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

How many kWh can they squeeze into the Model 3...?

This site may earn commission on affiliate links.
The 2170 project is not an exercise merely be different. I've read multiple detailed estimates where the energy density per VOLUME is improve by at least 10% for the cell alone. Less wrapper, more content. Furthermore the 5mm taller cells are said to fit standing up in an S/X pack. Nearly 8% increase in pack volume utilization right there. Consider that the current 90kWh packs already show charging behavior that suggest them being 100's running into some sort of limitation.
And all this is before we factor in a lower volume % in plumbing.
Nice post. I agree with your numbers. If you calculate the free vertical space gained by going from 18-65 to 21-70 you get a cylinder slice of 21-5, which, percentage-wise equates to about a 10.5% gain in space over the 18-65 format.

Subtract any packing inefficiencies from the increase in horizontal dead-space, and I would believe something on the order of an 8% improvement in overall density for a given pack size.
 
  • Like
Reactions: Cloxxki
I was incorrect,
P90DL is effectively a 220-240kW car at the Nurburgring. Still below half the claimed power figures which are available for mere seconds.
the car struggles for 200kph after entering a straight well over 100kph. Not remotely ludicrous, just a heavy car that's not racey.
Now this same drivetrain in a smaller lighter car, that could make a lot of difference. from 8:50 to 8:20? The last 50 seconds would need to come from cooling (or risk) that enables more like 400kW to remain on tap. How to do this? A big P motor also in the front? Itt would not help the weight but it would perhaps increased the long term maximum load to 300kW? Perhaps investing 100kg extra in cooling would net more time gains.
 
Nice post. I agree with your numbers. If you calculate the free vertical space gained by going from 18-65 to 21-70 you get a cylinder slice of 21-5, which, percentage-wise equates to about a 10.5% gain in space over the 18-65 format.

Subtract any packing inefficiencies from the increase in horizontal dead-space, and I would believe something on the order of an 8% improvement in overall density for a given pack size.
Thanks, but we seem to have differing details in mind.

The space freed up is effectively 5mm on top of a 18650 which is 7,7%.
The 2170 being fatter and taller makes that less volume (and weight) by % is used for packaging, and more for energy storage. According to people better informed than I, this is 10% on a power to volume density. So he 2170 is up to 46% bigger (2070 would be 33%) than the 18650, so proportionately fewer cell fit inside a given volume. Of course, +/- a different plumbing layout/system which might free up some space for extra cells. These cells are placed upright in the same fashion as 18650's in the past, and the 5mm extra is made to somehow fit.

This bring the capacity of the S/X pack, based on a full 100kWh today (or announced), to x1,10 x1,08 to about 115-120kWh.
 
I don't have all too high hopes that Model 3 is some featherweight with long range. If Model S is a porker for a large luxury sedan, then 3 will be a porker for a mid to large hatchback.

To me, a 'featherweight' vehicle begins at something below 2,800 lbs. None of the direct competitors to the Tesla Model ☰ qualify as such. The BMW 3-Series used to show the weight of the M3 as over 4,100 lbs (maybe that was a convertible?), but mysteriously it is now listed at about 3,500 lbs instead. The AUDI A4 is a few pounds under 3,700 lbs. The Mercedes-Benz C-Class is 3,800 lbs.

The creation of ever-more-stringent crash test requirements has led to cars getting bigger and bigger to resist such unlikely situations as a rollover that somehow exerts five times the weight of the vehicle on its rooftop. That has resulted in heavier vehicles that get worse and worse fuel economy. The gunslit windows in vehicles that meet those requirements no longer allow clear vision from the driving position, so rear view cameras and displays had to be added for safety, along with sensors to warn of the presence of nearby unseen objects. All of this leads to so-called 'Compact' cars that weigh as much as or more than Midsize cars did 25 years ago.

Yes, the Model S is a porker. And versions of all of their direct competitors' cars are similarly porkers. So what? It amazes me how weight is always a primary consideration the instant American cars are compared to European brands. But somehow, automotive journalists repeatedly claim that the overweight German cars in particular make up for it with 'superior handling' and 'driver engagement' that allows you to 'feel the road'. All the same trite expressions I've been reading for close to thirty years now. Somehow, when a German coupe is over 4,000 lbs that leads to a 'feeling of stability' and a 'command of the road'. Meanwhile, an American car that is 3,500 lbs is referred to as 'portly' and 'in need of a diet' whether it is a sedan, two seater, or two-plus-two.

I believe that the maximum fully equipped weight of a Model ☰ will be significantly less than that of a barebones version of the Model S. I also believe that a fully equipped Model S will weigh 800-to-1,000 lbs more than a fully equipped Model ☰. Thus, the Model ☰ will very likely weigh in a range of 3,700 lbs to 4,100 lbs dependent upon trim level.

A slightly shorter wheelbase on the 3 is not going to force Tesla to half the pack volume compared to Model S, surely? And there is no rule to keep a pack exactly a certain thinness. A 1-2cm thicker pack can add dozens of kWh. Here's to hoping Tesla reserved a space nearly as big as on Model S/X to pack 90-100kWh from the get-go. It needs it to stun the world. For a new Tesla, the bar is set high.
The battery pack on the Model ☰ will be plenty large enough. Remember, the battery pack in the 56 kWh Tesla Roadster held more battery cells than the one in a Tesla Model S 85. JB Straubel expects a similar improvement in energy density -- more energy held within the volume of fewer battery cells -- with the Model ☰ as compared to Model S. He puts it at 'about 40%'. I am inclined to believe him.

Thus, if someone believes, as I do, that the Model S will someday offer anywhere from 120 kWh to 170 kWh within the formfactor of its current battery pack... Then it is not hard to accept that a Tesla Model ☰ may be able to contain around 80% of that amount. Thus, 100 kWh to 135 kWh, perhaps as a maximum limit either initially or eventually.
 
  • Like
Reactions: Cloxxki
Volumetric energy density is not the problem in the Model S. And the increase in energy density they have talked about is in gravimetric energy density, which is the limiting factor on the Model S.

Now, for the 3, it could be that volumetric energy density is a problem, but likely the switch to 2170's fixes that issue, and the problem is still specific energy/gravimetric energy density. At best, we're probably talking 180 Wh/kg with the next generation of cells and battery pack. At 100 kWh, that's 556 kg, or 1,225 pounds, or roughly equal to the original 85 kWh pack and the current 90 kWh pack. Going to 120 kWh would mean 667 kg, or 1,470 pounds. That's not happening with the next increment in specific energy.

That also rules out 100 kWh in the Model 3 at first... 1,225 pounds for the battery is a lot in the small form factor.
 
I'm finding a number of references online to Elon saying back in 2014 that the energy density at the cell level was 260 wh/kg. Again, this is at the cell level, in 2014. I can't find Elon's quote, but I see a number of people using that number.

In that vein, Randy Carlson at Seeking Alpha just wrote a humdinger of an article today titled "Tesla, Already There". http://seekingalpha.com/article/4001134-tesla-already

You have to either register, or if you have Safari you can view each page in "Reader Mode". Just increment the page numbers.

He takes a deep dive into battery chemistry and reverse engineers the numbers from the 60/75 software upgrade as well as the forthcoming 100kWh pack -- and concludes that Tesla is using new battery chemistry allowing for a higher voltage that makes both of those pack sizes possible.

His conclusions:
- Tesla is using a higher per cell voltage with new battery chemistry
- This new chemistry allows for a 60/75 software upgrade as well as making the 100kWh pack possible
- This new chemistry reaches 385 wh/kg at the cell level
- A 53kWh pack on the Model 3, combined with a low drag coefficient will yield a 257 mile range

If there's a better definition of Tesla battery porn, I'd like to hear it. Those numbers look ridiculous and if true, would be an epic mic drop on the entire industry.

Hoping someone else can fact check his numbers and assumptions.

[As a disclaimer or other point of reference, he also wrote a long article a few months back on how Tesla might use air-cooling on the Model 3 battery, which is possible but seems unlikely.]

2u09HGk.jpg
 
Last edited:
  • Like
Reactions: Troy
I have a feeling that Randy Carlson is a pretty cool guy. I do tend to disagree with his numbers, which suggest it is possible to do a lot more with less than I imagine. Maybe, if the Model ☰ were going to be a Volkswagen Golf shaped econobox squeezed into a form factor similar to the Fiat 500, that miraculously weighed under 2,800 lbs...? Then these numbers could work. Maybe. If it used tires with a 165 mm wide tread that were inflated to 60 psi. And you had a tailwind, and were going downhill.

EPA range is calculated based upon energy used in total and through five different cycles of testing -- not at a constant speed. 53 kWh in a battery pack would likely be treated as if it were actually 62 kWh of energy expended. So the EPA range would end up being only 218 miles instead, which Tesla Motors would round down to 215 miles -- as they already have done when claiming 'at least' 215 miles as an EPA range for Model ☰.

Or, to look at it another way... 257 miles using 53 kWh of energy comes to about 206 Wh expended per mile on average. And 290 miles using 60 kWh of energy comes to about the same. But for a Midsize vehicle that will probably weigh in the 3,700 lbs to 4,100 lbs range, I find this extremely unlikely. Especially if it is to achieve Performance figures that equal or that surpass other vehicles in class.
 
  • Like
Reactions: Cloxxki
Or, to look at it another way... 257 miles using 53 kWh of energy comes to about 206 Wh expended per mile on average. And 290 miles using 60 kWh of energy comes to about the same. But for a Midsize vehicle that will probably weigh in the 3,700 lbs to 4,100 lbs range, I find this extremely unlikely. Especially if it is to achieve Performance figures that equal or that surpass other vehicles in class.
I totally agree with you on the optimistic nature of his numbers. This guy is really bullish on the company.

So, given his predictions of battery chemistry improvements he's also predicting that the Model 3 will weigh less than a comparable BMW 3-series car. He doesn't give a number, but that would likely put it less than 3,500 lbs -- which is how he gets to those range numbers.

The key point of his analysis seems to center on the curb weight differences of the new 60/75D vs the old 70D, versus the battery counts of each vehicle done by Tesla Tap. He's saying the weight differences don't add up, and that a chemistry change must be in the mix.

He also concludes that the 60/75 car and the 90kWh cars are using different battery chemistries (aka running at different cell voltages).

That the 90kWh cells are the original cells running at 4.2V and that Tesla is using their new chemistry cells in the 60/75 vehicle, running at 4.8V. The 100kWh pack will get the new 4.8V chemistry (giving it a 342 mile EPA range).

The Model 3 will get the new chemistry plus the size improvements, putting it at 5.0V per cell.

Although, these voltage numbers seem to contradict his Model 3 slide...

Anyway, it'll be interesting to see how it pans out. His hypothesis should be testable soon, once the 100kWh pack appears. If the gain in range exceeds the 11% gain in energy, then we can be pretty sure there's a change in weight due to a change in chemistry.
 
Last edited:
  • Like
Reactions: Red Sage
I am not going to read the SA article. Claims of cells capable of being charged to greater than 4.2 V need to be substantiated. Government and industry research on high voltage electrolyte-cathode combinations has so far discovered many options that do not work. Maybe Panasonic has discovered a breakthrough, but if it really is in production 75 kWh Tesla packs, there would be patents and other literature available. Without an announcement from Panasonic or third party testing of the cells, these claims are likely just BS.

GSP
 
I was incorrect,
P90DL is effectively a 220-240kW car at the Nurburgring. Still below half the claimed power figures which are available for mere seconds.
the car struggles for 200kph after entering a straight well over 100kph. Not remotely ludicrous, just a heavy car that's not racey.
Now this same drivetrain in a smaller lighter car, that could make a lot of difference. from 8:50 to 8:20? The last 50 seconds would need to come from cooling (or risk) that enables more like 400kW to remain on tap. How to do this? A big P motor also in the front? Itt would not help the weight but it would perhaps increased the long term maximum load to 300kW? Perhaps investing 100kg extra in cooling would net more time gains.

Car in that video is P85D
 
  • Informative
Reactions: Red Sage
I like the combination of Brownian motion and confirmation bias in the way a lot of people think. It tests limits.

I think it is easier to get area than half cell voltages that are safely significantly different. 85kWh at introduction. More later.

Just a beginner at speculative thinking. But working on it. Lot of good teachers here.
 
  • Like
Reactions: Red Sage
I am not going to read the SA article. Claims of cells capable of being charged to greater than 4.2 V need to be substantiated. Government and industry research on high voltage electrolyte-cathode combinations has so far discovered many options that do not work. Maybe Panasonic has discovered a breakthrough, but if it really is in production 75 kWh Tesla packs, there would be patents and other literature available. Without an announcement from Panasonic or third party testing of the cells, these claims are likely just BS.

GSP
I'm not sure we should rely on US government research for the advancement of BEV's.
And what if it were Tesla making the discovery? Would they even need to patent it, considering no-one will be able to produce near their cost for the next few years?

But I agree, the evidence in the article is shaky.
Next time someone wants to break open a Model S pack, let it be the 75kWh to count the cells :)
 
Last edited:
I said "government and industry research," not just the US government. That said, the US DOE annual budget for battery research is about $100M.

Tesla is not doing battery research as far as I can tell. They are relying on their supplier Panasonic to do all of the research, design, and all manufacturing of the cells. However, Tesla has contributed to the design of the can that holds the active cell materials.

GSP
 
  • Like
Reactions: Cloxxki
Tesla is not doing battery research as far as I can tell.
Not so sure about that... Maybe not at the level of the US DOE or Panasonic, but they sure know what to order from Panasonic, and as you was touching into they have Panasonic to produce a custom cell for their own use. Yes, we know about the contribution to the cell design, but it was also hinted about some "spacial sauce" in the electrolyte? It may hint at some research at Tesla. Anyway are they working very close with Panasonic on this. As Panasonic said - Tesla is not just an customer any more, their cooperation has developed well behind that.

But anyway - I'm not claiming that they do any advanced battery research, just will not exclude the possibility.
 
Maybe I misunderstood something but I've been under the impression that the power limiting is due to battery heat, how would bigger front motor help with that?
I was under the same impression until recently. A Tesla connaisseur said it's actually also the motor stator that heats up, which makes sense. The battery has a large volume, it might take longer to reach sub optimal heat levels.
I'd like to know, which can be pressed 100% longer in a straight line, 90 or P90?
 
Another comment on the SA article. The author references a Gigafactory retooling as evidence of some kind of chemistry change... I believe this in mentioned during the recent tour as an upgrade to one of the manufacturing stages devised by Tesla & Panasonic engineers that consumes 80% less power.

Also the numbers he's throwing around seem strange. Li-on cells operating already at 4.9V? Says who?

God bless this guy's enthusiasm but o_O

Batteries have improved cost/density an avg. of 8 - 9 % year on year and the leading experts in the field expect this trend to continue - yet one does love to let the imagination run wild!