The Model S battery

The Model S battery

I just heard Elon say that the Model S will feature a 90 kWh battery (300 mile version i presume) and that the gravimetric energy density has incresed about 30-35%.

I also heard him say that the cells in the Model S uses a quarter the amount of cobalt compared with the roadster cells.

Based on this, and the roadster specs, the weight of the Model S battery will be about 535-556 kg, and the chemistry used in the cells is probably NCA (LiNiCoAl02 cathode and graphite anode).

Panasonic has NCA cells scheduled to begin production in 2012

The cycle life of these cells seem the be very good if you cycle them between 4.1 and 3.0 Volts (80% capacity after 5000+ cycles)

So what i am trying to get at is what voltage range the cells will be kept in and how many cells the battery consists of.
If kept between 4.1 and 3.0 Volts the cells capacity will be about 85% of the rated capacity.
A 90 kWh battery built with Panasonic 3.4 Ah 18650 cells will then require about 8650 cells.
If you would use the rated capacity, 12.24 Wh per cell, then the number of cells required would be about 7350.

If you assume the ratio between cell mass, and total battery pack mass is the same as the roadster, about 0.66, then the number of cells would be 7820 (using 545 kg as total weight).
That gives a 11.5 Wh capacity per cell, or 94% of rated capacity.

So that do you think? are my assumptions right about the NCA chemistry? how many cells does it use? do you think they have improved (incresed) the total cell mass/battery mass ratio? what voltage range will the cells be kept between?

Timo | 15 mars 2011

I think the Model S battery pack packaging is quite a bit lighter than Roadster battery pack in comparison. Instead of having additional structure to carry heavy battery pack weight which has separated structure of its own, Model S battery pack *is* the structure that carries the weight.

peterbjarnholt | 15 mars 2011

So you think the ratio is higher than 0.66?
Since the battery pack augments the chassi in such a way that the rigidity of the car increses, i was tinking the battery must have som rigidity of its own, so maybe the ratio is not so much higher, but personally i think its between 0.75 and 0.66. In fact 0.68 is my guess because that makes the number of cells about 8000 =)
At 0.75 its 8885 which sounds like a bit to much.

William13 | 15 mars 2011

I think you should check out Eberhardt's posting in under model s where he found a PDF. File from panasonic on their batteries. Note pages 17 and 18 for cells labeled NCR~18650A. These appear to be the new Model S batteries.

The packaging appears to be more efficient in the model s than the roadster. There is some concern raised that there may be additional packages of batteries to get the extra kWh.

I would like to note that with larger capacity batteries we get longer lifetime as fewer cycles occur. That is a 160 pack would go through nearly twice the number of cycles as a 300 pack over 100,000 miles. Of course there is mention of a calendar lifetime to these cells also which we can not know until the future occurs.

William13 | 15 mars 2011

I think they are using 8000 cells that are nominal 3.1mAh

William13 | 15 mars 2011

90,000Wh/(3.1mAh x 3.6 V)= 8065 batteries

Timo | 15 mars 2011

nitpick: Ah, not mAh. 3100mAh.

William13 | 15 mars 2011

Thanks Timo I was looking at a poor copy and mistook a , for a . and forgot to check units.

peterbjarnholt | 16 mars 2011

I read it, but i dont think they will use a 3.1 cell in the model s (300 mile version), because then they would need to use the whole capacity of the cell to get 90 kWh at around 8000 cells.

The roadster cells are kept between 4.15 and 3.0 Volts and thats about 93% of rated capacity. I dont think the model s cells will use a wider range than that, in fact i think they will use less, to get better calander life (at high voltage level the cells degrade faster)
So 4.15-4.10 to 3.0 is my best guess because that would give 90 kWh out of 8000 3.4 Ah 18650 cells.

Timo | 17 mars 2011

I think you have misunderstood something and are arguing with yourself. 3.6V is the voltage that is used to give 3100mAh. That is between 4.15 and 3.0 volts.

peterbjarnholt | 17 mars 2011

I assumed when William13 said 3.1 Ah cell he meant the cell rated at 3.1 Ah in the pdf-document above. My point was that it cant be that cell if you want to keep the voltage between 4.1-4.15 and 3.0 volts and the number of cells at 8000 or a weight ratio near 0.68.

A cell rated at 3.4 Ah and 3.6 V nominal voltage discharged between 4.1-4.15 and 3.0 equals a cell rated at 3.1Ah and 3.6 V nominal (can store same amount of energy) =)

peterbjarnholt | 17 mars 2011

William13, i liked your last point there about double cycle life on 300 mile version vs 160 mile version (assuming same driving habits of course). I have thought about that to =).Also, for best overall battery life, you should probably keep your car between 80%-20% SOC as often as possible, and live someplace where is moderately cold (5-15 degrees celsius). The former will probably be easier on the 300 mile version.

Timo | 18 mars 2011

I think in that case you got the weight ratio wrong. Adjust that. 3.1Ah * 3.6V * 8000 cells gives ~90kWh, which is pretty much confirmed by Tesla for 300 mile battery pack. 8000 cells weight around 368kg. Roadster battery pack weights 450kg and has quite a bit less cells 6831 to be exact. Each cell weights around 45g so that is 307kg for cells alone. Capacity is 56kWh

56kWh/307kg = 182Wh/kg
3.1 * 3.6 / 46g = 242Wh/kg

242/182 = 1.33 = 33% increase in energy density.

There is no indication whatsoever in your initial post about entire battery pack weight, only cell weight. You have to guess the entire battery pack weight. My guess is that difference is even better when you compare entire packs & additional structures required to carry those packs.

peterbjarnholt | 18 mars 2011

Okey so you think Elon means 30-35% gvavimetric energy density on the cells themself and not not the entire batterypack? Because i think me meant the totalt battery pack, sorry if that was not clear in my post.

But since i think the ratio on the model s is close the roadsters ratio, which is about 0.67, it makes no difference.

But if the ratio is higher on the model s, like you think, then it makes a difference.

So if you are right about the ratio beeing higher on the model s, then maybe i can agree that it is possible that they will use a cell like the one in the pdf rated at 3.1Ah, if it was not for the fact that the number of cells whould have to be closer to 9000.

These calculations are based on that the cells are kept between 4.15-4.1 and 3.0 volts.

8960*3.6*3.1*0.90 = 90 kWh

8170*3.6*3.4*0.90 = 90 kWh

So if it is the cells rated at 3.1 Ah then:

8960*0.046 = 412 kg

6831*0.045 = 307 kg

412/307 = 1.34 = 34% increase in energy density.

And since the number 8000 was pretty much confirmed by Tesla as you said, i can not se how they would be able to use that 3.1Ah cell in the pdf.

Timo | 18 mars 2011

Where did you get that 0.9 there? From discharge curve? Or do you have some more detailed data-file than Panasonic industrial pdf?

Based on this:

With typical 0.2lt discharge 3.0V is reached at about 3000mAh, not 2800mAh.

peterbjarnholt | 18 mars 2011

The difference in capacity going from 4.2-2.5 V and 4.13-3.0 V is about 10%, hence the factor 0.90. I know this from experimental testing, but you can also read up on it at

That 0.2C discharge curve is from 4.2 to 2.5 V, and that gives a capacity of about 3.1 Ah at 3.6 V nominal = 11.6 Wh

If you were to charge it to lets say 4.13 V and then discharge it to 3.0 V then that would give about 0.9*10.8 = 10.0 Wh

10.0/3.6 V nominal = 2.8 Ah

90000/10.0 = 9000 cells...

Timo | 18 mars 2011

AFAIK Tesla uses battery pretty much max capacity in range mode, not much safety buffers in either end of the charge curve, however even if they don't I'm pretty sure that they still get that 300 miles with 3.1Ah batteries and ~8000 cells. That would mean that they use nominal capacity for battery size measurements but don't use it completely to get that result.

It also looks to me that using that nominal discharge of 590mA you really do get that near 3100mAh capacity even with 4.13 and 3.0V limits. At least with that battery type (Nominal capacity is bit misleading way of telling how much capacity you actually have IMO, looks like it varies quite a bit with different discharge loads and temperatures. Kind of like EPA mpg ratings).

peterbjarnholt | 18 mars 2011

Oh really? That is great if you are right, because that means you have the ability to use the entire capacity if you want to =)
But that doesnt mean you have to i hope (i hope they let you choose a narrower voltage range if you want to).
It also means that they would get even better range with the 3.4 Ah cells :D

It might look like it on that graph, but im pretty confident in that you loose about 10% of the capacity going from 4.2-2.5 to 4.13-3.0 Volts. But lets say its 4.15-3.0, that would give about 93% of full capacity.

90000/(3.1*3.6*0.93) = 8671 cells

That is still a bit much don't you think?

But if they use the entire capacity as you say, then im with you on that they probably use a 3.1 Ah cell in the 300 mile Model S.

peterbjarnholt | 18 mars 2011

I should add that all that is based on 90 kWh, if it happens to be 85 kWh, then

85000/(3.1*3.6*0.93) = 8190 cells

and that is feasible.

Straight Shooter | 18 mars 2011

Cool stuff indeed. As for battery options, Tesla can't offer a myriad of batteries in range vs weight configurations. I can see this being an option in the longer term, like a few years from now, but not in the short term, as it is too costly. The Model S won't really benefit (from an accelleration and handling perspective) from it anyway as it is already very heavy. That said, if they offered it in longer ranges, I'd see 95% of the customer base always taking the longer range. These are family cars after all, and they are meant to travel long distances.

The Roadster can be a huge benefit and it is light and taking off 250-400kg will result in massive performance improvements. Accelleration will reach supercar levels. Handling, if all the weight disappears from top 1/2 of the battery will be way better too.

Tesla knows this and you'll probably see it as an option on the 3.0 version of the Roadster, but for now effort is on the Model S and the SUV crossover.

I don't think you'll ever see a "light weight" battery option in the SUV, as that is not the purpose of the vehicle. Elron has recently stated that the SUV will target SUV owners, but he is also trying to make this vehicle look (function) like a mini-van. That is the "gold mine" of vehicles to aim for and 100,000s of these are sold every eary in the USA alone. Who in their right mind would order an electic mini van with anything less then the biggest possible battery in it? They are the most travelled and heavily abused vahicles on the road going from coast to coast to coast to boarder countless times. Expect to see a 700mile and 1000mile battery in these.

William13 | 18 mars 2011

The roadster is rated 56 kWh nominal with 53 usable. Elon said 90 kWh for the 300 mile pack and 8000 cells were mentioned as battery size. WhenElon talks about the size of the roadster battery he says 56 kWh. Nominal means nominal thus: 3.1* 3.6*8000=90 kWh. I also would like more as in a 3.4 size but I think this is one more year away. Tesla needs to be comfortable about endurance and they test the batteries themselves. There will always be a lag before the new cells will be available.

William13 | 18 mars 2011

Also the batteries are pated tighter/ more efficiently in the model s.

Listen to the three engineering videos.

William13 | 18 mars 2011

Packed tighter.

Tim10 | 19 mars 2011

I believe there was investors presentation where they mention the 300 mile pack would use 1 generation old batteries. If 3.4AH are current generation when the S is launched that would fit with 3.1 cells being 1 behind.

William13 | 19 mars 2011

Precisely Tim10. Elon wants to test batteries prior to instillation.