Model S Battery Pack - Cost Per kWh Estimate

[Zzzz...]

If Tesla is paying $1.70/cell, that is ~$152/kWh. The 85kWh pack would have a cell cost of ~$13,546 and probably a total cost of under $16,000.

In a very recent interview, like a year or year and a half ago, Elon Musk said that he think that prices below $200 per kWh could soon be archived. But reaching that price level would take some bettary advances...
That pretty much clears that Tesla not YET paying less then $200 per kWh. Part of problem - they need to certify process of manufacturer so cells are produced to exact specs and enforce strict quality control. Another part - warranty and liability clauses, they add to cost. And on top of that - Tesla needs lot of volume. So dealing with majority of small companies are out of the question.

Bottom line, it is very unlikely that TM paying $200 or less per kWh atm. But I'm confident that they are pressing cells suppliers for lower prices and looking at every opportunity that could lower cells cost - that price is crucial for their business and success.

PS. I also quite confident Tesla is paying less then $300 per kWh atm. But really hard to say how much less.

 

[stopcrazypp]

I can't directly dispute that because I haven't attended battery conferences.

I haven't attended them either but you can find some of the slides on the internet. The one I found was for the 29th International Battery Seminar and Exhibit (2012) and it had data of 4 Billion Li-ion cells, $9 Billion in sales for 2011 for an average cell price of about $2.25/cell. Average capacity works to about 7Wh per cell (~1900mAh). This does include cells for all applications (not just 18650), but they do break it down in graph format, the cells and sales for the laptop li-ion cells.

 

[CapitalistOppressor]

This is not "information" but it is flavor. An interview in 2011.

Toyota sees Tesla EV battery cost at 1/3

"If (Tesla's battery structure) works, we won't have to wait for a breakthrough in battery technology to develop a relatively cheap electric vehicle," Executive Vice President Takeshi Uchiyamada, who heads Toyota's research and development, told Reuters in an interview at the Detroit auto show on Tuesday.


"It could be as low as one-third of the cost of batteries being developed by car makers, because (laptop) batteries are produced in massive volumes."

 

[lolachampcar]

probably one of the most telling quotes yet.

 

[CapitalistOppressor]

probably one of the most telling quotes yet.

What's our estimate for the cost of a Rav4 battery? $11,000 if the cell cost is $200/kWh? I constantly hear estimates from Wall St types of $20k+, but they seem to just base that on the. $400/kWh meme.

Toyota has been drastically discounting the RAV4 lately, and the assumption has been that they are losing a bundle. But the RAV4 is a compliance car, and Toyotas goal was only to sell the minimum number needed for the ZEV quota, at the highest sale price possible.

It seems likely to me that the real reason the car is uneconomical is because they are spreading the development costs over so few cars.

 

[CapitalistOppressor]

Some new data, and a nice resource to track pricing trends. Not completely on point, but close enough for government work -

Reduced cell cost suggests the upcoming era of large capacity cells | EnergyTrend
Battery Price | EnergyTrend

If I read their chart correctly, they are pricing $/Ah and the Panasonic NCR18650A 3100mAh cells would fall into the <3.2aH category, with a Q3 estimate of $0.74/Ah or ~$2.29/cell. That works out to ~$205/kWh (for the cells) and a total pack cost of ~$21,000 for the 85kWh pack that I've simulated.

Unfortunately there are many problems with the market price listed. First, it's the average selling price for cylindrical cells in general.

The common 18650 cells are just a subset of the cylindrical cells, and based on my understanding they tend to cost less than the other formats because they are produced in much greater quantities. Thus I'd expect a lower average price for 18650's compared to what is quoted here. In addition, this does not account at all for the many cost saving features of the Tesla cells, or potential discounts because they are the largest purchaser in the world (which they almost certainly are now) or because of the business relationship they have with Panasonic (with Panasonic being a part owner of Tesla). On the downside, the Panasonic NCR18650A cells tend to sell for a higher price than competitors for a given cell size.

With all of the moving parts this doesn't get us anything like firm pricing, but I think it strongly supports the data we have pointing to a Tesla pack cost that is in the ~$200/kWh range. And again, we have many sources mentioned in this thread which have claimed that Tesla would deliver the Model S pack for under $200/kWh. We also have a wealth of other data pointing to that, including the I.E.K. report listing 18650 cell costs in the $120-$200/kWh range at the end of 2012 and the many advertisements on wholesaler websites also pointing to relatively low prices.

Other tidbits in these new links are confirmation of a general slowdown in 18650 production overall thanks to the movement towards ultra-thin devices, with manufacturers being forced to produce low end cells at or below cost. We have academic research pointing to per cell costs in the range of $1.50-$1.80 for the 18650 cells, without factoring in profits.

However, Tesla is now big enough that they appear to be reversing the trend in 18650 production -

Panasonic to increase lithium-ion battery production capacity - BUSINESS - Globaltimes.cn

And we also know that Tesla is also working with Samsung as an alternate supplier. With all of this activity, Tesla is certain to take over an increasingly large share of the 18650 market, and they will do so as other players are exiting the market, allowing Tesla to purchase batteries from existing plants whose costs have already been fully amortized. In the long term though, GENIII requirements will substantially outstrip existing production capacity (which I estimate is enough to supply ~120,000 Model S class batteries if you use a nextgen cell with 4.0mAh).

So we'll get a sense of Tesla GENIII sales expectations in a year or two when the major battery manufacturers start announcing additional production capacity to support it. A likely doubling or tripling of global cylindrical battery production is not the sort of thing that will fly under the radar.

 

[DonPedro]

Other tidbits in these new links are confirmation of a general slowdown in 18650 production overall thanks to the movement towards ultra-thin devices, with manufacturers being forced to produce low end cells at or below cost. We have academic research pointing to per cell costs in the range of $1.50-$1.80 for the 18650 cells, without factoring in profits.

However, Tesla is now big enough that they appear to be reversing the trend in 18650 production -

Panasonic to increase lithium-ion battery production capacity - BUSINESS - Globaltimes.cn

And we also know that Tesla is also working with Samsung as an alternate supplier. With all of this activity, Tesla is certain to take over an increasingly large share of the 18650 market, and they will do so as other players are exiting the market, allowing Tesla to purchase batteries from existing plants whose costs have already been fully amortized. In the long term though, GENIII requirements will substantially outstrip existing production capacity (which I estimate is enough to supply ~120,000 Model S class batteries if you use a nextgen cell with 4.0mAh).

So we'll get a sense of Tesla GENIII sales expectations in a year or two when the major battery manufacturers start announcing additional production capacity to support it. A likely doubling or tripling of global cylindrical battery production is not the sort of thing that will fly under the radar.

This is clearly a risk for GENIII production ramp up. I assume that GENIII will use 18650 cells and that new plants to manufacture such cells take some time to build (18 months from the demand is seen and until cells are delivered in the US?). In that case, if GENIII sales exceed the sales expectations of battery suppliers, the production will be constrained by the availability of 18650 cells for a period of at least 18 months. Tesla could hedge against this by providing some sort of incentives or option fees to the producers to build enough capacity but this could be disastrous for the company if GENIII isn't a success right from the start.

 

[lolachampcar]

I can understand sourcing cells for MS but I see a strong argument for doing it in house for G3. You're simply using too many cells. You go from the tail to the dog.

 

[JRP3]

You think Tesla is going to invest in cell production lines of their own? I doubt it.

 

[drinkerofkoolaid]

Although I don't know anything about battery production, I'd imagine since Tesla is the only company securing contracts for high volume production of the 18650 cells, it would be very cost prohibitive or impossible for other auto manufacturers to compete without working with Tesla. Tesla already has a 5 year head start with research and development. Even if other automobile manufacturers can produce an EV that can compete with Tesla, it would be an additional 3-4 years before they could achieve the kind of battery production they would require for mass production.

By the time the other automobile manufacturers are ready to produce a competitive EV, Tesla will already be selling 2-4 million Teslas annually, will have contracts that give Tesla a huge first mover advantage on cell prices.

 

[evme]

How about another way of looking at it? If Tesla becomes big enough, they can just have Panasonic build a factory locally.

 

[vfx]

You think Tesla is going to invest in cell production lines of their own? I doubt it.

Elon likes to do everything in house. After hearing him at TESLIVE worried about battery production constraints coming, it looked like he was thinking about it.

And why not do it in CA?

 

[Tim]

Doing some back of the napkin math I figure Tesla's demand for batteries right now is nearly the wh equivalent of 1/4 the worldwide supply used in cell phones(78KWh*25000/(5wh per phone)/1.7billion phones sold. I'm not surprised that Panasonic is running into supply constraints with Tesla probably asking them to double it.

 

[GSP]

Some new data, and a nice resource to track pricing trends. Not completely on point, but close enough for government work -

Reduced cell cost suggests the upcoming era of large capacity cells | EnergyTrend
Battery Price | EnergyTrend

If I read their chart correctly, they are pricing $/Ah and the Panasonic NCR18650A 3100mAh cells would fall into the <3.2aH category, with a Q3 estimate of $0.74/Ah or ~$2.29/cell. That works out to ~$205/kWh (for the cells) and a total pack cost of ~$21,000 for the 85kWh pack that I've simulated.

Unfortunately there are many problems with the market price listed. First, it's the average selling price for cylindrical cells in general.

The common 18650 cells are just a subset of the cylindrical cells, and based on my understanding they tend to cost less than the other formats because they are produced in much greater quantities. Thus I'd expect a lower average price for 18650's compared to what is quoted here. In addition, this does not account at all for the many cost saving features of the Tesla cells, or potential discounts because they are the largest purchaser in the world (which they almost certainly are now) or because of the business relationship they have with Panasonic (with Panasonic being a part owner of Tesla). On the downside, the Panasonic NCR18650A cells tend to sell for a higher price than competitors for a given cell size.

With all of the moving parts this doesn't get us anything like firm pricing, but I think it strongly supports the data we have pointing to a Tesla pack cost that is in the ~$200/kWh range. And again, we have many sources mentioned in this thread which have claimed that Tesla would deliver the Model S pack for under $200/kWh. We also have a wealth of other data pointing to that, including the I.E.K. report listing 18650 cell costs in the $120-$200/kWh range at the end of 2012 and the many advertisements on wholesaler websites also pointing to relatively low prices.

Other tidbits in these new links are confirmation of a general slowdown in 18650 production overall thanks to the movement towards ultra-thin devices, with manufacturers being forced to produce low end cells at or below cost. We have academic research pointing to per cell costs in the range of $1.50-$1.80 for the 18650 cells, without factoring in profits.

However, Tesla is now big enough that they appear to be reversing the trend in 18650 production -

Panasonic to increase lithium-ion battery production capacity - BUSINESS - Globaltimes.cn

And we also know that Tesla is also working with Samsung as an alternate supplier. With all of this activity, Tesla is certain to take over an increasingly large share of the 18650 market, and they will do so as other players are exiting the market, allowing Tesla to purchase batteries from existing plants whose costs have already been fully amortized. In the long term though, GENIII requirements will substantially outstrip existing production capacity (which I estimate is enough to supply ~120,000 Model S class batteries if you use a nextgen cell with 4.0mAh).

So we'll get a sense of Tesla GENIII sales expectations in a year or two when the major battery manufacturers start announcing additional production capacity to support it. A likely doubling or tripling of global cylindrical battery production is not the sort of thing that will fly under the radar.

CO,

Thanks for the great analysis and links. This is great work.

When you say "Tesla pack cost in the ~$200/kWh range," I think you mean Tesla's cost for the cells to make a complete pack, without the pack enclosure, assembly costs, BMS, cooling system, quick connect coolant and power connectors, contactors, pre-charge circuit, etc. Do I understand that correctly?

I would guess that the non-cell costs could be about $8500, which would be an additional $100/kWh for the 85 kWh pack.

GSP

 

[deonb]

CO,

Thanks for the great analysis and links. This is great work.

When you say "Tesla pack cost in the ~$200/kWh range," I think you mean Tesla's cost for the cells to make a complete pack, without the pack enclosure, assembly costs, BMS, cooling system, quick connect coolant and power connectors, contactors, pre-charge circuit, etc. Do I understand that correctly?

I would guess that the non-cell costs could be about $8500, which would be an additional $100/kWh for the 85 kWh pack.

GSP

No. Pack cost. Look at the very first post of this thread.

 

[JRP3]

How about another way of looking at it? If Tesla becomes big enough, they can just have Panasonic build a factory locally.

Elon likes to do everything in house. After hearing him at TESLIVE worried about battery production constraints coming, it looked like he was thinking about it.

And why not do it in CA?

Both good points, though I think that would lock them into the 18650 format for a longer period of time, which may at some point become a disadvantage. If Tesla builds a factory or has Panasonic build a factory they have to guarantee a certain number of cells to make it pay back.

 

[CapitalistOppressor]

CO,

Thanks for the great analysis and links. This is great work.

When you say "Tesla pack cost in the ~$200/kWh range," I think you mean Tesla's cost for the cells to make a complete pack, without the pack enclosure, assembly costs, BMS, cooling system, quick connect coolant and power connectors, contactors, pre-charge circuit, etc. Do I understand that correctly?

I would guess that the non-cell costs could be about $8500, which would be an additional $100/kWh for the 85 kWh pack.

GSP

I've occillated between pack integration costs in the range of $2,000-$4,000. The various Tesla patents on their pack just don't look very complicated, and are highly amenable to mass production. I'd be shocked to my very bones if someone produced internal Tesla documentation pointing to a pack cost of ~$8,500.

Plus, if pack costs are higher than $4,000 you really start to eat into any credible economic case for GENIII unless you assume absurdly low battery costs.

- - - Updated - - -

Elon likes to do everything in house. After hearing him at TESLIVE worried about battery production constraints coming, it looked like he was thinking about it.

And why not do it in CA?

I didn't realize he had even mentioned constraints. But when I do the math it's obvious that even the rapidly expanding sales potential of just the Model S will put a strain on battery production. The potential of GENIII is clearly far beyond what the existing battery industry can immediatly support.

 

[ggr]

Although I don't know anything about battery production, I'd imagine since Tesla is the only company securing contracts for high volume production of the 18650 cells, it would be very cost prohibitive or impossible for other auto manufacturers to compete without working with Tesla. Tesla already has a 5 year head start with research and development. Even if other automobile manufacturers can produce an EV that can compete with Tesla, it would be an additional 3-4 years before they could achieve the kind of battery production they would require for mass production.

By the time the other automobile manufacturers are ready to produce a competitive EV, Tesla will already be selling 2-4 million Teslas annually, will have contracts that give Tesla a huge first mover advantage on cell prices.

But they're not the only high-volume purchaser. There may only be ten of them in every Lenovo laptop, but there are a lot more laptops than Model Ss. Tesla is big (now) but probably still not even half of the production.

 

[CapitalistOppressor]

This is clearly a risk for GENIII production ramp up. I assume that GENIII will use 18650 cells and that new plants to manufacture such cells take some time to build (18 months from the demand is seen and until cells are delivered in the US?). In that case, if GENIII sales exceed the sales expectations of battery suppliers, the production will be constrained by the availability of 18650 cells for a period of at least 18 months. Tesla could hedge against this by providing some sort of incentives or option fees to the producers to build enough capacity but this could be disastrous for the company if GENIII isn't a success right from the start.

This is a definitely a quality, first world problem.

My feeling is that they are going to have to set a very conservative production goal, and focus on just the US market to begin with. Then scale up and move into new markets as additional production capacity comes online.

At the same time though, I am coming around to the idea that Tesla might just want to aquire Panasonic, or maybe just their battery business. It's one thing to rely on commercial battery providers to take the risks in this space when you are a small player. But Tesla is looking at the possibility that they could be purchasing 50%+ of global 18650 production by 2016 before they even start production on GENIII.

The fact is that once automotive production goes electric the consumer battery market will be a tiny fragment of total battery production. The big players are going to be the automakers.

GM and the rest have been making this argument for years to justify their expensive proprietary batteries, while Tesla has sprinted ahead with more advanced and less expensive commodity cells. But in the long run I think GM has the economic argument right. Battery design and production is going to need to be a core competency, just as engine production was in the past. The scales required and competitive pressures involved require it.

- - - Updated - - -

But they're not the only high-volume purchaser. There may only be ten of them in every Lenovo laptop, but there are a lot more laptops than Model Ss. Tesla is big (now) but probably still not even half of the production.

They are ok for now. At a 24,000 Model S/yr production rate they are something like 27% of existing 18650 production. But 18650 production has been going down for the last couple of years, so they are probably less than a quarter of existing capacity. Maybe as low as 20%.

 

[lolachampcar]

JR,
I think I would be concerned if they were not considering it. If building 21K MS' a year (152MM cells) is putting a strain on cell supply can you imagine what 500,000 G3s would do? If that were 4000 cells per car you are looking at 2B cells per year. I'm thinking I would need that line in house to keep it from being my weak link.