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

Why not create a custom battery

This site may earn commission on affiliate links.
Hi,

Given that the Tesla Model S is designed custom from the ground up, it surprised me greatly to learn that the battery pack is comprised of thousands of standard battery cells. Wouldn't it be more power efficient to design a custom (larger cell) lithium ion battery? There would presumably be less metal used (each cell currently used is encased in a metal cylinder), and the power density would be higher (as would the cost too).
Can anyone shed some more technical detail on this? Thanks!

mark
model S
model x reserv
 
Some bullet points to consider:
- Economics of scale
- In-house v.s. outsourcing of development/testing
- Safety/certification
- New technology v.s. tested/proven reliable technology
- Focusing your efforts on your core buiseness (in Tesla's case building a car, not batteries)
- Freedom to switch supplier as technology evolves

Some good discussion on this topic:
Why is such high voltage needed for the battery pack?
Size and weight of the 40 KW, 60 KW and 85 KW battery (which are available now)?
It's the Batteries, Stupid! (this thread is looong but read it all and you will be much wiser after)
 
I agree in the economic considerations; is there anyone who can comment on the technical/engineering? Any insight on the power/weight potential increases with a custom designed cell? Thanks!

Actually, the big aspect is 'economies of scale' - literally tens of billions of the standard cell is supplied to the full TAM, and using these keeps the price low. If Tesla created a custom cell (without other industry players picking it up) which would be likely utilize only tens of millions each year (20K/yr x batts/car+some yield loss), and thus would be considerably more expensive in a permanent way.
 
It's also not clear a bigger cell would necessarily be better for Tesla's purposes:

- You need to cool the cell. Fatter cells would have less surface area to volume ratio, so harder to cool.
- You need to arrange that the battery is safe under failure conditions - one bad cell doesn't lead to the whole battery catching fire.
Bigger cells have more energy each, hence the energy in a single failure is bigger and so more difficult to stop it spreading to adjacent cells.
- You need to get the current out of the cell. Electrode terminations could become difficult, or electrode resistance become significant,
depending what shape you made the bigger cell.

So, standard cells might not be absolutely optimum for Tesla's application, but they are probably fairly close.
 
I believe some other cars oriented batteries, such as the ones from A123 for example, are larger cells that are more of a flat/sheet-like shape. The 18650's are cylindrical, not all that unlike regular AA or AAA Duracells.

Panasonic2350.jpg
 
There's no doubt that moving from a smaller cylinder design to a larger flat or prismatic cell could make the pack more efficient, by packing space. As Arg stated though, there is an upper limit to thermal solutions as the pack gets thicker. I'm convinced that a cell triple or quadruple the 18650, especially the prismatic ones in that size could reduce the size of the pack or allow larger capacity for the same size. But once again, the TAM for 18650 is 50 thousand times larger than prismatic designs.

I'd love to know if that opinion matches anything Tesla engineers have actually gone through.

David
 
Just to add onto B5's economic reason post, using standard cells allows Tesla to switch battery suppliers very easily if Panasonic tries to jack prices or falls behind in innovation. Changing suppliers would be a simple matter of testing the new cells - no redesign of the car or battery required.
 
I agree in the economic considerations; is there anyone who can comment on the technical/engineering? Any insight on the power/weight potential increases with a custom designed cell? Thanks!
The standard Panasonic NCR18650A lithium-cobalt (a more advanced lithium nickel cobalt aluminum oxide to be precise) cells being used by Tesla right now for the 85kWh Model S are already most energy dense cells in the market right now by far at about 245Wh/kg (675Wh/L). It makes the Model S pack (and Roadster pack before it) the most energy dense battery pack in the industry despite using thousands of smaller cells (rather than hundreds of larger cells like other EVs):
http://www.teslamotors.com/roadster/technology/battery

The problem with large format cells (AKA automotive grade cells) is that they have thus far used battery chemistries (like lithium manganese or lithium iron phosphate) optimized for power density/safety/longevity, making their energy density is not nearly as high as smaller standard cells. Part of that is necessary because a larger format cell is a bigger danger when it fails (there's more energy in each cell so when one fails there's more potential damage). So it'll take sometime before battery manufacturers find a chemistry that is both energy dense AND safe enough to use in a large format cell.

Now, is it conceivable that Tesla can make a larger custom format cell using the same chemistry as the standard cells and optimize it for both energy density and safety? Yes, that is possible, but that would also involve all new R&D and tooling, which is expensive (likely more expensive that what Tesla has spend on the Model S thus far). And it likely won't result in a battery pack that is that much more energy dense (Tesla's pack is already the best in the industry). Maybe in the future it might make sense for Tesla to build their own cells, but right now it's not making a lot of sense.
 
Another reason. Packaging.

The battery box shape can be wide (S), thin, tall (Roadster), diagonal, or round. If they want to trim a corner off the flat battery pack small cells can fill the shape more efficiently. The RAV4 and Smart Car, and A Series battery boxes can be any shape are not limed by the battery form.