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

Amazing Core Tesla Battery IP - 18650 Cell

This site may earn commission on affiliate links.
I've been trying to model the battery for the Model S, mainly to get an idea of what it costs Tesla to manufacture it. As a part of that I've been doing a deep dive into Tesla battery IP. I actually considered putting this into the investors section because of how impressed I am, but the thread is intended to explore the batteries from a technical standpoint (assuming anyone else finds it interesting, lol).

At first blush, I am extremely impressed by the scope of the patents that Tesla has devoted to the 18650 cells they use (though the IP itself is not limited to the 18650 form factor).

Core Design Changes

Here are the patents that likely describe the core changes Tesla made to the battery to make it less expensive to manufacture, and more suitable for use in EV's -

Cell cap assembly with recessed terminal and enlarged insulating gasket - Patent application
Battery cell with a partial dielectric barrier for improved battery pack mechanical and thermal performance - Patent application

Fundamentally, these patents dramatically simplify manufacturing, making the batter lighter, cheaper and more suited to use in a battery pack using massive numbers of these like Tesla does. I can't say that other manufacturers are blocked from using small li-ion cells, but I suspect these patents (if enforced) could easily trip them up.

Keep in mind that batteries with this design will probably explode in any other application than a highly controlled pack like Tesla uses, so there might not be competing IP, and it might not be considered "obvious". Patenting an otherwise non-functional design might be a literal bar for anyone attempting something similar, beyond the Roadster level (which probably has it's own IP to trip folks up) which just used conventional batteries. These batteries can probably only function in a macro-pack like Tesla uses (or with additional modifications). (Edit: removed reference to air cooling)

I'd be interested in hearing from anyone who knows of another auto-maker bringing commodity format cells to market so we can see possible strategies to get around this setup. Or maybe super basic, exploding batteries are an obvious improvement? Or maybe there is related IP from the olden days?

Also, I haven't finished researching the macro-level pack patents, but I suspect Tesla has made similar efforts to keep automakers from doing massive arrays of commodity batteries.

Regardless, the primary patent (the first one linked) basically removes most (all?) of the safety features in the commercial 18650. Here are diagrams of the interiors -

First (conventional commercial battery design) -
Cell cap assembly with recessed terminal and enlarged insulating gasket - diagram, schematic, and image 02

Second (Tesla's amazing, exploding, non-functional battery design) -
Cell cap assembly with recessed terminal and enlarged insulating gasket - diagram, schematic, and image 01

It should be obvious how much this simplifies the manufacturing process. Many complicated features and manufacturing steps are just deleted, and the safety systems are handled by the central battery control system. The battery is also considerably lighter because of the ability to use aluminum (the cost difference between that and steel should be negligible at this scale). Besides making it cheaper and lighter, these changes also facilitate Tesla's manufacturing process and pack integration. I'm impressed.

The second patent deletes the plastic film on the exterior of the battery, leaving bare metal. This cuts both weight, and improves the ability of Tesla to do thermal management on the battery (a plastic film is bad for preventing fires apparently). This change also makes the battery explode (technically, it will short circuit). Non-obvious improvement.

The National Geographic special showing a blurred interior with green batteries (which is the color of the commercial versions) is a complete fake. L.O.L.

There is a related patent that adds a gasket (basically a thin film at the cylinder edges) which is enough to keep the batteries from shorting out (which is one of the purposes of the plastic cover) in the very controlled environment that Tesla has inside of its pack. I didn't link it because its inclusion is almost immaterial in terms of cost, weight or functionality, though it does make the deletion of the cover possible. It also doesn't prevent the battery from exploding. Tesla's pack and manufacturing process does that. Non-obvious.

Tesla's Amazing Cell Level Thermal Management - Battery Fire All But Impossible

Here is the critical thermal management patent that is almost certainly being implemented, that also includes key information on how the cells are integrated into the pack, as well as the changes to the interior and exterior of the cell to prevent thermal runaway -

Cell Thermal Runaway Propagation Resistance Using Dual Intumescent Material Layers - Patent application

The key feature is that after pack assembly, the cells and entire interior of the pack are sprayed with 2 layers of intumescent material. For those not familiar with it, its basically a material that when exposed to a heat source will absorb that heat, and then undergo a chemical reaction causing it to expand.

An earlier patent by this research team used a single layer, and didn't describe how it was to be integrated into the manufacturing process. In that patent, the single layer would suck heat out of the battery (significantly delaying or stopping thermal runaway) and then expand, keeping the battery thermally separated from other components. Once it got hot enough, it would char and harden. Once this char formed it created a hard thermally resistant cylinder which would direct any heat which managed to burst through the battery shell vertically through the cylinder, and away from surrounding batteries.

In the "finished" patent this charring layer is the second layer, while the first layer is able to absorb heat quicker (thus increasing the chance of the battery not bursting) and then transfer its heat to the second layer, which would provide the final barrier effect described above.

In addition, they describe a manufacturing process which would coat the interior surfaces of the cell with their own intumescent barriers, which would hopefully halt thermal runaway before it ever gets to the exterior of the battery. Again, because this just adds a quick (and cheap) step to the manufacturing process (where the battery is dipped or sprayed) it seems highly likely to be in use.

The pack assembly process is detailed, and basically the individual cells are integrated into the pack, and then the two layers of intumescent material are sprayed on the whole assembly, coating batteries and all of the interior surfaces of the pack. This leaves the metal shell of the batteries still connected to the active cooling elements, while the rest are coated. Any thermal runaway is thus shunted directly into the cooling system, while every other surface is protected by 4 layers of intumescent material (the two touching the battery, and the two on any opposing surface). And that isn't counting the intumescent layers that might be inside of the battery as well.

It seems simple, cheap and entirely fireproof, considering the small size of each individual battery. I am extremely impressed with the way that Tesla is using simple chemical reactions to snuff out thermal events before they occur, which then also create a mechanical barrier to reactions that still manage to get out of hand. This is aside from all of the active controls and safety features built into the macro-pack, including other mechanical barriers, cooling systems and power electronics.

In addition, two related pieces of IP which might or might not be implemented -

Battery Cell with a Center Pin Comprised of a Low Melting Point Material - Patent application
Battery Cell with Center Pin Comprised of an Intumescent Material - Patent application

In these patents, the stock central core is replaced with a core that either melts (to deform inwards) and/or contains an intumescent material. In a stock battery the configuration of the central core makes it more likely that the battery will burst during a thermal event. These patents cause the material to be sucked inwards, and/or helps cool down the reaction in the first place.

Seems smart, and I see no reason either would substantially increase costs over a stock core, because its basically just a different material choice. But the method might not be necessary since it looks like they are dipping the entire innards of the battery in intumescent material, and otherwise containing any thermal events with the exterior application.

Defensive Patents

Here is a basic kitchen sink patent which Tesla threw at the patent office, with a huge menu of ideas that are then fleshed out in other individual patents, and which gives an idea of alternative ways to customize small format cells. You can read it to get an idea of how Tesla went about trying to ensure that nobody else will be able to use replicate their thermal management breakthroughs, by basically patenting a large number of alternative methods -

Method and apparatus for maintaining cell wall integrity during thermal runaway using an outer layer of intumescent material - Patent application

I don't want to link and describe all of the alternate methods of preventing thermal runaway. But they involve wrapping the batteries, both inside and out with all kinds of different materials, (including intumescent materials applied differently than in the core patents) whether steel, titanium, kevlar or whatever.

There are multiple patents and ideas, all of which seem workable, but inferior to what I've decided is likely the core thermal protection methodology. It's possible that Tesla actually used some of those methods instead of what I think they did, but I really doubt it. The core patent I identified has research notes in it that the other patents lack, showing real effort being put into that program. The rest appear to be set up to make it impossible for automakers (or anyone really) to make changes to small li-ion batteries that will help prevent them from blowing up.

Maybe it will work, or maybe it wont. But Tesla clearly put a lot of thought into the alternatives and seem to have a nice defensive portfolio to keep others from following the path that Tesla is following.

EDIT: Just to be clear the cells are only unsafe if you attempt to use them individually. As a part of the pack the entirety of the system is incredibly safe!

Edit 2: Since this post/thread has been referenced a number of times in the media, I just want to clarify a few points that are important in hindsight.

First, its not certain that intumescent materials are being used in the Model S pack. In the case of the pack they built for Toyota, RAV4, they do not appear to have used an external application directly on the cells (we do not know if they coated other surfaces, or the interior of the batteries).

That said, the fire events that have occurred after striking debris seem to point to there being a significant delay between the initial impact and a serious fire breaking out. This would be entirely consistent with intumescents slowing the reaction, just as described in the patents.

Second, all of the analysis in this post about the unlikelihood of a fire in the Model S pack was based on the notion of a spontaneous fire cascading through the pack after the (possibly) spontaneous failure of a single cell. Ramming a heavy piece of metal through the armor shield would obviously disrupt a sizable number of cells, and none of the features described here could do more than slow or mitigate the resulting fire.

At that point the metal barriers that segment the pack and the capability of the pack to vent the fire into the frame of the vehicle (which then shuttles the flames to the front of the car to keep the passenger compartment and access points safe) become the key features that maintain passenger safety. The ability of the pack to vent the heat away from the passengers was a post I always intended to write up, but never got around to. In retrospect, it is clear that these features (at a minimum) work as designed in the real world.
 
Last edited:

You need to read this thread first: Inside the Tesla battery pack

"Dual Intumescent Material Layers" This is not implemented, not at least in rav4 pack.
evseupgrade.com.jpg
 

You need to read this thread first: Inside the Tesla battery pack

"Dual Intumescent Material Layers" This is not implemented, not at least in rav4 pack.
View attachment 22946

Look at the closeup. You can clearly see the goo sprayed all over everything. Nothing, anywhere, says that this stuff can't be clear.

evseupgrade.com.jpg

Look at the lower right corner where the goo is built up on the cooling elements.

- - - Updated - - -

Note also the pictures of the tops of the cells. These look like the positive side, with the built up flange that prevents short circuits, and no "bump" that is characteristic of a conventional cell. I'm pretty sure that if these were the bottoms they would just be flat, with no flange (maybe someone else can comment?). It looks like the diagram included in the patent for the simplified battery. Also no outer casing, as discussed in the patents (just bare metal).



evseupgrade2.jpg


Notice the gasket between the cell case and the cap assembly, is described in the abstract of the diagram I originally linked. Its the triangular pieces of metal covering part of the cap that the wire is touching. Here is the link to the diagram again to compare (read the abstract and look at the diagram) -

Cell cap assembly with recessed terminal and enlarged insulating gasket - diagram, schematic, and image 01

And here is the diagram for the conventional cell -

Cell cap assembly with recessed terminal and enlarged insulating gasket - diagram, schematic, and image 02

I seriously doubt these are anything but the super simple, exploding cells covered in goo I discussed in the OP.

Also, to be utterly clear, these cells are only "exploding" on their own. The system they are a part of is extraordinarily safe!

Also, not to get ahead on the macro-pack systems, but I already know it has a fuse system that is somehow centrally fused for each individual cell, and is supposedly extraordinarily cheap to manufacture and super reliable. I'll post more after I finish the analysis of the rest of the systems.

- - - Updated - - -

Might as well pull this photo into the thread so the set is complete, lol.


evseupgrade3.jpg
 
Last edited:
So my question is: How much do you think it costs Tesla for a 85 kwh and 60 kwh battery pack?

I am still working to nail down my estimates. I'll likely post a detailed estimate next week in an investors thread. It's much less than anyone in the media or investment world seems to realize. Much less than most folks on TMC seem to realize as well.

- - - Updated - - -

Looking just at the manufacturing effort required to build these cells vs conventional cells indicates a major difference. Maybe 70% fewer tasks? (thats just a quick and dirty guess)

The cost of the chemical innards doesn't change, but these must be able to be manufactured at a substantial discount, especially since Tesla will be purchasing 147m+ of these this year, which is like 4% of global production of C3 cells (which includes all commodity cells, whether 18650 or not).

The pack enclosure looks simple too, especially now that I have pictures to help visualize it. From the IP I know this is all ultrasonic welding, instead of the standard processes, and that simplifies stuff too. I thought these were conventional cells, with a modified chemistry. I didn't realize how much simpler they were than standard cells, or that it would all be based on Tesla IP. I am extremely impressed.
 
CO: your work on this and on previous issues is AMAZING! Very appreciated both from a tech enthusiast's as well as an investor's perspective!!!

I really need to link more patents when I can get back on my laptop. Tesla literally cut the protective cover off of the Panasonic battery (which breaks it, but in a good way for Tesla) and then patented the broken battery. They also claim that any battery broken in a similar fashion is also their invention, lol.

I mean, can you imagine what Panasonic must be thinking when they see this, lol.
 
Very good read. So Tesla is protecting themselves from copycats. However, would they draw fire if they tried to sue someone for copying their pack after Elon's multiple quotes along the lines of "We want to get everyone building electric cars"? I fully support them defending the hard work they've done whilst the automakers sat on their behinds, I'm just wondering if his altruistic comments may come back to bite them in their own behinds.
 
Very good read. So Tesla is protecting themselves from copycats. However, would they draw fire if they tried to sue someone for copying their pack after Elon's multiple quotes along the lines of "We want to get everyone building electric cars"? I fully support them defending the hard work they've done whilst the automakers sat on their behinds, I'm just wondering if his altruistic comments may come back to bite them in their own behinds.

To alleviate that, Tesla just needs to come out and publish licensing terms for each patent that they hold.

Basically, "Here are all the various ways you can chose to give money to Tesla if you want to do business in the state of California :). Don't you wish you built that EV-2 when you had the chance?".
 
Awesome work and thank you. I have to say that this confirms my guess that Tesla is using conventional commodity (consumer) chemistry and unconventional packaging.

I figured that based on:

1) Given that the consumer market very much cares about energy density and cost, the odds are very high that the chemistry that has the best of both characteristics will be used in consumer cells. The economies of scale further reduce the cost. So if you want great energy density and cost, the winning solution is one that uses commodity parts and builds something reliable out of them. If you can do it. All the enterprise storage vendors do this with hard drives today so this is a known approach that works.
2) Elon has gone on record as saying that the cell was modified for automotive use: that the Model S cells were no longer straight up consumer cells. So if the chemistry wasn't going to change, that meant the changes had to be in the packaging. And the changes you're pointing out make perfect sense: consumer cells have no additional safety nets. If the cell burns, you've got a fire. Tesla cells can assume they will be used only in a Tesla battery pack.

I further suspect that the charging and discharge of the battery pack has to be reasonably tuned or the pack gets stressed which reduces its expected lifetime. You could try and make the pack responsible for keeping its charge/discharge rates in line but I'm sure Tesla leaves the pack vulnerable and has coded the engine control and charging software to ensure that the pack isn't stressed. I bet that if you take a Tesla battery pack and pair it with non-Tesla engine control and charging software that you wind up with prematurely aging and even toasted packs.

This is why I've said a number of times that Tesla has a fundamental technology lead on its competitors: at least 3-5 years. Probably more like 5-8 (even ignoring the patent issues).

I think most car manufacturers think of the drivetrain as distinct well-behaved components. And they think the same way when designing the pieces of their drivetrain.

Tesla thinks of its drivetrain, battery pack and charging system as one integrated architecture and they carry that approach into the design of the major drivetrain components too. Each piece may have less than ideal characteristics but the surrounding pieces take that into account and it all rolls up into something that works.

The advantage of this approach is you can make a silk purse of out very cheap sows ears.

The disadvantages are the interdependencies between the pieces of the system and the design complexity. If the design of one of the pieces changes, there can be ripple effects throughout the system that have to be addressed. And this approach is a royal pain to get to work. Everything has to fit together like a very big puzzle. It took Tesla years and at least two major architectural revisions to get this to work.

Replicating this would be hard and take a fair amount of time. You know that it's possible and have a rough idea of how to make it work. But with systems like this, the devil is truly in the details.
 
Last edited:
Bob Lutz said there is really nothing in terms of IP for Tesla in their Model S battery technology that will prevent others big car manufacturers from building a car of the same range. It is just they all don't see a volume potential to jump in at this time
 
Bob Lutz said there is really nothing in terms of IP for Tesla in their Model S battery technology that will prevent others big car manufacturers from building a car of the same range. It is just they all don't see a volume potential to jump in at this time

This is true. They just need enough kWH packaged in the car and it doesn't have to be commodity small cells packaged in the way Tesla packages them. However, the big advantage right now for Tesla is that the 18650 cells are widely used in other industries. Tesla gets to piggy back on the wider market demand taking advantage of the economies of scale. There are reports that 18650 batteries in quantity are $125-200/kWH raw cell cost (Q4, 2012). If Telsa is paying $200/kWH, then the raw cell cost in a 85kWH pack is $17k. Further, the cell integration into a battery pack is done in house at Tesla which hopefully will drive down costs. For some, the integration costs are double the cell costs - $400/kWH but I suspect that is for low volume manufacturing (see Volt costs below). Again, since Tesla is doing the integration themselves, they can focus on the cost reduction themselves. From the limited pictures of the factory, it seems that the process is mostly automated. I would guess that the Telsa 85kWH pack costs Telsa around $20k to make right now.

For GM, they've painted themselves somewhat into a corner. First, they decided to make the first Voltec powertrain car as a Chevy instead of a Cadillac. Now they have a very slow and hard to differentiate Cadillac ELR for a $20-30k price premium. Further, they understandably chose a series hybrid approach, but this means they are going to be wedded to a concept that is very difficult to execute properly. It has two power plants and so if you drive mostly in electric mode, you are hauling around the weight of a gasoline generator and gas tank. If you are in gasoline mode, you have a much too big and therefore heavy battery pack. The resulting performance and efficiency (both driving dynamics and passenger/cargo space) suffer. As for GM's costs, the Volt battery pack is written up here:

The Real Story On GM's Volt Costs - Forbes

$6k for the 16kWH ($350/kWH)
$4k in battery pack structure, cooling, inverter, motor

But the bigger issue is the chassis - they are using bulky prismatic automotive batteries and the packaging is difficult:

http://4.bp.blogspot.com/-YDbkjzW7QAI/UBmpgxzOqBI/AAAAAAAAQm4/EBtw7AWLcD0/s1600/chevy-volt-chassis-top.jpg

Look at how much room is used for the gas generator and how little room is leftover for the passengers.
 
Bob Lutz said there is really nothing in terms of IP for Tesla in their Model S battery technology that will prevent others big car manufacturers from building a car of the same range. It is just they all don't see a volume potential to jump in at this time

I'd take anything Bob Lutz says with a large grain of salt. I've seen him on Bill Maher and the guy is dumb as a post. Not to say he didn't have a distinguished career compared to the other auto exec nimrods who ran the American car industry into the ground.
 
Bob Lutz said there is really nothing in terms of IP for Tesla in their Model S battery technology that will prevent others big car manufacturers from building a car of the same range. It is just they all don't see a volume potential to jump in at this time

Crap, I just lost a long post.

Summary:
Other manufacturers are using flat large format cells.
Tesla's model gets earky access to new chemistries that can't otherwise be commercialized.
Easy to scale up manufacturing of cells and thus batteries.
All-in-one, 0-incident battery design easy to use in grid storage applications.
If Tesla sustains lead in cheap cells, cheap batteries, fastest, most extensive charging grid and can realize Gen 3 Elon happy to see competition because competition will license Tesla technology: Tesla is making cars to make others make cars.
Not going to support Chademo except where essential in Japan because that would help Nissan, the biggest competition.
June 20th: if Elon Musk announces that he's been testing 4.6, which fixes the vampire/wake-up issue and it will be included in the now-commenced global models and rolled out to US models before winter I shall cry/cheer/punch the air.

Thanks CapO. Your post explained a lot.
 
For GM, they've painted themselves somewhat into a corner. First, they decided to make the first Voltec powertrain car as a Chevy instead of a Cadillac. Now they have a very slow and hard to differentiate Cadillac ELR for a $20-30k price premium. Further, they understandably chose a series hybrid approach, but this means they are going to be wedded to a concept that is very difficult to execute properly. It has two power plants and so if you drive mostly in electric mode, you are hauling around the weight of a gasoline generator and gas tank. If you are in gasoline mode, you have a much too big and therefore heavy battery pack.

I thought this was no longer true. The original marketing for the Volt was a series hybrid to differentiate them, but when it actually came to market, the power train was a blended hybrid. It can run in series mode some of the time, but they couldn't put a large enough battery and motor to cover the upper end of the performance chart, so it also has to couple the gasoline engine output directly to the wheels. Which makes it even more of a compromised and complicated design.
 
Yes indeed, Elon wants other manufacturers to embrace EVs. But the way things are going, by the time they NEED to get to market, licensing Tesla's technology will be the only logical path available. I just have this vision of the car makers (and specialty truck makers) ignoring EVs until Tesla gets big enough that Tesla severely impacts their sales. The game will be over by then.