Notes: 1) The 78% figure for the packing density was conservative. By staggered the cells, the packing density can be increased to ~90%. However, CATL/BYD are in reality closer to 400 Wh/l max than 450 Wh/l max, so I was aggressive there. That is, the estimate I provided would still be a good template. The math itself could have been dialled in better, but the results (250 miles), is still a good metric (in my view). 2) Although my assumption is that Drew was implying that the structural battery meant the 4680 structural battery, it could also mean a prismatic based structural battery. It's also worth keeping in mind that Tesla frequenly changes direction with their development plans and their thinking may have developed since Battery Day.
Jordan, curious on your take. After battery day I made some estimates based on the presentation. I came up with the following. I modeled range based on a Model 3. 4680 Per Cell LFP - 80Wh Nickel Mag - 110Wh High Nickel - 120Wh 4680 Structural Pack - 960 Cells LFP - 77kW, 329 Miles, $2900 Cost, $37/kW Nickel Mag - 105kW, 448 Miles, $4180 Cost, $39/kW High Nickel - 115kW, 492 Miles, $5200 Cost, $45/kW It will be interesting once we see production cells and vehicles. I hope that the biggest surprise is going to be in charge time. I think Tesla is sandbagging on this for 4680. I am guessing we will see at least a 50% reduction out of the gate, with up to 75% possible with tweaks over time.
@@motofunk1 Nice! More aggressive than I would have been, but not out of line with the information we've received from Tesla and Sandy Munro. That is, it's a matter of interpretation, but I can't fault the stats above. I'm just thinking of all the challenges they'll face and the increasing prices we're seeing in raw materials.
@@thelimitingfactor True, rising material costs will have an impact, but I don't really count that. I just look at my cost estimates as 2020 dollars, time adjusted for inflation will happen. Thanks for the feedback.
I feel their structural pack are for all batteries moving forward. Elon said the best makers of products are toy manufacturers. He is trying to create less variations not more
Jordan's videos are technical in nature and most people don't bother with technical details. Other TH-camrs are more financial in nature which appeals to a larger audience. Personally, I'm interested in the technical details so kudos to Jordan. But I agree, his subscriber numbers should be in the multiple 100 thousands.
Another brilliantly detailed but completely understandable for a total pleb, like me. 😁 ....and the fact that, "the illustrious James Douma" supports your channel, is THE unparalleled nod. Love it ! Congratulations!
I love how you present your material without any kind of hype.You present your material in methodical logical manner knowing the breakthroughs are not quite there at this time.
EXCELLENT PRESENTATION with logical structured thinking. Other communicators are guessing, with very little thought, are often calling others idiots. My viewing is limited to those with original well scripted ideas & meaningful visual backing. YOU ARE TOP OF THE LIST..
Jordon you are a credit to the Tesla Community and Investors. I always enjoy your work and now feel that the lead Tesla has only continues to grow. If I have understood you well, Tesla has OPTIONS and that means FLEXIBILITY which also means they have even more control over their DESTINY that leads to their DOMINANCE so that my outlook for the company as a whole only cements, but it also shows how precarious legacy automakers are in deciding which form factor, chemistry and ability to scale in order to even be in this game. There is so much here, I may just watch it again! Thank you so much for all that you do and for your Patreon supporters that allow you to do such deep dives.
You can't easily make a structural battery with a prismatic cell; prismatic cells need cages to contain their (limited for LFP) swelling and the long format isn't conducive to torsional loading. A honeycomb structure with a depth relative to "diameter" (which may be why Tesla chose the 4680 ratio/format) allows for a true structural pack. How they physically lock the cells together, regardless of format, will be interesting though.
The trick is that the current incarnation of the Model Y "structural" pack, really isn't used for significant structure! The Berlin (and Austin) Model Y's are still a complete unibody, with essentially just the floorpan missing! The "structural" body in white is a fully-stressed form, with all the lateral and torsional stability of the old design. The structural pack adds the lateral cross-bracing for side-impact protection, but beyond that, it's really just a floorpan that is bolted into place versus welded/glued. Munro and others will inevitably look it over, but 4680s in the "structural" pack will probably NOT be part of the functional structural rigidity of the pack, and honestly, you could argue that a Model 3/Y pack is "structural" since the cell blocks are extremely rigid and directly fastened to the pack casing. Basically, the Tesla structural pack is a production optimization (seats, etc installed from below), not a significant effort to actually make the pack part of the vehicle's stressed structure. I always find in comical when Tesla fanbois state that Tesla are based on a "skateboard" platform when that is categorically incorrect. ALL current Teslas are unibody vehicles with separate drive units installed independently from below along with a battery pack hung underneath, which has ZERO to do with the actual skateboard concept. A fully-developed EV skateboard will be a structurally-complete system that is the drive units and battery as a complete unit, onto which varying body styes could be mounted.
@@awebuser5914 The Y has multiple modules which are then installed/wired into the pack. The structural pack is effectively a big module removing the concentric casing and with the cell honeycomb giving the stiffness. Your common timber bedroom door contains a weak honeycomb sandwiched between two planes and the doors are very rigid.
@@patreekotime4578 Ummm, I'm not sure what your point is, but the bottom line is that the 4680's are really a stop-gap concept and so-called "structural" unibody Model Y's (the ones with no floor) being built in Berlin and Austin WILL have alternate packs available in a year or so using exactly the same body, so the concept of the 4680 being critical to a structural pack is obviously incorrect. It's trivial to create a so-called "structural" pack with cells smaller than a 4680, it may be less mass-efficient, but that's essentially irrelevant.
@@patreekotime4578 Many motorcycles use the engine as a stressed part of the bike chassis (frame). There is no reason why 4680 cells stacked in staggered rows between cooling plates cant be extremely rigid. If the hollow box of a bike engine is rigid enough, the packed-in cells of a battery pack would be incredibly stiff.
Thanks Jordan, for another well researched and reasoned video! It's becoming increasingly difficult to find, reasoned, fact based discussion about any of this stuff out there as people seem to be hardening into uninformed, unmovable positions.
Amazing explanations even for someone like me, who has no educational background in any of these battery topics. It really helps to affirm the investment case of TSLA. Thank you!
If I understand it correctly, from a marketing perspective, LFP could then be positioned as a standard energy source providing great range but lacking ultimate performance. The nickel could be positioned as an up sell for its lighter weight and higher performance?
It could, but I'd think it would be better to use Nickel for the upsell. (as always, I don't claim to have a monopoly on the truth) You have the right idea: You can play around with form factor and chemistry to suit a particular need. There is no 'One Best' solution and it's about product-market fit.
Bravo! This TH-cam video presciently anticipated Tesla's generally unexpected use of LFP chemistry in the Giga Austin Model Y's structural 4680 Battery Cells.
Great video! It's likely that Tesla is already developing LFP energy density and chemistry. And as soon as they hit a energy density that is satisfactory, they will apply it to the 4680.
I'm interested in the use of lithium phosphate, from low grade spodumene, as a substitute for lithium carbonate in LFP batteries. Seems to remove a step, as well as make use of lower grade resources.
Thinking out loud: Why not use both chemisties in the same car? If LFP has a long cycle life but less density, have a 40kwh pack that is used first, then draw on a 40kwh NCA battery. Even better, have the LFP pack provide power for cruising and then draw on the NCA pack for hills, accelerating, etc. Effecively reduce the cycle time on the NCA by half and the LFP is cycled twice as much.
(1) Tesla 4680 manufacturing plants are certainly designed to handle LFP electrodes & electrolytes. (2) Tesla vehicle & storage production is effectively limited by cell availability. (3) Tesla has stated it will purchase as many cells as outside manufactures can provide . (4) It becomes a question of how fast Tesla can increase its in-house manufacturing.
I see two most probable paths for LFP in Tesla vehicles: 1. Tesla makes 4680 LFP in house 2. Tesla buys Prismatic LFPs from Chinese suppliers I don’t see the advantage of Tesla spending the r&d and more so the distraction of engineering resources when they can buy off the shelf with enough supply at reasonable prices. I agree both would end up in structural packs.
5:49: the picture here shows a situation when the cells are not in honeycomb order. The pictures we have from Tesla suggest a honeycomb order, which uses the available space better. Based on my calculation in this case the packing density can get close to pi/(2*3**(1/2) ) == 90.69 %. (only close, because at the borders things get a bit worse).
Excellent. Isn't there a recharging advantage with the 4680 design? Be interested if there is any difference in recharging speed between prismatic and 4680s.
Thank you for your great analysis. In particular when you showed how the energy consumption efficiency of Tesla gives them a competitive advantage to use the LFP chemistry. It is now so obvious - because of the way how you explained it.
Great video as always. Perhaps take look at C4V's batteries. Prismatic, Cobalt free LFP, structural packs apparently tab less and using dry battery electrodes. Sounds very familiar
i think the logical path would be for Tesla to offer LFP in structural 4680 packs. I think the reason they offer LFP in prismatic form now is because they aren't (yet) using structural battery packs (giga cast frames). Once they roll out giga cast frames, they'll use 4680 in all of them and forgo prismatic. So, for the model 3/Y, my guess is they will eventually offer the RWD version (previously SR+) with 4680 LFP and a range of around 325. The LR version will use NCA and have a range of around 400. Just my 2 cents worth.
Gonna sub to your Patreon soon as I get home from work. I don’t know anything about batteries but I love Tesla and I found this breakdown of their new battery fascinating.
In the recent call, Tesla clearly said that LFP will NOT come in 4680 form factor, due to fundamental issues of the cell chemistry. Also, all stationary storage will go LFP.
love your videos. Based on what Drew said I am still convinced that LFP will stay as prismatic, as they can be form dense and structural if engineered properly and dont require as much cooling. BUT, its possible that 4680 is easier to produce and if so would be could be a driving factor toward a preference in 4680.
I think if Tesla succeed to mine lithium themselves in Nevada, they will put it in their own 4680 as LFP and perhaps with some silicon doping as Jordan said or replacing Iron with Manganese as Elon has twitted.
I agree. CATL and BYD seem to be scaling prismatic LFP so Tesla has a plentiful supply. That can leave Tesla to focus 4680 for higher performance applications. Having a diverse battery supply chain only benefits Tesla and allows their suppliers to keep advancing LFP. Just because they can technically do 4680 LFP, doesn't mean they have to.
@@sparks869 Up-and-coming battery shortage coming this decade (consequently, suppliers increasing prices) speaks towards Tesla needing to bring LFP largely in-house, if they want to compete cost-effectively for the cheaper market segment that requires cheap LFP batteries.
I agree, I think Tesla will use LFP with 4680. I recall seeing a hint on Twitter from a Tesla engineer saying that 4680 + LFP would make 2022 a huge year for Tesla Storage. I recall them using the 👀. Great video as always and congrats again on the new digs and wheels bro! 😎
Oh hey man! That would be 🔥. These videos are alwayhs my best guess and mainly to share information, but I always do a little fist pump if I get it right, lol.
As always, great work, and much appreciated for the strategic comparisons. Overall, I don’t think it’s practical for Tesla to produce prismatic LFP, given the additional R&D and production costs. That’s what the CATL and potential BYD deals are for - LFP coverage for the near term, especially in China. In the meantime, Tesla can focus on its own 4680 production lines, and eventually add in LFP chemistry over time, with or without silicon. Plus, As you note, it fits with their design philosophy.
Thanks Jordan for another great video. This would be assuming that the maxwell dry battery electrode technology is compatible with the LFP chemistry...
If we suppose that different form factors are optimal for different battery chemistries, might not the diameter of the cell be the deciding variable? My hunch: Tesla's batteries will be all cylindrical (none prismatic) and 80mm high. This will make them all work in the same battery packs, and Tesla will be able to adapt its manufacturing techniques the easiest. The different chemistries will all have different optimal diameters though. What do you think? e.g. Nickel 4680, Manganese 5480, LFP 6680, Sodium Iron 7280. (I made those dimensions up for illustration purposes, as I don't have any insight into how to calculate the optimal form factor.)
Hey man! It's not my chart, if you look you'll notice a credit to Sawyer Merritt. Besides, neither Tesla nor Lucid are the gold standard for range. Lightyear One is.
Great presentation. I understand that LFP can be charged to near 100% where as the NCA batteries can only be charged to 80% without shortening the battery life. I think this is huge as the advertised range is more closely achievable for the LFP vs 80% for the NCA. I'm not sure that is addressed in the comparisons. Being able to use the full capacity of the battery as well as higher life cycles should make the LFP a clear winner. Now about other chemistries..........
Nobody's ever going to use the packing arrangement shown at 5:41. They're going to use a hexagonal packing lattice that allows 90% packing density. The 10% unused space is a perfect place to route temperature-control fluid (for cooling and warming up).
Actually, the surface/volume ratio of prismatic cells is greater than cylindrical. It is the packaging density in modules where the aggregate effectively densifies the block requiring another approach to cooling. Small but fundamental point.
According to the investor's call, Austin-built Model Y's already are running around with 4680 structural packs. It's just a matter of time until these are being sold to customers. My guess is that they are working out how to sell these along side the older Freemont-build Model Y's, since they could have significant weight and/or battery performance improvements.
@@roddlez Maybe a larger question is "What happens to the 2170 cells from Nevada if Texas starts with 4680 and Fremont pivots (temporarily?) to lower range LFP cars?" ....... Options. 1) Sell them. 2) Powerwall, until such time as and LFP alternative is available (If constrained, most LFP would go to commercial projects??) 3) Install a 4680 line in Nevada to serve the Semi, then as it ramps, staged replacement of the 2170 facilities.... ? Just thought of another one. 4) A "low range" 2170 vehicle, non structural, smaller 2170 pack?
@@rogerstarkey5390 interesting, but it will be at least a year before Texas produces enough vehicles to meet CURRENT monthly demand. Also we have every reason to believe that demand will be even higher a year from now. Just not sure how they will handle the different Model Ys in the interim.
Remember that a power train efficiency can only improve so much. The better it is, the harder it is to improve and the less it will have an impact on range.
You are really among the best. The amazing thing about Tesla is that they put it out there for anybody to see. Same with SpaceX. Not much hidden. Elon is true to himself and us when he says he wants the world to change. No other company or human has done or is doing innovation this way. If only other companies would learn from this. A good series might be to look for any companies emulating his methodology. Would be interesting. That being said, who could do what Elon does?
The 4680 according to Tesla is the ideal size for tabless Li-Ion battery from what I remember and yes because of the structural battery pack and the removal of framing weight because of it. It seems like a really good solution and when I looked as the prismatic I don't see any way to make a structural battery pack with it. What we don't know is if that same size would be ideal for LFP. If it isn't, they might have to make another format, but going smaller is not good, and going larger I think means losing more space for battery capacity.
Recycle LFP at 14:35 - The Lithium will be recyclable from LFP. It will be a water soluble component rather easy to separate from the other materials which are all insoluble. You would get uncontaminated Lithium back. Recyclers will need to separate incoming batteries into type so they apply an efficient extraction for each type. You would not want to process LFP batteries for say, Nickel extraction. perhaps an x-ray emission spectrometer testing each battery as it passes on a conveyer. This would identify which metals are present in each battery and allow efficient sorting.
One benefit of cylindrical structural batteries over prismatic structural batteries is: cylindrical cells offer structural support from all directions. They act almost exactly like a honeycomb structure. Prismatic cells like the BYD blade are much weaker to impacts perpendicular to the largest face on the cell (what most people would call the "side" of the cell).
Thanks again Jordan! As usual, you've taken my somewhat jumbled thought process, distilled it, dotted most of the i's, crossed most of the t's and I think largely confirmed my thoughts. Going for the second listen but first.... ..... Did you consider the "system efficiency" improvement from weight reduction of the whole vehicle using the structural pack. That may put the *vehicle* into the energy to weight band of the current Nickel pack based cars? ........ Would production rate /physical factory volume be a consideration? If the Wh per hour leaving a given size of factory is much lower for prismatic, that hands the advantage back to 4680?
Cooling the battery radially seems like a mistake. There is a sheet of copper that runs the length of the jelly role and thermally bonded to an end cap.... why not wick heat from there?
I'm curious how TESLA is managing the LFP battery in the cold temperature. As you mentioned in the video and as many have said, LFP is bad in the cold winter. Is Tesla managing it by preheating the battery when the preheating the cabin is turned on? Is that how they are able to warm the battery and use the battery in the higher temperature range, so it is not so much affected by the cold temperature? If this is so, LFP is such a viable option for cars even in cold climates.
Jordan, great job as always! One question: have you done any video on battery recycling? The alternative techniques, pros and cons, the players, etc? Would love to hear your take on it. I know getting info on the approach taken by private companies such as Redwood Materials probably is hard, but ABML, Li-Cyle and few others are public and probably share some details. Thanks!
Hi Jordan, interesting video as always. I belive that the tabless design might enable thicker LFP electrodes and higher energy densities, which should improve energy/volume density of a proposed 4680 lfp battery. Couple that with reduced cost from DBE, and it seems that this route might become the norm for Tesla for most applications
Jordan, at the recent earnings call, Elon mentioned manganese for batteries. There are Lithium Manganese Oxide rechargeable batteries, and manganese may be cheaper than iron phosphate. Is it possible that Tesla has a new chemistry up their sleeve?
Hi Jordan, your videos are awesome, great work, thankyou. If Tesla were to make LFP, using the 4680 format might be possible but I can't see it stacking up against a prismatic cell design. As you pointed out LPF materials are so cheap compared with NMC that recycling LFP is not really feasible. And if LFP is only 20% cheaper than NMC per kW/h due to the cheaper materials then more than 70% of the price related to cell manufacturing. From the battery investor day we know that DBE will substantially lower production costs. Separately the 4680 cell format also lowers production costs compared with the 18650 cell format and larger cells means fewer cells per vehicle. Prismatic cells are already quite large compared with the 4680 cell format, I don't think Tesla would want to put more cells in their vehicles. It's the tab-less electrode design that unlocked the large cell format with lower internal resistance and better heat dissipation. A tab-less prismatic cell design would allow for much larger cells to be made at a lower price per kWh. These benefits combined with the benefits of the DBE manufacturing process could see prismatic cells perform almost as well at the current 18650 cells. It is worthwhile for Tesla to invest in a new form factor for LFP? Yes, If nickel battery’s are projected to only represent 25% of the battery market in the next few years and it was worth while moving to the 4680 cell format, then surely they would see be benefit of optimizing the prismatic cell format for LFP which is projected to be 75% of the market.
@@thelimitingfactor Yes, that is the tweet. May be Tesla can only make Manganese and high Nickel batteries and just buy LFP from CATL. Especially if Manganese batteries are closer to LFP in terms of Cycle life and cost but have density advantages over LFP. Then they could stick with 4680 format.
High Josh! I made an error on that one. Staggering would increase the energy density. I actually covered this in my Big F'ing Cell video but forgot it. Regardless, the end result in reality would be similar because I was overaggressive with CATL/BYD energy density figures.
In the comparisons of blade to wound cells I'd like to see what the actual "active square footage" is of the 4680 battery material - ie unwind a cylindrical cell ( to get a biggish rectangle) and compare with the active area of a blade battery. Plus be good to see and compare the length of electrical path to the poles (tabless makes for very short paths) - so, how does a blade compared on this???
One limiting factor we kept a blind eye on is the intrinsic thermal handling with tab less battery 46800, regardless of MNC, LFP or any cell chemistry. While 18650 and 21700 can pass current via its welded tab but heat. Heat must conduct heat through the case cylinder wall just because it has a largest case area, while thermal resistance there remains too large in 18650 thermal management. Is a limiting factor because core heat must diffuse through multiple layers of metal-chemical-separator reaching to the side wall. How about this? Since the electrodes are metal, is it a good idea to sink core heat vertically via the case base, the flat portion of cell casing? Answer: yes and no. Yes? It is wise to sink heat vertically. No? A bristle wafer placed between anode-case junction conducts electricity but heat. Why not heat ? The bristle wafer is a good thermal insulator. That is a gating issue, otherwise called a show stopper. Consequences? Over heated chemical affecting cell longevity and safety against a fire or explosion. Remedy? In the first principle view we need a solid junction to pass heat and electricity and not a bristle wafer. In practice we can employ induction heated reflow soldering at the anode-case junction. Each cell takes about 5-10 sec. A carousel of N induction coils can reduce cycle time by a factor of N. Set back? A slight increase battery pack mass by weight difference from wafer to solder. Benefit? out weights the risk stated above.
A less discussed challenge is electrode matching. High nickel and silicon composite anode designs both have high initial irreversible capacities. LFP does not. This means an LFP / C6:Si design would likely require online pre-lithiation. This adds big costs via capital and throughput slowdowns.
How do prismatic and 4680 tabless cells compare with regard to internal resistance? Surely the 4680 has a significantly lower internal resistance. This surely means that not only that it can charge more quickly, while being able to remove that heat more easily. This is a huge benefit that nobody seems to be talking about. It's hard to see how you can make prismatic cells anywhere near as cheaply as cylindrical ones. Rolling the electrodes is always going to be quicker and easier to do at scale than separate plates. Since cost is a huge driving factor, it seems clear to me that cylindrical cells will eventually become the dominant form factor.
What about that Tesla model S whit a modified LFP battery from One power, what do you think we're the specs fill factor, density etc do you think they have sacrfied the thermal management to fit it in?
Notes:
1) The 78% figure for the packing density was conservative. By staggered the cells, the packing density can be increased to ~90%. However, CATL/BYD are in reality closer to 400 Wh/l max than 450 Wh/l max, so I was aggressive there. That is, the estimate I provided would still be a good template. The math itself could have been dialled in better, but the results (250 miles), is still a good metric (in my view).
2) Although my assumption is that Drew was implying that the structural battery meant the 4680 structural battery, it could also mean a prismatic based structural battery. It's also worth keeping in mind that Tesla frequenly changes direction with their development plans and their thinking may have developed since Battery Day.
Jordan, curious on your take. After battery day I made some estimates based on the presentation. I came up with the following. I modeled range based on a Model 3.
4680 Per Cell
LFP - 80Wh
Nickel Mag - 110Wh
High Nickel - 120Wh
4680 Structural Pack - 960 Cells
LFP - 77kW, 329 Miles, $2900 Cost, $37/kW
Nickel Mag - 105kW, 448 Miles, $4180 Cost, $39/kW
High Nickel - 115kW, 492 Miles, $5200 Cost, $45/kW
It will be interesting once we see production cells and vehicles. I hope that the biggest surprise is going to be in charge time. I think Tesla is sandbagging on this for 4680. I am guessing we will see at least a 50% reduction out of the gate, with up to 75% possible with tweaks over time.
@@motofunk1 Nice! More aggressive than I would have been, but not out of line with the information we've received from Tesla and Sandy Munro. That is, it's a matter of interpretation, but I can't fault the stats above. I'm just thinking of all the challenges they'll face and the increasing prices we're seeing in raw materials.
@@thelimitingfactor True, rising material costs will have an impact, but I don't really count that. I just look at my cost estimates as 2020 dollars, time adjusted for inflation will happen. Thanks for the feedback.
I feel their structural pack are for all batteries moving forward.
Elon said the best makers of products are toy manufacturers. He is trying to create less variations not more
It's π/4
Why are you not the number 1 you tuber regarding Tesla? Your analysis is literally amazing and informative.
Cause it ain’t clickbait, good or bad.
and it requires a basic level of engineering and technical competency that goes over a lot of peoples' heads.
Jordan's videos are technical in nature and most people don't bother with technical details. Other TH-camrs are more financial in nature which appeals to a larger audience. Personally, I'm interested in the technical details so kudos to Jordan. But I agree, his subscriber numbers should be in the multiple 100 thousands.
@Bob Howard
He is, for those who pay attention.
This is a science learning channel not a marketing channel like most other Tesla channels.
As always your coverage of batteries can't be excelled. The depth and accuracy of your research is beyond anything else out there. Thank you!
Another brilliantly detailed but completely understandable for a total pleb, like me. 😁
....and the fact that, "the illustrious James Douma" supports your channel, is THE unparalleled nod. Love it !
Congratulations!
🤜🤛🤠
I love how you present your material without any kind of hype.You present your material in methodical logical manner knowing the breakthroughs are not quite there at this time.
EXCELLENT PRESENTATION with logical structured thinking.
Other communicators are guessing, with very little thought, are often calling others idiots. My viewing is limited to those with original well scripted ideas & meaningful visual backing. YOU ARE TOP OF THE LIST..
Always great to hear from you Edward 🙌
I just recently found this channel…it’s pure gold
Another exceptionally thoughtful and erudite analysis by the Limiting Factor
I gotta say thank you to the author for such a high quality content and thank you for everyone supporting such a content. You guys are awesome!
Did not understand much but enjoyed the whole 30:15min presentation..
your voice is hypnotic...
Jordon you are a credit to the Tesla Community and Investors. I always enjoy your work and now feel that the lead Tesla has only continues to grow. If I have understood you well, Tesla has OPTIONS and that means FLEXIBILITY which also means they have even more control over their DESTINY that leads to their DOMINANCE so that my outlook for the company as a whole only cements, but it also shows how precarious legacy automakers are in deciding which form factor, chemistry and ability to scale in order to even be in this game.
There is so much here, I may just watch it again! Thank you so much for all that you do and for your Patreon supporters that allow you to do such deep dives.
Oh hey thanks man! I appreciate the kind words. Amen - Tesla has options. That's as important as anything else.
You can't easily make a structural battery with a prismatic cell; prismatic cells need cages to contain their (limited for LFP) swelling and the long format isn't conducive to torsional loading. A honeycomb structure with a depth relative to "diameter" (which may be why Tesla chose the 4680 ratio/format) allows for a true structural pack. How they physically lock the cells together, regardless of format, will be interesting though.
The trick is that the current incarnation of the Model Y "structural" pack, really isn't used for significant structure! The Berlin (and Austin) Model Y's are still a complete unibody, with essentially just the floorpan missing! The "structural" body in white is a fully-stressed form, with all the lateral and torsional stability of the old design. The structural pack adds the lateral cross-bracing for side-impact protection, but beyond that, it's really just a floorpan that is bolted into place versus welded/glued. Munro and others will inevitably look it over, but 4680s in the "structural" pack will probably NOT be part of the functional structural rigidity of the pack, and honestly, you could argue that a Model 3/Y pack is "structural" since the cell blocks are extremely rigid and directly fastened to the pack casing. Basically, the Tesla structural pack is a production optimization (seats, etc installed from below), not a significant effort to actually make the pack part of the vehicle's stressed structure.
I always find in comical when Tesla fanbois state that Tesla are based on a "skateboard" platform when that is categorically incorrect. ALL current Teslas are unibody vehicles with separate drive units installed independently from below along with a battery pack hung underneath, which has ZERO to do with the actual skateboard concept. A fully-developed EV skateboard will be a structurally-complete system that is the drive units and battery as a complete unit, onto which varying body styes could be mounted.
@@awebuser5914 The Y has multiple modules which are then installed/wired into the pack. The structural pack is effectively a big module removing the concentric casing and with the cell honeycomb giving the stiffness. Your common timber bedroom door contains a weak honeycomb sandwiched between two planes and the doors are very rigid.
@@patreekotime4578 Ummm, I'm not sure what your point is, but the bottom line is that the 4680's are really a stop-gap concept and so-called "structural" unibody Model Y's (the ones with no floor) being built in Berlin and Austin WILL have alternate packs available in a year or so using exactly the same body, so the concept of the 4680 being critical to a structural pack is obviously incorrect. It's trivial to create a so-called "structural" pack with cells smaller than a 4680, it may be less mass-efficient, but that's essentially irrelevant.
@@patreekotime4578 Many motorcycles use the engine as a stressed part of the bike chassis (frame). There is no reason why 4680 cells stacked in staggered rows between cooling plates cant be extremely rigid. If the hollow box of a bike engine is rigid enough, the packed-in cells of a battery pack would be incredibly stiff.
Thanks Jordan, for another well researched and reasoned video! It's becoming increasingly difficult to find, reasoned, fact based discussion about any of this stuff out there as people seem to be hardening into uninformed, unmovable positions.
Amazing explanations even for someone like me, who has no educational background in any of these battery topics. It really helps to affirm the investment case of TSLA. Thank you!
I finally found a channel to learn from it about batteries. Kudos.
Your channel is fantastic. Full of real info/facts so well explained that instead of being confused after watching, I feel informed. Thank you sir!
If I understand it correctly, from a marketing perspective, LFP could then be positioned as a standard energy source providing great range but lacking ultimate performance. The nickel could be positioned as an up sell for its lighter weight and higher performance?
It could, but I'd think it would be better to use Nickel for the upsell. (as always, I don't claim to have a monopoly on the truth)
You have the right idea: You can play around with form factor and chemistry to suit a particular need. There is no 'One Best' solution and it's about product-market fit.
Great video again, probably have to watch it a few more times to completely get it
Bravo! This TH-cam video presciently anticipated Tesla's generally unexpected use of LFP chemistry in the Giga Austin Model Y's structural 4680 Battery Cells.
I came here to say "but they said ...", but you beat me to it and covered it :) Another great video.
A large portion of what I know about batteries is from watching your videos. Thanks very much!
One of your best Jordan!
Can’t wait for your sodium ion series since it can match present day LFP density and is theoretically far more abundant and inexpensive!
Thanks Jordan! This video exactly answers my questions about 4680’s. Can’t wait for your video on sodium.
Great video! It's likely that Tesla is already developing LFP energy density and chemistry. And as soon as they hit a energy density that is satisfactory, they will apply it to the 4680.
I'm interested in the use of lithium phosphate, from low grade spodumene, as a substitute for lithium carbonate in LFP batteries. Seems to remove a step, as well as make use of lower grade resources.
Thinking out loud: Why not use both chemisties in the same car? If LFP has a long cycle life but less density, have a 40kwh pack that is used first, then draw on a 40kwh NCA battery. Even better, have the LFP pack provide power for cruising and then draw on the NCA pack for hills, accelerating, etc. Effecively reduce the cycle time on the NCA by half and the LFP is cycled twice as much.
As always I understand like 15% of your videos. But i am very greatful for them!
(1) Tesla 4680 manufacturing plants are certainly designed to handle LFP electrodes & electrolytes.
(2) Tesla vehicle & storage production is effectively limited by cell availability.
(3) Tesla has stated it will purchase as many cells as outside manufactures can provide .
(4) It becomes a question of how fast Tesla can increase its in-house manufacturing.
I see two most probable paths for LFP in Tesla vehicles:
1. Tesla makes 4680 LFP in house
2. Tesla buys Prismatic LFPs from Chinese suppliers
I don’t see the advantage of Tesla spending the r&d and more so the distraction of engineering resources when they can buy off the shelf with enough supply at reasonable prices. I agree both would end up in structural packs.
5:49: the picture here shows a situation when the cells are not in honeycomb order. The pictures we have from Tesla suggest a honeycomb order, which uses the available space better. Based on my calculation in this case the packing density can get close to pi/(2*3**(1/2) ) == 90.69 %. (only close, because at the borders things get a bit worse).
Please read the pinned comment. It would have saved you some time.
Great content, superb delivery! Thanks for spreading science and hope
👌👌 LFP 4680 with genius silicon doping for energy density is my guess as the ruling form factor from what you say Jordan. Luving it, thnx!
I always look forward to your analysis.
That is so well researched!
Excellent. Isn't there a recharging advantage with the 4680 design? Be interested if there is any difference in recharging speed between prismatic and 4680s.
Thank you for your great analysis. In particular when you showed how the energy consumption efficiency of Tesla gives them a competitive advantage to use the LFP chemistry. It is now so obvious - because of the way how you explained it.
Amazing video once again Jordan!
thank you! you have a great "radio" voice. easy to listen too.
Catl really revolutionizing LFP battery tech. I hope more variety and innovation for all companies.
Awesome videos. Happy to support your channel!👍
Great video as always. Perhaps take look at C4V's batteries. Prismatic, Cobalt free LFP, structural packs apparently tab less and using dry battery electrodes. Sounds very familiar
Another fine video with excellent explanations and well-sourced data. I love your channel!
i think the logical path would be for Tesla to offer LFP in structural 4680 packs. I think the reason they offer LFP in prismatic form now is because they aren't (yet) using structural battery packs (giga cast frames). Once they roll out giga cast frames, they'll use 4680 in all of them and forgo prismatic.
So, for the model 3/Y, my guess is they will eventually offer the RWD version (previously SR+) with 4680 LFP and a range of around 325. The LR version will use NCA and have a range of around 400. Just my 2 cents worth.
Gonna sub to your Patreon soon as I get home from work. I don’t know anything about batteries but I love Tesla and I found this breakdown of their new battery fascinating.
Love the vids man! Always look forward to them! Keep it up
In the recent call, Tesla clearly said that LFP will NOT come in 4680 form factor, due to fundamental issues of the cell chemistry. Also, all stationary storage will go LFP.
You didn't watch the video from what I can tell. I said stationary storage would go LFP.
@archie, yes i recall vividly Elon stating this.
Well of COURSE different batteries have positives and negatives. That's how batteries work! /s
Punny
You're officially discharged from this conversation.
I was SO close to commenting, but it was too bad even for me.
@@rogerstarkey5390 Roger, we appreciate your neutrality…
love your videos. Based on what Drew said I am still convinced that LFP will stay as prismatic, as they can be form dense and structural if engineered properly and dont require as much cooling. BUT, its possible that 4680 is easier to produce and if so would be could be a driving factor toward a preference in 4680.
I think if Tesla succeed to mine lithium themselves in Nevada, they will put it in their own 4680 as LFP and perhaps with some silicon doping as Jordan said or replacing Iron with Manganese as Elon has twitted.
I agree. CATL and BYD seem to be scaling prismatic LFP so Tesla has a plentiful supply. That can leave Tesla to focus 4680 for higher performance applications. Having a diverse battery supply chain only benefits Tesla and allows their suppliers to keep advancing LFP. Just because they can technically do 4680 LFP, doesn't mean they have to.
@@sparks869 Up-and-coming battery shortage coming this decade (consequently, suppliers increasing prices) speaks towards Tesla needing to bring LFP largely in-house, if they want to compete cost-effectively for the cheaper market segment that requires cheap LFP batteries.
Amazing analysis. Dang, you’re so brilliant.
I agree, I think Tesla will use LFP with 4680. I recall seeing a hint on Twitter from a Tesla engineer saying that 4680 + LFP would make 2022 a huge year for Tesla Storage. I recall them using the 👀. Great video as always and congrats again on the new digs and wheels bro! 😎
Oh hey man! That would be 🔥. These videos are alwayhs my best guess and mainly to share information, but I always do a little fist pump if I get it right, lol.
Woot woot! Also huge congrats on 80k subs!! 🍻
As always, great work, and much appreciated for the strategic comparisons. Overall, I don’t think it’s practical for Tesla to produce prismatic LFP, given the additional R&D and production costs. That’s what the CATL and potential BYD deals are for - LFP coverage for the near term, especially in China. In the meantime, Tesla can focus on its own 4680 production lines, and eventually add in LFP chemistry over time, with or without silicon. Plus, As you note, it fits with their design philosophy.
IMO Tesla bought this silicon battery start up in Colorado for one reason SI
Thanks Jordan for another great video. This would be assuming that the maxwell dry battery electrode technology is compatible with the LFP chemistry...
Great video Jordan also really liked your interview with Dave Lee investing.
Hey thanks man!
great video, dope intro song, left a like (thanks for putting intro song in description👌)
Hey! You're the first person who's noticed out of millions, lol. Glad it was useful.
If we suppose that different form factors are optimal for different battery chemistries, might not the diameter of the cell be the deciding variable? My hunch: Tesla's batteries will be all cylindrical (none prismatic) and 80mm high. This will make them all work in the same battery packs, and Tesla will be able to adapt its manufacturing techniques the easiest. The different chemistries will all have different optimal diameters though. What do you think?
e.g. Nickel 4680, Manganese 5480, LFP 6680, Sodium Iron 7280.
(I made those dimensions up for illustration purposes, as I don't have any insight into how to calculate the optimal form factor.)
You are so good at this. Thank you.
Excellent work. Question: where is Lucid in the 22:11 chart?
Hey man! It's not my chart, if you look you'll notice a credit to Sawyer Merritt. Besides, neither Tesla nor Lucid are the gold standard for range. Lightyear One is.
Great presentation. I understand that LFP can be charged to near 100% where as the NCA batteries can only be charged to 80% without shortening the battery life. I think this is huge as the advertised range is more closely achievable for the LFP vs 80% for the NCA. I'm not sure that is addressed in the comparisons. Being able to use the full capacity of the battery as well as higher life cycles should make the LFP a clear winner. Now about other chemistries..........
Nobody's ever going to use the packing arrangement shown at 5:41. They're going to use a hexagonal packing lattice that allows 90% packing density. The 10% unused space is a perfect place to route temperature-control fluid (for cooling and warming up).
Actually, the surface/volume ratio of prismatic cells is greater than cylindrical. It is the packaging density in modules where the aggregate effectively densifies the block requiring another approach to cooling. Small but fundamental point.
Isn't the Blade battery shape better? Better heat disapation and packing ability?
Oh, I'm waiting for them. Did apply to production Model Y already?
I don't understand that question. Not sure which them you are referring to and which region of Model Y production
According to the investor's call, Austin-built Model Y's already are running around with 4680 structural packs. It's just a matter of time until these are being sold to customers.
My guess is that they are working out how to sell these along side the older Freemont-build Model Y's, since they could have significant weight and/or battery performance improvements.
@@roddlez
Maybe a larger question is
"What happens to the 2170 cells from Nevada if Texas starts with 4680 and Fremont pivots (temporarily?) to lower range LFP cars?"
.......
Options.
1) Sell them.
2) Powerwall, until such time as and LFP alternative is available (If constrained, most LFP would go to commercial projects??)
3) Install a 4680 line in Nevada to serve the Semi, then as it ramps, staged replacement of the 2170 facilities.... ?
Just thought of another one.
4) A "low range" 2170 vehicle, non structural, smaller 2170 pack?
@@rogerstarkey5390 interesting, but it will be at least a year before Texas produces enough vehicles to meet CURRENT monthly demand. Also we have every reason to believe that demand will be even higher a year from now.
Just not sure how they will handle the different Model Ys in the interim.
@@rogerstarkey5390 I don't know what is a deal with Panasonic. It is possible that they have to meet some buyer's quota.
Great video as usual !
Remember that a power train efficiency can only improve so much. The better it is, the harder it is to improve and the less it will have an impact on range.
You are really among the best. The amazing thing about Tesla is that they put it out there for anybody to see. Same with SpaceX. Not much hidden. Elon is true to himself and us when he says he wants the world to change. No other company or human has done or is doing innovation this way. If only other companies would learn from this. A good series might be to look for any companies emulating his methodology. Would be interesting. That being said, who could do what Elon does?
The 4680 according to Tesla is the ideal size for tabless Li-Ion battery from what I remember and yes because of the structural battery pack and the removal of framing weight because of it. It seems like a really good solution and when I looked as the prismatic I don't see any way to make a structural battery pack with it.
What we don't know is if that same size would be ideal for LFP. If it isn't, they might have to make another format, but going smaller is not good, and going larger I think means losing more space for battery capacity.
Recycle LFP at 14:35 - The Lithium will be recyclable from LFP. It will be a water soluble component rather easy to separate from the other materials which are all insoluble. You would get uncontaminated Lithium back.
Recyclers will need to separate incoming batteries into type so they apply an efficient extraction for each type. You would not want to process LFP batteries for say, Nickel extraction. perhaps an x-ray emission spectrometer testing each battery as it passes on a conveyer. This would identify which metals are present in each battery and allow efficient sorting.
One benefit of cylindrical structural batteries over prismatic structural batteries is: cylindrical cells offer structural support from all directions. They act almost exactly like a honeycomb structure. Prismatic cells like the BYD blade are much weaker to impacts perpendicular to the largest face on the cell (what most people would call the "side" of the cell).
I think you would enjoy listening to Dr. Shailesh Upreti. Very good video with Ev reporter.
Thanks again Jordan!
As usual, you've taken my somewhat jumbled thought process, distilled it, dotted most of the i's, crossed most of the t's and I think largely confirmed my thoughts.
Going for the second listen but first....
.....
Did you consider the "system efficiency" improvement from weight reduction of the whole vehicle using the structural pack.
That may put the *vehicle* into the energy to weight band of the current Nickel pack based cars?
........
Would production rate /physical factory volume be a consideration?
If the Wh per hour leaving a given size of factory is much lower for prismatic, that hands the advantage back to 4680?
1) Nope - I was conservative.
2) Yup - But that's a 'should' be true rather than something I can demonstrate or have data on.
🤠
@@thelimitingfactor
1) That's what I thought.
2) I agree, seems likely, but time will tell.
👍
@@thelimitingfactor
P.S....
You're REALLY back in the US, aren't you 😁😁
🤠 👍
Cooling the battery radially seems like a mistake. There is a sheet of copper that runs the length of the jelly role and thermally bonded to an end cap.... why not wick heat from there?
I'm curious how TESLA is managing the LFP battery in the cold temperature. As you mentioned in the video and as many have said, LFP is bad in the cold winter. Is Tesla managing it by preheating the battery when the preheating the cabin is turned on? Is that how they are able to warm the battery and use the battery in the higher temperature range, so it is not so much affected by the cold temperature? If this is so, LFP is such a viable option for cars even in cold climates.
Preheating and dialling in their software. People have done videos on using a Tesla battery in like -30 C
Jordan, great job as always!
One question: have you done any video on battery recycling? The alternative techniques, pros and cons, the players, etc? Would love to hear your take on it. I know getting info on the approach taken by private companies such as Redwood Materials probably is hard, but ABML, Li-Cyle and few others are public and probably share some details. Thanks!
I know Fully Charged channel has done some quite OK work on this
Hi Jordan, interesting video as always. I belive that the tabless design might enable thicker LFP electrodes and higher energy densities, which should improve energy/volume density of a proposed 4680 lfp battery. Couple that with reduced cost from DBE, and it seems that this route might become the norm for Tesla for most applications
Jordan, at the recent earnings call, Elon mentioned manganese for batteries. There are Lithium Manganese Oxide rechargeable batteries, and manganese may be cheaper than iron phosphate. Is it possible that Tesla has a new chemistry up their sleeve?
Hi Jordan, your videos are awesome, great work, thankyou.
If Tesla were to make LFP, using the 4680 format might be possible but I can't see it stacking up against a prismatic cell design.
As you pointed out LPF materials are so cheap compared with NMC that recycling LFP is not really feasible. And if LFP is only 20% cheaper than NMC per kW/h due to the cheaper materials then more than 70% of the price related to cell manufacturing.
From the battery investor day we know that DBE will substantially lower production costs. Separately the 4680 cell format also lowers production costs compared with the 18650 cell format and larger cells means fewer cells per vehicle. Prismatic cells are already quite large compared with the 4680 cell format, I don't think Tesla would want to put more cells in their vehicles.
It's the tab-less electrode design that unlocked the large cell format with lower internal resistance and better heat dissipation. A tab-less prismatic cell design would allow for much larger cells to be made at a lower price per kWh. These benefits combined with the benefits of the DBE manufacturing process could see prismatic cells perform almost as well at the current 18650 cells.
It is worthwhile for Tesla to invest in a new form factor for LFP? Yes, If nickel battery’s are projected to only represent 25% of the battery market in the next few years and it was worth while moving to the 4680 cell format, then surely they would see be benefit of optimizing the prismatic cell format for LFP which is projected to be 75% of the market.
Can you comment on Manganese replacing LFP?
Already on it: twitter.com/elonmusk/status/1486821592533544971
As usual, there'll be a quarterly video.
@@thelimitingfactor Yes, that is the tweet.
May be Tesla can only make Manganese and high Nickel batteries and just buy LFP from CATL. Especially if Manganese batteries are closer to LFP in terms of Cycle life and cost but have density advantages over LFP. Then they could stick with 4680 format.
My take.
Manganese Nickel will be a thing, but still contains Nickel, with the associated costs and supply issues.
Yes my favorite channel!!!!!!!
Thanks a lot Again Jordan, your video are so clear and valuable Thanks to reply to many of my questions
Page 5:40 is a configuration not taken by Tesla’s structural pack. Tesla use a honey comb configuration with a least dead space.
Isn't max (hexagonal) packing density of cylindrical cells 0.9069 and not 0.78?
I wonder if his estimate took into consideration that the pack has to have less cell density to be used as a structural component. (?)
@@thelimitingfactor NOPE. en.wikipedia.org/wiki/Circle_packing Why not just look these things up before you misinform people?
I was thinking the same thing. Every mathmatician knows this: a honeycomb pattern is more spatially efficient than square-packing.
Just look at 7:50. Honeycomb.
High Josh! I made an error on that one. Staggering would increase the energy density. I actually covered this in my Big F'ing Cell video but forgot it.
Regardless, the end result in reality would be similar because I was overaggressive with CATL/BYD energy density figures.
In the comparisons of blade to wound cells I'd like to see what the actual "active square footage" is of the 4680 battery material - ie unwind a cylindrical cell ( to get a biggish rectangle) and compare with the active area of a blade battery. Plus be good to see and compare the length of electrical path to the poles (tabless makes for very short paths) - so, how does a blade compared on this???
Any idea what form factor the high energy density Gotion Hi tech LFP cells are? I assume prismatic, but might they be similar to the BYD Blade format?
Diy are already going hard on LFP for solar. They use server rack batteries. Will Prowse has a nice channel.
One limiting factor we kept a blind eye on is the intrinsic thermal handling with tab less battery 46800, regardless of MNC, LFP or any cell chemistry.
While 18650 and 21700 can pass current via its welded tab but heat. Heat must conduct heat through the case cylinder wall just because it has a largest case area, while thermal resistance there remains too large in 18650 thermal management. Is a limiting factor because core heat must diffuse through multiple layers of metal-chemical-separator reaching to the side wall.
How about this? Since the electrodes are metal, is it a good idea to sink core heat vertically via the case base, the flat portion of cell casing? Answer: yes and no.
Yes?
It is wise to sink heat vertically.
No?
A bristle wafer placed between anode-case junction conducts electricity but heat.
Why not heat ?
The bristle wafer is a good thermal insulator. That is a gating issue, otherwise called a show stopper.
Consequences?
Over heated chemical affecting cell longevity and safety against a fire or explosion.
Remedy?
In the first principle view we need a solid junction to pass heat and electricity and not a bristle wafer.
In practice we can employ induction heated reflow soldering at the anode-case junction. Each cell takes about 5-10 sec. A carousel of N induction coils can reduce cycle time by a factor of N.
Set back? A slight increase battery pack mass by weight difference from wafer to solder.
Benefit? out weights the risk stated above.
A less discussed challenge is electrode matching. High nickel and silicon composite anode designs both have high initial irreversible capacities. LFP does not. This means an LFP / C6:Si design would likely require online pre-lithiation. This adds big costs via capital and throughput slowdowns.
How do prismatic and 4680 tabless cells compare with regard to internal resistance? Surely the 4680 has a significantly lower internal resistance. This surely means that not only that it can charge more quickly, while being able to remove that heat more easily. This is a huge benefit that nobody seems to be talking about.
It's hard to see how you can make prismatic cells anywhere near as cheaply as cylindrical ones. Rolling the electrodes is always going to be quicker and easier to do at scale than separate plates. Since cost is a huge driving factor, it seems clear to me that cylindrical cells will eventually become the dominant form factor.
Also where is the Xpeng P5 & P7 on your chart?
I'd love to see a deep dive video on Na-ion batteries!
Great video! Do you know if there's any difference in the strength to vibration between cylindrical and prismatic cells?
An excellent question that I don't know the answer to. It's certainly a concern.
This was a great video, thanks a lot
Thank you for another fab fact video.
What about that Tesla model S whit a modified LFP battery from One power, what do you think we're the specs fill factor, density etc do you think they have sacrfied the thermal management to fit it in?
Model S won't ever use LFP is my view. It's a premium performance vehicle, not a robotaxi.
Have you had a look at the One battery that seems to be two types of chemistry used - in different modules- in one battery pack?
This shit will make me invincible when arguing with petrolheads. Thanks a lot for the super high quality video, Jordan!
Nah. There’s no way to spark logical thinking in neuroactively-challenged Petrolheads.
I'm curious as to what Elon and Tesla has in store in being able to produce their own batteries now... Awesome stuff, very insightful!
Isn't the production speed and therefore production capacity the main reason to go for the 4680 format?
It should, but should doesn't always translate to does
I bought a Model 3 highland SR (produced in Shanghai). Does it use Prysmatic cell?
Yup!
Hi Silicon anode is I hope going to be the next leap forward in battery technology. Might take another year but I think it is coming soon.