Tesla’s 4680 battery supply chain collapses

[BeFamousVideo]

it's like saying, I know it's BS, but I found the smell pleasing

Now that is a funny and true response. That source is garbage.

Sponsored

 

[HaulingAss]

The story is also being covered by other news outlets. There's no need to be limited to one specific take.
Basically it appears that Tesla is not buying the materials needed to make 4680 battery cells, which are currently used in the Cybertruck and potentially in future Tesla vehicles.

Incorrect. Tesla has robust supply chains which typically involves having at least two suppliers for everything. You can tell zilch about how much Tesla is buying by only looking at one supplier's sales without having more information from the other suppliers.

 

[HaulingAss]

Also, you should probably change the title of the thread to something less sensational and far more accurate. The supply chain didn't collapse, so "Tesla Suppliers' exclusive contract to supply cathodes ends, with very little revenue" would be more appropriate.

But it's Electrek, they have to try to make it sound disastrous!

 

[HaulingAss]

Wonder when will be able to buy a dry cathode cybertruck? Could the dry battery be much better than current wet 4680?

The main advantage of dry cathode is lower cost which means higher volume sales.

 

[HaulingAss]

No Semi has 4680, yet. What you are quoting is the roadmap. The long range Semis delivered use 2170s and the Cathode plant isn't really relevant to the deal discussed. That's where Elon says he wishes they wouldn't have gone with the dry cathode.

Tesla was testing early Semi prototypes in 2020 with 4680 cells but, due to 4680 production challenges, switched to 2170 cells in late 2022 for pilot production. They are still planning to use 4680 for regular production in 2026.

 

[Paul Scott]

The answer depends on which very into the truck. The long range does use the 4680s and the short range uses LFP.

Key Battery Details:

• Capacity:
  Long-Range: ~850 kWh (using 4680 cells).
  
  Short-Range: ~500 kWh (using LFP cells).
  
  
• Cell Technology:
  Primarily uses next-gen 4680 battery cells for high energy density, though LFP (Lithium Iron Phosphate) cells are used in the shorter-range version for cost and durability.
  
  
• • Efficiency:
    Extremely efficient, consuming less than 2 kWh per mile, enabling 500-mile range.
  • Charging:
    Supports Tesla's Megachargers for fast charging, adding up to 70% range in about 30 minutes.
  • Design:
    The large battery pack is integrated structurally, keeping the center of gravity low for stability, even when fully loaded.

No Semi has 4680, yet. What you are quoting is the roadmap. The long range Semis delivered use 2170s and the Cathode plant isn't really relevant to the deal discussed. That's where Elon says he wishes they wouldn't have gone with the dry cathode.

I wasn’t just looking at the roadmap I was talking to an engineer that works for Tesla on the semi. The original prototype had two stacked 100kw packs and late the shortage went to a custom 2170 pack and LR would use 4680 for additional energy density. Things can change but I don’t think the 10Q ever mentioned semi specifically. At battery day which I attended Elon stated the long range semi would use 4680.

 

[SCTesla]

I wasn’t just looking at the roadmap I was talking to an engineer that works for Tesla on the semi. The original prototype had two stacked 100kw packs and late the shortage went to a custom 2170 pack and LR would use 4680 for additional energy density. Things can change but I don’t think the 10Q ever mentioned semi specifically. At battery day which I attended Elon stated the long range semi would use 4680.

That's the plan on paper, but currently the 4680 isn't where they want it, yet. All Semis are still using 2170s for the long range Semi.

It's not due to shortage, the 4680 that Tesla makes, despite being twice as big has only slightly more energy than the 2170s. The thought is that Panasonic's 4680 will meet the needs of the Semi.

 

[Outdoors]

Stated at "battery day". Planning on. The thought is. More hopium.

Please tell me one announcement that has come true since that day in September of 2020.

So called move to dry cathode is nothing. Why isn't it in the cars. Que more: Plan is, thought is, makes sense if they. Tesla is going to do this as I talked to an Engineer.

 

[Paul Scott]

Stated at "battery day". Planning on. The thought is. More hopium.

Please tell me one announcement that has come true since that day in September of 2020.

So called move to dry cathode is nothing. Why isn't it in the cars. Que more: Plan is, thought is, makes sense if they. Tesla is going to do this as I talked to an Engineer.

The reason 16 550 is still used in SNX is due to contracts and the number of cells they agreed to purchase sometimes it’s not about the technology. It’s about the agreements that are made at the time.

 

[Outdoors]

The reason 16 550 is still used in SNX is due to contracts and the number of cells they agreed to purchase sometimes it’s not about the technology. It’s about the agreements that are made at the time.

Sure. The Panasonic product on the 18650 keeps getting better does more and more each revision.

The S&X don't even move the needle a bit. So even if you combine all of those with the CT it again doesn't move the needle. Where do all these awesome cells go?

Still waiting on the Battery day announcements that came true.

 

[Paul Scott]

Sure. The Panasonic product on the 18650 keeps getting better does more and more each revision.

The S&X don't even move the needle a bit. So even if you combine all of those with the CT it again doesn't move the needle. Where do all these awesome cells go?

Still waiting on the Battery day announcements that came true.

• 4680 Cell & Tabless Design: The goal was a larger cell (46mm x 80mm) with a tabless design for better power/range (16% range boost).
  • Progress: These cells are being produced and integrated into vehicles, though ramping up production has been challenging.
• Cobalt Reduction: Moving to high-nickel, low-cobalt cathodes for sustainability and cost.
  • Progress: Tesla is using more nickel-rich chemistries, reducing reliance on cobalt.
• Dry Battery Electrode (DBE) Manufacturing: A new process to eliminate solvents, saving cost, time, and increasing capacity.
  • Progress: Tesla acquired Maxwell Technologies for this technology; it's being implemented in their new cathode plant.
• In-House Lithium & Materialslans for direct lithium extraction from brine and high-capacity silicon anodes.
  • Progress: Tesla is exploring new lithium extraction methods and using silicon in anodes, boosting energy density.
• Cost Reduction Goal: Aimed for a 56% reduction in battery cost per kWh, enabling a $25,000 Tesla.
  • Progress: While costs have dropped, achieving the $25k car and massive cost-per-kWh goals is an ongoing, long-term effort.
• Model S Plaid: Unveiled with <2s 0-60 mph, 200+ mph top speed, and >520-mile range, featuring new powertrain tech.
  • Progress: The Plaid Model S was released with these specs, delivering on that performance promise.

In essence, many Battery Day concepts are being phased in, driving improvements, but the most ambitious cost and production targets are long-term goals still in progress

 

[YDR37]

• Model S Plaid: Unveiled with <2s 0-60 mph, 200+ mph top speed, and >520-mile range, featuring new powertrain tech.
  • Progress: The Plaid Model S was released with these specs, delivering on that performance promise

According to Tesla, the Model S Plaid does 0-60 mph in 1.99 seconds "with rollout subtracted". So OK, it meets the "<2s 0-60 mph" spec as claimed on Battery Day.

The Model S Plaid also has a top speed of 200 mph when equipped with the Track Package. So OK, it meets the "200+ mph top speed" spec.

But it doesn't meet the ">520-mil range" spec. Tesla's estimated range for the Model S Plaid is only 368 miles (or lower, depending on the wheels). Even the non-Plaid Model S only has an estimated range of 410 miles.

 

[Outdoors]

• 4680 Cell & Tabless Design: The goal was a larger cell (46mm x 80mm) with a tabless design for better power/range (16% range boost).
  • Progress: These cells are being produced and integrated into vehicles, though ramping up production has been challenging.
• Cobalt Reduction: Moving to high-nickel, low-cobalt cathodes for sustainability and cost.
  • Progress: Tesla is using more nickel-rich chemistries, reducing reliance on cobalt.
• Dry Battery Electrode (DBE) Manufacturing: A new process to eliminate solvents, saving cost, time, and increasing capacity.
  • Progress: Tesla acquired Maxwell Technologies for this technology; it's being implemented in their new cathode plant.
• In-House Lithium & Materialslans for direct lithium extraction from brine and high-capacity silicon anodes.
  • Progress: Tesla is exploring new lithium extraction methods and using silicon in anodes, boosting energy density.
• Cost Reduction Goal: Aimed for a 56% reduction in battery cost per kWh, enabling a $25,000 Tesla.
  • Progress: While costs have dropped, achieving the $25k car and massive cost-per-kWh goals is an ongoing, long-term effort.
• Model S Plaid: Unveiled with <2s 0-60 mph, 200+ mph top speed, and >520-mile range, featuring new powertrain tech.
  • Progress: The Plaid Model S was released with these specs, delivering on that performance promise.

In essence, many Battery Day concepts are being phased in, driving improvements, but the most ambitious cost and production targets are long-term goals still in progress

None of these did anything however.

And Maxwell technologies was bought in 2019 well before battery day happened.

Costs haven't produced a 25k car with 4680

No Model S plaid with 4680

In house is a joke compared to the cells they purchase.

Reduction of nickel why does it still need to use the 18650 with tons of nickel. Because the plaid needs high performance cells that allow high discharge rates and high charging rates beyond 80%.

Tesla cells are a joke. And so is the hype surrounding the cells.

So again where are the 4680's going? In the CT.

 

[henchman24]

That's the plan on paper, but currently the 4680 isn't where they want it, yet. All Semis are still using 2170s for the long range Semi.

It's not due to shortage, the 4680 that Tesla makes, despite being twice as big has only slightly more energy than the 2170s. The thought is that Panasonic's 4680 will meet the needs of the Semi.

This is incorrect in a few ways. First Gen 2 4680s have ~5x the energy of 2170s. This is a function of simply having a much larger format cell. Stating slight is a mischaracterization by a fairly large degree. Their energy densities are very close though. Gen 2 4680s are around ~285wh/kg and current Panasonic 2170s used in Semi/3/Y are around ~270wh/kg. The new 973 chemistry cells (kind of a gen 2.5 4680) are rumored over 290wh/kg, but no teardowns have verified this. The theoretical max is around 330, though I doubt we get over 310-315.

When it comes to Panasonic vs Tesla 4680s, Tesla cells will have a higher capacity, density, and energy output. The Gen 2 Tesla cells are over 25ah... very close to 26ah. The new chemistry cells are rumored low 26s (again, unconfirmed externally). Panasonic cells, we have statements 'over 20ah' and rumors have it around 23-24ah. The LG 4680s are just over 23ah.

The issue with Tesla 4680 cells is scrap rate, not performance, capacity, etc. And this gets into a key difference in strategies. Tesla is trying to incorporate a few technologies and improvements, while LG and Panasonic are utilizing tried and true methods. Tesla has changed the can design dramatically from 2170s, it is lighter, thinner, connected differently, etc. LG and Panasonic incorporate tabless design that is similar to tabless 2170s mass produced around the world with a thicker and more traditional can. This hurts density and total capacity a touch, but the failure rate is very low. What is the more dramatic difference is the chemistry change. Gen 2 Tesla cells were 955 NMC... meaning ~90% Nickel 5% Magnesium 5% Cobalt. The newest cells have moved to 973. Though a bit rougher with the numbers as Nickel is closer to 91% and Cobalt is under 3%. Panasonic and LG use different chemistries. Panasonic uses NCA and LG uses NCMA for their 4680s. Both are using around 80% Nickel (80-83%). The utilization of less Nickel simply lowers the capacity (in these chemistries, rule of thumb is more nickel, more capacity)... but since the chemistries utilized there are tried and true, they are very easy to produce and don't have much scrap.

Up until Tesla was producing the 4680s, nobody had mass produced 955 chemistry and same goes for 973. These chemistry changes cause issues. The 2170s had similar issues when they went to 811 chemistry (and similar NCMA), but after a few years, 80% nickel mixes are widely used now. The same will happen with 90% nickel mixes.

 

[CyberTrk]

This is incorrect in a few ways. First Gen 2 4680s have ~5x the energy of 2170s. This is a function of simply having a much larger format cell. Stating slight is a mischaracterization by a fairly large degree. Their energy densities are very close though. Gen 2 4680s are around ~285wh/kg and current Panasonic 2170s used in Semi/3/Y are around ~270wh/kg. The new 973 chemistry cells (kind of a gen 2.5 4680) are rumored over 290wh/kg, but no teardowns have verified this. The theoretical max is around 330, though I doubt we get over 310-315.

When it comes to Panasonic vs Tesla 4680s, Tesla cells will have a higher capacity, density, and energy output. The Gen 2 Tesla cells are over 25ah... very close to 26ah. The new chemistry cells are rumored low 26s (again, unconfirmed externally). Panasonic cells, we have statements 'over 20ah' and rumors have it around 23-24ah. The LG 4680s are just over 23ah.

The issue with Tesla 4680 cells is scrap rate, not performance, capacity, etc. And this gets into a key difference in strategies. Tesla is trying to incorporate a few technologies and improvements, while LG and Panasonic are utilizing tried and true methods. Tesla has changed the can design dramatically from 2170s, it is lighter, thinner, connected differently, etc. LG and Panasonic incorporate tabless design that is similar to tabless 2170s mass produced around the world with a thicker and more traditional can. This hurts density and total capacity a touch, but the failure rate is very low. What is the more dramatic difference is the chemistry change. Gen 2 Tesla cells were 955 NMC... meaning ~90% Nickel 5% Magnesium 5% Cobalt. The newest cells have moved to 973. Though a bit rougher with the numbers as Nickel is closer to 91% and Cobalt is under 3%. Panasonic and LG use different chemistries. Panasonic uses NCA and LG uses NCMA for their 4680s. Both are using around 80% Nickel (80-83%). The utilization of less Nickel simply lowers the capacity (in these chemistries, rule of thumb is more nickel, more capacity)... but since the chemistries utilized there are tried and true, they are very easy to produce and don't have much scrap.

Up until Tesla was producing the 4680s, nobody had mass produced 955 chemistry and same goes for 973. These chemistry changes cause issues. The 2170s had similar issues when they went to 811 chemistry (and similar NCMA), but after a few years, 80% nickel mixes are widely used now. The same will happen with 90% nickel mixes.

Quite an outstanding comment, I wonder how you come out with such detail information?

My otherwise source been basically watching Jordan Giesige on The Limiting Factor Youtube channel...

I was wondering if you may have any scepticism about the expected large volume production of the Rivian R2 using 4695 lithium-ion battery cells by opposition of the well established R1 using 2170 cells?

Thinking about Tesla previous short attempt of using 4680 for the Model Y...

Sponsored