@@thelimitingfactor Cell/Battery. See Brad Sloan's video from 31 October. Roller type components on the loading platform at 6.37. Seen again at 10.24 being driven away. Cathode. See Joe's 28 October at 14.28. The white and red equipment has been used. Either second hand from elsewhere, or has been removed from inside. I don't see much evidence of anything being produced in commercial quantities yet at the Cathode plant.
@@thelimitingfactor Brad Sloan's video 12 November. Another crate being shipped out from Battery/Cell. Speculating a lot of equipment for Battery was never installed as the processes have changed or matured.
I think by "any alternative" Elon was referring to apples-to-apples comparisons, e.g. 4680 NMC would beat buying NMC cells from outside sources. Hence their current desire to design LFP and other variants of the 4680 cells to compete with those as well.
That all sounds very promising. Though Elon has said they keep relying on existing suppliers. So I can't see them redoing the 3/Y lines with 4680s anytime soon.
Professional and well done! Thank you for making these videos. I would like a video on the ability of Graphene made here in U.S. to replace Chinese graphite in the battery. A U.S. company has developed a way to make graphene from food waste.
Sure thing! Graphene won't replace graphite because it's an additive rather than an energy storing medium. It increases power and reduces heat generation. But, carbon nanotubes seem to be a better option
Great video as usual! Jordan, do you think iterations on in-house 4680 for the foreseeable future is the main path Tesla will use or is there another new option you think they may lean on in the 2030's?
What a great analysis - probably the best I've seen thank you!!! Quick question: 1. why do automotive cells cost more? 2. from a performance perspective how far are we from medium-heavy duty vehicle electrification?
1. Automotive cells need to be high performance and light weight. Storage cells can save a lot of cost by being a bit heavier (energy density) and a bit lower performance (power density, fast charging and cooling capability). Storage cells are often physically a lot bigger (bad for cooling, good for cost). 2. what is a medium heavy duty vehicle to you? probably a bit of a language and cultural barrier here.
Thanks man! They use smaller cells that reduce manufacturing efficiency and they have less stringent requirements all around. Moving things are hard on things 🤠 We're already there for heavy duty (semi) in terms of tech. They just need to scale production
When they start building 4680 lines in Nevada you'll know they're happing with the dry cathode process and are ready to scale. Maybe a cathode plant there as well? Have they started building the battery production lines in Berlin yet?
I'm not sure what you're talking about. There are alternatives that are better in a single or two aspects, but the 4680 is better in multiple aspects. That is, all trade-offs considered, it's the best - if they can scale it... which is what we're waiting on.
I'm not convinced a cylindrical LFP cell makes sense, the volumetric density takes a hit compared to prismatic/pouch. BYD Blade is structural and makes better use of space, fewer larger cells means fewer connections and presumably faster, cheaper pack construction. Am I missing something?
Intuitively what you say about volumetric density makes sense. What I don't know atm is how thermal management is handled in prismatic cells. No show-stopper, I'm sure, but I'm not familiar. I must imagine Tesla has a good reason for continued iteration of cylindrical cells.
@@thelimitingfactor Cooling needs are lower for LFP and even more so in a low performance vehicle like the Robotaxi so I'm not sure cylindrical provides an advantage but you could be correct.
If you require hyper performance speed acceleration then you need to control the energy cells temperature... Most Chinese vehicles don't have hyper performance.
If indeed 272Wh/Kg is the energy density of the standard 4680, can we expect 120Wh/Kg for a Lipo4 version? If so...a 40 KWh pack would suffice at 5.5Miles/KWh to exceed 200 miles of range (40KWh x 5.5KWh/M = 220 Miles). This would be a 350ish cell pack.
AM I MISSING SOMETHING ? :: If LFP from China are $60.00 & NMC $75-80. Why is the IRA PTC added? It is a credit making :: $25 $40-45 : [ Corpus Christie etc the next step ]
Tesla said their figures were after deducting the production tax credit, so in order to get their production cost before the production tax credit, I had to add it back in So, yes, if you want to rededuct it to arrive back at the cost with the tax credit you can
So does this mean cybertruck will eventually reach the 500 mile range as previously estimated ? Or is that just imposible even for dry 4680 since estimates are off by 32ish % ? So they would need 32% better pack density and maybe a bit bigger of a pack to reach that 500 goal ?
I think it's going to be at least 5 years before we see a 500 mile range Cybertruck that achieves that range without a range extender in the bed I think they've fundamentally changed who they've geared the truck towards and the use case that they've geared the truck towards
Wonderful news! So, we can expect Tesla Semi using the 4680 eventually. Seeing the progress in the Semi factory, maybe they can have it ready in time to start the Semi mass production? Oh and I guess that also implies an eventual sleeper cab version could have even more range! 500 miles is enough to go 250 into the middle of nowhere and back, but if they need to go 350 miles into the middle of nowhere it could be possible if they achieved 700mile range for long haul sleeper cab Tesla Semi. But those are my expectations anyways. Also of course, cybercab with LiFePo 4680 sounds great, seems they will put only enough range to make it work and maximize production throughput. On another topic, about nuclear, I have been thinking Elon will need it for the Starship. SMR, MSR reactors seem perfect to fit there and since the fuel can be thorium, it's even better and that can be a permanent energy source additional to solar panels in Mars and Moon colonies. A nuclear powered Starship (having both regular methalox Raptors and nuclear propulsion systems) could go to Mars faster and only take about 2 months instead of 6 months. Elon did say he was interested in making a matter-antimatter propulsion system, but he hasn't mentioned the topic again. Maybe now that Trump won, the political and bureoucratic slowdows can be removed and things can be worked out.
Anyone know the rationale behind making a non prismatic LFP cell? Why would someone sacrifice packing efficiency while using an already low energy density cathode?
I don't think they have any plans for solid state. That's because it's fundamentally more expensive due to the lithium rich separator (rather than a plastic separator) The primary way around that is a much thinner solid state separator. But, thinner often means using slower coating tech that's more precise - which means greater production expense. It'll get there, but they still have a good 5-10 years on the liquid electrolyte roadmap
@@thelimitingfactor Appreciate the reply. Do you have any insight who is leading the race? (QS, Solid Power, Toyota, Panasonic or BYD) $/KwH will certainly be another decade at least
@@thelimitingfactor Is there any tangible indication that Tesla will ever make LFP 4680s? Also, given that in Battery Day Musk said that LFP will be 80% of batteries produced, why hasn't Tesla gone at all that route? The only explanation I can come up with is that there are patents in the way. Patents Chinese have blocking the West (not the other way around!)
China subsidized the cost of all of their industry to a ridiculous level so yeah it only cost us a lower amount but giving china all the production capacity isnt safe.
good details! China went a little too far subsidizing batteries. 200 miles of range and 5.5 miles/kWh? That's the numbers most are waiting for. Looks like it will weigh about 1.5t. Sadly, not a big deal. It could be shorter and have thinner wheels to weigh about 1.3 t without safety issues (that's about to 10x of 2 passengers weight which is optimal for comfortable and efficient ride). BUT looking at changes made to all models through time, it may up its efficiency as more cabs will be made
you can't add inflation. We don't have inflation data for batteries. You can't just apply standard core inflation. Maybe you can guess it my using the price of energy in the extraction/mining sector and energy in general
Pretty sure china is selling LFP for profit, sure there is more profit in a module or pack. Most of the money comes large volume contracts, not 1 or 2 cells for low cost, but millions. Everything tracks with what I've seen on costs and price estimates.
@thelimitingfactor I'm not so sure. The Chinese have a burgeoning population, a mandate for electric vehicles, the manufacturing capacity _and_ absolutely no respect for IP. It's not a race, it's a grind.
For Robotaxi, I could see Jeff Dahn's million mile NMC battery chemistry. Outperforms LFP in energy density somewhat and lifespan is well beyond LFP. "Lorna Casselton Lecture May 2023" on TH-cam.
As of now id expect tarrifs on Chinese cells to grow exponentially, but who knows with Elon's bestie..... 🙄 For sure i wouldn't go pissing off the 800 pound dragon in the room. If they decide to call on trillions of trade deficit January 21st we are well and truly screwed.
@@thelimitingfactor It has no path forward to being able to equal the range to gasoline cars. 160wh is the best it can do, current gen nca's can do 250wh/kg and with silicon 280wh/kg or more. The lfp weight penalty becomes a non starter rally quickly.
This was an awfully good one. Didn't realize automotive cells cost more, but it makes sense.
Glad to hear it 😊 I was rushed on this one because I'm trying to assemble a solar + battery system, lol
They shouldn't cost more though, at all. Last I checked, the same 70-105Ah cells being used in automotive applications are also used in grid storage.
Thank you, Jordan 🙏🏾 - please keep them coming
Thank you for the manual closed captions
Sure thing! You're the first person that's every thanked me!
I'll be the second. Thank you!
Thank you very much for all this research Jordan ❤👍 you are stud of this community 😊👌
😂 You're most welcome
Thanks!
Sure thing man!
🙋♂️JORDAN,GREAT UPDATE ON THE 4680 🔋🔋🔋
Great video !
Thanks man! Hope you had a good Christmas!
Thanks, that was a great one
Glad to hear it! 🤠
Once again, very nice analysis and summary.
Thank you!
17:00 GT Cathode and Cell plants. Also worth noting that equipment has seen to be removed from both Cathode and Cell in recent videos.
Interesting!
@@thelimitingfactor Cell/Battery. See Brad Sloan's video from 31 October.
Roller type components on the loading platform at 6.37. Seen again at 10.24 being driven away.
Cathode. See Joe's 28 October at 14.28. The white and red equipment has been used. Either second hand from elsewhere, or has been removed from inside.
I don't see much evidence of anything being produced in commercial quantities yet at the Cathode plant.
@@thelimitingfactor Brad Sloan's video 12 November. Another crate being shipped out from Battery/Cell.
Speculating a lot of equipment for Battery was never installed as the processes have changed or matured.
I think by "any alternative" Elon was referring to apples-to-apples comparisons, e.g. 4680 NMC would beat buying NMC cells from outside sources. Hence their current desire to design LFP and other variants of the 4680 cells to compete with those as well.
You know Tesla is getting close to the mark when you hear the President of CATL saying that the 4680 won’t be successul
Great video as usual
Nice lineup for upcoming shows
That all sounds very promising.
Though Elon has said they keep relying on existing suppliers. So I can't see them redoing the 3/Y lines with 4680s anytime soon.
This is a great detailed fact-based analysis. Very useful and informative. Thanks.
thanks for the updates
Cheers mate.
Outstanding report, thanks heaps.
4680 LFP for the win :)
How about natrium/sodium?
Amazing work!
Professional and well done! Thank you for making these videos. I would like a video on the ability of Graphene made here in U.S. to replace Chinese graphite in the battery. A U.S. company has developed a way to make graphene from food waste.
Sure thing! Graphene won't replace graphite because it's an additive rather than an energy storing medium. It increases power and reduces heat generation. But, carbon nanotubes seem to be a better option
@@thelimitingfactor Thanks - it is great to hear from an expert!
Great video as usual! Jordan, do you think iterations on in-house 4680 for the foreseeable future is the main path Tesla will use or is there another new option you think they may lean on in the 2030's?
It depends on the country and the specific vehicle. Most new vehicles wil use the 4680 in the U.S.
Thanks mate good content
So what about Panasonic's 4680? where are those cells going? Model Y? Model 3?
No one knows
Yes
What a great analysis - probably the best I've seen thank you!!!
Quick question:
1. why do automotive cells cost more?
2. from a performance perspective how far are we from medium-heavy duty vehicle electrification?
1. Automotive cells need to be high performance and light weight. Storage cells can save a lot of cost by being a bit heavier (energy density) and a bit lower performance (power density, fast charging and cooling capability). Storage cells are often physically a lot bigger (bad for cooling, good for cost).
2. what is a medium heavy duty vehicle to you? probably a bit of a language and cultural barrier here.
Thanks man! They use smaller cells that reduce manufacturing efficiency and they have less stringent requirements all around. Moving things are hard on things 🤠
We're already there for heavy duty (semi) in terms of tech. They just need to scale production
When they start building 4680 lines in Nevada you'll know they're happing with the dry cathode process and are ready to scale. Maybe a cathode plant there as well? Have they started building the battery production lines in Berlin yet?
They make cell parts at Berlin but not the whole cell. Shipping part to the U.S. gets them IRA credits
seems like there are much better alternatives already that will make this extint once introduced. How much longer can they ride with wave?
I'm not sure what you're talking about.
There are alternatives that are better in a single or two aspects, but the 4680 is better in multiple aspects.
That is, all trade-offs considered, it's the best - if they can scale it... which is what we're waiting on.
Why do automotive cells cost more than batteries used for stationary storage?
Tha's been answered in another comment. 🤠
Hoping the workhorse is Natrium based, as it seems like a better alternative.
Not enough energy density
I'm not convinced a cylindrical LFP cell makes sense, the volumetric density takes a hit compared to prismatic/pouch. BYD Blade is structural and makes better use of space, fewer larger cells means fewer connections and presumably faster, cheaper pack construction. Am I missing something?
I covered that in the video. And if you'd like more in depth analysis, check out the '4680 vs Qilin' video
Intuitively what you say about volumetric density makes sense. What I don't know atm is how thermal management is handled in prismatic cells. No show-stopper, I'm sure, but I'm not familiar. I must imagine Tesla has a good reason for continued iteration of cylindrical cells.
@@thelimitingfactor Cooling needs are lower for LFP and even more so in a low performance vehicle like the Robotaxi so I'm not sure cylindrical provides an advantage but you could be correct.
If you require hyper performance speed acceleration then you need to control the energy cells temperature...
Most Chinese vehicles don't have hyper performance.
It's likely a manufacturing thing
If indeed 272Wh/Kg is the energy density of the standard 4680, can we expect 120Wh/Kg for a Lipo4 version?
If so...a 40 KWh pack would suffice at 5.5Miles/KWh to exceed 200 miles of range (40KWh x 5.5KWh/M = 220 Miles). This would be a 350ish cell pack.
AM I MISSING SOMETHING ? :: If LFP from China are $60.00 & NMC $75-80. Why is the IRA PTC added? It is a credit making :: $25 $40-45 : [ Corpus Christie etc the next step ]
Tesla said their figures were after deducting the production tax credit, so in order to get their production cost before the production tax credit, I had to add it back in
So, yes, if you want to rededuct it to arrive back at the cost with the tax credit you can
Is the tax credit for every cell made? I don't even grasp what you are saying...you should work of simplicity, please@@thelimitingfactor
So does this mean cybertruck will eventually reach the 500 mile range as previously estimated ? Or is that just imposible even for dry 4680 since estimates are off by 32ish % ?
So they would need 32% better pack density and maybe a bit bigger of a pack to reach that 500 goal ?
I think it's going to be at least 5 years before we see a 500 mile range Cybertruck that achieves that range without a range extender in the bed
I think they've fundamentally changed who they've geared the truck towards and the use case that they've geared the truck towards
Wonderful news! So, we can expect Tesla Semi using the 4680 eventually. Seeing the progress in the Semi factory, maybe they can have it ready in time to start the Semi mass production? Oh and I guess that also implies an eventual sleeper cab version could have even more range! 500 miles is enough to go 250 into the middle of nowhere and back, but if they need to go 350 miles into the middle of nowhere it could be possible if they achieved 700mile range for long haul sleeper cab Tesla Semi. But those are my expectations anyways.
Also of course, cybercab with LiFePo 4680 sounds great, seems they will put only enough range to make it work and maximize production throughput.
On another topic, about nuclear, I have been thinking Elon will need it for the Starship. SMR, MSR reactors seem perfect to fit there and since the fuel can be thorium, it's even better and that can be a permanent energy source additional to solar panels in Mars and Moon colonies. A nuclear powered Starship (having both regular methalox Raptors and nuclear propulsion systems) could go to Mars faster and only take about 2 months instead of 6 months. Elon did say he was interested in making a matter-antimatter propulsion system, but he hasn't mentioned the topic again. Maybe now that Trump won, the political and bureoucratic slowdows can be removed and things can be worked out.
Did you catch in your radar the new Tesla patent for a cathode with a mix of LFP & NMC?
It could be the workhorse cell for cybercab
x.com/LimitingThe/status/1854536554322256000
Does it make Sense to use 4680 LFP in stationary energy storage ?
In my view, not really
Anyone know the rationale behind making a non prismatic LFP cell? Why would someone sacrifice packing efficiency while using an already low energy density cathode?
I did an entire video on this
Just look up '4680 structural versus byd'
Doesn’t LFP require it to expand and contract? How can they do that with a 4680 form factor?
Not sure what you're talking about. Sounds like you're talking about silicon
I want an Xl version of a cybertruck with 569 miles of range and a solar wrap on it ..... gotta have dreams they say.
💯
How are Tesla's solid state efforts coming along? Understandably a 2030+ effort?
I don't think they have any plans for solid state.
That's because it's fundamentally more expensive due to the lithium rich separator (rather than a plastic separator)
The primary way around that is a much thinner solid state separator. But, thinner often means using slower coating tech that's more precise - which means greater production expense.
It'll get there, but they still have a good 5-10 years on the liquid electrolyte roadmap
@@thelimitingfactor Appreciate the reply. Do you have any insight who is leading the race? (QS, Solid Power, Toyota, Panasonic or BYD) $/KwH will certainly be another decade at least
So there is no visibility on Tesla matching in the Future the LFP prices CATL offers in China *today*.
I'm not sure what you're trying to say
@@thelimitingfactor Is there any tangible indication that Tesla will ever make LFP 4680s? Also, given that in Battery Day Musk said that LFP will be 80% of batteries produced, why hasn't Tesla gone at all that route? The only explanation I can come up with is that there are patents in the way. Patents Chinese have blocking the West (not the other way around!)
China subsidized the cost of all of their industry to a ridiculous level so yeah it only cost us a lower amount but giving china all the production capacity isnt safe.
I wonder why I only see this vid now, while others have reacted 5 days ago.
Because I release my videos a week early for Patreon supporters and X Subscribers
Is it true that the 4680 is produced using equipment from a Korean company?
If we take their word, some of it is
@thelimitingfactor Do you know korean company's name?
good details! China went a little too far subsidizing batteries.
200 miles of range and 5.5 miles/kWh? That's the numbers most are waiting for. Looks like it will weigh about 1.5t. Sadly, not a big deal. It could be shorter and have thinner wheels to weigh about 1.3 t without safety issues (that's about to 10x of 2 passengers weight which is optimal for comfortable and efficient ride). BUT looking at changes made to all models through time, it may up its efficiency as more cabs will be made
I think it'll easily beat 5.5 miles/kWh 🤠
Yeah, 1.5t sounds about right.
Don't forget adding inflation to old cost targets
you can't add inflation. We don't have inflation data for batteries.
You can't just apply standard core inflation.
Maybe you can guess it my using the price of energy in the extraction/mining sector and energy in general
Pretty sure china is selling LFP for profit, sure there is more profit in a module or pack. Most of the money comes large volume contracts, not 1 or 2 cells for low cost, but millions.
Everything tracks with what I've seen on costs and price estimates.
Not according to people I've talked to in the industry. This is common practice in China. Devour the market by selling below cost.
@thelimitingfactor I'm not so sure. The Chinese have a burgeoning population, a mandate for electric vehicles, the manufacturing capacity _and_ absolutely no respect for IP.
It's not a race, it's a grind.
For Robotaxi, I could see Jeff Dahn's million mile NMC battery chemistry. Outperforms LFP in energy density somewhat and lifespan is well beyond LFP. "Lorna Casselton Lecture May 2023" on TH-cam.
Yeah, could be, I just view it as lower likelihood for the reasons I gave in the video
'VolksTaxi' mien Herr? Or maybe 'Peoples-Cab' (not likely. the anagram).... to soon?
As of now id expect tarrifs on Chinese cells to grow exponentially, but who knows with Elon's bestie..... 🙄
For sure i wouldn't go pissing off the 800 pound dragon in the room.
If they decide to call on trillions of trade deficit January 21st we are well and truly screwed.
We are the 800 lb gorilla
We've just been drinking too mini soy lattes 🤠🇺🇸
Why does your voice modulate...the way it does?...not saying you sound like AI but your modulation just sounds strange and random lol
That's my voice. That's it, that's the explanation.
LFP is a dead man walking in automotive its just too darn heavy for long range cars. The bog standard 4680 will best it in time
Nonesense, LFP output is only expanding.
I'll take less range and lower cost over longevity any day
@@thelimitingfactor It has no path forward to being able to equal the range to gasoline cars. 160wh is the best it can do, current gen nca's can do 250wh/kg and with silicon 280wh/kg or more. The lfp weight penalty becomes a non starter rally quickly.