Julian Cox now on TMC

[Julian Cox]

Kevin, BPC Equity replied with this:

"So why would the battery manufacturers continue to make "cheap" battery cells and divert production and resources away from higher margin products? The production lines for 18650 will be utilized for higher margin prismatic and other advanced cells decreasing supply of 18650s. Julian is grasping at straws with his assumption."

The answer is obvious. Any factory that can continue to produce for a new and growing market using plant and machinery that is long ago bought and paid for in the service of a historical market requirement is on to a very good thing.

- - - Updated - - -

But Julian, I don't see clearly what the solution is? Will the existing 18650 plant sufficient for Tesla' use? for how long? Will Tesla continue to us 18650 forever, or for what period of time?

Just to use Petersen's numbers for sake of ease. He is claiming that there is current and existing global 18650 cell supply capacity to produce > 2 Million Model S vehicles. He is also indicating, correctly in my opinion, that the electronics industry is migrating away from the need to use this capacity.

There are many benefits and nothing wrong with using 18650s in EV battery pack design. What JP is unwittingly helping to identify is a super-trend of price decrease and supply availability imposed on top of the commercially motivated efforts of Panasonic to increase production.

Note that Panasonic has dismissed JP's argument that Panasonic is taking a loss in supplying Tesla. This piece of business has been highlighted as a profit center for Panasonic, turning around previous losses.

Regards to the future of 18650s in Tesla vehicles. Straubel has stated that they keep evaluating cost per kWh vs functionality of cells on the market and keep concluding that 18650s are the way to go. What JP has done is simply to help identify a reason why this analysis is likely to continue to produce the same result.

Note. In a recent JP article on the subject, some nonsense about being crushed by battery supply constraints, JP was effectively expressing paranoia that Tesla was in real and present danger of suffering a bottleneck at 100 ~ 200,000 vehicle sales. Nice problem to have. However with this latest piece it seems that JP is highlighting a route to 2 Million vehicles with existing supply on a trend to be demand-vacated by the consumer electronics industry. Very cool.

Note also for context that Axion's CEO has dismissed JP's contention (since 2010) that BMW is just about to place a large order with Axion - to paraphrase Axion's CEO on the Q2 Q&A BMW and the like have taken up enough of Axion's valuable time for maybe orders on an undefinable time frame. Beware of repeated claims the contrary.

 

[kevin99]

Kevin, BPC Equity replied with this:

"So why would the battery manufacturers continue to make "cheap" battery cells and divert production and resources away from higher margin products? The production lines for 18650 will be utilized for higher margin prismatic and other advanced cells decreasing supply of 18650s. Julian is grasping at straws with his assumption."

The answer is obvious. Any factory that can continue to produce for a new and growing market using plant and machinery that is long ago bought and paid for in the service of a historical market requirement is on to a very good thing.

...

Regards to the future of 18650s in Tesla vehicles. Straubel has stated that they keep evaluating cost per kWh vs functionality of cells on the market and keep concluding that 18650s are the way to go. What JP has done is simply to help identify a reason why this analysis is likely to continue to produce the same result.

Note. In a recent JP article on the subject, some nonsense about being crushed by battery supply constraints, JP was effectively expressing paranoia that Tesla was in real and present danger of suffering a bottleneck at 100 ~ 200,000 vehicle sales. Nice problem to have. However with this latest piece it seems that JP is highlighting a route to 2 Million vehicles with existing supply on a trend to be demand-vacated by the consumer electronics industry. Very cool.

Ok now I am your acting agent and voice too: :smile:

Tesla Motors Inc (TSLA): Putting Tesla Motor's Gargantuan Battery Supply Problem Into Perspective - Seeking Alpha



Man I don't want to be a messenger forever. But I have the obligation to close the loop:

Julian Cox

http://bit.ly/165jvea

The answer is obvious....

I hope they don't keep coming back at this.


So the answer is the 18650 cell will be good for a long long time? till Gen III? How does the 4V new cell come into picture? ie. what retrofitting on the plant is required to build higher capacity of 18650 cell? At which point you think the 4V would be used?

 

[Julian Cox]

Ok now I am your acting agent and voice too:


Man I don't want to be a messenger forever. But I have the obligation to close the loop:

Julian Cox

http://bit.ly/165jvea

The answer is obvious....

I hope they don't keep coming back at this.


So the answer is the 18650 cell will be good for a long long time? till Gen III? How does the 4V new cell come into picture? ie. what retrofitting on the plant is required to build higher capacity of 18650 cell? At which point you think the 4V would be used?

Kevin, I think that is a 4Ah cell (not a 4V cell).

The answer is not necessarily any change to plant and machinery at all. Improved cell properties come primarily from chemical and materials-science advances for the coatings on the cell windings, etching and welding of current collectors and reductions in bulk and increase in the ionic permeability of separator materials. at the end of the day the machines for slitting, materials transport, winding, welding and so on to automatically pack all this into a cell produce, by definition, a cylindrical object with the dimensions 18650.

 

[JRP3]

And as we've discussed in the battery costs thread some, maybe most, of those machines can easily do other sizes, and do prismatic cells as well. Since Tesla will be using the majority of cells produced they can still have economy of scale if they switch to a larger cylindrical cell or even a prismatic cell.

 

[Julian Cox]

And as we've discussed in the battery costs thread some, maybe most, of those machines can easily do other sizes, and do prismatic cells as well. Since Tesla will be using the majority of cells produced they can still have economy of scale if they switch to a larger cylindrical cell or even a prismatic cell.

Not really. Slitting, coating and calendering. That's about all the commonality with a prismatic line.

All of the stacking and winding equipment for prismatics, the tab welding, tab seal, pouch seal, dry room configuration etc etc etc makes for a completely different operation and process flow.

With regards to actually making a battery pack for a vehicle, the mechanical stability of an 18650 is really useful in handling and fixture of welds. Dealing with aluminum tabs is a nightmare and any tab, even a bi-metal and nickel tab combination has all kinds of mechanical considerations that are very much harder to deal with than the robotic pick and place of a discrete component. A prismatic cell needs to be fully supported on at least two faces, ideally with a calibrated and sprung surface pressure to accommodate expansion and contraction in different charge states while preventing delamination, whereas an 18650 given its external can structure can be supported on just two points which can double as the electrical join, and the internal winding can function as a watch-spring.

Traditionally only downside of using an 18650 in an automotive pack is the proportional weight of the can vs functional electrochemistry - where by the high proportion of can material to active materials by weight can negatively affect overall energy density. Theoretically a high capacity prismatic lends itself to improvements of the ratio of active to passive components by weight. However from what I have read the cells purchased by Tesla contain a design specification with respect to the wall thickness of the can to balance containment and structural integrity towards an optimization in in the interests energy density.

 

[JRP3]

Not really. Slitting, coating and calendering. That's about all the commonality with a prismatic line.

All of the stacking and winding equipment for prismatics, the tab welding, tab seal, pouch seal, dry room configuration etc etc etc makes for a completely different operation and process flow.

I've definitely seen winding machines that claim to do both cylindrical and prismatic winding.

A prismatic cell needs to be fully supported on at least two faces, ideally with a calibrated and sprung surface pressure to accommodate expansion and contraction in different charge states while preventing delamination, whereas an 18650 given its external can structure can be supported on just two points which can double as the electrical join, and the internal winding can function as a watch-spring.

I don't think the LEAF, Volt, Spark, Fisker, etc do anything more than strap them together with end plates. That's certainly all we do with our large format prismatics, no issues. Some don't even strap them.
Nissan LEAF 4 cell stack, no special strapping or pressure plates that I see:

A123 pack:

Traditionally only downside of using an 18650 in an automotive pack is the proportional weight of the can vs functional electrochemistry - where by the high proportion of can material to active materials by weight can negatively affect overall energy density. Theoretically a high capacity prismatic lends itself to improvements of the ratio of active to passive components by weight. However from what I have read the cells purchased by Tesla contain a design specification with respect to the wall thickness of the can to balance containment and structural integrity towards an optimization in in the interests energy density.

I'd like to see Tesla at least go to a cylindrical cell that's twice the diameter of the 18650, assuming cooling needs can be met. A "D" cell can holds about 2.5 the volume of an 18650 can, and it's shorter, so a potentially thinner pack with less than half as many cells, which is something Elon has mentioned he'd like.

 

[Julian Cox]

I've definitely seen winding machines that claim to do both cylindrical and prismatic winding.
I don't think the LEAF, Volt, Spark, Fisker, etc do anything more than strap them together with end plates. That's certainly all we do with our large format prismatics, no issues. Some don't even strap them.
Nissan LEAF 4 cell stack, no special strapping or pressure plates that I see:

View attachment 30010

A123 pack:

I'd like to see Tesla at least go to a cylindrical cell that's twice the diameter of the 18650, assuming cooling needs can be met. A "D" cell can holds about 2.5 the volume of an 18650 can, and it's shorter, so a potentially thinner pack with less than half as many cells, which is something Elon has mentioned he'd like.

JRP - I was the commercial lead, applications technology developer and the ultimate gate-keeper for selling A123 the prismatic cell technology you are quoting.
They are not the authority on the subject. Note, as a matter of fact that Nissan Leaf sardine can looks as though it has a sprung pressure plate stamped into the surface.

Here is a more obvious illustration of the principle:

If I could fault what Tesla has done, I would. The 18650 has a lot of commercial (non-technical) advantages as a common standard. In my opinion a D-Cell is too large to be optimum when seeking to isolate a point source of failure with an energetic material like LiNiCoAlO2 (Model S NCA) and too great a mass to be retained by a sensible sized electromechanical connection verses vibration. LiFePO4 and LiMn2O4 are a heck of a downgrade in terms of cycle life and energy density to trade-off in return for passive safety in large format prismatic use or in a D-Cell.

Julian.

 

[JRP3]

Looks to me like the Nissan can is rather loose, and, the stamping would not provide any pressure where the raised sections are, so I'm not sure. I'd guess the stamping is just to add stiffness to the thin metal of the can.
The Rimac setup is the only one I've seen using springs like that.
The Volt pack does not appear to have any expansion built into it.

As for a larger can such as the D cell or prismatic, I was thinking more of future chemistry, LiS/LiSi, presumably more stable, less need of cooling.

 

[Julian Cox]

Well I did say ideal rather than impossible to do without.

There are a number of non-terminal conditions that can happen to a prismatic cell that can expose plate surfaces to delamination.

In a technical sense, a large high-discharge prismatic is almost always (or should be) a large stack of parallel cells within a pouch rather than a single continuous winding.

The separator should be applied in a Z pattern or in a stack and alternate wind pattern. A bubble from evolved gases or a non-perforating dendrite effectively reduces the capacity of one or more of the cells within the cell and can denature the electrochemistry one affected plate at a time towards the extremes of charge or discharge leading to capacity loss and potentially safety issues. We found that on balance there is a large reliability advantage in maintaining a sprung pressure on the cell surface especially in high performance applications.

It stands to reason considering that there is nothing holding a prismatic cell together besides air pressure from the outside and a partial loss of vacuum is a calendar and cycle life limiting event. Under decent management this is somewhat reversible, but not if the effect cascades due to delamination in the absence of mechanical pressure that is spring to accommodate the expansion and contraction of the cell in differing charge states.


It has to be said that the LiSi nanotube concept looks really interesting. The idea with that is to stop the internal structure falling to bits owing to extreme changes in volume between charged and discharged states. I assume there is a price barrier to bring the tech no market, I am due for another look at it to find out what the matter is. That said there is nothing at all to stop such a cell material being applied to a standard 18650 line and therefore direct incorporation into Tesla's processes and designs without changing anything except to recalibrate the BMS.

 

[JRP3]

I do agree that keeping some pressure on the cells is important, and I do have mine strapped, I'm just not sure that pressure needs to be variable beyond the natural internal flexibility of cell components.
You are correct that the CALB, Thundersky, etc, large format prismatics are stacked or Z folded layers.
Professor Cui at Standford has done some interesting things encasing S and Si in an "egg shell" that allows them to expand inside without fracturing. If these can be made in a cost effective manner, using a non flammable electrolyte, and not containing an oxide, then potentially they could be safely used in larger format cells. If they have 4 times the energy density of current Tesla cells, and if you put them in a can twice as large, your pack can now be made with 1/8 as many cells.

 

[Norbert]

It seems some of you might be able to answer this question:

What is the cost of building a battery factory of a certain production capacity in kWh per year?

 

[JRP3]

This Panasonic plant in China cost about $100 million: http://www.zemotoring.com/news/2011/04/panasonic-to-build-new-366-million-lithium-ion-plant-in-china

It will spend at least 10 billion yen to build a new factory around 2012 next to existing Sanyo facilities in Suzhou, Jiangsu Province. Panasonic will invest another 5 billion yen or so to upgrade a Sanyo plant in Beijing.

Don't know the capacity but it's 49,000 m[SUP]2 http://panasonic.co.jp/corp/news/official.data/data.dir/2012/07/en120717-2/en120717-2.html
[/SUP]

 

[Julian Cox]

OK


This

Tesla Motors Inc (TSLA): Putting Tesla Motor's Gargantuan Battery Supply Problem Into Perspective - Seeking Alpha

and This

Tesla Motors Inc (TSLA): Putting Tesla Motor's Gargantuan Battery Supply Problem Into Perspective - Seeking Alpha

Has to stop.

The Author's primary source

This: http://rsc.li/19uT957


Immediately states that the environmental return on investment in embedded energy of Lithium Ion is 10:1

In road use, replacing and ICE with a maximum energy to kinetic efficiency of 25% turns this directly into 40:1

The remaining cycle life is entirely usable in distributed grid storage enabling access to the full >10:1 benefit even if the embedded energy was all Coal. It is not all coal and it is entirely feasible in real life that the embedded energy of battery production is > 50% if not greater than 90% renewable. In the case of Tesla and Solar City combined, and also in light of the trend of PV ownership or leasing accompanying Model S ownership, the carbon offset of the entire Tesla operation and the entire energy of the fleet is better than carbon neutral.

Petersen's arguments are dishonest on every level, they have been debunked repeatedly, and their continuation is offensive if not in fact an offense.

 

[Norbert]

This Panasonic plant in China cost about $100 million: http://www.zemotoring.com/news/2011/04/panasonic-to-build-new-366-million-lithium-ion-plant-in-china

Don't know the capacity but it's 49,000 m[SUP]2 http://panasonic.co.jp/corp/news/official.data/data.dir/2012/07/en120717-2/en120717-2.html
[/SUP]

Since it made the news, it's probably not the smallest of factories. So perhaps Tesla could start with $50 million "entry-level" factory.

There should be enough space in the Fremont buildings... (That's just a guess, though.)

 

[Julian Cox]

Since it made the news, it's probably not the smallest of factories. So perhaps Tesla could start with $50 million "entry-level" factory.

There should be enough space in the Fremont buildings... (That's just a guess, though.)

Norbert, it is not really necessary for Tesla to do any of this. It is enough just to threaten to do so. In other words, dear battery industry, get ready to take our orders or get ready to be cut out of the loop.


With regards to the embedded energy question.


There is a central fallacy in all embedded energy arguments leveled against EVs. The unscientific assumption that the source of energy to produce batteries and cars must come from conventional sources in conventional proportions (or worse, actually favor coal). This is abundantly untrue.

Batteries for transport are presently a relatively small percentage of the total transport system. The trend towards mass adoption of EVs mirrors the trend in mass adoption of Solar to power manufacturing, operations and energy for transport. In fact even the major players in EV batteries and solar photovoltaics are one and the same.

BYD Solar: PV | BYD
Samsung Solar: SolarModules - Products - Samsung
LG Solar: http://www.lg.com/us/commercial/solar
Panasonic Solar: Solar | Panasonic
'Tesla Solar': Solar Panels, Solar Power Energy Systems & Energy Efficiency - SolarCity


Here is the world's largest lithium ion battery producer's concept of embedded energy:

Here is an article stating Panasonic's desire to lead the Japanese solar energy industry.

Panasonic aims to be Japan No. 1 in solar energy | REVE

 

[JRP3]

I see Petersen is still pushing his false "400+ wh /wh of battery" which I previously deconstructed here: Garbage In, Garbage Out. Bad Data Leads To Wrong Conclusion About Tesla Battery Pack - JRP3 - Seeking Alpha
Anyone who still has comment privileges on his articles feel free to post the link.

 

[Norbert]

Norbert, it is not really necessary for Tesla to do any of this. It is enough just to threaten to do so. In other words, dear battery industry, get ready to take our orders or get ready to be cut out of the loop.

If they need to be threatened, why not relieve them of that burden?

 

[Julian Cox]

One of the best things about the Tesla business is to outsource the risks of rapidly evolving battery technology and concentrate the value of making future-proof cars in TSLA. A123 for example had a horrible risk profile precisely because it was an effort to commercialize a particular kind of battery technology - subject to being completely wiped out by any number of inventions both technical and commercial. A 2013 Model S can take advantage of every step change in battery evolution no matter who the battery winner of the day happens to be, as can a 2014 Model S, 2016 Gen III and so on. With battery manufacturing the risk and reward profile is exactly reversed right up until the point where there comes a risk of battery companies making cars that can compete with Tesla. So long as Tesla remains the world's most sought after customer for EV battery cells, Tesla is on the right track.

 

[Julian Cox]

New SA Article: Tesla Motors Inc (TSLA): Is Tesla Motors About To Take The Elevator Down? - Seeking Alpha

This is absolutely the worst kind of self-contradictory straw-clutching that neatly summarizes the profound illogicality of the Tesla bear case.

Professor Aswath Damodaran of NYU clearly stated on Bloomberg that he built a valuation model based on metrics that apply to traditional automakers. A model that fails to recognize the CCC inversion in Tesla's direct sales model. Damodaran stated that this stock did not trade on valuation, it traded on momentum, further stating that as a Tech stock TSLA was a buy.

The usual suspects (Archinbald et all) are the ones that get beaten every earnings call. They too generally fall into the trap of analogizing Tesla with an industry operating by a set of rules that Tesla is designed to re-write.

What makes this article the very worst kind of complete rubbish is that it goes on to treat excessive sales of electric vehicles as a problem (exactly the opposite of the concerns of the analysts). Then the author goes on to clutch at another straw concerning what seems to be the remarkable innovation of a motor built without rare earth magnets - something in common with the far more elegant liquid-cooled induction motor designed into the Model S from day one.

Running out of cash? The author was clearly suffering from selective hearing with respect to the GAAP vs Non-GAAP debate. Q2 $750M cash at bank with fairytale GAAP losses on paper only affecting a maximum of 4% of 2016 sales volume by any sensible estimate - the first date at which residual guarantees have any relevance. The real cash money (Non-GAAP) is in the bank heading into guaranteed profitable Q3 earnings.

One simply cannot rationally call a stock price collapse in any name when the underlying business has a line of customers out of the door willing to pay cash up front for a product with no competition as fast as the company can physically produce that product, and simply no evidence for any trigger for such a collapse from here to the distant horizon.

Back in reality, this stock is worth precisely what it says it's worth on the Nasdaq (that being what investors are willing to pay) and more than likely it will get bid up to the $184 bond-conversion level in relatively rapid due course.

Watch out for a return of Musk on the offensive regards Tesla after the 14th September. By all appearances he's about to bring a first-stage of a Falcon 9 1.1 back from an orbital launch and hover the thing on re-entry. An extraordinary spectacle whether or not he pulls it off and incredible if he does, which would amply explain his relative quietness. Meanwhile Tesla has been working on new cell supply arrangements, accelerating the European roll-out and clearing the dealer association obstacle.

As Musk clearly and correctly stated in his often misquoted CNBC interview, the stock does not trade on current or past valuation metrics. It trades on the balance of confidence in Tesla's future prospects. While it is incumbent upon Tesla to "hit it out of the park" it is incumbent on investors to question if anything is in evidence that materially affects the base assumption of Musk driving the market cap to $43.2bn and deliver a cumulative 300,000 vehicles as he is contracted to do. So far nothing. In fact it looks like he is on track to achieve that by 2018 at the absolute latest putting him roughly 4 years ahead of schedule.

So there we have it, again, another glorious opportunity to short TSLA into another round of good news.

 

[Julian Cox]

Tesla Motors Inc (TSLA): Tesla's Current Market Price Is Very Much Exaggerated - Seeking Alpha

Here is a SA article that most eloquently and comprehensively summarizes missing the whole point about Tesla, neatly supported by a lot of the acknowledged background risks disclosed in Tesla's 10-Q filing.


Quote: " if we look at the long-term, 10-year period, Tesla needs to achieve 37%-39% annual growth. These are obviously high hurdles, considering that both high growth rates and high net margins are difficult to achieve at the same time. "


That's it. The solution to the eternal question of capitalism - how to grow and profit simultaneously is whole deal about cash up front reservations and direct to consumer sales. This definitely includes the benefit of the Tesla Finance program (no matter what, Non-GAAP profit is cash in the bank for growth in the case of Tesla).

Tesla is completely unique in being able to structure sales around a highly profitable, cash flow positive sales model that guarantees the ability to grow and to yield profits simultaneously in a way that defies analogy with the legacy auto industry. This is the primary reason why every model that relies on analogy with the metrics of established automakers is simply wrong.

Unlike an ordinary manufacturing business whose main outgoings are inescapable, the big numbers when considering the balance of growth and net profit for the Tesla business are variables. They are not a risk factor, instead they are a control lever: For example it is entirely at the discretion of management to decide how fast to roll out charging and service capacity vs how much net profit to expose. The same goes for R&D expenditure and the rate of accumulating equipment to expand production capacity. Unlike a Ford or a GM, BMW, Toyota, Audi, Lexus, Mercedes etc etc etc there is no giant negative number on the books for finished goods inventory. In the case of Tesla that number represents a few service loaners and store demonstrators. The difference is not a matter of degree, it is night and day.

We saw Tesla use this lever to engineer a non-gap profit in Q2 despite being unable to record the first batch of European production as sales in Q2 owing to transit delay (this batch or production of highly spec'd "Euro-Sigs" will be credited as sales in Q3 in addition to the kinds of figures we saw in Q2).

There is precisely zero chance of the underpinnings of future confidence expressed in the TSLA price suffering a reversal going into Q3. It is far more likely that the artificially imposed concern regarding cell-supply capacity will be resolved to great fanfare with the announcement of global capacity added from Samsung and possibly BYD too for the Chinese market where there are huge financial and market-access incentives for Tesla to supply Model S with Chinese cells to the Chinese market.