Rumor: Model 3 to use new 4416 battery cell

[Cloxxki]

4416 / 8 would mean 552 per module.
How to divide that and come to a practical module voltage?
If 3.7V, 2042/6 = 340V. That's 92 in series, 6x parallel. 4.1V makes it 376V
552 divided by 5 doesn't work.
138 x 4 makes 500+V per module.

Or are some modules linked in series, coming to a lower module voltage than reaches the motor(s)?

New Hyundai Ioniq EV roughly the same size as Model 3 gets as low as 140Wh/km on the highway. That would be 520km highway range with the Model 3! Also Model 3 can be more efficient than the Ioniq. Very eager to see results.

While Ioniq will likely lose on highway air drag versus Model 3, it does have a (weaker) pernanent magnet motor. This alone could improve consumption considerably, Tesla motors are NOT that efficient in the market. They go far because of big batteries and slick aero. Now a Model 3 with Ioniq battery and drivetrain... Even if just (doubled) 61kWh, it would charge at 140kW peak. Probably beat Model S 100D on most road trips, if it could use superchargers, or a future CCS network.
Not a huge factor at 120kph, but Ioniq 31kWh is 1420kg. Don't be shocked if the 75kWh Model 3 is over 1700, even 1800kg.
But the highway cruising will be likely awesome with Model 3. How that translates into convenient supre-range road trips will largely depend on charging speed. At the same, say, 100kW, the Model 3 adds range 10-15% quicker perhaps. But will it add kW's as fast? Model S/X need to remain top dog, and Tesla is known to limit convenience/value through software. It would not be cool if Model 3 were to prove quicker over long distances. Like that US coast to coast thing. Or Norway's South to North Cape.

 

[emir-t]

4416 / 8 would mean 552 per module.
How to divide that and come to a practical module voltage?
If 3.7V, 2042/6 = 340V. That's 92 in series, 6x parallel. 4.1V makes it 376V
552 divided by 5 doesn't work.
138 x 4 makes 500+V per module.

Or are some modules linked in series, coming to a lower module voltage than reaches the motor(s)?

Lithiom ion cells' nominal voltage is 3.6-3.7V always. (LiFePo4 is 3.2) So BMW i3, Renault ZOE, Model S-X with large packs always use 96 in series to reach the same voltage of 355 nominal, 400 100% SoC. BMW i3 uses just 96 cells in series. ZOE does 2 parallel and then 96 series for a grand total of 192. Model S connects 6 groups in series, calls them a module and connects 16 modules in series; 16 * 6 = 96!

So like Elon has always been saying Tesla won't reinvent the wheel with this. 96 in series would mean, with 8 modules, 12 groups in a module. So 4416/96 = 46 is in parallel. At 4700mAh 46 parallel would make a 216Ah pack. 355V nominal means 216 * 355 = 76,6kWh

And for the lower spec just removing 2 modules, making the voltage lower and the pack much cheaper makes sense and further differentiates S in terms of range, charging speed and performance. So I think Model 3 base will have 270V nominal voltage. That would make fastest supercharging speed for Model 3 55 between 75 to 80kW. And that would quickly taper down to 60kW. Although this looks like half of what current fleet can manage, it is the same % per unit time.

 

[Cloxxki]

Nice to see that the 4416 number does seem to work dandy for 8 modules and 96 in series. Increases likelihood for it to be spot on.
I like the idea that the smaller pack would just get 2 modules less rather than special grated ones. Removing two outer ones will reduce number of modules lost in a crash. Or, implement half modules to keep the mass more evenly distributed and get more crash proofness?

80kW for a 55kWh pack (1.45C?) would be right up there with the S85 pack that peaked I think around 117kW for 81.5kWh true capacity. Despite the more cumbersome form factor? Even the 102.4kW pack now on sale doesn't get much over 1.12C (pre-throttled, it seems).
55x1.12 is 61.6kW. Much more in line with the "current" market for somewhat affordable BEV's.
75x1.12 is 84kW. Very reasonable compared to Model S. Bigger range than S75, but spend similar time supercharging on road trips.

We saw a likely 75 kWh Model 3 charging at 70.6kW after well over 7 minutes and around 30% SOC. That much have been close to maximum, right? I'd say <80kW, listening to only my gut feeling. The smaller pack, might well get way less. Or simply 3/4.
Extrapolating 75/55 * 80 would mean up to 109kW for the 75 pack. Around 400mph. The 100D doesn't even get that. Unlikely, especially with no direct plans to swap S/X over to 2170's.
Tesla should improve C rates, especially now with 2170's, but I doubt the Model 3 launch will be the moment.

 

[scaesare]

A 65 versus 75 kWh version doesn't seem to offer much in the way of range differentiation.

Well, if you assume the M3 is slightly more efficient than the S, the change as a percentage of overall range may not be that much different than the 15kW steps typically seen thus far for the S.

My Model S 85 (RWD) consumes ~275Wh/mi. If we assume the smaller 3 can do 250Wh/mi, then that's an additional 40 miles of range for an additional 10kW of real pack capacity.

If the base 65 model gets 215 miles of range, and the 75 gets 255, then that's a difference than that's an additional 19%. Even if the base 65 were able to manage 240 miles of range (beating the Bolt), than a 280 mile 75KWh model represents 17% more range.

When you could still get a 75 and a 90 Model S, the ranges for the RWD models were listed as 259 and 294 miles respectively. That's only a 35 mile difference, which is only 14% greater range.

 

[emir-t]

Tesla should improve C rates, especially now with 2170's, but I doubt the Model 3 launch will be the moment.

I agree. 31kWh nominal capacity Ioniq can sustain 60-70kW CCS charging at 100kW CCS stations. That is >2C. Considering its efficiency that is Tesla level.

While Tesla starts above 1C it drops down to less than 1C easily. So I think for Tesla overall Supercharging average speed is 0,75C while BMW i3, Hyundai Ioniq and all the rest can do above 1C at least til 80%. For example even my '14 i3 with 20kWh capacity can sustain 38-40kW til 80%. Could a Model S sustain 120kW till 80%?

But anyways, we're digressing.

 

[Cloxxki]

Indeed. While Tesla is monopolist in long range BEV's, they can get away with these charge rates. They were unfathomable just a few years ago, yet they brought. Maybe they've been complacent, or overplaying their hand. The 102.4kWh car not charges at a lower peak speed than the original 81.5kWh version. I know the curve makes it faster over long charges, but it's a symptom.
2 or 3 years from now, Tesla could be know for the slow chargers they offer, and the cars that can't keep up with others at the startions. Keep the stalls occupied way longer than modern cars.
CCS 350kW is on its way to our charging stops, and with the BEV volume on the roads ramping up like crazy the next years, short stops will be crucial. With batteries ever cheaper, the race will not be for longest range, but fastest to do 1500km in a day, starting on a nice overnight charge. Most cars will manage 500km on a full battery, but how will they travel and how long will they need to stop to make it to the destination with range to spare in case there's no charger there? By 2020, Teslas could be held up at their own Superchargers (for cost reasons) twice as long or worse than Audis, Mercs, VW's, Koreans, etc. They were so proud of their cheap cheap cells in 2017...

I do suspect Tesla are on the case, and if I were a benevolent overlord, the fast cells would be an upgrade for the P model cars currently grossly overpriced. And I'd make a light hot hatch with small range but really nice performance out of them. Imagine half the power and range of an EP9 or Concept 1 in a Civic R EV.

 

[wdolson]

4416 / 8 would mean 552 per module.
How to divide that and come to a practical module voltage?
If 3.7V, 2042/6 = 340V. That's 92 in series, 6x parallel. 4.1V makes it 376V
552 divided by 5 doesn't work.
138 x 4 makes 500+V per module.

Or are some modules linked in series, coming to a lower module voltage than reaches the motor(s)?


While Ioniq will likely lose on highway air drag versus Model 3, it does have a (weaker) pernanent magnet motor. This alone could improve consumption considerably, Tesla motors are NOT that efficient in the market. They go far because of big batteries and slick aero. Now a Model 3 with Ioniq battery and drivetrain... Even if just (doubled) 61kWh, it would charge at 140kW peak. Probably beat Model S 100D on most road trips, if it could use superchargers, or a future CCS network.
Not a huge factor at 120kph, but Ioniq 31kWh is 1420kg. Don't be shocked if the 75kWh Model 3 is over 1700, even 1800kg.
But the highway cruising will be likely awesome with Model 3. How that translates into convenient supre-range road trips will largely depend on charging speed. At the same, say, 100kW, the Model 3 adds range 10-15% quicker perhaps. But will it add kW's as fast? Model S/X need to remain top dog, and Tesla is known to limit convenience/value through software. It would not be cool if Model 3 were to prove quicker over long distances. Like that US coast to coast thing. Or Norway's South to North Cape.

They could do it with 24 cells in series per module and then 23 sets in parallel. The modules then are arranged with 4 in series in two sets in parallel. That gets a pack right around 400V.

 

[Cloxxki]

Well, if you assume the M3 is slightly more efficient than the S, the change as a percentage of overall range may not be that much different than the 15kW steps typically seen thus far for the S.

My Model S 85 (RWD) consumes ~275Wh/mi. If we assume the smaller 3 can do 250Wh/mi, then that's an additional 40 miles of range for an additional 10kW of real pack capacity.

If the base 65 model gets 215 miles of range, and the 75 gets 255, then that's a difference than that's an additional 19%. Even if the base 65 were able to manage 240 miles of range (beating the Bolt), than a 280 mile 75KWh model represents 17% more range.

When you could still get a 75 and a 90 Model S, the ranges for the RWD models were listed as 259 and 294 miles respectively. That's only a 35 mile difference, which is only 14% greater range.

You mention 250Wh/mi, which IMHO is conservative, at least for an official Model 3 EPA rating.

Then you mention 215 miles for a 65kWh. Even with 62kWh available, it would be (x4) 248 miles.
10kWh extra for the 75kWh would normally add 40 miles indeed, 288 miles for +16%.

 

[Cloxxki]

They could do it with 24 cells in series per module and then 23 sets in parallel. The modules then are arranged with 4 in series in two sets in parallel. That gets a pack right around 400V.

96 in series, interesting.
Would 23x24 fit comfortably in a rectangular module? Perhaps if the overhead were stuck on the short end.

 

[wdolson]

96 in series, interesting.
Would 23x24 fit comfortably in a rectangular module? Perhaps if the overhead were stuck on the short end.

The cells in series don't need to all be in one line. They could make the rows say 16 cells and one string of 24 could be 1 1/2 rows. This would leave a space with 8 cell spaces on the last row, but that could be dedicated to the electronics controlling the module.

 

[Waiting4M3]

From the quarterly letter:

"Model 3 drive units as well as battery packs made with our proprietary 2170 form factor cells are being built on new lines at Gigafactory 1"

Can we put this to bed now?

 

[JVL]

@SCaesar - I am averaging 325 KW per mile over the last 25k miles. You average 275? I guess I never really thought about it - but what is average? Do we know? (I also have RWD S85)

 

[Cloxxki]

@SCaesar - I am averaging 325 KW per mile over the last 25k miles. You average 275? I guess I never really thought about it - but what is average? Do we know? (I also have RWD S85)

Wh/mi surely ;-)

 

[Cloxxki]

Well, if you assume the M3 is slightly more efficient than the S, the change as a percentage of overall range may not be that much different than the 15kW steps typically seen thus far for the S.

My Model S 85 (RWD) consumes ~275Wh/mi. If we assume the smaller 3 can do 250Wh/mi, then that's an additional 40 miles of range for an additional 10kW of real pack capacity.

If the base 65 model gets 215 miles of range, and the 75 gets 255, then that's a difference than that's an additional 19%. Even if the base 65 were able to manage 240 miles of range (beating the Bolt), than a 280 mile 75KWh model represents 17% more range.

When you could still get a 75 and a 90 Model S, the ranges for the RWD models were listed as 259 and 294 miles respectively. That's only a 35 mile difference, which is only 14% greater range.

Turns out, the long range Model 3 gets 40.9% more range.
Some website snoops have found 237Wh/mile as consumption, not sure for SR or LR, although it shouldn't differ too much.
The 14% (rather than +20%) between 75 and 90 is largely down to the 75 being true, the 90 a lie. 85.8kWh is just +14% indeed.

 

[scaesare]

@SCaesar - I am averaging 325 KW per mile over the last 25k miles. You average 275? I guess I never really thought about it - but what is average? Do we know? (I also have RWD S85)

That's to get rated range.

My lifetime average is a bit higher, as I have reasonably cold winters.

With moderate weather I can hit rated range on most trips of reasonable distance.

 

[Trancela]

One thing can be said, and that is that 4416 is divisible with both 96 and 8. That speaks for this being a possibility. You'd have a 96s46p architecture, and potentially modules with 552 cells. This is a scenario that's a lot more probable than the cell format being 4416.

That is pretty plausible. My own estimation based on 51/54 kWh indicated a pack of 3s32s43p, assuming 3 modules, as per Tesla. The 96s is pretty standard. You propose 8 modules for 8s12s46p. As much as I prefer this architecture where modules are low-voltage, I think we should take Tesla word that the T3.220 consists of 3 modules (official).

At 237 Wh/mi (official), we get 52.14 kWh for the T3.220. This would expand to about 55.25 kWh which includes the top and bottom buffers. And 96x46x3.7x3.4 = 55.55 kWh. Note the cell capacity at 3.4 - either these are 3.4 Ah sandwich stuffed in a 2170 or an 18650 stuffed in a 2170 can. Note 96x46x3.7x4.7 = 76.8 kWh, i.e. the T3.310. Another possibility I would speculate for the T3.220 is 96s33/46p x 4.7 Ah. 33/46 indicates that only 33 cells are populating a 46 cell bay.

 

[Trancela]

211Ah * 355V = 74,9kWh with 72,5kWh usable. Yielding 285mi EPA.

How did you arrive at 285 EPA from 72.5 kWh?

I doubt that the T3.220 will be a 300 V machine. I am pretty sure they will stick to 400 V, or they will have to change a lot of electronics.

Thus if the T3.310 is 8s12s46p, then the T3.220 is 8s12s33p rather than 6s12s46p (which has too much range).

 

[Trancela]

I agree. 31kWh nominal capacity Ioniq can sustain 60-70kW CCS charging at 100kW CCS stations. That is >2C. Considering its efficiency that is Tesla level.

With pouch cells, it is possible to achieve higher C rates for charge and discharge. I believe this may be due to the superior cooling capacity of pouch cells. The Leaf, Volt, and i3 all use pouch cells. In fact only Tesla and a few odd-lot companies use cylindricals.

 

[Peter Egan]

Have you guys seen this post from Teslarati? Thoughts?

Tesla could be using a "4416" battery cell in Model 3 and next-gen S, X fleet

During the investor call Elon said the M3 has Gigafactory 2170 cells while MS, MX have japanese 18650 cells.

 

[Cloxxki]

If we look at the volume increase from the form factor, using exactly 21 and 18mm, 70 and 65, that's +46%.
18650's are 12.67Wh or 12.40Wh depending on source I just googled.
If chemistry for 2170 is 100% unchanged, that brings us to at least 80.2kWh for 4416 cells.
But, JB and Elon were saying they saw 10-15% density improvement in the chemistry itself.
If the 4416 figure holds true, then either:
A- the chemistry update was not ready in time, they may use the original 2012 chemistry even.
B- there is hidden capacity.

B seems most plausible in the ight of this new battery narrative. With so much emphasis on the range figures, perhaps degradation over the limited warranty offered is included? Each time some degradation occurs, charging only takes longer. You need say 96 in stead of 95% real capacity to reach 220/310 rated miles. Such a buffer, especially at the toop, seems prudent if they're expecting heavy supercharging to 100% for people trying to make it through the week on a charge, having no access to overnight charging. You know people will go shopping around the corner from the supercharger, topping off, before going to bed. By the masses. If in fact they are only getting 90% true capacity, that's not half as detrimental for the battery, I understand?

But,
C- then Elon may have been a bit untruthful about capacity.

Who's ready to bet when we'll first get to see inside a LR pack? I have yet to find a torn apart Powerwall 2. I wonder how many cells it has, and what each is capable of...