Tesla's Model 3 2170 cells=same energy desity as 18650's

[Topher]

Why do you think they would use 18650 cells with new chemistry? (Genuine question.)

Because the 18650s FIT in the Model S and X packs (we don't know what it would take to fit 2170s). Improving the chemistry, improves the range etc. Why WOULDN'T they do it.

2170 is what they built the Gigafactory for, right?

Nope. They built the Gigafactory to produce battery cells; just that. Of whatever size, chemistry, pack configuration they need at any given time. If state of the art in batteries changes, do you expect them to demolish it, and build a new one? Neither do I.

I would expect them to retire the 18650's when possible.

Why would they do that? They may or may not have a use for 18650s. They certainly have a contract with Panasonic to buy 18650s for some time into the future. The 18650 may in fact be a better cell size for the Model S (or not).

As for the added mass, why would 2170s necessarily mean added mass?

No, More energy, requires more cells, means more mass. Absent chemistry changes, that is the only way to increase total pack energy. 2170s vs 18650 is pretty much irrelevant to that point.

the new cell design allows for more optimal cooling

Seems unlikely. Cooling a volume occurs at its surface, and is thus dependent on surface area. The new cell design has a greater volume to surface area ratio (cubed-square law). Thus it makes cooling MORE DIFFICULT. Whatever Tesla optimized for it wasn't (abstract) ease of cooling. There could of course be other factors, for example, limitations on minimum cooling tubing wall thickness, which made previous cooling designs inefficient (by dint of working too well). But absent evidence of that, it generally makes sense to reduce component size to increase cooling, hence margaritas.

Thank you kindly.

 

[Cloxxki]

Before wading through the rest of the thread, OP voices what what've been feeling and writing about, in wonderful detail.

My main open question is regarding the actual measurements of the cells. Both of them. There is uge need for caution, with Tesla's tendency to taking rounding up Olympic.
We need to optain each cells and take PROPER measurements. Submerging (but not sinking) in water on a scale will show the density (1 gram = 1 cm³ displaced). More practical however will be length and diameter.
Other 21700's on the market are reportedly 20.3mm. A humongous difference for volume, 7.0% to be exact. At the same time, more of these would fit in a given confinement. if you're to use a given cooling loop though, looping fewer cells will reduce cooling density losses within the total pack.

The seemingly sandbagged range figures leave a lot of questions. Is the energy there or not? Will it be used to match (offset) degradation?

 

[R.S]

Before wading through the rest of the thread, OP voices what what've been feeling and writing about, in wonderful detail.

My main open question is regarding the actual measurements of the cells. Both of them. There is uge need for caution, with Tesla's tendency to taking rounding up Olympic.
We need to optain each cells and take PROPER measurements. Submerging (but not sinking) in water on a scale will show the density (1 gram = 1 cm³ displaced). More practical however will be length and diameter.
Other 21700's on the market are reportedly 20.3mm. A humongous difference for volume, 7.0% to be exact. At the same time, more of these would fit in a given confinement. if you're to use a given cooling loop though, looping fewer cells will reduce cooling density losses within the total pack.

The seemingly sandbagged range figures leave a lot of questions. Is the energy there or not? Will it be used to match (offset) degradation?

Heureka, someone should reread Archimedes.

By submerging them in water you get the volume, not the mass. Just putting it on a scale will give you the mass and then you can submerge it to get the volume and then you can calculate it's density.

 

[transpondster]

Other 21700's on the market are reportedly 20.3mm.

you are mixing things, these cells are Panasonics from Japan, before Tesla decided to follow with Samsung 21700, remember originally GF supposed to produce 20700

 

[Cloxxki]

The bigger the cylinder the worse the packaging will be, so 18650s should be easier to package.

Already we see that Model 3 has larger modules, around 20kWh each (between total and usable). The 100 S/X pack modules sit at 7kwh. That's a lot more module circumference (the interface between stacked circles and a straight module wall).

Main advantage of larger diameter cells emerges IF the 2 cells would both work well with a given fixed size cooling "snake". 16% narrower cell rows makes for 16 fewer cooling snakes. And those add up to a few extra rows of cells. You end with a higher surface percentage of cells, lower of cooling.

If I were JB, I'd be pushing for larger modules in S/X packs, and stuffing in so many cells that the power output, peak and continuous, overall remains at least as good as existing cars. So the 2170 120 pack matches the 18650 100 pack for power output, while beating it in range and charge speed.

 

[transpondster]

Heureka, someone should reread Archimedes.

By submerging them in water you get the volume, not the mass. Just putting it on a scale will give you the mass and then you can submerge it to get the volume and then you can calculate it's density.

although submerging in quicksilver would work , scales are for n00bs

 

[Cloxxki]

you are mixing things, these cells are Panasonics from Japan, before Tesla decided to follow with Samsung 21700, remember originally GF supposed to produce 20700

I am only showing than until we put digital calipers to both, we are doing calcs on Tesla style nominal number. History is rich in example where Tesla used a nominal figure when they meant to say "about yea big" or "none of your $% business. Enough gospel surrounding Tesla. Show me the cell, in calipers!
Also, who knows the capacity per 2170 in Powerwalls and/or Powerpacks?

 

[Cloxxki]

Heureka, someone should reread Archimedes.

By submerging them in water you get the volume, not the mass. Just putting it on a scale will give you the mass and then you can submerge it to get the volume and then you can calculate it's density.

Are you trying to point out a grave error in my misrepresentation of physics, or a linguistic imperfection?
I'm pretty sure I wrote it down correctly. Due to a cm³ weighting exactly 1 gram (at 4ºC), a scale can directly be considered indicating cm³ in stead of grams, in the specific case of a correct submergence (displacement of the water).

 

[transpondster]

I am only showing than until we put digital calipers to both, we are doing calcs on Tesla style nominal number. History is rich in example where Tesla used a nominal figure when they meant to say "about yea big" or "none of your $% business. Enough gospel surrounding Tesla. Show me the cell, in calipers!
Also, who knows the capacity per 2170 in Powerwalls and/or Powerpacks?

you NOT SHOWING ANYTHING, you tossed some numbers without proof. That's how these 130 kWh pack threads begin.
The only known to public 21700 cell is 21mm diameter:
Test of Samsung INR21700-30T 3000mAh (Gray)
Battery test-review 18650 comparator

 

[Topher]

Main advantage of larger diameter cells emerges IF the 2 cells would both work well with a given fixed size cooling "snake".

But why would we think that even remotely likely? Larger cells produce more heat per area of surface connected to cooling apparatus. So more cooling area is needed, or cooler coolant, or faster flow. If they could accomplish any of those, why didn't they, on the previous modules?

Thank you kindly.

 

[Cloxxki]

If Tesla were to upgrade the 100 cars to 120, dividing the pack in fewer modules migh save as much weight as the extra cell mass needed to make 120kWh weighed extra. The fewer module walls should increase the surface utilization for cells. Using the exact same cooling loops would further increase the percentage of the pack floor occupiedd with cells. Then, the larger form factor adds the mentioned 2-3% extra in content vs cell packaging. And any length increase is a direct win.
The individual cells could not be (dis)charged as fast as 18650's with the same chemistry, but over the whole, both would still improve.

If I were Tesla though, I'd want to remain ahead. Focusing on ways to reduce the overhead weight of packs, why getting more cell volume in them, as well as increase volume of the packs themselves. Why? To accomodate faster charging chemistries which are bound to sufffer a bit in the energy density department. A 80kWh pack with 2C cells would beat a currrent 100 pack over a 200kWh road trip in time traveled, easily, by charging much fast those 120kWh (200-80) needed versus 100kWh (200-100) for the 100 pack. Even 1.5C would beat it.
Range of course it sexy, so keeping to 100kWh with faster cells would be a huge win.
The current 100 pack has a 1.17 (if that) charging peak. The original 85 had 1.44. Where did that charge performance go? Yes, I know, different charging strategies. But emplain that to the buyer pondering between ICE and BEV still. Tesla will need to do better, and keeping at current range/capacity figures and seemingly regressing chemistries, is not going to do it in the long run. Guess what the new 350-400kW charge stations are inspiring the other brands to as they design their packs from the ground up in 2017 rather than 2011?

 

[Cloxxki]

But why would we think that even remotely likely? Larger cells produce more heat per area of surface connected to cooling apparatus. So more cooling area is needed, or cooler coolant, or faster flow. If they could accomplish any of those, why didn't they, on the previous modules?

Thank you kindly.

I agree. But you need to keep the capacity of the pack the same AND the charging and discharging rates the same. If you use the increase of capacity for offsetting the heat to be dissipated by the larger cells, neither charge nor discharge need to suffer as much.
But obviously, an updated chemistry either for increased energy or power (and hypothetically both), will do much more.

Also, the 100 pack charges at 1.17C whereas the original 85 got 1.44. So there is some margin (pre-throttling) going on there, it seems. The 100 pack is 16% or so bigger than the 90 (not exactly the same cells perhaps), but barely charging any faster. A 100 pack from 2170's with 2012's chemistry, how fast would those safely charge with similar to 2012's cooling tech, not even the slimmer 100's?)

 

[Cloxxki]

you NOT SHOWING ANYTHING, you tossed some numbers without proof. That's how these 130 kWh pack threads begin.
The only known to public 21700 cell is 21mm diameter:
Test of Samsung INR21700-30T 3000mAh (Gray)
Battery test-review 18650 comparator

I did not even attempt to offer proof. If you know anything about Tesla, you know that if they give a number to go by, later it will turn out to be subject to A LOT of interpretation and rounding. I'll be happy to be surprised when verified 21.0mm Tesla cells show up. If they are 21.0mm though, it bring a bigger disappointed for the lack of density progress. JB an Elon spoke 10-15% of density improvements from chemistry alone. They were speaking about Model 3 if I'm remembering correctly. This 10-15% was likely not a lie, but it certainly has not made it to Model 3 if the cells truly are 21.0mm.

 

[R.S]

Are you trying to point out a grave error in my misrepresentation of physics, or a linguistic imperfection?
I'm pretty sure I wrote it down correctly. Due to a cm³ weighting exactly 1 gram (at 4ºC), a scale can directly be considered indicating cm³ in stead of grams, in the specific case of a correct submergence (displacement of the water).

The cell displaces volume. So you can see how much volume it makes up by catching the water, but not how much it weights. Or do you mean you put it on a scale and then look how much water it displaces and how much weight was gained?

 

[Topher]

If I were Tesla though, I'd want to remain ahead. Focusing on ways to reduce the overhead weight of packs, why getting more cell volume in them, as well as increase volume of the packs themselves.

I start with the assumption that neither current, nor past Tesla engineers were complete morons. So when Elon says "100kWh is the most that can be fit with current battery technology", that statement has been vetted by the actual engineer in charge. Any improvements can only come from actual improvements in technology somewhere along the line, or a re-examining of basic precepts (safety, structure, design trade-offs, etc.). So they think that they have already reduced the overhead weight of the packs, and already increased the volume of the packs themselves. Otherwise, they wouldn't have said that. Elon could just as easily have said, "we don't plan on increasing the pack size (energy) in the near future."

Thank you kindly.

 

[Topher]

If you use the increase of capacity for offsetting the heat to be dissipated by the larger cells, neither charge nor discharge need to suffer as much.

I can't make any sense out of this sentence. 'capacity offsetting heat'? What capacity? How would it 'offset' heat? What does 'offsetting' heat even mean?

Tesla thinks it is at the engineering limits in their current design. Do you have ANY evidence that they are wrong about that?

Thank you kindly.

 

[Cloxxki]

The cell displaces volume. So you can see how much volume it makes up by catching the water, but not how much it weights. Or do you mean you put it on a scale and then look how much water it displaces and how much weight was gained?

I'm confused. Why would I use water if I wanted the weight of a cell? Was I not speaking in volumetric units? Why do I feel ganged up on? I usually don't get that vibe here.

 

[Cloxxki]

I start with the assumption that neither current, nor past Tesla engineers were complete morons. So when Elon says "100kWh is the most that can be fit with current battery technology", that statement has been vetted by the actual engineer in charge.

Considering the 2170 is supposedly 5mm longer, then indeed these cells have worse density than the 18650's which already manage 102.4kWh.

 

[Cloxxki]

I can't make any sense out of this sentence. 'capacity offsetting heat'? What capacity? How would it 'offset' heat? What does 'offsetting' heat even mean?

Tesla thinks it is at the engineering limits in their current design. Do you have ANY evidence that they are wrong about that?

Thank you kindly.

- take S/X 100 as given power output (say 550kW battery power) and input (say 120kW charge rate)
- If you up the total capacity of 102.4kW (100/18650) to say 125kW (125/2170), a <1C rate suffices to add miles just as quickly Supercharging as on the current 100 cars. Increasing cell volume by 46.5%, affects C rate by how much exactly enjoying equal cooling provisions? Building a larger 2170 pack will at least aid in reducing charging speed losses. I don't care so much about the ludicrous plus party trick, but that will likely be affected similarly. In my opinion, P100DL's go plenty fast enough. I suppose doing slower in new versions will not be an option for Tesla, but at least using the cheaper cells to cram even more energy into the car will help not suffer the downside of said cheapness.

 

[R.S]

I'm confused. Why would I use water if I wanted the weight of a cell? Was I not speaking in volumetric units? Why do I feel ganged up on? I usually don't get that vibe here.

Sorry, I think there was a misunderstanding. The density of water in your comment got me confused. I think you wanted to weigh the displaced water, to get to the volume right?

IMO just putting it in some graduated jug halve filled with water should be good enough. No need to measure the displaced water. But your idea seems fine, too.