this is my first piece on (NASDAQ:TSLA), the company so justly famous for its Electric Vehicles. I've gradually pulled this article together over the last half year at a time when few articles on SA even mentioned the Gigafactory. After the Model X launch however, focus has finally shifted to Tesla's biggest endeavor.
The Gigafactory is a fascinating subject to write about. Mr. Musk told investors 2014 that he wants it to become the largest building in the world, even that it will revolutionize an entire industry. These fantastic claims seem to mesh well with the fact that it is only our underpowered, overpriced current battery technology that is holding back the Electric Vehicle Revolution. Radically cheaper batteries would also make home energy storage economically sensible, perhaps even personal electric flight. Like many Tesla bulls do, I also believe those aren't merely dreams; I'm sure we'll have those revolutions, and soon. Yet after thorough research, I've come to the conclusion that the project is at far greater risk of failure than Tesla's stock valuation allows for. These risks include the mundane, like construction mistakes and accidents, supply-side risks where Tesla might be forced to pay exorbitantly for inputs, to the catastrophic like total obsolescence of thin-film Lithium Batteries tech.
In researching for this article, I set out to truly understand the whole battery industry, including everything from the science of batteries to investor perceptions and industry trends. For the latter, I find it to be very helpful to view those through a historical lens - we have been experiencing industry upheavals, like the EV revolution promises to do, for over 200 years after all. On account of the many different subjects I'll cover, I've split the article into 3 sections for readability:
Section 1: All about Batteries. A condensed history of their science up to the modern day, followed by an in-depth examination of the changing scientific and economical landscape of this crucial industry.
Section 2: The Gigafactory. A brief summary of Gigafactory developments to date, followed by my own research into possible Upsides and Risk Factors.
Section 3: Conclusion. How the Gigafactory project ties together with TSLA's corporate character, and why the Gigafactory, out of all of TSLA's daring ventures, has the greatest potential to bankrupt the company.
Early History of Batteries:
***Note: If you are familiar with how current thin-film Lithium Batteries work, please skip to the paragraph titled: 'Public Perceptions of Current Technology'.***
The earliest experiments with electricity occurred well over 2000 years ago and involved rubbing a piece of amber over wool. In simple terms, the amber physically scratches electrons away from the wool. Since amber is a bad conductor of electricity, the extra electrons stay near the surface, giving the amber a local negative charge.
Nothing changed until 1800AD, when the renowned Italian scientist Volta stacked Zinc and Copper plates over each other, separated by some cloth and bathed in a salty solution. His Voltaic Piles were the first batteries, and worked similarly to the one shown below:
Source: Slide 4 of Presentation about novel Paper Batteries.
The basic principle of separating reactants, letting ions flow from one side to the other and routing the energetic electrons through wires powers batteries to this day.
Hundreds of different kinds of cells were developed in the 1800's, until in 1896 Karl Gassner of Mainz, Germany (which is also the town where printing was invented) invented dry cells. These were the first practical batteries and represent the beginning of our modern era of portable electronics. Mr. Gassner's dry cell principle powers alkaline cells to this day.
But there's a problem: Batteries are heavy. Tesla's cars lug around over 1000lbs of them. So what makes things heavy? In atoms, 99.9% of mass resides in the nucleus. However, nuclei are not involved in chemical reactions. For example, Lead has 82 protons and 120+ neutrons that give it mass, which are surrounded by a vast, mostly empty cloud of electrons drawn to the positive charge. But only the outer few electrons play any role in chemistry. So to get more power for the least weight, the answer is to use Lithium, the third-lightest element after Hydrogen and Helium. The first mention of Lithium batteries was in 1911.
How it Works: Current thin-film Lithium Ion Batteries.
Experimental Lithium Metal batteries had existed in the literature for many decades, but were considered too dangerous because of their unfortunate tendency to explode. Finally, in 1981, Dr. John Bannister Goodenough of Oxford University had a breakthrough and invented modern Lithium Ion batteries, specifically the Cathode material which makes them possible.
***Side Note: The inventor is 92 now and still working. Here's a fascinating story about him, which I highly recommend reading as it reviews in detail the whole development of Lithium batteries up to 2015. The article is a valuable resource with which to place the Gigafactory, which is also mentioned, into historical context. Interestingly, in the article Dr. Goodenough's research team claims to have found another breakthrough which will drop the cost of Lithium batteries by 70%.***
Back to modern Lithium batteries: Sony commercialized the technology in 1991 when it paired the new cathode material with a charcoal anode. The revolutionary batteries became an instant blockbuster, and helped Sony sell a tremendous amount of equipment like hand-held video recorders. All modern Lithium batteries, including Tesla's, derive directly from this technology. In simple terms, it's a two part system:
The Cathode is the high-energy side which draws in Li-ions and electrons during discharging. The cathode is a porous crystalline solid of Cobalt, Nickel and other elements. Li-ions travel into this porous structure and chemically bond to it, a process which requires electrons. This bonding is what releases energy and 'pulls' electrons through our motors and electronics, doing work.
The Anode is a sort of 'storage tank' for lithium made from graphite. During charging, the applied voltage rips electrons and lithium ions away from the Cathode, forcing them to travel and combine with the Graphite Anode. When the battery discharges again, the Anode releases Lithium as electrons and ions again to complete the cycle.
The whole setup is bathed in an electrolyte so that the Lithium ions can dissolve first through the anode, the separator and then into the cathode. At no point is lithium metal formed, which makes the technology much safer, and explains why it is called Lithium ION. Here's a picture showing a Lithium battery discharging, please ignore that it says Silicon for the Anode:
Source: Nexeon, which has developed better Silicon Anode chemistry. In regular batteries, Graphite takes the place of Silicone on the negative side.
Shapes and Varieties of Current Batteries:
Shapes: Since lithium ions travel extremely slowly through the solid cathode and anode materials, those need to be made very thin. Therefore, modern batteries are made up of many layers of thin films. Like paper (imagine a stack of $20 bills or a continous strip), the films can be either stacked or rolled up, which explains the two major battery types: Flat or cylindrical. Flat packs have several advantages, such as more even material stresses, higher surface area for cooling and denser packing possibilities, but also negatives, like less resistance to puncture and volume expansion of the battery pack during use. Tesla is using cylindrical batteries, which are regarded as safer but less powerful. Overall industry trends have greatly favored flat-packs in recent years. Some of Tesla's competitors like VW have recently announced they'll be pursuing flat-pack technology; others will stick to cylindrical cells, like Tesla.
Chemical Varieties: To anyone interested, here's a link that explains the five minor variations of Cathode chemistries. All are themes of the basic Cobalt Mineral Cathode / Graphite Anode thin-film method. Compared to each other, they each have 20 - 30% advantages in some areas but weaknesses in others.
Today's batteries are very expensive:
There are many design challenges which make the Li-Ion technology costly. Examples:
- The ions diffuse very slowly through the solid cathode and anode materials, so the electrodes need to be extremely thin for reasonable amounts of power to flow. Working with these hair-thin, brittle and fragile films requires a complicated manufacturing process which is expensive to set up and operate. It is mainly here where economy of scale might be most effective.
Source A basic 18650 battery as used in present Tesla packs showing its rolled-up structure.
Source: 24M Tech. Shows their improved design next to the laminar structure of a regular battery.
- Each battery needs a micro-chip to control battery state, as well as vents and other safety equipment. This adds to the technology's cost and complexity.
- Lithium is expensive to produce, rare, with the market controlled by a few miners. Here's an excellent SA article on the risks this poses to TSLA. In addition, the other rare metals making up the Cathode (like Cobalt and Nickel) actually outweigh Lithium. This complicates the supply chain and exposes battery makers to numerous price / supply risks. Click to enlarge
- Source of Lithium Use Graph: Wikipedia
- Battery factories are expensive: There are over 40 steps necessary to assemble batteries from raw inputs, most of which involve handling fragile, toxic or extremely moisture sensitive materials. The factories are costly to build and tricky to get running smoothly.
- Lastly, Lithium batteries never fully discharge. Internal electronics shut down discharging at 30% to protect the fragile electrode structures. This obviously makes the technology heavier and less powerful than it otherwise would be, further raising price.
Improving Today's Batteries is Hard:
The rate of improvement for Lithium batteries averages at most ~8% per year. This rate is much lower than improvements in computing processor power, data storage or transmission, for example. Why? Engineers are faced with huge lists of problems. Here's a video of the manufacturing process. It can't be stressed enough how complex and finicky thin-film Lithium cell construction actually is.
Some specific examples of why thin-film batteries are so hard to improve: Painting and bonding the brittle electrodes onto the hair-thin current collectors (sheets of hair-thin aluminum and copper foil), then assembling packs from the coated sheets is difficult to get right. The cathode/anode coatings are brittle, the hair-thin metal films prone to oxidation, sticking, tearing and a dozen other problems. There are more steps like calendaring, annealing, filling in electrolyte under controlled atmosphere conditions, attaching valves, and for cylindrical batteries especially, soldering the current collectors together which are each difficult to optimize. Once a production line is running smoothly, operators may hesitate to shut it down for upgrades. It's a tough choice between producing robust, less-powered cells and investing into the latest, but risky optimization scheme.
From a science perspective, the actual cell chemistries have historically been even harder to improve than production processes. That is because new chemistry needs to work for thousands of charging cycles but also stand the test of time, something that's not always adequately simulated by rapid charge/drain cycles. Late-stage problems could be dendrites (crystals) forming which puncture and short the delicate films, material stresses, flaking, irreversible side reactions, thermal behavior, potential for run-away reactions etc. The biggest problem of all is electrode disintegration: As Lithium comes and goes from the cathode, it swells in size and contracts. Dr. Goodenough's major breakthrough was that his material could withstand the repeated cycles of expansion (called intercalation in industry jargon). Even so, current electrode technology is limited to about 1000 cycles before it irreversibly pulverizes. Here's a link (pdf) about failure modes of current batteries for those who wish to know more.
I hope the point is made that the underlying chemistry and physical construction of current thin-film Lithium batteries are hard to work with and to improve.
Theoretical Limits to Battery Power:
Let's take a step back from the rat race that is improving Lithium batteries. The fact is, since 1991 industry focus has been on making evolutionary improvements, not on producing revolutionary jumps in technology. In this aspect, the battery industry is even more conservative than trends happening in computing soft / hardware.
But Silicon processor chips have hit a hard engineering ceiling: due to quantum tunneling effects, transistors just can't be made any smaller. The same isn't true for batteries. To understand why, it is instructive to know what the theoretical limits of battery power are:
- Future TSLA Gigafactory Li-Ion pack (optimistic): 300 Wh/kg
- Lithium metal with Oxygen: 3800 - 5200 Wh/kg
- Lithium metal not counting external Oxygen: 11100 Wh/kg
- Aluminum metal with Oxygen: 4800 Wh/kg
- Reversible H2O2 Fuel Cell: 2580 Wh/kg
What jumps out is that we are 10X to 30X below theoretical limits. That is because the physical structure of current batteries contains relatively little Lithium. How little? A typical 18650 battery spreads just .75g of Lithium over an entire 44g battery. Another example: my fairly large laptop is powered by about 3 M&M candy pieces in weight of Lithium. In the laptop, that tiny mass of metal is spread over films with the area of a pool table. Another fact: For a $20000+ Tesla car battery, the input cost of Lithium metal barely comes to $150. No wonder, as the 1000+lbs battery pack contains less than 50lbs of it. Read this paper from Stanford University for an introduction to the maths and construction behind batteries.
Public Perception of Current Batteries:
Before I move on, there is an important point to consider that's directly relevant to Tesla and the Gigafactory. Why do we think current batteries are so great? Mostly, I think investors are simply unaware of how weak batteries are from a theoretical standpoint. Another point to consider might be how society got to know Lithium batteries. Here's my personal experience: About a decade ago, I started using Li-ion batteries together with LED flashlights for the first time. The quantum leap from 2 hours of dim yellow light to 20 hours of bright blue radiance dazzled me for years. It seemed like magic, and left me with a deep impression that Lithium batteries were an 'enabling' technology that made new things possible.
Experiences like mine may be an important factor affecting the valuation and fate of the Gigafactory. For example, having a huge cellphone be powered all day by a little flat square is amazing to anyone who was around during the 90s and 80s. Such impressions could have colored the perceptions of the even the savviest financiers who bought bonds for the Gigafactory in 2014, or will buy TSLA stock in the latest offering. It's easy to be dazzled by modern Lithium power tools, the Model S's acceleration etc. With that mindset, investing to fund a gigantic battery factory may seem like a good idea.
But let us not be fooled: Current tech has hit a very hard ceiling, yet no physical law forbids batteries that have 10X more power, charge in seconds or are 10X cheaper. Batteries which use cheap metals or organics and still handily beat our weak, expensive technology. And since Chemistry encompasses literally Trillions of possibilities, it is unthinkable that there aren't Millions of ways to build an ideal, safe battery. This situation is fundamentally different from the problems facing silicon chips or the hard jump from cycle-based software processing to continuous computation.
Public Perception of Battery Breakthroughs:
Another factor important to the current valuation of the Gigafactory and Tesla are the three decades of experience we've had with thin-film Lithium batteries. Improving them has proven to be fiendishly difficult, something which has influenced the entire industry's opinion on battery innovations in general.
But since about the time the first iPhone got released, research in battery tech has really taken off. Research funds are flowing like never before, and of course the increased pace and interest in battery innovation is reflected in the media: For years now, news stories are beating the drum about the next big technology, so much so that this Tesla MB poster wrote in response to a post about Aluminum-Air batteries:
..So it hasn't actually been demonstrated in the real world yet; the article notes they're still trying for a pilot project. In other words, it's no different than any other miracle battery technology we've read about over the past few decades. Yawn. Wake me when I can buy one.
Such dismissive attitudes demonstrate a bias well known to behavioral finance: Once a pattern is established (in this case, news of fantastic battery break-throughs never making it to market), we expect it to continue. For the last decade that has been true. With the Gigafactory tooling up to exclusively produce current technology, Tesla has in effect taken the bet that for the lifetime of the Gigafactory, no radical breakthrough will occur to make cylindrical thin-film Li-Ion cells obsolete. This is a crucial point to any investment thesis regarding Tesla.
Yet situations in which even the leaders of high-tech industries underestimate the innovative potential of their own fields are all too common, even when they haven't been conditioned by 30 years of living with a frustrating technology. For example, in the '80s/early '90s, Steve Jobs himself didn't think personal computing would take off the way it did. Likewise, many of the early inventors and biggest proponents of the internet doubted at the time the world wide web would ever be popular outside of academia. Those were rational, highly intelligent men and women. The error, not that it mattered much to them, was that they simply extrapolated industry trends from their own limitations, then were taken by surprise by the speed of unanticipated outside innovation.
The same kind of positive surprise with improvements in battery power would boost Tesla IF it hadn't tied itself to a method of production. In my opinion, Tesla's Gigafactory project would be like an early Microsoft building a huge plant to make cheaper floppy disks, or like internet entrepreneurs betting the farm on improved dial-up modems to bring the web to the masses. The way I see it, Tesla has left its core field of competence (building / marketing great EV cars) to produce what everyone, even if just subconsciously, knows is a clearly underpowered product. Historically, similar endeavors have usually ended badly.
Market Ready New Technologies:
Source: Gizmag Magazine, Nov. 28 2015.
Before examining scientific research I would like to give a sense for what is already out there. This article lists 43 established companies pursuing new technology. It's a mere snapshot of the largest or most prominent companies, the tip of the iceberg so to speak. Here are a few competitors and some facts about them:
- Alevo Inc: New methods for inorganic Lithium Batteries for grid storage, 32 member research team.
- Boston Power (NYSE:BSX): Makes Li-Ion batteries for EVs. Reportedly building out to Giga-watt capacities.
- Electrovaya: Nontoxic Lithium battery manufacturing process.
- (NYSE:GE): Spends heavily on research into alternative battery technologies. Here's a quote (Feb. 2015): "General Electric is significantly scaling back production of its sodium-ion Durathon batteries, a move that comes amid what the company says is a slow-to-develop market for grid-scale energy storage -- or, at least, not as fast-growing as it hoped it would be by now." GE's view of the market directly contradicts TSLA's claim that demand for its Power products is roaring. More on that later.
- Imergy: Vanadium Flow batteries for home energy storage. Already selling, cheaper than Lithium tech and enjoying a greater yearly rate of improvement than thin-film Lithium.
- Pellion: Magnesium Battery. Quote from their website: "Pellion's demonstration devices now perform at energy densities >1000 Wh/L. We build test cells by the thousand and prototype cells up to the 2Ah capacity level."
- Sakti3: Solid state batteries.
- Faradion: Sodium-Ion Chemistry. A quote from their website: "We are in the process of optimizing a wide range of materials that give similar performance to Li-ion materials, but at a fraction of the cost."
- 24M: 50% cheaper Lithium-Ion manufacturing process. Quote: "Viswanathan adds that 24M's new battery design "could do the same sort of disruption to lithium ion batteries manufacturing as what mini-mills did to the integrated steel mills.""
- Amprius: Better Silicon Anode for much improved power density / reduced cost. Major investments by VC funds and even the former US science secretary.
- Siemens (OTCPK:SIEGY) This German industry titan is spending heavily on secret research it claims will revolutionize batteries in the near term (H2 2016).
- Dual Carbon Batteries: Invented last year in Japan. Claims are that it charges 20X faster, has the same power as Li-Ion, is drastically cheaper, lasts for 3000 cycles and has double carbon electrodes. The company is very secretive and may be a fake, but the technology was developed at a major university.
The point of the above list isn't to showcase individual competition for the Gigafactory, though each competitor has of course a serious potential to outcompete Tesla in ways current investors probably don't take into account. Rather, it is to give a feeling for the breadth and pace of the industry. It would take 3 of my long articles just to list all the battery competitors I've researched! Incidentally, I believe that there exist unique investment opportunities in the battery field. Quite possibly, the investor sentiments I've described drastically undervalue the potential of technological advancements, while at the same time overvaluing endeavors like the Gigafactory. Picking the right chemistry before it becomes a big success has the potential to be quite profitable.
In regard to market-ready technologies entering the mainstream, it has to be mentioned that the field of Li-Ion battery improvements is especially vast and late stage. Over the past 2-3 years, many papers have come forward reporting durable ways to improve power density / reduce costs from 2X to 5X within the existing thin-film Li-Ion framework. A few years ago, those same improvements required expensive laboratory-grade materials. Now we hear news like about these organic Li-Polymer batteries: "The Li-PBQS battery exhibits superior energy density (734 W h kg-1) to that of commercial inorganic cathodes, stable long-term cycling performance (1000 cycles, 86 %) and outstanding fast-discharge/charge capability (5000 mA g-1, 72 %)". The paper further states that cheap, economical routes for synthesis of the required materials have also been worked out.
The Role of Secrecy:
A point to keep in mind: As is true for the whole field of commercial high-tech developments, the most valuable and groundbreaking research is almost always kept quiet. Also, a good portion of university researchers, after publishing initial success with a new method like the above, are hired into private industry where they are forbidden from further publishing. This limits disclosure and gives the appearance of one-off papers, where a promising method is reported and then not heard of again. It's likely that there exists a whole field operating in the dark, which could release explosive news of success at any time.
One thing is certain, however: There are now thousands of 3rd party confirmed methods which achieve drastic improvements over existing thin-film Li-Ion battery technology. What are the odds that every one of them will fail? Many of these improvements are 'drop-in' types, like better and cheaper electrode formulations that can be used with existing equipment. Again, what are the odds that TSLA/Panasonic will manage to license the best tech for more than a decade?
Of course Tesla also has its own iron in the fire. The company has hired the famous Dr. Jeff Dahn who has been instrumental in perfecting the current NCA Cathode chemistry. Tesla hopes his research will save costs by gradually replacing the Graphite Anode with Silicon. Graphite itself is a very cheap material, but the precise nano-structured kind required for batteries is expensive. However, even if using Silicon saves money, Dr. Dahn's chemistry is theoretically unable to provide the 2 - 5X power increase other published methods demonstrate. Incorporating Silicon Anodes promises only a small, step-wise increase in power density, which is what Tesla has announced will be the plan. Of course there are problems: One is that Silicone swells over 300% in volume when absorbing Lithium, causing the Anode film to pulverize itself. Also, the technology is not unique to Tesla: Silicone Anodes have been disclosed in the literature for 50 years now, and several companies like Nexeon and others offer improved Silicon technologies for license to anyone. Tesla and Panasonic do not have, and are not planning to produce batteries with any revolutionary improvements.
Research into new Battery Technologies:
It's become a vast field; I recommend phys.org as a wonderful resource which reviews the whole field of science in an approachable way. Searching their website for 'lithium battery' or just 'battery' pulls up thousands of articles. Likewise, Google Scholar shows more than 20.000 new papers per year on the subject. Grants for research have exploded in recent years, funding everything from huge collaborative efforts to a small army of post-docs and grad students working on their own projects. At this point I would like to remind everyone just how little Lithium is used in modern batteries, and just how far below maximum power we are. I truly think the field of battery tech may be one of the last where a lone researcher can still single-handedly discover and build something that vastly eclipses the current state of the art.
And the possibilities for success have never looked brighter. Better Algorithms, backed by 10 times the computational power as of a few years ago search through millions of chemical possibilities, while better measurement techniques like Neutron scattering and more powerful X-ray crystallography are making it possible for the first time to study the nano-processes underlying the many failure modes batteries go through.
The following is my personal experience, so take it with a grain of salt: I walked through the physical chemistry wing of my old university recently and stopped to chat. Several teams are working on battery tech there (nano-tubes as better electrode material, and very interesting research into a novel Boron-organic liquid flow cell in regular battery size formats, powered by cheap micropumps). Compared to a few years ago the number and size of teams have grown, and an atmosphere of excitement lay in the air that had been absent before. I was told that another researcher in the department had the rights to her technology bought. She still worked there, but not even she knew what was happening to her break-through (a regenerative Anode material for Lithium batteries made from anchored carbon-nanotubes holding a graphene / Silicon foam). My personal impression is that the field is experiencing a renaissance.
In regards to the DOE's slide, I would like to add that patent applications can be kept unpublished until the very day of Allowance. The patenting process typically takes 2 - 4 years, so it is anyone's guess what's in the pipeline.
Finally, here's a small selection of well-known, published chemistries which each vastly outclass the Gigafactory's technology. I've numbered them (purely my subjective judgment) according to the above DOE's 1 - 7 stage process.
- Lithium - Sulfur (stage 6)
- Reversible Aluminum - Air (stage 5 - 7)
- One way Aluminum - Air (stage 6)
- Sodium Ion (stage 7)
- Magnesium Ion (stage 7)
- Fuel Cells (many designs) (stage 1 - 7)
- Flow Batteries (many designs) (stage 1 - 7)
- Aluminum Graphite Ion (stage 7)
- Super Capacitor Batteries (stage 1-3)
- Graphene Batteries (stage 2)
- Solid State Batteries (many designs) (stage 1 - 5)
- Iron Sulfide Battery (stage 1)
- Carbon-Carbon Battery (stage 7)
- Compressed Air Energy Storage (stage 5 - 7)
- And many more
I could go on and list 100s of promising papers and technologies, but instead I urge everyone who has a financial interest in Tesla to confirm the surge in research for yourself. It's especially instructive to examine the differences between old papers (2009 and before) and newer ones. Whereas before there were many caveats disclosed, such as that the experimental batteries required expensive gold nanoparticles or rare elements, current research now has the luxury to inform us that it's made from common, inexpensive inputs.
Conclusion about the Battery Industry:
Even more than in the late 80s when our current technology was first commercialized, the world hungers for better batteries. Likewise, our fundamental understanding of Lithium batteries is finally reaching a satisfying level (link is to an inspiring undergrad's research project). And there are no hurdles to adoption for new tech: Batteries are the ultimate drop-in ready technology.
That's good, because everyone agrees on some level that things need to change: Phones barely make it a full day, drones drop from the sky after a few minutes and Tesla's cars are limited to ~200 miles. And just for that feat, the Model S requires over a thousand pounds of very expensive batteries. I agree 100% with Tesla bulls that the sooner the world switches to EVs, the better. But I don't think that thin-film Lithium Ion tech, like the Gigafactory plans to produce, will lead that charge. Rather, it will be a revolutionary scientific advance, probably one that has already been discovered, which will make our EV dreams come true.
And we are further along toward making a better battery than is realized by most people. Numerous technologies like Magnesium and Sodium are already sold, and their improvement curves are 30 to 50% per year. Yet a number of trends has conditioned us to ignore or belittle claims of battery advancements even when they are extensively 3rd party confirmed and their economics make sense. The fact that many of the most successful researchers are joining the private industry and stop publishing is another factor abetting our collective blind-spot.
Historically, many industries have been blindsided by abrupt changes, though in almost all cases there was ample evidence at the time that things were changing. Even Nikolas Tesla, the company's namesake, can provide us with a great example: The invention of alternating current made it possible to distribute electricity cheaply, with none of the losses that limit direct current to a 1-mile radius around the source of power. Nevertheless, proponents of the gas-lighting industry pushed ahead with massive investments in their industry even as the nascent technology's cost function made it clear to everyone that gas lighting was doomed. The gas lighting industry simply assumed the new technology would be like old direct current tech. Investing massively to produce thin-film Lithium Ion batteries with so much potential for an even bigger revolutionary jump in technology has all the hallmarks of a very unwise decision. Thin-film Lithium Ion batteries are the Whale Oil of modern days: it's clear to everyone that this store of power is simply not suitable in the long run. An alternative will be found.
Section 2: The Gigafactory
Source: 2015 Tesla Presentation
History of the Gigafactory:
First announced Sep. 06 2013 in an off-hand remark by Elon Musk, the original plan was to build a 10mm square feet Gigafactory at a cost of 5 Billion Dollars, 1.5 to 2B of which were to be funded by Panasonic. TSLA raised roughly 2.3 Billion dollars through convertible bond sales early in 2014, debts specifically advertised as being intended to finance the Gigafactory's construction. The bonds carry very low interest rates; even at the time the company was applauded for locking in great financing. Immediately following the issue, JP Morgan doubled TSLA's price target to $340.
I will disregard TSLA's finances for the most part except to note two points: 1) Despite promises, capital raised from the bonds was mainly used for SG&A and 2) Significant collateral needed to be posted, such as property deeds, machinery etc. In the event of a price collapse or worse, shareholders are last in line to receive any compensation. This also means Bondholders likely have an outsized influence with the company. If they ever needed to convert to equity, history shows existing shareholders would face serious dilution.
Reno, Nevada was chosen as the site's location due to generous incentives. Construction started in earnest in 2015, with the shell of a building only 17% as large as planned completed early Fall 2015. Originally forecast to go into production in 2017, the greatly reduced project is now advertised to start manufacturing in 2016.
TSLA has not shared clear reasons for shrinking the project. Nevertheless, the building still represents a massive undertaking, with over 1 million sqft of space erected. Sunken costs already exceed 200 million and the rate of spending is accelerating, as the present news cycle makes clear.
Panasonic is Tesla's current supplier of battery cells and the partner for the Gigafactory. According to what has been disclosed of their contract, Tesla is solely responsible for erecting the building, putting in plumbing etc. When that is finished, Panasonic will install and run (with its own employees and technology) the actual machinery to produce the cylindrical battery cells. TSLA will then assemble battery packs from those cells directly in the same building. Those packs will go into cars or constitute the Power products the company plans to sell in volume. As of November 2015, Tesla has started moving cell-to-pack assembly into the building.
How the Gigafactory hopes to make Cells Cheaper:
Fact: A typical 18650 cell costs about $1.35, with roughly 30c spent on construction costs and the rest for materials.
The plan for Tesla's Gigafactory is to take the very complex battery production process and replicate it hundreds of times in the same building. The thinking goes that extreme economy of scale could squeeze out the last bits of margin the other manufacturers don't capture. These savings would naturally apply only to the 30c currently spent on construction, not input costs. This is understandable as there are almost no discounts to be had for increased material purchases: For example, 100 tons of copper will cost nearly the same per kilo as 500 tons of copper.
Some materials may even become more expensive with the Gigafactory's increased demand, as manufacturing / mining capability for them doesn't exist yet. More on that later, but at face value a 30% reduction in price for the above $1.35 battery would mean 40c savings. That obviously exceeds the current labor costs altogether. There are varying estimations for battery costs, but in no case are manufacturing costs more than 30% per battery. Clearly, Tesla expects not only massive savings in battery assembly that no one has so far suspected, but also in other places as well.
The only other place to save so much would be to save on materials or handling. I won't get into speculation of exactly which flavor of chemistry and tricks Panasonic is rumored to to be using for the cells, except to repeat what Tesla has definitely disclosed: The cells the Gigafactory plans to make will contain the usual cylindrical electrode structure with industry-standard chemistry. The cells will look almost identical to 18650s, but are a bit larger and have 30% more volume. Saving on materials will be hard.
So those extra savings will have to come from further efficiencies of scale in the procurement chain, transport costs, or both. I've summed up the Positives below:
Upsides of the Gigafactory:
- It promises lower shipping costs for US based product assembly as compared to shipping cells from Asia. But since early 2014 when the project got under way, container shipping rates (as measured by the Baltic Dry Index) have fallen precipitously from ~2000 to as low as 279. They are presently in the 500 range. Hence, what would have been sizeable savings on heavy volumes of hazmat shipping have disappeared.
- Tesla and Panasonic have the ability to reduce/increase production at will, allowing Tesla to match demand with supply. Why is that important? Currently, Tesla signs contracts guaranteeing certain volumes to suppliers. But batteries start going stale the second they are produced, so if today Tesla signs for too much volume, it is faced with an excess of batteries that rapidly go stale. Selling the cells at a discount to move them, in Power products for example, is obviously hugely unprofitable. With its own factory, TSLA can better match production with demand (if the contract with Panasonic allows it).
- Efficiency of Scale: Mr. Musk mentioned 30% savings over the lifetime of the plant due to efficiencies of scale relative to the competitive environment of early 2014. This explains the catchy name 'Giga' factory, which refers to the amount of battery electrical storage expected to be produced. However, multiple rivals like BYD, LG and even Boston Power have announced their own construction of 'Giga' factories. Nevertheless, TSLA might enjoy a few years of low-cost battery supplies from the current setup.
- Quality Control: It is open to discussion how damaging the few battery fires were to Tesla's reputation. Interestingly, that drama played out exactly during the time Tesla announced plans for the Gigafactory. I've always been of the opinion that the whole issue had been vastly overplayed. So if current packs are indeed safe, building the Gigafactory to ensure better QC is unnecessary.
- Reducing cost by housing cell production and pack assembly under one roof. Costs are already very small to assemble packs from batteries, which is a process so simple the average layperson can do it.
- Control over Currency Risks. As we all know the Dollar has had a wild ride since 2014. This has significantly cheapened imports, such as battery cells, but is making US based exports more expensive for the rest of the world. If the plan had been to save on import costs / hedge against currency risks, then those calculations have been thoroughly upended. The USA remains Tesla's largest market, and not only are competitors now able to source cheaper cells from abroad, but Tesla's US based manufacturing makes it harder for the company to profitably export to other countries. Exports of US high tech products have been dropping dramatically over the last year, a development not reflected in Tesla's stock price.
- Gaining Panasonic's organizational expertise at building batteries. This is hard to quantify but could well be a factor if Tesla plans to become an independent battery producer. It's a far-fetched argument, but these hard-to-quantify soft factors can play major roles in the destiny of a company.
- The GF is too much, too soon. Companies usually require decades of experience before they can master hugely elaborate projects. Take Ford: Before building the first true assembly line in 1913, the Ford company worked for more than a decade to perfect the complex technique in gradual steps. Negative Example: (SZYM), a young company which grows algae in a sugar solution to produce tailored oils. Their gigantic brewing plant named Moema, advertised 2 years ago to "bring economy of scale" and expected to ramp up without a hitch, failed miserably to enter into full operation. Sadly, there are many similarities between it and how Tesla is conducting the Gigafactory project, including PR tactics, fundraising, use of contractors, rapid construction, the level of prior experience, the role of partners etc. SZYM's failure is definitely worth a look as an educational lesson:
- Equipment Malfunction can Idle the Whole Plant: With such a vertically integrated facility, there's an 'all eggs in one basket' problem. With so many processes running in one place, a shut-down anywhere in the chain would have outsized consequences, potentially reducing operational output relative to expectations in a lasting way.
- Experimenting with Technological Improvements. In highly interconnected, very large operations it's harder to idle a line or two to try out something different, according to a study by Siemens. This leads to a paradoxical situation: The largest, most modern industrial facilities are harder to optimize and update than medium or just plain large facilities.
- Mistakes are Time-consuming and Costly to Fix. Let's assume problems emerge in the robots wiring the cells together. With 100s of the machines, all with the same error, it will take much longer to fix than with a modular, organic expansion. That way, mistakes are caught sooner, employ workers more evenly and result in less down-time.
- Other battery makers flood the market with below cost merchandise. This would destroy TSLA's margins and is at high risk of happening, as the whole of Asia places deep importance on manufacturing prowess. Specifically, China has announced in its latest 5-year plan that it will foster high-tech manufacturing and energy products, so producing premium batteries is doubly important to it.
- New battery technology suitable only for stationary energy storage undercuts in price. Without the expected production for this market, the Gigafactory's economy of scale disappears. This cheaper stationary storage technology already exists, like CAES or flow cells, with super-fast charging Aluminum Ion batteries or cheap Sodium ion batteries and dozens more technologies either already selling or planned to go public in the next 3 - 9 months. Those technologies likewise enjoy faster yoy improvements of 20 to 50%, far above thin-film Lithium's 6 to 8% / year rate.
- Risk of Accidents: TSLA says production will ramp continuously until 2020. This implies a multi-year construction project inside a structure where toxic, explosive chemicals are handled. At the very least, extra teams are necessary to synchronize the heavy construction with the ongoing battery production just feet away, costing money. Construction in active industrial facilities is a major cause of accidents.
- Panasonic. Why have the Gigafactory plans been scaled back and pulled forward? What is TSLA's working relationship with Panasonic? These unknowns create explosive risk to the stock in case it was Panasonic which has gotten cold feet. Once revealed, the reasons for the drastic scale-back from 10mm square feet to 1mm are unlikely to be positive for investors.
- Two operations under one Roof: This will require two different companies, speaking two different languages to synchronize two separate manufacturing processes in one building, a difficult proposition at the best of times, which involves juggling twice the number of supply chains etc. As stockholders, we expect things like factories to simply work; to us, they are numbers on a page, or noble concepts. But reality is gritty: Especially where complex interdependencies and massive scale are involved, minor problems can snowball into major disasters with depressing ease.
- Use of Contractors/Subcontractors. For years now, Tesla has blamed delays in manufacturing on faulty contract work. What sets more experienced companies apart are also their deeper industry connections. Should there be faulty work with the rushed construction of the Gigafactory, a similiar situation which tanked SZYM stock, it will be TSLA who bears the fall-out.
- Other EV makers like Nissan (OTCPK:NSANY) will remain flexible in regards to their battery suppliers. The Gigafactory ties Tesla to Panasonic and the cell chemistry it owns. Purchasing outside cells en masse turns the GF into a gargantuan write-off by default. Source: LG Chem, which TSLA recently invited to supply cells for the Roadster. Not all is well in the Panasonic partnership.
- Supply Shortages may slow down the running of the Gigafactory. It's not just Lithium and Cobalt which are already in short supply and whose price could go parabolic if demand for them tripled; should the Gigafactory truly scale up to become the biggest battery producer of the world, then it'll require vast amounts of other specialized inputs. For example, there currently don't exist the manufacturing capabilities to produce all the specialized graphite necessary. The same is likely true for other material inputs. When demand outstrips supply, prices rise.
- Tesla will need to convince its suppliers to invest in their own expansion, especially for materials like Graphite and battery-cell electronics. And unlike for Lithium which has many other uses, specialized input materials would be specific for the Gigafactory. Tesla hopes to sign exclusive deals with suppliers where it guarantees certain volumes for a span of years, usually 5 to 10. These types of arrangements therefore serve as an excellent barometer to show how 3rd parties view the Gigafactory project and Tesla's stability as a partner. Worryingly, Tesla has not announced those partnerships. The news Tesla releases are of prospective partnerships, where it is clear that very little monies have been spent. This June 2015 WSJ article claims that Tesla is now planning to go into some specialized manufacturing itself. If true, it would mean Tesla will divert even more capital towards the Gigafactory, take on larger risks and that it doesn't enjoy the confidence of industry insiders.
- But the main risk is that one of the many, many battery breakthroughs already reported works out commercially. Such news could come any day. It may even go unrecognized for a time, until reality becomes unavoidable.
Gigafactory Break-even only in 7-10 Years:
Execution risks span the time-spectrum from construction delays to long-term obsolescence. But especially the threat of new technologies matters because Tesla itself projects 7 - 10 years for the Gigafactory to achieve breakeven. So for the next decade, no major change to the basic technology must happen for the project to start paying off. Amazingly, in another decade thin-film Lithium batteries will be 50 years old.
But assuming industry trends will remain static can be catastrophic. Here's an example: Erickson Aircranes (NASDAQ:EAC), a thirty years old and well-respected helicopter services company, fundraised heavily in 2013 to radically increase their fleet. Everything looked good, and EAC promised efficiencies of scale that would dominate their industry for decades. Just as with TSLA, the debt was secured by the company's core assets. EAC's sector, which was doing OK when the deal was struck, went into a sharp decline not two years after the "good times", something few foresaw at the time. Thanks to its ill-advised, debt-fueled expansion EAC is near bankruptcy now, see below: Click to enlarge
The situation is not a perfect match to Tesla's but there are important similarities: Tesla is committed to the same spirit of positive assumptions EAC probably made about growth in their respective market. And just like EAC, Tesla is expanding radically for "Efficiency of Scale" although no other industry player is doing so. As EAC shows, assuming that expected demand will materialize can land otherwise healthy companies into a lot of trouble, especially when excessive debt is in play. EAC shareholders have seen their holdings reduced from $20 to $2 as other Investors hold back on giving the company money, rightfully fearing that their investments could suffer a total loss. And EAC is no fluffy bubble stock: Their brand was the most respected in their entire industry, they are highly diversified and still bring in hundreds of millions in revenues each year. A good brand-name alone cannot save a company from ruin when the numbers don't work.
In my opinion, cases like EAC's are worth to be studied for the many lessons their failures can offer, including how rapidly our markets can price previously popular companies into a kind of liquidity death spiral. As applied to Tesla, those lessons would be that speculative expansions, especially when they are made on assumptions of fantastic growth, can turn out very badly when excessive debt is involved. Conditions change, and especially the battery industry promises massive upheaval sooner rather than later. Tesla has voluntarily tied its fate to that of the Gigafactory, and downside risk could be catastrophic no matter how assured current shareholders are. Simply because the brand is the best in the business and sales are booming is no guarantee for future success.
Tesla Batteries for Home Storage:
A point I've noticed which is relevant to TSLA bulls is the fervent belief that home energy storage will soon be big business, so I'll address it here. It's hoped this future demand guarantees vast new markets for the Gigafactory's products. But simple economics show price arbitrage with electricity (charge at night when rates are low, use during the day) is impossible: An average homeowner might save $100 to $300 a year, but would need many thousand dollars worth of batteries and installation costs to do so. Since batteries in daily use don't last more than 5 to 7 years, there is absolutely no way to break-even.
Let me make the situation even clearer: In some place around the world (for example in certain parts of Australia), very high differences between day/night electricity rates do exists that could just barely push rate arbitrage into the breaking-even range over a decade. But industrial-scale storage is maturing even more rapidly than home-batteries, the idea of which has been around since the 1860s by the way. In view of cheaper industrial scale storage, how likely is it that economically sensible price differences will still exist in a decade? Also, the current adoption of easy-on easy-off gas turbines tends to flatten energy prices. As fossil gas prices look to stay cheap for a long time, utilities keep investing heavily in them. Solar Power adds the finishing touch: It produces during the day when demand is highest. More solar = flatter rates. It is safe to say that since the economics don't work out, customers won't buy these types of products. There are also numerous legal issues with utilities running equipment in private homes. For these and many more reasons I think the current excitement about home-energy storage is nothing but a bubble in sentiment.
Someone might point out that there exists at least a mature market for off-grid self-sufficiency (Solar panels + Batteries). And though it's a neat idea I'm even looking into for my own home, that market is tiny. Yet another point makes it seem unlikely that the Gigafactory will ever supply the storage market the way Tesla projects: For stationary storage there are fewer requirements for power density or resistance to high-impact collisions than for automotive grade batteries. In your garage, it won't matter if one battery is a bit bigger or heavier than the next; what matters is cost and saftey. Other battery chemistries on the market are already cheaper, longer-lasting and inherently safer than Tesla's planned batteries even though they are barely starting to be mass-produced. Their rates of improvements are also far steeper, on the order of 20 - 50% per year. Investors are clearly in a euphoric bubble of sentiment in regards to the potential of Gigafactory automotive cells to profitably dominate the home electricity storage market.
It helps to have observed prior bubbles in sentiment in order to recognize when one is going on presently. So for your edification, I refer to a nearly identical situation in farming four years ago: Folks believed that growing cover crops like switchgrass or miscanthus would soon provide boundless feedstock for biofuels and electricity generation. In a way, Tesla taps the exact same secular interests (revolution in powering cars and electricity). Switchgrass seed companies like (NASDAQ:CERE) and speculative ethanol producers rode what some at the time called a bubble, which predictably popped when it became clear that the economics had no chance to work out. The whole sector then crashed 90 - 99%. Btw if you are interested to read about the kind of stock I am long in the farming / environmentally friendly category, look into Arcadia Biosciences (RKDA).
Another example of a bubble in a tech stock that burst due to over-estimated demand: 3D printing. I consider this an even more promising new industry than anything to do with electricity, but again, if the economics don't work, those overbought valuations eventually deflate. Practical Example: If you had put $230 into (NYSE:DDD) stock anywhere near the top, all you'd be left with now would be a measly $20. This isn't saying those aren't fabulous ideas; but good timing is a must for any kind of investing, even long-term. Here's what I mean: Click to enlargeSource. Where is Tesla on this Chart?
Quick List of Possible Gigafactory Catalysts:
There are many variables to consider in regards to the Gigafactory, so I've made two brief lists to sum it all up. I've be using them to track the news-flow of the Gigafactory and evaluate my predictions.
- Partners spend serious money to update their own capabilities to produce specialized inputs for the Gigafactory.
- Tesla enters binding supply deals for Lithium, Cobalt etc.
- Tesla discloses firm orders for battery packs.
- Panasonic installs machinery on schedule.
- First cells produced by fall 2016.
- Trouble-free ramp up.
- Products work well.
- Major manufacturer announces they're producing a new, superior type of battery.
- Credible news of battery advances.
- Competition fields better / cheaper batteries.
- Panasonic reveals how far it has scaled back investment.
- Panasonic reveals why it has scaled back.
- More aspersions cast on Panasonic's Li-Ion technology.
- Other reasons revealed why project has been shrunk.
- Supply Chain Difficulties: No binding contracts announced.
- Secondary to raise capital doesn't go well.
- Construction Delays.
- PR of reduced cell production forecast.
- Slower ramp than expected.
- The actual products are sub-par.
- Freight costs stay low, permanently removing a major price advantage compared to importing cells.
- Dollar stays strong, making competitor's imports even cheaper in the US and significantly hindering Gigafactory exports.
- Rival Gigafactories go online (already happening).
- Price War in Batteries.
- Battery prices drop faster than projected, further undermining the Gigafactory's hoped-for margins (such unanticipated price drops are starting to happen, the reasons are still unclear to industry watchers).
- Tesla's order backlog forces it into unprofitable sales.
- Industrial storage market settles on other technology, ending the current period of experimentation and closing the utility markets to Tesla's automotive batteries.
- Home storage market doesn't materialize.
Chart at the time I took my first short position in TSLA.
Tesla is a young company headed by a charismatic CEO. They've just released the Model X, a luxury 700hp SUV with falcon wing doors and 155mph top speed aimed at women. So they're bold. Perhaps too bold. Since April 2013 Tesla has been valued for the dream it pitches; this could conceivably go on until earnings catch up. But doubt is growing: After technically becoming cash flow positive some years ago, the company is now more than ever reliant on massive fundraising to stay afloat. If just one of Mr. Musk's bets turns sour (Model X, Gigafactory, Model 3), it would cast doubt on Tesla's whole financial structure.
Of all its daring ventures, the Gigafactory might be Tesla's riskiest (though the Model X's unusual design features give it a good run for its money). Slim margins on battery sales are a poor reward for show-stopping risks like total obsolescence of the company's biggest asset once new technology comes out. The Gigafactory's great complexity also brings with it a smorgasbord of execution risks. And finally, red-hot competition in the commoditized battery industry mean that if indeed the Gigafactory is ever successful, the business model would soon be copied, nulling any advantage.
It would deeply damage investor trust in Tesla's other ventures if the Gigafactory doesn't live up to expectations. As a failure of the Gigafactory also implies huge financial losses, Tesla stock could easily fall to pre-rally levels as investors may become unwilling to give money to Tesla at the current stratospheric valuations. Like the example of Erikson Aircranes shows, having the industry leading brand is no insurance against problems with debt. With Tesla's very high R&D expenses, obligations and debt levels, the Gigafactory's failure could even lead to a self-reinforcing situation of dropping equity prices, reduced operational freedom and finally loss of sales, especially if the competition uses better batteries. Customers care a great deal about range, and lighter batteries will bring exponential benefits (such as less weight to accelerate, hence smaller engines, meaning still less weight and so on..).
I truly wish Tesla's autocratic CEO would call off this nonsense for the sake of the speed with which our society does the inevitable switch to EVs. Wouldn't it have been better to spend those Billions on battery research or on just staying #1 in the pure EV sector? Another smart idea would be to keep more money in reserve: Service centers, suppliers, interest payments etc all need capital even if sales lag for a year. A recession is bound to happen eventually. Alas, it is too late to change course. I suspect building the Gigafactory will turn out to have been a very ill-timed way to spend the goodwill directed at the company. And when a CEO dismisses fuel cells as "fool cells", it makes me wary they may not fully grasp what risks lie ahead.
Which brings me to my last point: Like many others, I've also noticed the frequent allusions to gambling when Tesla's CEO is mentioned. People like to say: "Don't bet against Elon Musk." But no winning streak lasts forever, and his Gigafactory gamble against scientific progress in the field of batteries may be the one that breaks it.
My present trading tactic is to short Tesla in line with its cycles of publicity. I also own long-dated puts on TSLA as a portfolio hedge against a broad market downturn; in those, it's often the speculative names that get hammered most. Nevertheless I admire the company, just not that it is betting the farm that the status quo in the battery industry will stick around forever.
Disclosure: I am/we are short TSLA.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: I am invested in a private company developing an alternate battery technology which I discovered in the course of researching this article. The company is not mentioned in this article. All views and analysis are my own.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.