First of all, special thanks to Stopcrazyapp for some help on technical information on batteries, which really helped clarify a number of things for me.
Hey Kids, Don't Try This At Home!
I've been working lately on modeling costs for the battery pack for the Model S, both to get a sense of how it impacts Tesla's margins, as well as to understand how it's going to affect the CES concept that they appear set to roll out.
It turned out to be quite a bit more difficult than I expected, and the information I am going to present is far from definitive. So feel free to poke holes in this as you see fit. This is a working thread, and this post is just a starting point.
First, lets just get one point out of the way. There is no catalog that lists the cost of the Tesla battery. For that matter, there isn't a handy guide for the individual cell costs either (or maybe I just missed it). To generate a credible estimate you have to gather data, perform some analysis, and spice it up with speculation.
I will present some data in a moment, but I'll ruin the surprise a bit by stating that the available data (to me anyways) is highly variable and contradictory to the eyes of a layman like myself. So it's helpful to try and establish boundaries, and see if we can narrow the range of where we are likely to find our answer. As a layman, the first way I try and do that is to see if folks who may be smarter than me have studied the issue and said or written something about it.
Setting Boundaries - Media Style
For instance, its often been reported in the media that the cost per kWh for the Model S is $400/kWh. By itself, that is an impressive number, because it is much cheaper than any other automotive battery. At the time the Model S came out, the battery for the Leaf was estimated to cost $18,000, or ~$750/kWh. So obviously the Model S is amazing with it's $400/kWh cost. But how did the media get that cost?
I wont cite it, because its not important, but they got the cost by looking at the cost difference between the 60kWh and the 85kWh Model S. The cost difference is $10,000, which works out to $400/kWh for the 25kWh of storage.
Even as recently as last week media reports were citing this number. The problem is that the cost difference between the two cars is a retail price, and tells us nothing about the cost to Tesla. Nevertheless, the cost difference is a very useful metric to bound the data that we will come back to.
Goldman Sachs
Beyond the media, we've had highly paid financial analysts estimate the pack cost. In 2012 Goldman Sachs estimated that the pack cost Tesla $22,500 -
Is Trouble on the Way for Elon Musk and Tesla?: Video - Bloomberg
With an 85kWh battery that works out to ~$265/kWh. Goldman Sachs is a reasonably credible source, but we don't know how they arrived at the number, and its from 2012, so its slightly dated.
Elon and Company
In 2010-2012 various Tesla related sources (including Elon) have said they expect the pack to be delivered for under $200/kWh -
Cheap batteries will make the Tesla Model S profitable at 20,000 annual deliveries - egmCarTech
Endless-sphere.com View topic - Tesla pack costs $200/kwh? Tipping point price?
An 85kWh pack would cost around $17,000 at this price. Note also, they are saying pack cost, not cell cost.
International Trade Groups
The IEK did a report at the end of 2012 stating that 18650 cells (not limited to Panasonic) were running between $120-$200/kWh during Q3 when Tesla was first beginning production -
Global Li-ion Cell Shipments to Increase 9% in 2013: IEK-ITIS | CENS.com - Taiwan Industry Updates | Industry In-Focus | HTML |Ta1-CaE-Dy2012/12/03-Id42230
Smarter Media Method
We can do some speculative bounding in a few different ways.
First, going back to the media estimate of $400/kWh, we can refine that methodology in a couple of ways. For instance, we now know that $2,000 of the cost difference is the SuperCharger fee. Doing media math we get $320/kWh, based on the actual $8,000 price difference.
Again though that is a retail cost. The retail rule of thumb is that the retail cost is twice wholesale. That gets us to $160/kWh.
Science!!
Or, turning away from speculation we can go all scientific on the problem. Here is a definitive paper from 2000 -
http://www.transportation.anl.gov/pdfs/TA/149.pdf
And a must read writeup for the layperson that references the above paper -
http://www.tms.org/pubs/journals/jom/0809/daniel-0809.html
Bottom line is that in 2,000, the reference design was a 18650 1350mAh cell, with a materials cost of $1.28, and a production cost of $0.42, with total cost being $1.70/cell. The big problem that the authors identified was the need for the material and production costs to come down, which they expected to happen as mass production got rolling.
If you take that estimate, and compare it to information contained in this terrible (but must read) story by a clueless fool, you can maybe use it to (very) roughly project the costs from the 2,000 study into 2013 -
Lithium-ion batteries for autos | Future growth same as historical
This is the most important graph -
Looking at the fall in costs between 2,000 and 2012, we can estimate that costs are approximately 1/6th of what they were. If you figured that power density has (very) roughly tripled from 1350mAh to 3100mAh, then you can (very) roughly guess that costs must have fallen by half, to around $0.85/cell. Doing that gets you battery costs that are (very) roughly 1/6th of the conditions studied in 2,000. A bastardized methodology to be sure.
Moving on...
Because the initial estimate was just the basic costs to assemble a battery (as opposed to run a business) we will apply a simple rule of thumb markup and double the cost to find a (very) rough wholesale estimate of $1.70/cell for our current generation 3100mAh cell.
FYI, a thread somewhere says that Elon has estimated the lowest possible cost of the li-ion cells to be around $0.80 right now. I don't feel like running down the citation, but it's out there. If some bastardized method of analysis gets me within shouting distance of something Elon said, I'll take it.
Time To Build A Pack
Ok, we've set some boundaries, whether per/kWh or at the cell level. The next step is to simulate a pack so we can go shopping for cells and see how much it costs. The first problem with this is that there is an active controversy about many aspects of pack design. Some folks think that Tesla is using ~7,000 NCR18500B 3400mAh cells. Other's think Tesla is using some other number of cells.
For purposes of my simulation I will be using 7,968 modified NCR18650A 3100mAh cells for the 85kWh car. That works out to ~88.9kWh of storage, with the normal operating SOC being between 5%-95% (which on your indicators correspond to 0-265 rated miles).
This simulated pack is consistent with how the Roadster pack operated, characterizes the amount of emergency range you have below zero (thank you Broder), and leaves the upper 5% forever untouched. From this simulation, and knowing that power is reduced by half when the Model S enters emergency mode, we can guess that the Model S might be capable of traveling 20-30 miles after being computer limited to ~45mph in an emergency. Again, this is consistent with what Broder experienced. You are welcome to simulate your own pack, but I have a reasonable level of confidence in my pack design, so I'm going to go with it.
So knowing how many cells I am looking for, now I need to find a per cell cost. And with an ~89kWh car we need to modify the "promised" price of the Model S pack to ~$17,800 ($200/kWh).
Data, Data, Data!!
Knowing I want 7,968 NCR18650A 3100mAh cells, I did like Denzel in American Gangster and hopped onto a metaphorical Google flight to China to buy the product at the source, and eliminate as many middle men as possible. In various back alleys of Shenzen, where Panasonic appears to do most of its assembly, I found a massive variation in pricing. Many vendors are quoting prices for these batteries at $1-$10 depending on how many millions you are prepared to buy.
Theoretically, these prices are dependent on the quantity you want. You get the best price by purchasing large numbers on a regular basis. So a contract to purchase 1 million per month for the next year might let you purchase the item for $1/cell, while if you buy just the minimum purchase it's more like $10/cell.
However, as Stopcrazyapp pointed out when I linked him an example, many of these price ranges were probably placeholders for related models. When I attempted to contact the sellers, nobody was willing to give me pricing information. Maybe some of you can try and succeed where I failed.
So I dug deeper by clicking on every vendor I could find, and then going through their whole product line to see if they had other prices advertised with better differentiation. I also ignored the possibility that these batteries cost $1, because when I sifted enough prices, I found that batteries costing that much were probably older cells (like 2400mAh), while I started to find credible hints as to pricing for our target cells.
Here are numerous examples of links that I think are good bets to represent actual wholesale pricing for the 3100mAh cells, because in most cases, these prices were differentiated from other prices in that distributors product line. I also included a few 3400 cells for flavor (and because they had a different price from the same distributors 3100 cells) -
Original Japan li-ion rechargeable 18650 battery with 3100mah
18650 cells 3100mAh Li-ion Pana sonic 18650 battery rechargeable battery
NCR18650A 3100mah 3.7v lithium battery with cheap price, View NCR18650A 3100mah 3.7v lithium battery, panasonic Product Details from Shenzhen Victpower Technology Co., Ltd. on Alibaba.com
Panasonic 3100mah NCR18650 3100mAh 3.7V protected Rechargeable Li-ion Battery, View Panasonic 3100mah, Panasonic Product Details from Shenzhen Listman Technology Co., Limited on Alibaba.com
panasonic cgr 18650 3.7V 3100mAh
Panasonic NCR18650A 3100mAh 3.7V
18650 panasonic high power battery 3400mah
3.7V NCR18650b lithium ion 3400mAh rechargeable battery
Li-ion 18650 battery panasonic NCR18650A 3100mAh
NCR18650A 3100mAh 18650 Lithium-ion battery cell, View 18650 protected li-ion battery, other Product Details from Shenzhen Julibao Battery Co., Ltd. on Alibaba.com
In general, the most credible target price I could find (without someone willing to just tell me) showed most 3100mAh pricing right around $2/cell, and a number of those distributors were offering 3400mAh cells for $2.50/cell. There were more distributors offering 3400 cells for $3-$4/cell though, mainly I think because it is a newer cell and is selling for a premium in many markets.
When we translate those numbers into per/kWh, we see prices of around $179/kWh for the 3100mAh cells, and $204/kWh for the 3400mAh cells (only a couple of entries showed differentiated prices at this level. Most were more like $250-$300/kWh). $179/kWh is around where we need to be, considering that Tesla has all but promised to deliver a finished pack for under $200kWh.
Also, 7968 cells will cost ~$15,936 @ $2/cell, so that leaves us with nearly $2,000 to build the pack and come in under the $200/kWh budget.
However...
The $2/cell cost is what any schmoe off of the street who is willing to buy a few thousand cells gets to pay. Tesla is purchasing ~4% of all commodity C3 li-ion cells this year.
When you cut away non 18650 cells, generic (non-Panasonic) 18650 production, and look only at Panasonic production of NCR18650A 3100mAh cells (there are many other brands who manufacture in the 18650 format), you see that Tesla is one of the largest purchasers of these cells in the world, and will likely purchase ~147 million of them in 2013.
So Tesla is not just some schmoe. In fact, in 2010 Jeffery Straubel was invited to the new Fab in Japan, where he was personally handed the very first NCR18650A 3100mAh cell to be produced. By the President of Panasonic Energy Company. Panasonic has a significant investment in Tesla, and has a seat on their board.
Panasonic's 3.1Ah batteries to be used in the Tesla Model S, have highest energy density yet
Not even close to being some schmoe.
But That's Not All!
In addition, as I detailed in this thread -
Amazing Core Tesla Battery IP - 18650 Cell
Tesla has designed a radically simplified cell. In comparison to the stock cells, these should be absurdly easy to manufacture. There should be really significant savings from such a radically simpler design.
So if you look back at our "scientific" method, we came up with a projection that was consistent with a basic manufacturing cost in 2012/2013 of around $0.85/cell, and we doubled that to $1.70 to find a proposed wholesale cost. The actual wholesale cost appears to be in the neighborhood of $2/cell, which is broadly consistent with this model. So Tesla's cost should lie somewhere between the basic manufacturing cost, and the wholesale cost, and they get bonus points for having a lower start value on the manufacturing cost.
For my part, the more I massage the moving parts, the more I like a smart version of the initial media calculation I presented at the beginning of this post. Tesla is charging $8,000 for 25kWh, or $320/kWh. A normal retail markup for the pack would imply a finished pack costing ~$160/kWh, or $14,240 for the 89kWh pack that I am simulating.
So My Assessment Is -
Tesla is using an ~89kWh pack that requires 7968 cells.
If you assume Tesla is paying ~$1.50/cell, that gives you a total cost for the cells of $11,952.
If you spend ~$2,000 to integrate it into a finished pack, it gets you right in range of the $14,000 pack cost of the smarter media calculation.
BTW, if you think $2,000 is too low to mass manufacture a few stamped and milled pieces of aluminum, wiring, cooling and electronics for the pack enclosure, then I think you are high
Feel free to make your own assessment.
Cost of a New Replacement Motor + Transmission For A BMW: ~$35,000-$40,000
Short Engine. S85B50A. Short Engine. BMW # 11000443598
Rmfd-6 speed gearbox. GS6-53BZ - THA9. Manual Transmission GS6-53BZ. 2012 BMW M5 # 23007606343
Sorry, couldn't help myself after watching the Bloomberg video that I linked with the Goldman Sachs reference. I just couldn't believe how shocked they were at the costs they were imagining for Tesla's battery, which is by far the most expensive part of the Tesla drivetrain
Big Edit 12/15/2013: Appending this post from down-thread to give the current best estimates for what we now believe to be a 7,000 cell pack using the modified 3.4aH Panasonic cells (instead of the 3.1aH cells in the OP). This change results in a higher pack cost (because the cost/kWh of the cells is higher) but a lower weight. These estimates are based on more information (not to mention debate) than available in the OP, but they are still just a rough estimate.
I am just putting this here for quick reference, but I recommend users with the time read the thread because there is a ton of good information and discussion that goes beyond just the price of the pack (especially in terms of the requirement and costs costs to Tesla to build a giga-factory). For that matter, this is still a working thread with new information coming in that we are actively debating.
Hey Kids, Don't Try This At Home!
I've been working lately on modeling costs for the battery pack for the Model S, both to get a sense of how it impacts Tesla's margins, as well as to understand how it's going to affect the CES concept that they appear set to roll out.
It turned out to be quite a bit more difficult than I expected, and the information I am going to present is far from definitive. So feel free to poke holes in this as you see fit. This is a working thread, and this post is just a starting point.
First, lets just get one point out of the way. There is no catalog that lists the cost of the Tesla battery. For that matter, there isn't a handy guide for the individual cell costs either (or maybe I just missed it). To generate a credible estimate you have to gather data, perform some analysis, and spice it up with speculation.
I will present some data in a moment, but I'll ruin the surprise a bit by stating that the available data (to me anyways) is highly variable and contradictory to the eyes of a layman like myself. So it's helpful to try and establish boundaries, and see if we can narrow the range of where we are likely to find our answer. As a layman, the first way I try and do that is to see if folks who may be smarter than me have studied the issue and said or written something about it.
Setting Boundaries - Media Style
For instance, its often been reported in the media that the cost per kWh for the Model S is $400/kWh. By itself, that is an impressive number, because it is much cheaper than any other automotive battery. At the time the Model S came out, the battery for the Leaf was estimated to cost $18,000, or ~$750/kWh. So obviously the Model S is amazing with it's $400/kWh cost. But how did the media get that cost?
I wont cite it, because its not important, but they got the cost by looking at the cost difference between the 60kWh and the 85kWh Model S. The cost difference is $10,000, which works out to $400/kWh for the 25kWh of storage.
Even as recently as last week media reports were citing this number. The problem is that the cost difference between the two cars is a retail price, and tells us nothing about the cost to Tesla. Nevertheless, the cost difference is a very useful metric to bound the data that we will come back to.
Goldman Sachs
Beyond the media, we've had highly paid financial analysts estimate the pack cost. In 2012 Goldman Sachs estimated that the pack cost Tesla $22,500 -
Is Trouble on the Way for Elon Musk and Tesla?: Video - Bloomberg
With an 85kWh battery that works out to ~$265/kWh. Goldman Sachs is a reasonably credible source, but we don't know how they arrived at the number, and its from 2012, so its slightly dated.
Elon and Company
In 2010-2012 various Tesla related sources (including Elon) have said they expect the pack to be delivered for under $200/kWh -
Cheap batteries will make the Tesla Model S profitable at 20,000 annual deliveries - egmCarTech
Endless-sphere.com View topic - Tesla pack costs $200/kwh? Tipping point price?
An 85kWh pack would cost around $17,000 at this price. Note also, they are saying pack cost, not cell cost.
International Trade Groups
The IEK did a report at the end of 2012 stating that 18650 cells (not limited to Panasonic) were running between $120-$200/kWh during Q3 when Tesla was first beginning production -
Global Li-ion Cell Shipments to Increase 9% in 2013: IEK-ITIS | CENS.com - Taiwan Industry Updates | Industry In-Focus | HTML |Ta1-CaE-Dy2012/12/03-Id42230
Smarter Media Method
We can do some speculative bounding in a few different ways.
First, going back to the media estimate of $400/kWh, we can refine that methodology in a couple of ways. For instance, we now know that $2,000 of the cost difference is the SuperCharger fee. Doing media math we get $320/kWh, based on the actual $8,000 price difference.
Again though that is a retail cost. The retail rule of thumb is that the retail cost is twice wholesale. That gets us to $160/kWh.
Science!!
Or, turning away from speculation we can go all scientific on the problem. Here is a definitive paper from 2000 -
http://www.transportation.anl.gov/pdfs/TA/149.pdf
And a must read writeup for the layperson that references the above paper -
http://www.tms.org/pubs/journals/jom/0809/daniel-0809.html
Bottom line is that in 2,000, the reference design was a 18650 1350mAh cell, with a materials cost of $1.28, and a production cost of $0.42, with total cost being $1.70/cell. The big problem that the authors identified was the need for the material and production costs to come down, which they expected to happen as mass production got rolling.
If you take that estimate, and compare it to information contained in this terrible (but must read) story by a clueless fool, you can maybe use it to (very) roughly project the costs from the 2,000 study into 2013 -
Lithium-ion batteries for autos | Future growth same as historical
This is the most important graph -
Looking at the fall in costs between 2,000 and 2012, we can estimate that costs are approximately 1/6th of what they were. If you figured that power density has (very) roughly tripled from 1350mAh to 3100mAh, then you can (very) roughly guess that costs must have fallen by half, to around $0.85/cell. Doing that gets you battery costs that are (very) roughly 1/6th of the conditions studied in 2,000. A bastardized methodology to be sure.
Moving on...
Because the initial estimate was just the basic costs to assemble a battery (as opposed to run a business) we will apply a simple rule of thumb markup and double the cost to find a (very) rough wholesale estimate of $1.70/cell for our current generation 3100mAh cell.
FYI, a thread somewhere says that Elon has estimated the lowest possible cost of the li-ion cells to be around $0.80 right now. I don't feel like running down the citation, but it's out there. If some bastardized method of analysis gets me within shouting distance of something Elon said, I'll take it.
Time To Build A Pack
Ok, we've set some boundaries, whether per/kWh or at the cell level. The next step is to simulate a pack so we can go shopping for cells and see how much it costs. The first problem with this is that there is an active controversy about many aspects of pack design. Some folks think that Tesla is using ~7,000 NCR18500B 3400mAh cells. Other's think Tesla is using some other number of cells.
For purposes of my simulation I will be using 7,968 modified NCR18650A 3100mAh cells for the 85kWh car. That works out to ~88.9kWh of storage, with the normal operating SOC being between 5%-95% (which on your indicators correspond to 0-265 rated miles).
This simulated pack is consistent with how the Roadster pack operated, characterizes the amount of emergency range you have below zero (thank you Broder), and leaves the upper 5% forever untouched. From this simulation, and knowing that power is reduced by half when the Model S enters emergency mode, we can guess that the Model S might be capable of traveling 20-30 miles after being computer limited to ~45mph in an emergency. Again, this is consistent with what Broder experienced. You are welcome to simulate your own pack, but I have a reasonable level of confidence in my pack design, so I'm going to go with it.
So knowing how many cells I am looking for, now I need to find a per cell cost. And with an ~89kWh car we need to modify the "promised" price of the Model S pack to ~$17,800 ($200/kWh).
Data, Data, Data!!
Knowing I want 7,968 NCR18650A 3100mAh cells, I did like Denzel in American Gangster and hopped onto a metaphorical Google flight to China to buy the product at the source, and eliminate as many middle men as possible. In various back alleys of Shenzen, where Panasonic appears to do most of its assembly, I found a massive variation in pricing. Many vendors are quoting prices for these batteries at $1-$10 depending on how many millions you are prepared to buy.
Theoretically, these prices are dependent on the quantity you want. You get the best price by purchasing large numbers on a regular basis. So a contract to purchase 1 million per month for the next year might let you purchase the item for $1/cell, while if you buy just the minimum purchase it's more like $10/cell.
However, as Stopcrazyapp pointed out when I linked him an example, many of these price ranges were probably placeholders for related models. When I attempted to contact the sellers, nobody was willing to give me pricing information. Maybe some of you can try and succeed where I failed.
So I dug deeper by clicking on every vendor I could find, and then going through their whole product line to see if they had other prices advertised with better differentiation. I also ignored the possibility that these batteries cost $1, because when I sifted enough prices, I found that batteries costing that much were probably older cells (like 2400mAh), while I started to find credible hints as to pricing for our target cells.
Here are numerous examples of links that I think are good bets to represent actual wholesale pricing for the 3100mAh cells, because in most cases, these prices were differentiated from other prices in that distributors product line. I also included a few 3400 cells for flavor (and because they had a different price from the same distributors 3100 cells) -
Original Japan li-ion rechargeable 18650 battery with 3100mah
18650 cells 3100mAh Li-ion Pana sonic 18650 battery rechargeable battery
NCR18650A 3100mah 3.7v lithium battery with cheap price, View NCR18650A 3100mah 3.7v lithium battery, panasonic Product Details from Shenzhen Victpower Technology Co., Ltd. on Alibaba.com
Panasonic 3100mah NCR18650 3100mAh 3.7V protected Rechargeable Li-ion Battery, View Panasonic 3100mah, Panasonic Product Details from Shenzhen Listman Technology Co., Limited on Alibaba.com
panasonic cgr 18650 3.7V 3100mAh
Panasonic NCR18650A 3100mAh 3.7V
18650 panasonic high power battery 3400mah
3.7V NCR18650b lithium ion 3400mAh rechargeable battery
Li-ion 18650 battery panasonic NCR18650A 3100mAh
NCR18650A 3100mAh 18650 Lithium-ion battery cell, View 18650 protected li-ion battery, other Product Details from Shenzhen Julibao Battery Co., Ltd. on Alibaba.com
In general, the most credible target price I could find (without someone willing to just tell me) showed most 3100mAh pricing right around $2/cell, and a number of those distributors were offering 3400mAh cells for $2.50/cell. There were more distributors offering 3400 cells for $3-$4/cell though, mainly I think because it is a newer cell and is selling for a premium in many markets.
When we translate those numbers into per/kWh, we see prices of around $179/kWh for the 3100mAh cells, and $204/kWh for the 3400mAh cells (only a couple of entries showed differentiated prices at this level. Most were more like $250-$300/kWh). $179/kWh is around where we need to be, considering that Tesla has all but promised to deliver a finished pack for under $200kWh.
Also, 7968 cells will cost ~$15,936 @ $2/cell, so that leaves us with nearly $2,000 to build the pack and come in under the $200/kWh budget.
However...
The $2/cell cost is what any schmoe off of the street who is willing to buy a few thousand cells gets to pay. Tesla is purchasing ~4% of all commodity C3 li-ion cells this year.
When you cut away non 18650 cells, generic (non-Panasonic) 18650 production, and look only at Panasonic production of NCR18650A 3100mAh cells (there are many other brands who manufacture in the 18650 format), you see that Tesla is one of the largest purchasers of these cells in the world, and will likely purchase ~147 million of them in 2013.
So Tesla is not just some schmoe. In fact, in 2010 Jeffery Straubel was invited to the new Fab in Japan, where he was personally handed the very first NCR18650A 3100mAh cell to be produced. By the President of Panasonic Energy Company. Panasonic has a significant investment in Tesla, and has a seat on their board.
Panasonic's 3.1Ah batteries to be used in the Tesla Model S, have highest energy density yet
Not even close to being some schmoe.
But That's Not All!
In addition, as I detailed in this thread -
Amazing Core Tesla Battery IP - 18650 Cell
Tesla has designed a radically simplified cell. In comparison to the stock cells, these should be absurdly easy to manufacture. There should be really significant savings from such a radically simpler design.
So if you look back at our "scientific" method, we came up with a projection that was consistent with a basic manufacturing cost in 2012/2013 of around $0.85/cell, and we doubled that to $1.70 to find a proposed wholesale cost. The actual wholesale cost appears to be in the neighborhood of $2/cell, which is broadly consistent with this model. So Tesla's cost should lie somewhere between the basic manufacturing cost, and the wholesale cost, and they get bonus points for having a lower start value on the manufacturing cost.
For my part, the more I massage the moving parts, the more I like a smart version of the initial media calculation I presented at the beginning of this post. Tesla is charging $8,000 for 25kWh, or $320/kWh. A normal retail markup for the pack would imply a finished pack costing ~$160/kWh, or $14,240 for the 89kWh pack that I am simulating.
So My Assessment Is -
Tesla is using an ~89kWh pack that requires 7968 cells.
If you assume Tesla is paying ~$1.50/cell, that gives you a total cost for the cells of $11,952.
If you spend ~$2,000 to integrate it into a finished pack, it gets you right in range of the $14,000 pack cost of the smarter media calculation.
BTW, if you think $2,000 is too low to mass manufacture a few stamped and milled pieces of aluminum, wiring, cooling and electronics for the pack enclosure, then I think you are high
Feel free to make your own assessment.
Cost of a New Replacement Motor + Transmission For A BMW: ~$35,000-$40,000
Short Engine. S85B50A. Short Engine. BMW # 11000443598
Rmfd-6 speed gearbox. GS6-53BZ - THA9. Manual Transmission GS6-53BZ. 2012 BMW M5 # 23007606343
Sorry, couldn't help myself after watching the Bloomberg video that I linked with the Goldman Sachs reference. I just couldn't believe how shocked they were at the costs they were imagining for Tesla's battery, which is by far the most expensive part of the Tesla drivetrain
Big Edit 12/15/2013: Appending this post from down-thread to give the current best estimates for what we now believe to be a 7,000 cell pack using the modified 3.4aH Panasonic cells (instead of the 3.1aH cells in the OP). This change results in a higher pack cost (because the cost/kWh of the cells is higher) but a lower weight. These estimates are based on more information (not to mention debate) than available in the OP, but they are still just a rough estimate.
I am just putting this here for quick reference, but I recommend users with the time read the thread because there is a ton of good information and discussion that goes beyond just the price of the pack (especially in terms of the requirement and costs costs to Tesla to build a giga-factory). For that matter, this is still a working thread with new information coming in that we are actively debating.
Back to cost and other issues, I've been pecking away at costs using a 7,000 cell model.
My quickie maximum cost simulated pack using our known constraint (less than a quarter of the cost of most models) is this -
60kWh Pack
- 4992 cells 3.4aH cells
- 61.1kWh
- $3 Cell Cost
- Pack Cost - $2,500
- Total Cost - $17,476
85kWh Pack
- 7,104 3.4aH cells
- 86.95kWh
- $3 Cell Cost
- Pack Cost - $2,500
- Total Cost - $23,812
This is a maximum price simulation that is consistent with our constraint. Current Trendforce wholesale price data points to a per cell cost of ~$3 for 3.4aH cells ($0.90/aH) but we've established plenty of reasons to suppose that Tesla is paying less than wholesale price.
If you assume a pack cost of $2,500 (which I am quite comfortable with) the lowest cell cost consistent with our constraints is $2.50, which happens to have been our best guess for 3.4aH cells based on Alibaba prices discussed up thread (edit: in the OP actually). That is a per cell cost of ~$205/kWh and a total cost in the ~$233/kWh range under this scenario.
Under those assumptions, every model except the base 85kWh (with no options) has a pack cost under 25% of the cost of the vehicle, which satisfies the constraints discussed by Straubel in the MIT piece. In addition, the "best case" is broadly consistent with the original IEK data discussed in the OP as well as a profuse amount of speculation and research.
Therefor, I think a better simulation is this -
60kWh Pack
- 4992 cells 3.4aH cells
- 61.1kWh
- $2.50 Cell Cost
- Pack Cost - $2,500
- Total Cost - $14,980
85kWh Pack
- 7,104 cells 3.4aH cells
- 86.95kWh
- $2.50 Cell Cost
- Pack Cost - $2,500
- Total Cost - $20,260
Edit: marked it up to add links
Last edited: