I'm very curious to find out how they've packed in the additional cells in the P100D. We're probably talking about an additional 800-1000 cells, which is a lot in an already compact design.
I have a couple of theories about this. There are 16 modules within a Tesla Model S 85 battery pack. So, that means each module likely held something on the order of 5.3125 kWh of energy. If all the battery cells in the Model S 90 are of the same type, that means that each module was upgraded to 5.625 kWh of energy each. That's a 5.88% improvement. I rather doubt that is enough to change every battery cell in the pack, because of 7,104 of them, the actual improvement would be a mere 0.703829 Wh per battery cell. That ain't none too much.
That is why I believe the improvement from 85 kWh to 90 kWh may have been by changing the battery cells in a single module or two within the battery pack instead. That would mean that perhaps two of the modules were upgraded to hold 7.8125 kWh... Or that only one of the modules was upgraded to be 10.3125 kWh in capacity... Either way, the balance of the modules would be filled with the older style battery cell design that totaled 5.3125 kWh per each module.
Thus, if all sixteen modules were eventually upgraded to the 10.3125 kWh capacity you could have a maximum of 165 kWh some day. Or, if all sixteen modules became the 7.8125 kWh variety, you could reach a 125 kWh battery pack capacity. Taking this notion further, if adding 5.0 kWh of capacity requires an improvement to battery cells within four modules, that means the Models S
P100D and Model X
P100D each use twelve 6.5625 kWh modules and four 5.3125 kWh modules. So, the total the Generation II vehicles could reach would be 105 kWh.
The most important question is if the new pack architecture is transferable to the Model 3, and there's really no way of knowing without opening one of the new packs up. It's entirely possible the new architecture utilizes the extra 5 mm of space above/below the cells for cooling. But that this space will be absorbed by the 21-70 cells, so you need the old cooling solution for the 21-70 based packs. If so, going to the 21-70 format may yield no additional space for cells. And a 21-70 based Model S/X pack may only get a <5% capacity increase.
5 mm is less than 1/4". So, less than the thickness of the outer casing that forms the battery pack itself. The height of the 2170 battery cells is not a real problem. If Tesla Motors wants to use those cells in Generation II vehicles, Model S and Model X, they will be able to do so. The battery pack is reportedly 4" thick, which is 102 mm. Subtract 7 mm for top and bottom leaves 88 mm. Subtract another 3 mm for top and bottom of each module within the battery pack and you have 82 mm. So, whether the battery cells are 65 mm tall or 70 mm tall, there is plenty of room left over. About 12 mm, which is roughly half an inch. Plenty of room for wiring and components.
Also, as there was no improvement due to chemistry, that may increase the likelihood that we'll see a capacity increase due to chemistry next year. Since 2012, we've had a single ~6% increase due to chemistry. That's an average of 1.5% per year. Not extremely good.
Depends upon how you look at it. Tesla Motors has never said they intend to roll out advancements in energy density on a yearly basis, as they happen in the lab. Honestly, it would be dumb of them to even try. What they have shown is a likelihood to show advancements in battery technology with the introduction of a new Generation of vehicles. They have also shown their willingness to upgrade older technology from previous Generations to accommodate better battery cells when possible.
A 5% increase in 2017 would increase this to 2.2% per year, at least.
JB Straubel says you will see a 40% increase over 2012 with the Generation III vehicles to be released in 2017. That works out to 8% per year. I believe him. You don't have to.