Tesla Buys Nearly 2 Billion Panasonic Cells

hard to believe tesla chose to go with hundreds of small Panasonic NCR18650A 3100mAh cells
must be labour intensive to piece all those cells together but i imagine it's an automated process
small format cells have smaller rate of failure than larger ones but pain to work with
 
davec said:
hard to believe tesla chose to go with hundreds of small Panasonic NCR18650A 3100mAh cells


NCR18650A is nothing like what modern Tesla cars use, and would make a very poor choice of EV cell. They run a purpose-made for EV applications 18650 cell that uses a NCA (Nickel Cobalt Aluminum) based cathode material.

Why they hook thousands of tiny cells together is a bit odd and very challenging to do right, but they did find how to do it right and make it work well, and now they get to buy the cheapest $/Wh and leverage mature 18650 manufacturing processes.
 
Didn´t ive had read that robots put the packt togetha ?
And that they use smart BMS with cluster cell conection that when 1 or more cell´s is dud. teh pack keeps operating like nothing happened ?
 
So... crashed Tesla will be a source for future cells ? :mrgreen:
 
cal3thousand said:
So... crashed Tesla will be a source for future cells ? :mrgreen:


One crashed model S could provide cells to make >100 ebikes. :)
 
Elon Musk and his team of engineers have taken some criticism in the beginning, and outsiders have speculated that the packs made from thousands of cells were a compromise. However, it my understanding that the Tesla cars have more than one full-voltage sub-pack to them (I could be wrong). If a cell in a string begins to get hot, the computer cuts that string out of the pack-grid, long before it can become an issue...and the car can still drive on with slightly less range to the pack.

A dealer who has a Tesla come back in because of a "check battery pack" code, can quickly identify which string of cells has the bad one. It is a quick and simple matter to swap in a new string, and send the customer on his way. I now think this is a pretty damn good way to do it.
 
Billions of these lovely cells moving through the Tesla facility in Fremont -- 20 miles from me -- over next couple years?

Has to be good news.

First thought is it's time to start buying lunch for some of the guys on the Receiving dock :lol:
 
I imagine 2 billion cells is enough to make a significant impact on Panasonic's manufacturing. That's real volume.

20 billion Wh, or 20 million kWh... WTF is that even right?

That would translate to ~300k 60kWh cars...
 
footloose said:
Billions of these lovely cells moving through the Tesla facility in Fremont -- 20 miles from me -- over next couple years?

Has to be good news.

First thought is it's time to start buying lunch for some of the guys on the Receiving dock :lol:

My thoughts exactly! :lol: :lol: Keep us posted on how the dumpster diving goes.
 
My thinking is that they will have a recycling scheme with as closed a loop as they can manage, both for cost and public image reasons.

I always reckoned that their choice to use a pack with thousands of cells was more about fault tolerance and thermal management than it was about reducing cost to the utmost degree. It seems self-evident to me that at a given production scale, packaging thousands of tiny cells would be inherently more costly and less weight-efficient than using a small number of large format cells to achieve the same capacity. They must be buying some kind of engineering benefits with those tradeoffs.

Tesla Motors is not the sort of outfit to spec their battery arbitrarily just because it's what happens to be cheapest at the moment they design the car.
 
The chance of defect goes up with the size of the cell. By using smaller cells when you have a defect you only have to scrap a small "area"

Imagine one large format cell unrolled flat. Imagine flat "1 layer" battery cut into 10 smaller cell layers. Now if 1 of 10 smaller cells roll area is defective you have to scrap the whole 10 cells because in reality it is not cut up. If you just had those ten smaller cells made individually you only have to scrap 1 cell if there was the same defect rate per area.

The Tesla pack is made 100% placed and wire bonded by robots so there is little time advantage of going to a large format cells that has less connects. Also if you have a 24s large format pack lets say for example. If one cell goes bad you lost 1/24th (4.2%) of your capacity and maybe the whole pack might need to be pulled. If you had a one of 96 18650 in parrallel go bad on a 24s96p pack you would lose 1.0% capacity. Double check my math
 
flathill said:
The chance of defect goes up with the size of the cell. By using smaller cells when you have a defect you only have to scrap a small "area"

Imagine one large format cell unrolled flat. Imagine flat "1 layer" battery cut into 10 smaller cell layers. Now if 1 of 10 smaller cells roll area is defective you have to scrap the whole 10 cells because in reality it is not cut up. If you just had those ten smaller cells made individually you only have to scrap 1 cell if there was the same defect rate per area.


Conceptually this is true. In reality, fallout on mature processes for both 18650's and large format pouches is well below <0.1% for any respectable mfg right now, so scrapping a large cell vs scrapping a small cell makes no practical difference.


flathill said:
The Tesla pack is made 100% placed and wire bonded by robots so there is little time advantage of going to a large format cells that has less connects. Also if you have a 24s large format pack lets say for example. If one cell goes bad you lost 1/24th (4.2%) of your capacity and maybe the whole pack might need to be pulled. If you had a one of 96 18650 in parrallel go bad on a 24s96p pack you would lose 1.0% capacity. Double check my math


Yes, if a Tesla pack runs say 100p100s (just using make up numbers for the example, though that's not far off of reality), they can have any cell fail anywhere in the pack, and it only costs them 100cells worth of capacity (~1% pack energy) and the pack still functions fine. If they lose an additional cell in the same parallel group they lost the first cell, it costs them another 100cells worth of capacity (~1% pack energy), yet if they have another cell fail in any other group, it costs them no additional pack energy. In this respect, a pack made from a ton of tiny cells is pretty robust to single cell failures (when using Tesla's unique clever architecture at least).
 
:shock: :shock: :shock:
ver2_electric_bike_2aegb.jpg

Trying to see if I could fit that someway on my bike :p :lol: :lol: Shared Battery /frame possible. 8)
 
I can tell you from experience, with just 5kWh on a bicycle, I can tire of sitting on a bicycle seat long before the pack is empty. :)
 
speedmd said:
Charging once a week, several times a month may not be all bad. As long as it was not too heavy/ costly it would be a game changer. :wink:


With a 10-20x improvement in battery energy density, we will be there. :) It's inevitable that it will come, but I think it's likely my beard will be closer to grey than brown when it happens.
 
spinningmagnets said:
Elon Musk and his team of engineers have taken some criticism in the beginning, and outsiders have speculated that the packs made from thousands of cells were a compromise. However, it my understanding that the Tesla cars have more than one full-voltage sub-pack to them (I could be wrong). If a cell in a string begins to get hot, the computer cuts that string out of the pack-grid, long before it can become an issue...and the car can still drive on with slightly less range to the pack.

A dealer who has a Tesla come back in because of a "check battery pack" code, can quickly identify which string of cells has the bad one. It is a quick and simple matter to swap in a new string, and send the customer on his way. I now think this is a pretty damn good way to do it.
I Know! That's why sometimes thinking smaller aka micro is better!
Elon Musk is my hero!
 
davec said:
hard to believe tesla chose to go with hundreds of small Panasonic NCR18650A 3100mAh cells
must be labour intensive to piece all those cells together but i imagine it's an automated process
small format cells have smaller rate of failure than larger ones but pain to work with
It makes perfect sense to me, as pretty much everything Tesla has done to date to design, build & sell the most amazing vehicle ever put on the road. Its a long-term cost saving measure. See, a life-time of battery is built into the price, so Tesla really owns them. They've got a niffty battery swap machine at each of the their solar charging stations, so the batteries will be coming off on a regular basis. I'm confident that they'll then charge & test. If any cells go bad, as they certainly will, then you've got the smallest amount of battery to replace. In other words, they've atomized the battery capacity, so as to reduce costs over the long-haul. By locking in a contract with Panasonic, they have once again shown their genius.
footloose said:
Billions of these lovely cells moving through the Tesla facility in Fremont -- 20 miles from me -- over next couple years? Has to be good news.
First thought is it's time to start buying lunch for some of the guys on the Receiving dock :lol:
Keep us in the loop please and if you sneak in an off-the-books purchase (heehe), triple the number for us screamers.
 
I think Chalo is right, they are going to try to keep a very closed loop system...
but there is always some leakage.

Even armories have difficulty accounting for, say, all the "stuff" they are supposed to keep track of.
Can a manufacturer keep track of 2B cells?

A friend has a Tesla. Terrific car. And I saw one last year with its "skin" off. Very clean! Take away engine, transmission, cooling systems, drive shaft, differential, etc etc... Picture what is left. Body pan with smallish motor on each corner.
I drooled.
 
“This will be a giant facility. We are talking about something that is comparable to all of the lithium-ion battery production in the world — in one factory,” Musk said during a conference call with analysts to discuss the third-quarter earnings. “It’s big.”

While Musk said he didn’t want to divulge too many details about the battery factory plan, he did have this much to say:

*The factory will most likely be in North America.

*It could be a “soup-to-nuts” factory that takes in raw materials and produces the finished battery packs (as opposed to producing only cells and hiring another company to assemble the packs off-site, for example).

* The factory will have eco-friendly features, such as the use of solar power and a recycling program that makes use of old battery packs to make new battery packs. Musk, incidentally, is the chairman of SolarCity SCTY -15.79%, a solar panel installer in California.

*Tesla may tinker with the format of the battery cells. Or not. The current cells from Panasonic are the common cylindrical type. Musk said Tesla would likely stick with the same size and format. But if it does make design changes, then the cells wouldn’t be much bigger. “We have yet to see a situation where large format cells are actually cheaper on a cost per kilowatt-hour basis,” he said.
 
spinningmagnets said:
A dealer who has a Tesla come back in because of a "check battery pack" code, can quickly identify which string of cells has the bad one. It is a quick and simple matter to swap in a new string,

i dont think they will have the ability to disassemble and repair packs at dealer level.
Replace the entire pack more likely.
Can anybody confirm ?

...if a Tesla pack runs say 100p100s (just using make up numbers for the example, though that's not far off of reality), they can have any cell fail anywhere in the pack, and it only costs them 100cells worth of capacity (~1% pack energy) and the pack still functions fine.
I can see where losing 1 cell in a 100P string would loose 1% capacity in that "STRING" , but why would it cause the loss of the whole string ?
I thought each cell has individual protection, such that it can isolate itself and allow the rest of the parallel string to function at 99% capacity. ?
 
Back
Top