Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tech Improvements Unveiled on Battery Day

This site may earn commission on affiliate links.

Cosmacelf

Well-Known Member
Supporting Member
Mar 6, 2013
12,642
46,693
San Diego
I'm using this post/thread to summarize all the improvements Tesla announced on Battery Day to get to a total 56% $/Kwh reduction. You can watch the full Tesla video at the link at the bottom.

1. Cell Design - Tabless Cell

From a 21mm x 70mm size to a 46mm x 80mm

This gives them an immediate 17% cost reduction just from the bigger size, and a 7% energy density improvement. But they couldn't have built this bigger cell without some sort of heat dissipation improvement since the bigger cell would have heated up too quickly and reduced total power output (meaning much slower super charging times).

The tabless cell is what gives them a huge improvement in heat dissipation. Electrical path length goes from 250mm to 50mm even though the cell is much bigger. While they showed a picture of a tabless cell top, there was much they didn't show or explain.

End result is that the bigger tabless cell gives you 4x energy, +16% range, and 6x power of the 2170. However note that this new cell is 5.4x the volume of the old 2170, so what is really going on here? Why only 4x the energy, not 5.4x the energy? They ended up with a lower volumetric or gravimetric (which one wasn't stated) energy density? This might have to do with further below when Elon said the casing was now steel rather than aluminum. They most likely used steel because it is now a structural element (see last section).

The summary chart on this innovation shows a 14% overall $/kWh reduction.

2. Factory - Process and Machine Innovations

The big one here is the dry electrode process they got from Maxwell (and have been refining ever since). 10x reduction in footprint, 10x reduction in energy use. Close to working. Works at low yield, needs more machine designs to get high yield.

They increased line efficiency to get 7x output per line.

Cell formation improvements. Sounds like instead of charging each cell individually, they do them in batches plus other innovations. Ends up with an 86% reduction in formation equipment investment and a 75% formation footprint reduction.

Overall these factory improvements give you 75% investment reduction per GWh and a 10x smaller footprint per GWh.

So, on a battery pack basis, this results in a 18% $/kWh reduction.

3. Silicon Anode

Silicon is cheaper than highly processed graphite, and can have a theoretical 10x increase in energy density.

There are three known industry methods to get silicon into the anode. Silicon structured in SIO glass which they said costs $6.6/kWh for the anode. Silicon structured in graphite, costing $10.2/kWh, and Silicon nanowires, costing more than $100/kWh.

So Tesla came up with their own idea which only costs $1.2/kWh. Use raw silicon metal, powder it, encase it in an ion conducting polymer and use a highly elastic binder to account for its expansion when being charged.

This ends up giving you 20% longer range and it is cheaper than what Tesla currently uses. So, far cry from a theoretical 10x energy density improvement, but you still get a bump up from current anodes AND it is cheaper. In the end, Tesla cut through the BS, didn't go for the holy grail for silicon anodes, and went with a good enough improvement over what they have now. Classic Tesla.

OK, so this ends up giving you a 5% overall cost reduction in terms of battery pack $/kWh reduction.

4. Cathode

Tesla has made a cobalt free, high nickel cathode using novel coatings and dopants that gives you a 15% reduction in cathode $/kWh (since cobalt is much more expensive than nickel and the other ingredients).

They also going to use three different kinds of cathodes.

Iron based cathodes for stationary storage and cheaper or low range cars. Nickel + Manganese (2/3 nickel, 1/3 manganese) for long range cars, and high nickel for long range mass sensitive vehicles like the Cybertruck and the Semi.

FYI, a pure nickel cathode versus iron cathode, nickel is 100% better (2x) than iron, but at the pack level, it is only 50% to 60% better.

Cathode process improvements. Instead of using metal sulfates from the mines, they are getting raw metal now. That alone gives them a 66% reduction in investment, 76% reduction in process cost.

For lithium (also part of the cathode), they will again skip the sulfate process step, and use their own processing to take raw lithium ore, process it the way they want to extract the lithium metal. This will results in a 33% reduction in lithium cost. This is a Tesla invented process that uses sodium chloride (ie table salt) to extract lithium from the ore. Wow.

Also, they obtained mineral rights for a 10,000 acre lithium clay deposit in Nevada. Tesla will mine their own lithium.

All this results in a 12% $/kWh reduction.

5. Using cells as a structural element

Use a single piece casting for front body and a single piece for rear body. Had to create own aluminum alloy for this. The reason for this (other than cost) is that they can now use the battery cells themselves as a structural element. Current battery uses a filler between cells that is a flame retardant. New battery architecture uses a filler than is a structural adhesive as well as a flame retardant. This glues the cell to the top and bottom sheets and allows for sheer force transfer. Ends up with a very stiff structure - stiffer than what they have now, which will be nice for driving dynamics. Elon used the word "steel shell case" to describe the cells, so I guess they use steel (as opposed to lighter aluminum) as the casing material for structural reasons.

You end up with three body pieces. A front casting, a rear casting and joining them together is the battery case structure (which is also mostly a casting).

Bottom line is that these innovations give you 10% mass reduction, 14% range increase "opportunity", and 370(!) fewer parts.

So this innovation will result in 7% in a total battery pack cost $/kWh reduction.

Add all those together to get a 56% battery cost $/kWh reduction, which is HUGE.

Finally, note that Elon said there were other improvements that they didn't discuss as well as details that they are keeping under wraps ("secret sauce").

It'll be probably three years to fully realize all these cost savings. Will start using these savings in 1.5 years.

 
Great summary! This slide complements what you summarized.

Vert Int.JPG
 
I was unsure exactly what that slide said. Does it mean you get a range increase AND a cost decrease together? I guess it does, which is even more mind blowing.

Indeed mind blowing.

A 54% range increase with a 56% cost per kWh battery (pack) reduction and a 69% reduction on capital investment per GWH. If we assume each horizontal line is $20 / kWh in the below graph, Tesla is anticipating their battery (pack) cost in 2025 will be less than $50 / kWh.

Less than 50 kwh.JPG
 
I'm using this post/thread to summarize all the improvements Tesla announced on Battery Day to get to a total 56% $/Kwh reduction. You can watch the full Tesla video at the link at the bottom.

1. Cell Design - Tabless Cell

From a 21mm x 70mm size to a 46mm x 80mm

This gives them an immediate 17% cost reduction just from the bigger size, and a 7% energy density improvement. But they couldn't have built this bigger cell without some sort of heat dissipation improvement since the bigger cell would have heated up too quickly and reduced total power output (meaning much slower super charging times).

The tabless cell is what gives them a huge improvement in heat dissipation. Electrical path length goes from 250mm to 50mm even though the cell is much bigger. While they showed a picture of a tabless cell top, there was much they didn't show or explain.

End result is that the bigger tabless cell gives you 4x energy, +16% range, and 6x power of the 2170. However note that this new cell is 5.4x the volume of the old 2170, so what is really going on here? Why only 4x the energy, not 5.4x the energy? They ended up with a lower volumetric or gravimetric (which one wasn't stated) energy density? This might have to do with further below when Elon said the casing was now steel rather than aluminum. They most likely used steel because it is now a structural element (see last section).

The summary chart on this innovation shows a 14% overall $/kWh reduction.

2. Factory - Process and Machine Innovations

The big one here is the dry electrode process they got from Maxwell (and have been refining ever since). 10x reduction in footprint, 10x reduction in energy use. Close to working. Works at low yield, needs more machine designs to get high yield.

They increased line efficiency to get 7x output per line.

Cell formation improvements. Sounds like instead of charging each cell individually, they do them in batches plus other innovations. Ends up with an 86% reduction in formation equipment investment and a 75% formation footprint reduction.

Overall these factory improvements give you 75% investment reduction per GWh and a 10x smaller footprint per GWh.

So, on a battery pack basis, this results in a 18% $/kWh reduction.

3. Silicon Anode

Silicon is cheaper than highly processed graphite, and can have a theoretical 10x increase in energy density.

There are three known industry methods to get silicon into the anode. Silicon structured in SIO glass which they said costs $6.6/kWh for the anode. Silicon structured in graphite, costing $10.2/kWh, and Silicon nanowires, costing more than $100/kWh.

So Tesla came up with their own idea which only costs $1.2/kWh. Use raw silicon metal, powder it, encase it in an ion conducting polymer and use a highly elastic binder to account for its expansion when being charged.

This ends up giving you 20% longer range and it is cheaper than what Tesla currently uses. So, far cry from a theoretical 10x energy density improvement, but you still get a bump up from current anodes AND it is cheaper. In the end, Tesla cut through the BS, didn't go for the holy grail for silicon anodes, and went with a good enough improvement over what they have now. Classic Tesla.

OK, so this ends up giving you a 5% overall cost reduction in terms of battery pack $/kWh reduction.

4. Cathode

Tesla has made a cobalt free, high nickel cathode using novel coatings and dopants that gives you a 15% reduction in cathode $/kWh (since cobalt is much more expensive than nickel and the other ingredients).

They also going to use three different kinds of cathodes.

Iron based cathodes for stationary storage and cheaper or low range cars. Nickel + Manganese (2/3 nickel, 1/3 manganese) for long range cars, and high nickel for long range mass sensitive vehicles like the Cybertruck and the Semi.

FYI, a pure nickel cathode versus iron cathode, nickel is 100% better (2x) than iron, but at the pack level, it is only 50% to 60% better.

Cathode process improvements. Instead of using metal sulfates from the mines, they are getting raw metal now. That alone gives them a 66% reduction in investment, 76% reduction in process cost.

For lithium (also part of the cathode), they will again skip the sulfate process step, and use their own processing to take raw lithium ore, process it the way they want to extract the lithium metal. This will results in a 33% reduction in lithium cost. This is a Tesla invented process that uses sodium chloride (ie table salt) to extract lithium from the ore. Wow.

Also, they obtained mineral rights for a 10,000 acre lithium clay deposit in Nevada. Tesla will mine their own lithium.

All this results in a 12% $/kWh reduction.

5. Using cells as a structural element

Use a single piece casting for front body and a single piece for rear body. Had to create own aluminum alloy for this. The reason for this (other than cost) is that they can now use the battery cells themselves as a structural element. Current battery uses a filler between cells that is a flame retardant. New battery architecture uses a filler than is a structural adhesive as well as a flame retardant. This glues the cell to the top and bottom sheets and allows for sheer force transfer. Ends up with a very stiff structure - stiffer than what they have now, which will be nice for driving dynamics. Elon used the word "steel shell case" to describe the cells, so I guess they use steel (as opposed to lighter aluminum) as the casing material for structural reasons.

You end up with three body pieces. A front casting, a rear casting and joining them together is the battery case structure (which is also mostly a casting).

Bottom line is that these innovations give you 10% mass reduction, 14% range increase "opportunity", and 370(!) fewer parts.

So this innovation will result in 7% in a total battery pack cost $/kWh reduction.

Add all those together to get a 56% battery cost $/kWh reduction, which is HUGE.

Finally, note that Elon said there were other improvements that they didn't discuss as well as details that they are keeping under wraps ("secret sauce").

It'll be probably three years to fully realize all these cost savings. Will start using these savings in 1.5 years.

Dude, you nailed it! I think you covered everything. I was going to rewatch the presentation and try and break it down for myself. I owe ya a beer. Maybe a pitcher.
 
This might have to do with further below when Elon said the casing was now steel rather than aluminum. They most likely used steel because it is now a structural element

<snip>

5. Using cells as a structural element

<snip>
New battery architecture uses a filler than is a structural adhesive as well as a flame retardant. This glues the cell to the top and bottom sheets and allows for sheer force transfer. Ends up with a very stiff structure - stiffer than what they have now, which will be nice for driving dynamics. Elon used the word "steel shell case" to describe the cells, so I guess they use steel (as opposed to lighter aluminum) as the casing material for structural reasons

While not as strong aluminum conducts heat better than steel. Current pack cooling uses inter-cell liquid channels.

The new pack cooling may cool from the top/bottom of the cell vs inter-cell as is the case for the aluminum cased cells today.

Reason : with steel case, the cell will transmit heat less to adjacent cells, and this will make heat have an easier path to transmit the thermal energy to the cell top/bottom.
 
  • Informative
Reactions: navguy12
While not as strong aluminum conducts heat better than steel. Current pack cooling uses inter-cell liquid channels.

The new pack cooling may cool from the top/bottom of the cell vs inter-cell as is the case for the aluminum cased cells today.

Reason : with steel case, the cell will transmit heat less to adjacent cells, and this will make heat have an easier path to transmit the thermal energy to the cell top/bottom.

Yes, while they didn't say it, the whole point of a tabless cell is to wick heat from the top and bottom (or just one will work) instead of the sides. Other EVs already do this, so it isn't a surprise that Tesla is finally doing it too. Telsa just had to figure out how to do it with a cylindrical cell.

BTW, as Sandy Munro just said in a livestream, the reason Tesla uses cylindrical instead of prismatic or pouch like others do is manufacturing cost. When Tesla started in 2008 with the Roadster, cylindrical cell cost was lower than the pouch cells that other EV manufacturers were using. In theory, you can optimize pouch cells to make them as cheap or even cheaper than cylindrical, but, unlike the other OEMs, Tesla couldn't wait for "eventually", and decided to use the cheapest at the time, which was cylindrical.

Tesla finally has gotten around to doing the final bit of vertical integration they had avoided all this time and that was battery cells. I don't know why it took them 12 years to get to this point. Was it that they didn't have to until now? They didn't have the IP yet? Who knows, but there is no doubt that other OEMs are probably wondering what hit them since none of them were even thinking of doing the kind of vertical integration that Tesla has outlined.

Specifically, buying different feedstock from the mines is unique to Tesla. Building their own cathode plant is unique to Tesla. Buying mineral rights and doing their own mining is unique to Tesla. Tesla's silicon anode is proprietary. Building their own machines to go into the cell manufacturing line is unique to Tesla (they bought Grohmann and Hibar companies to do this). The dry electrode process is unique to Tesla. Huge capex saving there. And finally the battery cell to car integration, getting rid of the battery pack and using the cells as structural is unique to Tesla.

As the Tesla Daily podcast said, Tesla just gave us a roadmap to how they will become the biggest company in the world by 2030.
 
Who knows, but there is no doubt that other OEMs are probably wondering what hit them since none of them were even thinking of doing the kind of vertical integration that Tesla has outlined.

I just want to point out, Ford used to have this kind of vertical integration but moved away from it. At one point in time they mined, transported, mixed, formed, assembled.. everything except the actual sale. There are important efficiencies to hiring someone else to make things for you. Do you think GM is leaving any fractions of a penny on the table?

Contrary to popular opinion, there are intelligent people at other car companies.
 
Yes, while they didn't say it, the whole point of a tabless cell is to wick heat from the top and bottom (or just one will work) instead of the sides. Other EVs already do this, so it isn't a surprise that Tesla is finally doing it too. Telsa just had to figure out how to do it with a cylindrical cell.

BTW, as Sandy Munro just said in a livestream, the reason Tesla uses cylindrical instead of prismatic or pouch like others do is manufacturing cost. When Tesla started in 2008 with the Roadster, cylindrical cell cost was lower than the pouch cells that other EV manufacturers were using. In theory, you can optimize pouch cells to make them as cheap or even cheaper than cylindrical, but, unlike the other OEMs, Tesla couldn't wait for "eventually", and decided to use the cheapest at the time, which was cylindrical.
Yes, but is it still the cheapest way to go? Seems like making and using a can is expensive.
 
The Chevy Bolt was an LG drivetrain, infotainment car assembled in US by GM, sold in a few tens of thousands and discounted 10K to get it off dealer lots. Yeah, GMLG or LGGM is really good, we're all worried.
Don't discount GM (or VW's) desire to improve their EVs. They aren't going to sit back and loose market share to Tesla. They have their own R&D programs and will be a force in the market. Ford and BMW? Maybe not so much....
 
I just want to point out, Ford used to have this kind of vertical integration but moved away from it. At one point in time they mined, transported, mixed, formed, assembled.. everything except the actual sale. There are important efficiencies to hiring someone else to make things for you. Do you think GM is leaving any fractions of a penny on the table?

Contrary to popular opinion, there are intelligent people at other car companies.
It's the cycle bundling the whole manufacturing pipeline saves a lot of money, then unbundling it or rather rebundling it along different axes, saves even more. Most of the other car companies are currently in the rebundled stage where they have a whole supply chain that makes them a few steps removed from the place where innovation happens. Tesla's innovating in ways that the other car companies unfortunately can't easily because of the way their supply chain is structured. Ford when it started did the same thing, innovating in ways that it probably can't now.
 
  • Like
Reactions: navguy12
Yes, but is it still the cheapest way to go? Seems like making and using a can is expensive.

I suspect Tesla's current battery pack and a GM or VW or BMW battery pack costs at the pack level are similar if for no other reason than Tesla has been working on optimizing the pack architecture for a lot longer than the others. But yesterday's announcement was big since the can is now part of the car structure meaning you don't have to use metal elsewhere for the car structure. Only Tesla can pull off that optimization since they can spec out the battery cell to match how they're building the car.

How aspirational is it versus how real? Don't know.
 
I just want to point out, Ford used to have this kind of vertical integration but moved away from it. At one point in time they mined, transported, mixed, formed, assembled.. everything except the actual sale. There are important efficiencies to hiring someone else to make things for you. Do you think GM is leaving any fractions of a penny on the table?

Contrary to popular opinion, there are intelligent people at other car companies.

BYD started as a rechargeable battery manufacturer. Now it makes vehicles.

Companies that are just going to outsource batteries and cells are just waiting to be eaten by an aspirational supplier.
The battery is the most important component.
 
  • Like
Reactions: navguy12
I love my 2018 Model 3 AWD and was thinking about buying a Model Y for my wife next year. But now with all the improvements coming up, we are going to delay our next Tesla purchase. If we wait a year or two we can get a much better Model Y for the same $63K. In the meantime, my wife will keep driving her 2016 Toyota Sienna :)
 
  • Helpful
Reactions: Ron Burgundy
I love my 2018 Model 3 AWD and was thinking about buying a Model Y for my wife next year. But now with all the improvements coming up, we are going to delay our next Tesla purchase. If we wait a year or two we can get a much better Model Y for the same $63K. In the meantime, my wife will keep driving her 2016 Toyota Sienna :)

There will always be a better Tesla vehicle six months from now. Its how they roll. I'm on my second Model X because they had updated it sufficiently to make it worthwhile. No reason to wait, you can upgrade later.
 
  • Like
Reactions: navguy12
Great summary, thanks.

What I missed in Teslas presentation (or didn't get it) was actual figures about current price per kWh
and planned capacity for the plaid Model S.

Those graphs without units on one or both axis are more or less sales presentations/castles in the sky.
Not much worth, 50% of what...?

It seems Tesla has some great ideas (and is still far ahead of copmpetition), but the battery day had
a bit more entertainment character than being a "facts day" (->see what we have now!).
And when I hear that FSD is totally rewritten "and will work definitely" I know why I don't buy that feature.

Therefor it was a nice show and Tesla has great plans with it's terrawatt factories.
But all in all it was not overwhelming.
 
  • Informative
Reactions: TSleuths
Great summary, thanks.

What I missed in Teslas presentation (or didn't get it) was actual figures about current price per kWh
and planned capacity for the plaid Model S.

Those graphs without units on one or both axis are more or less sales presentations/castles in the sky.
Not much worth, 50% of what...?

It seems Tesla has some great ideas (and is still far ahead of copmpetition), but the battery day had
a bit more entertainment character than being a "facts day" (->see what we have now!).
And when I hear that FSD is totally rewritten "and will work definitely" I know why I don't buy that feature.

Therefor it was a nice show and Tesla has great plans with it's terrawatt factories.
But all in all it was not overwhelming.
I just watched the 90 minute post battery day online chit chat that had Sandy Munro as one of the talking heads.


He stated that the Model Y his team tore down and reverse engineered came in at $108.00 per kWh (pack level).
 
What I missed in Teslas presentation (or didn't get it) was actual figures about current price per kWh and planned capacity for the plaid Model S.

Those graphs without units on one or both axis are more or less sales presentations/castles in the sky. Not much worth, 50% of what...?

Correct. It was a presentation to show how Tesla is leaving no stone unturned in their efforts to reduce battery cost. It was also a roadmap to show Tesla will be one of the biggest companies by sales in ten years. Like all corporate roadmaps for general consumption they left out important information that could have helped direct competitors. To a large degree the effect the presentation has on you depends on how much you believe Tesla can achieve what they say they will achieve.

It seems Tesla has some great ideas (and is still far ahead of competition), but the battery day had a bit more entertainment character than being a "facts day" (->see what we have now!).

I think that's a bit harsh. Maybe Elon is too much of an engineer, but to an engineer like me, I ate it up. He presented things that no one thought they were working on (silicon anode, cobalt elimination, structural pack, direct metal mine offtake) and explained the things people did think they were working on well (dry electrode process and elimination of drying ovens). I (dimly) understand how hard it is to do each of the innovations Elon described, so if Tesla achieves most of what was shown, it will leave the competition even more in the dust than they are now.

And when I hear that FSD is totally rewritten "and will work definitely" I know why I don't buy that feature.

Elon totally got over his skis on FSD. I am looking forward to this new update due by the end of the year or so because there are very real enhancements. But I can understand the skepticism.

Therefore it was a nice show and Tesla has great plans with it's terrawatt factories. But all in all it was not overwhelming.

Well, not immediately overwhelming, and the stock market agrees with you. But it is a nice roadmap to continued EV domination.