and tested as far as I know. Love for someone to point me there.

Nope, nothing like that exists yet publicly. Hopefully as customers get Austin MY's we can collect more information.

 

[JRP3]

1 million mile battery is a rather vague metric, a 400 mile range pack only needs 2,500 cycles to hit it, and that's already possible with some cell chemistries.

 

[theothertom]

Sandy Munro stated that the 4680 will last a gazillion years, and he would know.

IMO, Sandy does well on construction techniques but is not a chemist. The lifetime of a battery is heavily dependent on battery chemistry, and AFAIK, 4680 uses the same chemistry as 2170.

 

[scottf200]

Aside: Good overview article I just ran across.
Title: What Batteries Is Tesla Using In Its Electric Cars?

What Batteries Are Tesla Using In Its Electric Cars?

Tesla is the world's largest electric car manufacturer and many wonder what batteries the company uses. Let's take a look at what we know.

insideevs.com

 

[MontyFloyd]

Visual size comparison thread.

4680 mock up

I wanted to see what a 4680 battery would look and feel like in real life so I turned on quickly on a lathe. No wonder it is going to be a game changer on energy density. I put it next to common things so you can see the dimensions to compare. Any requests let me know. The only 2170 I have...

teslamotorsclub.com

 

[tangible1]

Advantages of the 4680 battery fall into two categories; production and vehicular use.


The #1 production benefit is the dry electrode process. It is hard to overstate how much better this is over making a slurry with solvents, then having to dry the electrode material and recover the solvents. The dry electrode process which Tesla acquired when they bought Maxwell Technologies is the primary reason why the 4680 production uses 1/10th the area compared to wet electrode battery production.

View attachment 779175


The other production advantage of the 4680 is the "tabless" design, without that tab in the way the automation process is faster and easier.


View attachment 779176

View attachment 779177


The continuous tab design is key to improved vehicular performance as well. The longer and narrower current path through that tiny tab limited the size of the cell. The 4680 continuous tab design can handle the current of a larger cell, and has a better thermal heat dissipation path for the inside layers of the jellroll, as well as generating less heat due to that shorter and wider current path.

Here you can see even though the 4680 has only 5x the energy, it can generate 6x the power.
View attachment 779178

That's a 20% increase in power for the same capacity. This bodes well for for increase performance overall, or at least better performance as lower states of charge.

I hope and expect that Tesla will increase the energy storage for the Model 3 and Model Y by about 10%, to about a 88-90 KWh capacity pack. This means we could easily have a 4680 3 or Y Performance model with 600 HP.

The larger battery reduces part count and allows using the can as a structural member leading to further weight savings.

View attachment 779181


Both the production methods and cell-to-structural-pack nature of the design lead to significant cost savings. I expect 4680 continuous tab designs to emerge a clear winner in the future.

Keep in mind that as they continue to make dry electrode process production gains you can expect to see this form factor with various chemistries like LiFePo, higher silicon, and someday solid state.

View attachment 779184

But Tesla hasn't solved the production connundrum yet. Two years on.....
If they had, the MY would be in full production, the CT wouldn't be delayed, and the SEMI would already be on the road.

And is it any wonder that Panasonic has been reluctant to take on the 4680 as a product?
Not only will they only get marginal sales to Tesla, to fill what Tesla can't do themselves, but because Tesla entirely understands the costs, Panasonic profits will be limited as well.

 

[MontyFloyd]

But Tesla hasn't solved the production connundrum yet. Two years on.....
If they had, the MY would be in full production, the CT wouldn't be delayed, and the SEMI would already be on the road.

What conundrum?
Lets say they have to slow production to increase the yield, then to get the throughput make more battery machines.
In the long run more machines is better anyway.

The only issue the longer ramp up time, but still in lead.

 

[tangible1]

What conundrum?
Lets say they have to slow production to increase the yield, then to get the throughput make more battery machines.
In the long run more machines is better anyway.

The only issue the longer ramp up time, but still in lead.

two years in, and they can't make enough to supply a MYLR production run.
says volumes about the production process being immature at best.
ramping up the number of machines just increases the rate of scrap, raising the net cost. that's a manufacturing no-no.

 

[MontyFloyd]

two years in, and they can't make enough to supply a MYLR production run.
says volumes about the production process being immature at best.
ramping up the number of machines just increases the rate of scrap, raising the net cost. that's a manufacturing no-no.

Again:
Reduce rate to increase yield, reduce scrap.
Increase throughput with more machines.
In mfg there are limits to how fast one can make something.

 

[tangible1]

Again:
Reduce rate to increase yield, reduce scrap.
Increase throughput with more machines.
In mfg there are limits to how fast one can make something.

you might need to change your thumbnail to a flying pig. just kidding
watching the Austin factory build-out daily, Tesla isn't adding more machines yet, and they've got at least 5 lines running now.
which says volumes about their process maturity and costs.

They'll get there, but on Elon time.

 

[nativewolf]

Yes this is actually one of the giant fubars of Tesla. Something like 5 years ago they realized they had "made it". Demand was virtually unlimited and they needed more battery capacity. 5 years ago it was abundantly clear that they could build more 2170 plants, have partners build them, use alternative (lifepo something) chemistry and forms, or look at the problem and build an alternative. They did 2 things, signed agreements for alternative vendor products and then started work on their own battery form.

The 4680 is a really neat idea. It is just scaling very very slowly and much much slower than Tesla had expected. You don't build two new car plants whose key and limiting input is a specialized product unless you have very high confidence in delivering.

Still a bit surprising to me, the delay and manufacturing challenges I mean. When it was clear Berlin was going to use the 4680 I thought that meant they'd solved the issue. Oh well. No CT in 2023 for me.

I don't think there is anything keeping Panosonic from selling 4680 form battery to others. So with demand going crazy I would guess Tesla is not as much in the driver seat as they'd like to be. Ideally they thought they'd have manufacturing first mover advantages but at this rate (if Pano launches next year) it is not much of an advantage. The global demand for batteries will be unmet for at least another 8 years. If (IF) they solve the manufacturability than it is a great product and will see huge demand ..Tesla or others.

 

[Wikoogle]

Does tesla own a patent on the chemical composition inside their 4680 cell? That would prevent panasonic from selling it to others

LFPs are not the only batteries that can last a 100 years. Tesla just this morning published a paper on a nickel based low power battery that can last a 100 years and a million miles…

Tesla Research Partner Publishes Details About 100-Year Battery

Clearly, no one really needs an EV battery that lasts 100 years, but that won't stop Jeff Dahn and his team from being leaders in battery tech and innovation.

insideevs.com

Is there any chance that this chemistry is whats being used in 4680s. It would explain why tesla is pushing 4680 production so hard and planning to move everything over to 4680s in the coming years.

 

[JRP3]

Is there any chance that this chemistry is whats being used in 4680s.

Highly unlikely. They sacrifice energy density for longevity, more suitable for grid storage.

 

[tangible1]

Panasonic refused to commit to their making 4680 cells until it became clear that not just Tesla would want them.
Tesla essentially sought to replace Panasonic as a vendor; the relationship has been strained for years over Giga Nevada investment commitments and cost issues.
So Tesla determined that they'd do it themselves. It's taking a lot longer, as their endless delays on CT, Semi, etc reveal.

Yes this is actually one of the giant fubars of Tesla. Something like 5 years ago they realized they had "made it". Demand was virtually unlimited and they needed more battery capacity. 5 years ago it was abundantly clear that they could build more 2170 plants, have partners build them, use alternative (lifepo something) chemistry and forms, or look at the problem and build an alternative. They did 2 things, signed agreements for alternative vendor products and then started work on their own battery form.

The 4680 is a really neat idea. It is just scaling very very slowly and much much slower than Tesla had expected. You don't build two new car plants whose key and limiting input is a specialized product unless you have very high confidence in delivering.

Still a bit surprising to me, the delay and manufacturing challenges I mean. When it was clear Berlin was going to use the 4680 I thought that meant they'd solved the issue. Oh well. No CT in 2023 for me.

I don't think there is anything keeping Panosonic from selling 4680 form battery to others. So with demand going crazy I would guess Tesla is not as much in the driver seat as they'd like to be. Ideally they thought they'd have manufacturing first mover advantages but at this rate (if Pano launches next year) it is not much of an advantage. The global demand for batteries will be unmet for at least another 8 years. If (IF) they solve the manufacturability than it is a great product and will see huge demand ..Tesla or others.