Tesla Battery Director Discusses Evolution From 18650 To 2170 + More – Video

Tesla Model 3


Check out a video from last month that captures a speech and presentation from Tesla Director of Battery Technology, Kurt Kelty from the International Battery Seminar in Florida.

The discussion includes the following main topics:

  • Evolution from 18650 to 2170 cell
  • Tesla Powerwall 2
  • Tesla Powerpack 2
  • Tesla Model 3
  • Solar homes

And more…

The audio quality is a bit iffy, so crank the volume a bit to catch all 41 minutes of what Kelty has to say.

Source: Teslarati

Categories: Battery Tech, Tesla, Videos

Tags: , ,

Leave a Reply

24 Comments on "Tesla Battery Director Discusses Evolution From 18650 To 2170 + More – Video"

newest oldest most voted
M3- Reserved - Niro TBD

Video pretty much verifies same battery, same technology, same architecture– between PP, PW, EV.

Really hard not to argue for V2H application of using the EV instead of PW2.– except Tesla wants to sell you the PW2 14kw system extra on top of the 60-100kw battery you already have sitting in the garage.

Except that the battery chemistry is different. That’s kinda important.

M3 - reserved Niro-TBD

Do you have knowledge that the chemistry is different between Tesla PW and Tesla EV packs?

Doesn’t sound like that they are different on the presentation. Presentation all points to efficiency of scale and ubiquitous on the battery itself. The programming on rates maybe different which would make sense, but not the composition/manufacturing of the battery itself

Stationary applications were reported to use NMC, and mobile applications continue to use NCA. Makes sense.

“Really hard not to argue for V2H application of using the EV instead of PW2…”

Actually it’s pretty easy to argue against Vehicle to Home (V2H) application, unless it’s done rather rarely. Despite your claim here about the “same architecture”, PowerWalls are engineered and built to be cycled daily, and Tesla guarantees that they won’t drop below a certain capacity within a certain number of years. BEV battery packs are not designed to be cycled that frequently, and almost no EV maker guarantees that an EV battery pack will retain X percentage of its original capacity over Y number of years.

BEV battery packs need to be designed to put out high power, for heavy acceleration and for DC fast charging. Home energy storage battery packs do not need to output or input high power, but do need to be designed to be cycled every day.

The arguments against wearing out your BEV battery pack by using V2H frequently are the same as those against using your BEV battery pack for V2G (Vehicle to Grid). Just read the comments to this recent InsideEVs article:


“almost no EV maker guarantees that an EV battery pack will retain X percentage of its original capacity over Y number of years.”

Two of the biggest BEV available today both do it.

Bolt EV and Nissan LEAF both have capacity warranty. So does BMW i3 and VW E-Golf as well as Kia Soul EV.

The Mercedes B250e and Smart EV also has some kind of capacity promise as well.

So, I would say “almost no”. Some do and some don’t is more accurate in my opinion.

I wouldn’t say “almost no”.

“Bolt EV and Nissan LEAF both have capacity warranty.”

Nissan does not guarantee an exact percentage for the Leaf, they only guarantee it won’t lose more than X number of “bars” on the display. As has been exhaustively discussed elsewhere, that’s rather difficult to translate into an actual percentage, as some of the bars represent greater percentages than other bars. There were also serious problems with the display not matching the actual battery charge level, at least in early Leafs. In fact, for awhile Nissan was falsely claiming that premature battery fade due to overheated batteries was just an instrumentation problem, which could be corrected by resetting the calibration of the display.

I hope that latter problem has been fixed in later production Leafs, but Nissan’s false claim in that regard certainly does not give me much confidence in the accuracy of the “bars” display. It also sounds like a Nissan service department would have a built-in excuse not to honor the guarantee, by merely claiming the display was mis-calibrated.

* * * * *

Those wishing to read more about the controversy over Leaf battery fade vs. display calibration problems can find the subject exhaustively detailed here:


“Bolt EV and Nissan LEAF both have capacity warranty. So does BMW i3 and VW E-Golf as well as Kia Soul EV.”

Okay, thanks for the correction (altho, as noted above, I don’t agree on the Nissan Leaf). Apparently the industry standard has changed while I wasn’t looking.

M3 -reserved Niro-TBD

Watch the video, not my claim; the Tesla engineer giving the talk says it over and over and over again.

Same battery, same tech going into the Model 3. Guess your next argument is that the Model 3 won’t be designed to discharge its battery as regularly as Model S/X.

A 5 Ton AC draws 17kw, that’s a decent amount of draw that the PW2 will need to withstand and designed to deep cycle daily. Read up on the PW2 and what V2H’s goal are and what they are designed to do — you may actually get what they are doing there.

Tesla (and all the other Energy divison/car companies) just doesn’t want to do V2H when they have people lining up willing to shell out $5500 more for it.

“M3 -reserved…” said: “Watch the video, not my claim; the Tesla engineer giving the talk says it over and over and over again.” I’ll have to take your word for it. I tried listening to the audio (no actual video to watch, at least not in the portion I listened to). I mostly couldn’t understand what was being said, and of what I did understand, it wall all old hat. “Same battery, same tech going into the Model 3. Guess your next argument is that the Model 3 won’t be designed to discharge its battery as regularly as Model S/X.” No, but I agree with Tech01x: They are surely using different chemistry in the PowerWall cells than the EV cells, even if they guy giving that talk claims otherwise. It simply doesn’t make sense to use the same chemistry in both. Panasonic isn’t locked into one single chemistry; it makes cells for consumer electronics, and also for EV makers other than Tesla. Why would Tesla buy EV batteries from Panasonic and then put them in a home energy storage system? It would be, frankly, a waste of money to do that. Sadly, in my experience, businesses routinely lie to the… Read more »

I suffered through the entire #&*@ thing on your recommendation, and never once heard him say “same cells”. He did say “similar cells” once.

Many say Tesla uses NCM for Powerwall and Power Pack. Kelty didn’t confirm or deny this, but I got a strong feeling they are NCA only. My best guess is they tweak their NCA formulation for Powerwall and Power Pack to trade off some power density in exchange for longer cycle life.

~21:00-29:00 He talks about the battery and supply composition. He shows NCA as their construction and spends a good deal of time talking about the components –specifically how Tesla uses 30% less cobalt (the main HIGH price cost in their graphs) than their competitors

This is the difference between NCA and NCM composition and cobalt and supports 2170 is NCA which they say will supply their PP2, PW2, and Tesla 3 packs.

“A 5 ton air conditioner uses 17 kw”.

For a house? Not anymore they dont. Most dont use 1/3 of that.

SEER has to be 13 now to be sold in the USA, they haven’t sold any more of those 5 eer units in decades.

Was looking at a 2 1/3 ton (28,000 BTU/Hour), a cheap frigidare unit, INCLUDING THE INDOOR AIR HANDLER CONSUMPTION, of 3110 watts. If you bought 5 tons worth of this model, it would be nameplate rated at 6.664 kw.

Right, my ground source heat pump is >25 SEER and it usually draws less than 2 kW.

While there is a big battery in the car, when the car is not home then you still might want to capture all that solar generation. Power wall is the answer for that. Power wall is stationary in the building capturing energy. Car is out and about, so not really the same scenario at all. I think the idea is to capture energy, transfer it to the car, house, etc. V2H is for emergency use.
Range of use for car battery, super slow discharge in heavy stop\start traffic, high outputfor acceleration, charge\discharge many times per day when tripping. Really can’t see how V2H is a problem on any measure.
I’m sure of you cycle power wall more than designed then it will have less life time.

At around 20:40 he claims the M3 cooling architecture has been changed.

We saw how the P100D cell arrangement changed from the earlier cell arrangement Here:http://insideevs.com/new-tesla-p100d-battery-module-walkthrough/.

Basically all they did was tweek the old module config to add more cells by changing to 2 parallel loops per modue from 1.

Here they are implying something different. I’m dying to find out what they did.

I really want to see the new cell specs and this new module design too.

This was May. You’ll notice that he said they had NOT started manufacturing EV batteries on the new cell…all new cell production has been going to energy storage. And he said ‘in a month or two’ they will start but they are having a hard time getting everyone and everything up and running. You’ll notice he said they are importing batteries from Japan until they can get the US up and running for EVs on the new cell.

March, not May. That implies starting Model 3 cell production at the Gigafactory in April or May, which is just about right.

Its still a 21700 cell, call it a 2170 if you love to follow whatever tesla tells you.

*Shrug* Call it what you like; the Thought Police won’t come and arrest you.

But companies get to name their own products, even if that means breaking with tradition. You can of course use a different label, but that just adds to confusion. Language is useful only to the extent that we all agree on what words and terms mean. (Humpty Dumpty was wrong.)

I’ve been told that the extra “0” in the “18650” form factor (18 mm diameter, 65 mm tall) was a holdover from a naming convention for watch batteries. Since watch batteries are so flat, it made sense to have an extra significant digit to indicate tenths of millimeters of thickness.

But with the 2170 (or “21700”, to use your label) battery being 70 mm tall, using an extra digit to indicate increments of 0.1 mm of height seems rather unnecessary.

He talks about this at 19:00 – He addresses the nomenclature and why. 2170 v 21700

Actually, if you reseach Litium cells, you find very few that have any digit 1-9 in that last position, and while there are cells that are 18mm x 65mmm, there are also cells that are, for instance, 14mm x 40mm, too.

Personally, when I bought my fist EV back in 2006, and started to learn about batteries and cells for a personal need, I found that last ‘0’ on ‘18650’ cells, just to be some stuborn element of confusion! It seems you are stuck with that same stubborn streak that those battery makers sold you!

I would have simply called them 1865 cells, as they are, generally, 18mm x 65mm, but some are in fact, 18.2mm x 65.5mm, due to manufacturing variances!

Some of them can’t be used in a ‘18650’ cell holder that is cylindrical, and tightly designed for an 18mm cell diameter, or a 65mm cell length, due to these various tolerances!

A good digital Vernier or Digital Micrometer is your best friend in that discovery!