My take on the new 100kWh pack

[emir-t]

So I'm really in to figuring out Tesla's battery architecture and I love reading stuff about it. Now we know that the new 100kWh pack is a completely new architecture, including cell count and cooling system. There will definitely be people with more knowledge on batteries and will be able to make better guesses here on this forum, all in all whatever I learnt I read it online and I'm no engineer, not an authority by any means.

But here goes nothing.

Here's what we know;

• Cell chemistry is the same.
• Pack is 74 cells in parallel. That is a group. 6 groups connected in series is a module and 16 modules make the pack. 7104 cells in total.
• We know that original 85 pack had 3.1Ah cells and their nominal voltage is 3.7V. 4.2V fully charged, 3.2V cutoff.
• 90 pack is a 6% increase thanks to silicon on the cathode. That puts it in ~3.3Ah. So 3,3Ah * 3,7V * 7104 = 86740Wh --> 90 pack. With a 97% allowed DoD we end up with 84kWh usable energy. I think that would be in line with people's experience thus far with 90 packs.

So, with these assumptions 96,37% of 90kWh is called a 90 pack. Same went with the 85 back in the day. So for the 100 pack, actual capacity should be 96,37kWh. With a 315mi EPA estimation that's 297Wh/mi. Makes sense right?

So same chemistry cells, 3.3 * 3.7 = 12,21Wh. 96370/12,21 = 7892 cells for the new pack. 788 new cells, that's an addition of 78lbs just cell weight assuming ~45 grams per cell.

We also know that the drivetrain is the same. So 96 cells in series is 3,7 * 96 = 355 nominal volts for the pack. That should be the same. So for a 96.370Wh pack, you need 96370/355 = 271Ah pack. Compared to 244Ah of 90D. To achieve that you need to connect 82 cells parallel. And we assumed ~7892 cell count for the new pack, ~7892/82 = 96 cells needed in series. Same as before.

So Tesla somehow managed to put 8 more cells per group of 74 and group, module count didn't change. Elon said they were at the theorotical limit and that they've changed the entire cooling system. So my money's on them making cooling ducts thinner and overall packaging much tighter, allowing them to put somehow 788 new cells into the pack.

Another logic supporting my theory which would make thinks simpler for Tesla since not changing the entire layout is probably easier to do is;

We know that P90D's reach around 450kW peak power with a full charge in Ludicrous mode. Full charge is 403V. 450000/403 is 1116 Amps current. For the 244Ah 90D pack that's 4.5C. Assuming the same C rate for the new pack wouldn't be nonsense since chemistry didn't change. That would put it on 1220Amps. 1220* 403 = ~492kW.

However this logic fades away after the fact that recently delivered 90D packs achieved these power outputs and we found out that they are not software limited 100 packs.

What do you guys think? It could also be that they've changed it entirely so there's no way we can figure it out. Decreased the amount of cells in a group, decreased group count, increased module count... Who knows? We can't know before a teardown probably. Maybe I'm completely wrong about this but I just love speculating about Tesla's engineering feats.

 

[Adrian]

We know that P90D's reach around 450kW peak power with a full charge in Ludicrous mode.

They actually develop 510-512kW peak per this thread and data: Newer P90DL makes 662 hp at the battery!!!

 

[hiroshiy]

So I'm really in to figuring out Tesla's battery architecture and I love reading stuff about it. Now we know that the new 100kWh pack is a completely new architecture, including cell count and cooling system. There will definitely be people with more knowledge on batteries and will be able to make better guesses here on this forum, all in all whatever I learnt I read it online and I'm no engineer, not an authority by any means.

But here goes nothing.

Here's what we know;

• Cell chemistry is the same.
• Pack is 74 cells in parallel. That is a group. 6 groups connected in series is a module and 16 modules make the pack. 7104 cells in total.
• We know that original 85 pack had 3.1Ah cells and their nominal voltage is 3.7V. 4.2V fully charged, 3.2V cutoff.
• 90 pack is a 6% increase thanks to silicon on the cathode. That puts it in ~3.3Ah. So 3,3Ah * 3,7V * 7104 = 86740Wh --> 90 pack. With a 97% allowed DoD we end up with 84kWh usable energy. I think that would be in line with people's experience thus far with 90 packs.

So, with these assumptions 96,37% of 90kWh is called a 90 pack. Same went with the 85 back in the day. So for the 100 pack, actual capacity should be 96,37kWh. With a 315mi EPA estimation that's 297Wh/mi. Makes sense right?

So same chemistry cells, 3.3 * 3.7 = 12,21Wh. 96370/12,21 = 7892 cells for the new pack. 788 new cells, that's an addition of 78lbs just cell weight assuming ~45 grams per cell.

We also know that the drivetrain is the same. So 96 cells in series is 3,7 * 96 = 355 nominal volts for the pack. That should be the same. So for a 96.370Wh pack, you need 96370/355 = 271Ah pack. Compared to 244Ah of 90D. To achieve that you need to connect 82 cells parallel. And we assumed ~7892 cell count for the new pack, ~7892/82 = 96 cells needed in series. Same as before.

So Tesla somehow managed to put 8 more cells per group of 74 and group, module count didn't change. Elon said they were at the theorotical limit and that they've changed the entire cooling system. So my money's on them making cooling ducts thinner and overall packaging much tighter, allowing them to put somehow 788 new cells into the pack.

Another logic supporting my theory which would make thinks simpler for Tesla since not changing the entire layout is probably easier to do is;

We know that P90D's reach around 450kW peak power with a full charge in Ludicrous mode. Full charge is 403V. 450000/403 is 1116 Amps current. For the 244Ah 90D pack that's 4.5C. Assuming the same C rate for the new pack wouldn't be nonsense since chemistry didn't change. That would put it on 1220Amps. 1220* 403 = ~492kW.

However this logic fades away after the fact that recently delivered 90D packs achieved these power outputs and we found out that they are not software limited 100 packs.

What do you guys think? It could also be that they've changed it entirely so there's no way we can figure it out. Decreased the amount of cells in a group, decreased group count, increased module count... Who knows? We can't know before a teardown probably. Maybe I'm completely wrong about this but I just love speculating about Tesla's engineering feats.

Nice calculation, but how do you know the cell is the same? They could use 3.6Ah cells.

 

[ArtInCT]

Tesla said it was a new internal pack design with the same cells as in the current 90 pack. I think you are onto something, that Tesla has added more cells and done other internal packaging/system magic to increase the pack capacity and output and range. Do you expect the rated range at 100% to be near 290 rated miles? The 90 pack gave P90DL's 260 - 265 rated miles when new so increasing that by 10% should get you in the 285 - 290 rated miles range or perhaps higher (if they have come up with a new factor).

I did think I read that Tesla indicated that charge time for the 100DL pack would be about the same as the 90 pack, so would adding more cells fall in line with that statement?

Also unknown to all of us, is the new P100DL pack the exact same pack as the facelift P90DL pack? A photo of the P100DL pack label will be the tell I suppose. The Ludicrous pack could have three variants if this is true:

the early P90DL (pre facelift),
the facelift (P90DL)
the latest P100DL.

Emir-T, BTW I hope things are going better in Instanbul, it is a beautiful city full of wonderful folks and I hated to hear about the recent events.

 

[garsh]

Nice calculation, but how do you know the cell is the same? They could use 3.6Ah cells.

Electrek reported:

For the 100 kWh battery pack, Tesla is using the same battery cell, but a new module and pack architecture, new cooling system and electronics. CTO JB Straubel described the upgrade as a “significant change”.

I assume that this is information they obtained during that 30 minutes press Q&A session.

 

[DIL]

How much more does the 100 pack weigh, I wonder?

 

[hiroshiy]

Electrek reported:I assume that this is information they obtained during that 30 minutes press Q&A session.

Thank you, maybe my English skills were not sufficient but I read that statement as "the same (18650-sized) cells". My bad.

 

[emir-t]

Emir-T, BTW I hope things are going better in Instanbul, it is a beautiful city full of wonderful folks and I hated to hear about the recent events.

Thank you. Unfortunate times for the whole world. Trying to get by.

 

[llavalle]

We know that P90D's reach around 450kW peak power with a full charge in Ludicrous mode. Full charge is 403V. 450000/403 is 1116 Amps current. For the 244Ah 90D pack that's 4.5C. Assuming the same C rate for the new pack wouldn't be nonsense since chemistry didn't change. That would put it on 1220Amps. 1220* 403 = ~492kW.

You're forgetting about the voltage drop and it's HUGE.

My 85D is capped at 1150amp at the pack
PXXD are limited to 1300amps
Ludicrous updated packs are at 1500, some are at 1600.

Here's a video I've done in which I embedded Voltage and Amp information coming from the pack's BMS on the CAN bus :
My 85D :

A friend's P85D ,prior to ludi upgrade (1300amp limit)

 

[emir-t]

You're forgetting about the voltage drop and it's HUGE.

My 85D is capped at 1150amp at the pack
PXXD are limited to 1300amps
Ludicrous updated packs are at 1500, some are at 1600.

Here's a video I've done in which I embedded Voltage and Amp information coming from the pack's BMS on the CAN bus :
My 85D :

A friend's P85D ,prior to ludi upgrade (1300amp limit)

Wow I never saw these before. Thanks! Very valuable data. This is actually in line with data Tesla provided during Ludi upgrade announcement. OK then applying the same logic, 1300Amp P85D x 315V after drop = 410kW. For Ludi that is 472kW. In line again with what people provided.

This means Ludicrous packs can do 1500/244 = 6,14C!! That's humongous. 271Ah 100kWh pack with 6,14C limit is 1664Amps. That's 524kW with 315V after drop. That could very well be 2.5sec 0-60mph because DragTimes reported 2.65sec with ~500kW output on the new P90D. How they achieved to get 500kW out of a P90D is beyond me though. Maybe the limit isn't the batteries and 6C is not even a difficult thing for Tesla? Maybe they upgraded the fuses and the wiring again, the same way they would do a new P100D car. Only the battery is different only for capacity purposes, not performance.

 

[seclinton]

It's still the same 4.2V chemistry, however the addition of Si is usually added to the anode ( graphite) not the cathode. That is because silicon can intercalate with 8 Li ions vs 4 for carbon. So tweaking the additive a little probably helped raise the capacity of he battery. Increase the electrode loading density toward the USL spec limit and maybe a little more would likely be the result of further chemistry optimization between Tesla and Panasonic.

Improvements in cooling and thermal management improve the discharge curve profile of the cell banks and that's likely where the thermal improvement comes from.

Changing the active material loading in the cell to larger will likely increase impedence and slightly affect the rate of discharge. However the change to inconel and other between cell interconnects could mitigate this increased impedence.

The above combined points toward a 3 pronged approach: chemistry, thermal management and mechanical assembly. Highly refined design.

Innovation isn't always visible

 

[Matias]

"Tesla is using the same battery cell"

Tesla claims ‘Quickest Production Car in the World’ title with new 100 kWh battery pack: 0-60 in 2.5s & 315 mile range

 

[emir-t]

It's still the same 4.2V chemistry, however the addition of Si is usually added to the anode ( graphite) not the cathode. That is because silicon can intercalate with 8 Li ions vs 4 for carbon. So tweaking the additive a little probably helped raise the capacity of he battery. Increase the electrode loading density toward the USL spec limit and maybe a little more would likely be the result of further chemistry optimization between Tesla and Panasonic.

Improvements in cooling and thermal management improve the discharge curve profile of the cell banks and that's likely where the thermal improvement comes from.

Changing the active material loading in the cell to larger will likely increase impedence and slightly affect the rate of discharge. However the change to inconel and other between cell interconnects could mitigate this increased impedence.

The above combined points toward a 3 pronged approach: chemistry, thermal management and mechanical assembly. Highly refined design.

Innovation isn't always visible

Thank you, very informative. How does one learn about batteries like yourself? Is it your profession? Are you a scientist?

A few questions;

• By additive do you mean anode material(graphite) or electrolyte?
• Also what is USL spec limit?
• By active material loading do you mean cells getting more dense per volume?
• How do you think they got around Si expanding and harming the cell when charged vs. discharged? Do you think they were not sure how it would affect the pack so they limited Supercharger capability for 90 packs, making them slower than 85 packs upon delivery and after enough data collected they updated it so 90's supercharge way faster now. (remember news from a few weeks ago?)

However for the 100kWh pack discussion might I add news outlets say cells are the same. Addition of Si to the anode(thanks for correcting) was only on switch from 85 to 90 packs. As far as I interpret that made the cells go from 3100mAh to 3300mAh. So this was pack level innovation more than cell level. If the new pack was the same architecture but with more addition of Si resulting in 3700mAh cells, it would've been an amazing advancement.

However I'm fairly confident that they'll do amazing feats with the chemistry of 21-70. Current 18650's volume will increase 40%. With the current 3300mAh chemistry that takes them up to 4600mAh. With further advancements in cell chemistry they probably could achive 5Ah cells or maybe more.

With their new compact cooling solution and architecture and needing half the cells thanks to 21-70, they get their construction costs wayyy down before applying economies of scale thanks to Gigafactory, and they need less space for same capacity pack. Amazing times to live.

 

[sandpiper]

You're forgetting about the voltage drop and it's HUGE.

My 85D is capped at 1150amp at the pack
PXXD are limited to 1300amps
Ludicrous updated packs are at 1500, some are at 1600.

Very interesting. Do you know what factors differentiate the 1500 packs from the 1600 packs? Is it another fuse / contactor upgrade? Or is it a cell limitation? Or something else? I've been considering an L upgrade for my P85D and I'm curious what current limit it would be running.

 

[jmsurpri]

Anyone know why the 85D ramp to max power is so much slower than the P85D in those videos? I wonder if there is a bit more acceleration available to us 85D owners if Telsa ever decided to tweak the firmware someday.

 

[David99]

The new cooling system could help improve Supercharging a lot.

 

[SomeJoe7777]

Anyone know why the 85D ramp to max power is so much slower than the P85D in those videos? I wonder if there is a bit more acceleration available to us 85D owners if Telsa ever decided to tweak the firmware someday.

The small rear motor in the 85D has a lot less torque than the P85D's large rear motor. This results in slower acceleration. Since power = torque * speed, and you have both lower total torque as well as a slower ramp up to speed, you correspondingly get a slower ramp up to max power.

 

[Saghost]

Tesla said it was a new internal pack design with the same cells as in the current 90 pack. I think you are onto something, that Tesla has added more cells and done other internal packaging/system magic to increase the pack capacity and output and range. Do you expect the rated range at 100% to be near 290 rated miles? The 90 pack gave P90DL's 260 - 265 rated miles when new so increasing that by 10% should get you in the 285 - 290 rated miles range or perhaps higher (if they have come up with a new factor).

I did think I read that Tesla indicated that charge time for the 100DL pack would be about the same as the 90 pack, so would adding more cells fall in line with that statement?

Also unknown to all of us, is the new P100DL pack the exact same pack as the facelift P90DL pack? A photo of the P100DL pack label will be the tell I suppose. The Ludicrous pack could have three variants if this is true:

the early P90DL (pre facelift),
the facelift (P90DL)
the latest P100DL.

Emir-T, BTW I hope things are going better in Instanbul, it is a beautiful city full of wonderful folks and I hated to hear about the recent events.

Makes me wonder if the 75 is going to be replaced with an 83? (more likely called a new 85, I suppose,) when they start using the 100 pack technology and modules across the line...

 

[sorka]

Anyone engaging in cell chemistry discussion should watch this:

He also consulted with Tesla on their batteries.

 

[glhs272]

So when I look at the version 1 Tesla Modules I have been playing around with, I see some room for improvement. I posted a few picks. Notice the dead spaces where there are no cells, like the near some of the corners where the coolant loop turns around. They would have to re-arrange/repackage to fill these voids. With some other packaging enhancements or different coolant loop design, perhaps relocate the BMB board, they found room for added cells.