Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Model S Battery Pack - Cost Per kWh Estimate

This site may earn commission on affiliate links.
Panasonic Corporation today announced that shipments of its automotive-grade lithium-ion battery cells for Tesla Motors' premium, all-electric Model S sedan will surpass 100 million units by the end of this month.
Let's presume that they are being "conservative" the press release by only talking about customer vehicles (i.e. not test cells and alpha/beta vehicles).

Napkin calculation (upper limit of VIN):
100,000,000 / 15,000 ~= 6666 cells

For 7000 cells per vehicle, it's 14285 cars.

Seems to line up pretty well with the VINs we've been hearing lately.
 
To be quite clear it would not be that difficult to design a machine that aligns with the pack support bolts and removes them all in a few seconds. Probably how it's installed at the factory. Or a couple of people with impact drivers doing the same thing in less than a minute. Realist is talking gibberish.
 
Let's presume that they are being "conservative" the press release by only talking about customer vehicles (i.e. not test cells and alpha/beta vehicles).

Napkin calculation (upper limit of VIN):
100,000,000 / 15,000 ~= 6666 cells

For 7000 cells per vehicle, it's 14285 cars.

Seems to line up pretty well with the VINs we've been hearing lately.

Diamlers, B-Class and Smart
Toyota
Gridstorage for Solarcity
 
I guess there is no clear word on the time it takes to swap such a battery regarding the ZEV credits.

I really cannot think of battery swap system for this car. What I can think of is some kind of hybrid attachment to it in order to charge the battery.

Don't get me wrong guys. I have been monitoring the Model S for years and I really have deep respect for the Kind of Engineering these people have done. This is a tremendous achievement.

In fact the greatest cars in history have only rarely been money printing machines as well.

You are flat out wrong. Batteries are in fact replaced in service centers and it currently can be done with a group of people in less than half an hour. Although the battery is an important part of the structural rigidity while driving, it is simply bolted to the rest of the frame and easily removable. This has been documented by Tesla numerous times and is in fact apparent simply looking at the frame and battery that they display in their showrooms.
 
FWIW

LG had a presentation in China a few years ago http://img545.imageshack.us/img545/5350/liion.gif (full link now broken), which included commodity cell pricing.

It collaborates the $200/kWh as at 2010, assuming Tesla's NCA cells are more expensive than usual energy cells, but that prices have kept coming down, it would seem that first pass, $200/kWh is reasonable, with Tesla having the option to reduce that cost today if they were happy to use a lesser tier manufacturer and/or a lower energy density but safer chemistry (ie Mn2O4 blend).

NCA is about the best for energy/power/or combinations thereof, but its not as safe some other chemistries.

Future high energy chemistry are likely to have less power, so a hypothetical future 150kWh battery may only recharge at 60kW compared to a current 85kWh battery recharging at 120kW.
 
With reference to the newest Berger study there’s no indication that we are anywhere close to the sub 200$. In fact they think major innovation in CAM technology will be necessary to bring the cost down to 200$ by 2018.

If you are talking about the Berger Oct. 2012, "Lithium-ion batteries - The bubble bursts" report, first it uses his 2011 projected costs. He models the "typical 96 Wh PHEV cell" which, from his "Market Overview" presentation, is a NCM chemistry for which we already know is a much higher cost per kWh than Tesla's current solution. Interestingly enough, the overview report dues show a 2015 cell price breakdown for NCM chemistry that is $1.50 per cell for a 2.2Ah battery. Therefore, $2/cell for a "older tech" NCA battery is not out of the realm here. In either case, I think the market is gotten lower than these reports indicate.

Manufacturing. If you assemble such a large battery pack you are working with a high voltage component storing a lethal amount of energy. You need to implement high levels of safety. The same applies for logistics and shipment. It’s not an easy and simple process. It never will be.

Which is why I am modeling at $5,000 integration cost per pack. Again, look at the National Geographic Tesla factory tour video, specifically the battery assembly portions. Can Tesla drive these costs down? Quite possibly because this part is under their direct control.

If that battery has to be changed the cost will be horrendous. On a Lotus Evora a gearbox change goes to the complete dismounting of the chassis with a cost over 15.000$ thanks to the massive amount of working hours. I have no idea how Tesla wants to change such a high voltage pack which is completely integrated into the car’s chassis. A battery swao system for this car is impossible and fatuous!

You really went off the rails here. Again, look at the National Geographic Tesla Factory Tour video and look at how the battery is attached to the chassis. Removal is quite easy. I can understand that there is some confusion because the popular image is of the "skateboard" where they show just the battery and the suspension. There are two pictures in this photostream that better reflect the reality:

Flickr: Beauty in Metal's Photostream
Tesla Model S Chassis, Battery Detail | Flickr - Photo Sharing!
Tesla Model S Chassis, Battery Detail | Flickr - Photo Sharing!

The battery is bolted to the space frame.
 
Musk is not a genius, but obviously a smart business man.

Don't think there has been any public disclosure of his IQ, so hard to judge. He obviously is very good at what he does.

Regarding links. I cannot provide as many links as I want because I am junior member, watched by the admins. Yet all the sources are credible.
So let’s go back to TESLA’s presentation of 2012. They don’t give a precise number but on the graph you clearly see an indication above 350$.

Do you really think that the admins in this forum would be unwilling to allow a demonstrably well-researched and referenced post?

With reference to the newest Berger study there’s no indication that we are anywhere close to the sub 200$. In fact they think major innovation in CAM technology will be necessary to bring the cost down to 200$ by 2018.
It’s just not that easy. The cells used on the Model S are identical to a laptop cell in the same way a Smart Two Seater is identical to a Mercedes S-Class, since both vehicles have 4 wheels and an engine.

Right now the biggest player for Li-ion High Energy cells in the automotive sector is Toyota.
But of course since the cells are mainly used for hybrids they use relatively thin active layers. But for a 85kwh battery pack you cannot use that. You need thicker active layers in the electrode with higher internal resistance. So already on the automotive level there are great differences.

You mean a vehicle of the type: let's build a car with n wheels and n=4? Tesla started out looking at standard 18650 cells. There is quite a lot on Tesla's IP, including patents, in this thread, all quite well referenced:

=> Amazing Core Tesla Battery IP - 18650 Cell

Furthermore it’s simply absurd to believe that Tesla is able to lower the cost by simplifying the safety and thermal management in comparison to a laptop battery just because they use more cells. In a car you have a completely different environment. The battery management system is a lot more complex. Some cells for example might experience a capacity fade and manifest an increase in impedance. So what you need than is a system that balances the cells to maintain near equilibrium voltage across the entire battery string.

Thermal management. Have you ever seen a laptop heating it’s battery when turned off?

This is Tesla's core IP, how to deal with a complex battery system using cheap simplified cells.

Laptop batteries don't last very long and are a relatively cheap component to replace compared to the cost of a good buissness notebook. Using an expensive, complex, fluid-based and heavy battery architecture in a wheight and price constrained Laptop doesn't make any sense. The cost of such a battery architecture doesn't just scale down linearly from a huge Tesla battery.


Manufacturing. If you assemble such a large battery pack you are working with a high voltage component storing a lethal amount of energy. You need to implement high levels of safety. The same applies for logistics and shipment. It’s not an easy and simple process. It never will be.

These batteries are not assembled fully charged. You only deal with the high Voltage with a completely assembled battery, when in use. Obviously, there need to be high safety measures built into the battery architecture. Again, this is Tesla's own core IP and as Boing spectaculary demonstrated very difficult to get right, which gives Tesla a huge headstart compared to other manufacturers.

It is true that an accident damaged EV battery can be a serious hazard and service personnell and emergency services will need to deal with that.

A few words on the chassis design and construction. As I have mentioned the cost for each Model S is 95.000$, according to Q1 2013. That is by any comparison extremely expensive. Especially if you take a look at the chassis design.

You have, yet, to justify your numbers. Furthermore, do you think that the production cost of the Model S at this stage is representative of the price in let's say Q4 2013?

The Model S is basically a very simple one box rolling chassis design. It’s not a Space Frame like in the Audi A8. Instead it’s a rigid structure with the body parts attached to it. It’s similar to the Lotus VVA (Vehicle Versatile Architecture). This is not a surprise since Tesla chief engineer Peter Rawlinson is a former Lotus Engineering guy.

If you google pictures of the Lotus Evora you will find instant similarities to the Model S chassis. In detail the Model S chassis is even simpler and less complex with standard parts regarding wishbones, steering, brakes, even the air suspension. The Model S does neither have any expensive pre active safety feature. Just take a look at the doors and the seats and you know that this car is basically about saving weight and cost. Which is fine, because with such a low center of gravity and no engine in the front where is no need for any active dynamic tricks. This is a very nice straightforward design.

Agreed.

There is only one problem. Such a structure is not really working in a car weighting more than 2 tons. The chassis is not rigid enough. Not until you use the battery pack and it’s safety cell as a stressed member. In fact it is the central part of the chassis! That is a very intelligent construction, but of course not if you want to change that battery on a later date. And that is exactly the Achilles of the Model S.

If that battery has to be changed the cost will be horrendous. On a Lotus Evora a gearbox change goes to the complete dismounting of the chassis with a cost over 15.000$ thanks to the massive amount of working hours. I have no idea how Tesla wants to change such a high voltage pack which is completely integrated into the car’s chassis. A battery swao system for this car is impossible and fatuous!

The battery pack has been designed to be swappable from the start. Even though it is a structural element, it is part of the final assembly of the car and bolted into the rest of the chasis from the outside. It is in fact installed "in seconds":

=> The Tesla Factory: Birthplace of the Model S | Blog | Tesla Motors

This aspect creates 2 problems:
1. The insurance cost for the car will go up, because in case of an accident the battery pack might need a replacement and
2. The resale price will drop dramatically. Elon Musks 3-year resale price guarantee is in fact an explosive burden for Tesla shareholders as the resale value collapse is pretty inevitable.

1. It remain to be seen, how easily the battery is damaged in an accident. Looking at this forum there have been a couple of accidents reported and experienced and so far the general consensus seems to be that the Model S is built like a tank.
2. That is pure speculation. "inevitable" - You haven't brought any convincing arguments to the table, why such a value collapse should occur, even when batteries might have to be replaced more readily after accidents. Either the battery is damaged and needs to be replaced or repaired, or it is whole, in which case the value of the car shouldn't suffer. Unless the Model S tanks in a major way, because of so far unknown serious short-comings, low wear and tear will garantee at least a reasonable resale value. Even if there is a share holder risk imminent in this guarantee, it is not necessarily catastrophic. Keep in mind that Tesla not only will be able to demand a higher resale price for a Model S in connection with extended manufacturer guarantee, than 3rd party vendors. But they also don't need to make any profit on this secondary market transaction, but just break roughly even.

I don’t want to create “Anti-Tesla” Gibberish here but just look at the facts. In England Lotus produces a low volume Alu- composite car for less than 40.000$ and sells it for 65, whereas Tesla produces a high volume Model S for 95 and sells it for 105. How does that fit to a battery pack cost < 12.000$.?

Tesla up-front battery replacement of $ 12.000 can be redeemed after 8 years only without paying a penalty, which is from 2020 onwards. Battery prices have a way to go before then. On the other hand the Blue Star 3rd Gen vehicle will have been out for a few years, producing further economies of scale.

I guess wishful thinking is an understatement here.

I really don't think you are doing your user name justice with the imminent pessimism displayed in this and your other posts. If you produced good sources to back all of this up, I might be more willing to listen to you.
 
The 95.000 $ number can easily be calculated using Teslas Q1 2013 numbers.

Battery Swap. I have not seen a picture of the battery been attached to the Chassis.

What I can see is the Body Panel structure been bolted on the Rolling Chassis. That is indeed very easy but also something completely different.

Furthermore, attaching the battery to the chassis in a matter of seconds may no work the other way round.

If Tesla proves otherwise my arguments are false. Agreed.
 
It is truly a shame that the high quality flow of well thought through analysis has been interrupted by Realist's comments and the retorts that follow. Please allow CO and others to return this thread to its origins. Realist should loose interest in posting if no one responds (he obviously is not reading or understanding the content of the other posts).

As for CO and others contributing, thank you; this thread is a pleasure to read.
 
Regarding links. I cannot provide as many links as I want because I am junior member, watched by the admins. Yet all the sources are credible.
So let’s go back to TESLA’s presentation of 2012. They don’t give a precise number but on the graph you clearly see an indication above 350$.

No. They do not. The explicitly state that they are not providing the cost of their battery. The "indication" you are talking about is a thought-bubble containing the citation, which is clearly not a part of the graph. How you see that as a disclosure that they proactively affirm they are not making escapes me.

It is becoming increasingly mysterious to me as to why I should take you seriously when you refuse to present evidence to support your case while relying on an investor graphic which explicitly refutes your point.

Tesla_Presentation_-_Spring_2012-20.jpg
 
I love this thread. But rather than get hung up on the $/kWh can we discuss the rate of change? It seems this is the real headline here. The real question is what rate of decrease do we model for the next few years? It seems that there may have been a disruptive step-change in the cost too, which is incredibly important if true. Two causes of a "collapse" that have been mentioned here are:

1) Tesla becoming one of the worlds largest buyers and
2) Laptop demand soft.

That could lead to battery makers lining up to switch their lines to make the "simple cap" Tesla cells. But, even using some conservative assumptions (gleaned from CO article and TMC consensus):


10%annual reduction in cost/kWh
89kwh/pack
8000cells/pack
$2,000 pack cost, excluding cells
And using these two cost points:

yearapprox cost/kwhcomment
2009250per CO article
2012170per CO article
(forgive me if I misquote, I am picking and choosing some numbers to use)

That works out to be about 10% per year. Then you get this sort of trend, barring any future disruptions:

View attachment 23597


So a pack cost of 16.5k now, and a 12.6k 89kWh pack in 2016 for gen III ($9.5k for a 200mile). You can see why EM can say they can get there without major breakthroughs.

What does a BMW 3 series drivetrain (engine, transmission, etc) cost? We could be within a few % of cost parity, even with some problems with model assumptions. It starts to appear that in the next 5 years Teslas REAL problem is economy of scale and tooling up to not lose first-mover advantage, not battery stuff.

Thoughts?

First thought is, nice post!

The critical issue is of course cost. I'm pretty sure you can make a Gen III right now using the Panasonic NCR18650B 3400mAh cells if the cost was right. The problem is that there must be some lower bound that is based on the combination of materials costs and manufacturing costs. And they are kinda locked in on the materials that are used for that battery, and there probably isn't a ton of improvement to be had on manufacturing it.

So if the current "collapsed" price is not cheap enough, they probably need a more efficient design to get the cost savings they need. When I look at the research there is a lot of talk about improving the films and separators that are used in order to reduce costs, and there are potential cost savings to be had from a chemistry which uses inexpensive ingredients.

Getting all of those moving pieces to fall into place with a battery that otherwise matches (or exceeds) the performance characteristics of the 3400mAh cell is the core expertise of companies like Panasonic. And it seems very likely that they are very advanced into the prototyping stages for the cell that Tesla will use (prototyping both the cell and the manufacturing/logistical processes).

There isn't really any need for moonshot batteries that are being talked about in R&D labs. This is just an engineering and management problem at this point.

Or, there is the alternative scenario where the 3400mAh cell has fallen to ~$1.50/cell already. That seems improbable, but at that price you can probably build a Gen III with a true 300 mile range for ~$50k (using 6,432 batteries) and a 240+ mile car using 5,184 batteries for ~$40k.

I've posted ads for these batteries that listed $1.8-$3.2 and even $1-$3. The implied price from the 2012 IEK report would be ~$2.50/cell, if you assume that this was the most expensive cell in their sample. So the question is whether they've fallen from that level (or heck, the price might have gone up and the ads are a worthless indicator).