Exclusive: Tesla Takes Delivery Of 1.74 Million Samsung Battery Cells


Bill Of Lading

Bill Of Lading

It’s come to our attention, via an April 19th 2016 bill of lading for the Port Of Oakland, that Tesla received a massive shipment of Samsung SDI ICR18650-22V lithium-ion battery cells.

The weight of the shipment is listed at 113,183 kilograms or 249,526 pounds.

This volume is far more than would be required for testing purposes, so the logical conclusion is that Tesla has indeed signed a battery cell supply deal with Samsung SDI.

Previously, we had reported that Tesla was in talks with LG Chem, Samsung SDI and SK Innovations for battery cells. It was believed that Tesla was trying to secure another battery partner (aside from Panasonic) to shore up its supply line.

It now seems that Samsung SDI was already the chosen second partner. Further confirming this is Tesla’s receipt of a much smaller (3,390 kilograms) shipment of the same Samsung SDI cells on April 6, 2016. That shipment may have been for testing purposes.

The actual cell that Tesla took delivery of is Samsung’s ICR18650-22V, which is an unknown class of cell to us (provided of course the labeling was correct to contents).

Here (below) are some specs on the SDI ICR18650-22F cell (and here’s a link to the specs for a Samsung SDI ICR18650-22P cell):

Manufacturer – SAMSUNG
Rechargeable battery type – Li-Ion
Battery size – 18650, MR18650
Rated voltage – 3.6V
Capacity – 2200mAh
Body dimensions –  Ø18.25 x 65mm
Maximum current – 4.4A
Gross weight: 0.13 kg
Collective package: 2 pcs

And here’s an image of a Samsung ICR18650-22F cell:

SAMSUNG ICR 18650-22


These particular cells are typically sold in units of 2 with a listed weight of 0.13 kilograms (or 0.065 kilograms per cell). If we do some math, we find that Tesla’s purchase of 113,183 kilograms of cells equates to 1.74 million cells.

The question is, where are of all these cells going to be used? For an EV application? Tesla Energy for Powerwalls or Powerpacks? Something else entirely? Surely these cells won’t sit around waiting for the Model 3, so that’s out of the equation, but Powerwalls or Powerpacks seem to be the most likely recipients.  It is fun to speculate.

Below is a description of just a few of the containers on the ship loaded with Tesla’s Samsung SDI cells.

Container Contents

Container Contents

Hat tip to BillK!

Categories: Tesla

Tags: , ,

Leave a Reply

64 Comments on "Exclusive: Tesla Takes Delivery Of 1.74 Million Samsung Battery Cells"

newest oldest most voted

SO…how many Model S’s is that?

250, or something like that…

None actually since those cells don’t have enough energy density for a Model S, or X.

Biggest guess is powerwall, nothing else makes much more sense.

Yeah, 2200mAh isn’t dense enough for a EV. Probably something that is less sensitive to volume and weight, which is stationary storage.

Sounds good. Should free up a lot of Tesla’s own battery production for use in cars.

Agreed. I believe the Panasonic cells Tesla uses are 3100mAh.

The Panasonic NCR18650B cells that Tesla originally used in the Model S were 3400 mAh and they must be better now for the 90 kWh option.

Keep in mind that 220mAh spec is for the 22F and these are 22V cells, so their capacity is unknown.

ALL Samsung cells ending on 22x has a capacity of approx. 2200 mAh.
If it had been e.g. a 3500 mAh cell it would have ended on 35x instead of 22x.
The last letter after 22 tells something about the maximum discharge rate/internal resistance.

SDI ICR18650-22F = 2C
SDI ICR18650-22P = 5C
SDI ICR18650-22V = ?C

I would think the trailing letter was a generation indicator rather than a performance indicator of any kind. But yeah 22 is the capacity.

It is a performance indicator. Samsung has been using the same suffixes for many years, consistently.

I’m not sure its a market mover, but here’s more anecdotes of demand activity in stationary:

The more hostile Australia and utilities like Buffet’s get, the more people will pay through break-even, to sever ties.

I could be for energy storage at Supercharger stations to avoid demand charges and for peak shaving.

Nope. Powerwall cells are made to be cycled once a day. These batteries mentioned are made to be cycled 500 times.Thus once a week. They also are not much. I do not think the picture is of the actual cell. There is a V on the bill and an F on the picture. Besides the energy density and the number of cells are just enough to make 230 packs of 60kWh each. I suspect that these cells are for the M3. To make 200 or so M3 before the end of this year seems right!

I’m beginning to wonder of “Alaa” is just pranking us. How can every single post from him be so very wrong? You’d think he’d be right just by chance once in awhile.

The Model 3 will use the new, larger Gigafactory cells, Alaa. Not the 18650 cells specified in this shipment.

Yeah, using them for stationary storage seems likely.

And Alaa, if you’re not just pretending ignorance, there are at least two things you don’t appear to understand here:

1. DoD (Depth of Discharge) has a significant effect on battery life. All Tesla has to do is use a shallower charge/discharge cycle to get more than 500 cycles out of these cells. Quite possibly a lot more.

2. The rating for how many cycles a li-ion cell can withstand is an indication of how many before it gets down to 80% capacity. But there’s no reason stationary storage batteries have to retain 80% capacity in order to function properly. In fact, it would be a waste of money to not use them down to 50% capacity… perhaps even a bit lower before replacement.

Look at this
10 years that is 3650 times at DoD 100%. The cells mentioned here could not possibly be the type that they use in the powerwall or powerpack.

I have a feeling that the Bill of Lading has miss leading data except for the weight and maybe number of cells. The weight of the shipment and number of cells are maybe the only important thing for ship and clearing the customs. But what is actually inside the cells? Cobalt Magnesium who knows?

Well okay, fair points for you Alaa, the Tesla website does actually say “depth of discharge — 100%”.

Now, I rather strongly suspect that was written by the PR people rather than the engineers, and what that actually means is that the unit is cycled at 100% of the 6.4 kWh it’s rated at… that is, that’s the usable kWh rather than the actual capacity. Either that, or everything I know about li-ion batteries is wrong.

But fair points to you, it does actually say 100% DoD.

I leave as an exercise for the reader, the physical impossibility of daily cycling li-ion batteries at 100% DoD for 10 years, yet retaining a significant percentage of full capacity.

Musk just said that the average car cycles the equivalent of once a week.

If these cells were for Tesla Energy it seems like they would be sent to the gigafactory. Perhaps the are building large powerpacks at freemont

Thank you.

No, Tesla is assembling the PowerPacks and PowerWalls at the Gigafactory.


Pushmi-Pullyu said:
“How can every single post from him be so very wrong? You’d think he’d be right just by chance once in awhile.”

If that isn’t the pot calling the kettle black, I don’t know what is. ¯\_(ツ)_/¯

You’re very condescending when you accuse Alaa of “pranking us” and “pretending ignorance,” especially when you constantly spout off about how Tesla is profitable by “pretending ignorance” about the difference between gross profit and operating profit, even though countless people have explained the difference to you numerous times. Tesla is NOT profitable. I think you just like to insult people and pick fights with them by calling them derogatory names like FUDsters, pranksters, shills, stock shorters, etc. Are you the world’s older cyberbully?

“Tesla is NOT profitable.”

Now you’re dropping the blanket statement.

Not knowing the Model 3 uses something other then 18650 cells while making related claims is very different than super semantic arguments about Tesla’s profitability.

It costs less to make the Model S then they sell it for. That’s what people mean when they say “profitable”. They are not taking about GAAP profitability.

Well, sven, you’re not pranking us. You know quite well that what you post about Tesla isn’t true. Your FUD is a deliberate disinformation campaign, probably aimed at manipulating stock prices, just like other Tesla short-sellers who post here. Certainly you’ve shown you have a deep interest in Tesla’s finances, even more than mine. I’m not an investor… and it seems you are.

If Tesla wasn’t investing huge wodges of cash in growing the company, they’d certainly be cash flow positive, and then all you FUDsters wouldn’t have an excuse to pretend Tesla “isn’t profitable”.

Alaa, you forget that the cells do not have any public spec sheet and the 22F and 22P specs could be vastly different in terms of cycles and charge/discharge current.

2.2Ahr battery at 4.4A discharge is only 2C. Ugh! Glad SparkEV isn’t using them. Otherwise, it’d have less than 40 kW (53 HP) instead of 103 kW (137 HP) that it has now.

Maybe they are better suited for power-wall application.

How do you know discharge rate?

From article, “Maximum current – 4.4A”

The Spark EV doesn’t use 18650 cells.

1.74 million cells x 3.6V x 2.2Ah = 13,791 kWh
good for 153 90kWh packs

More likely 135 x 100 kWh PowerPacks.

“The question is, where are of all these cells going to be used? Model S? Model X? Tesla Energy for Powerwalls or Powerpacks?”
Isn’t the chemistry slightly different between the car batteries and the Powerwall batteries?

These cells are labeled ICR

That probably means their chemistry is LiCoO2 (LCO).


Those are old formula cells as far as I know. Currently I think Tesla use NCA in the Model S. and LG uses NMC in the Bolt. I think the Roadster used the same chemistry.LCO

I’d say they were for very large scale energy storage that gets fewer cycles because AFAIK those old chemistry cells only have around 500 cycle life.

To add to your post, the 6.4 kWh (formerly 7 kWh) Powerwall uses NMC, while the abandoned/defunct 10 kWh Powerwall used NCA. The Tesla Powerpacks apparently use NCA.

Using “13,791 kwh” from above, they could be for on-site GF storage. Tough to gauge power consumption and reserve targets of so big a place. BMW uses a 10MW CHP set-up, in its Spartansburg, SC facility. That would be for instantaneous needs.

I guess this is one way Tesla could turn up the heat on Panasonic.

7104 x 2.2Ah = 15.6288 KWh!!!

These are much too low density to go into the cars. But even assuming they’re for the powerwall, that means you’ll almost 3000 of them for the basic 6.4kwh powerwall!

I imagine their cycle-life must be phenomenal.

oops. nevermind. I forgot the voltage:
7104 x 3.6v x 2.2Ah = 56.26KWh

same conclusion though.

Sorry to correct you, but its 7104 x 2.2Ah x 3.6V.

Still just 56 and some kWh, so too little for a car. But 800 of those could make a nice 6.4kWh powerwall.

56.26 kWh would be just about right for Model 3 test mules, and if they are going to start making the new 20700 cells in the Gigafactory not for about a couple more months or later, these cells could be for early Model 3’s for getting a jump on testing sub systems in the car, and putting miles on the cars for fit stability and structural integrity tests, while they await the new cells.

Imahine if the Model 3 could get 215 miles range on older cells, how mutch range it could get from the new ones?!

These cells could also be part of a new Supercharger energy storage addition, not for sale, but for Tesla to avoid some demand charges!

Or even stranger, the new Tesla eBikes I heard some comments alluding to!

Good luck with that line of thinking. 7104 cells is the model S pack size. Not going to happen.

I agree it could be for the test mules for the model 3or,

for energy storage at the gigafactory itself or vehicle assembly itself.

If these are comparable to the Panasonic in Ah than that would be enough for about 220-225 cars. Panasonic was listed at 2.7Ah per cell with an output of .00972kWh/cell according to panasonic web site. Currently at 75kWh for model S and the X, would require about 7700 cell per car, or about 220 cars, more or less.

Roadster 3.0 packs?

Unlikely Roadster 3 cells, as these do match up with original Roadster pack cells… about 2,200 mAh 18650’s.

They could use these in remanufacturing existing Roadster packs, but buying 250 cars worth? That doesn’t make much sense.

There was a Roadster battery replacement plan purchased by some Roadster owners. We are coming up on 8 years since Roadster deliveries began, so that plan should be kicking in.

Right. I remember reading somewhere when the Roadster3.0 plan was announced that the new packs will be made with Samsung cells. It’s fair to presume that this plan is now in execution.

I recall this ^ also

My bad, I was confused. Actually it seems that the upgrade Roadster 3.0 program is based in LG Chem batteries:http://www.wsj.com/articles/tesla-gets-boost-from-korean-battery-maker-lg-chem-1446007554.

Yeah, if it was high power applications they would have used the 25R which is 2500mAh at upto 8 Amps.

Iirc all the car cells are 3000+ mAh, so this wouldn’t make sense anywhere else. According to the Spec sheet

These have a 1C charge and 2C discharge, that would match up very well with PowerWall applications. I think they’ve struck a large Industria PowerWall deal.

Yeah, Musk mentioned that they were starting to big project in Hawaii in conjunction with SolarCity for quite a few PowerPacks. They’ll need quite a few more shipments, as that is 52 MWh.


Looking at the bottom of the page link under applications might give you a better idea !

The suffix V isn’t listed anywhere I can find. That last letter, as Seth mentioned, denotes the output capability of the cell.(2200 is the mah rating) Where did this report and pic come up with the “F” suffix?
My guess is Powerwall test cells.

Per “Where did this report and pic come up with the “F” suffix?”, as they said, strickly for a baseline reference of known cells, as they could find no quick references to specs for the ‘V’ variant of these cells.

Perhaps they have an internal need to buffer a solar panel array, and it’s more cost effective to buy lower capacity cells instead of diverting their own production for internal use.

Could be to allow some super chargers to use more solar and less grid power.

My guess – these are cells for the 50MWh Powerpack install in Hawaii. The shipment size is about right. JB talked about this project at the shareholder’s meeting earlier this week.

Cell specs seem consistent with the older Roadster (although not for 3.0 version).

It could be for powerwall/powerpack, but the cells specs are consistent with a backup power application, and Tesla cancelled their backup version of the powerwall.

Impossible to be for S/X given the low density.

I somehow think if Tesla got a shipment of 50 million batteries they could most likely eat though that with in a few days or a week.

As far as i know all 18650 cells are around 45g! not 65

‘FlightSpeed’ batteries for Radio Flyer.

What do I win?

Yeah, this could be the upgrade to the original Roadster. I know it was going to use non Panasonic batteries.

“These particular cells are typically sold in units of 2 with a listed weight of 0.13 kilograms (or 0.065 kilograms per cell).”

The listed weight of 0.13 kg is the gross weight including packing for the two cells.
I have never seen an unprotected 18650 li-ion cell weighing more than 50 gram according to specs. Usually the weight is somewhere between 44-50 g.
And if you open the datasheet it http://www.tme.eu/en/Document/71b610093f14a53a8417944008628183/ICR18650-22F.pdf
It says: “Cell Weight 44.5g max”.

Of course the container shipment also has some additional weight for pallets and packing.