I’ve been writing about energy storage for several months because it will be a fundamental enabling technology for cleantech, the sixth industrial revolution, and I believe it is destined to be a major investment theme for the next 20 to 30 years. I’ve written about an emerging consensus that sales in the energy storage sector will grow from $25 billion to over $100 billion by 2020. I’ve also written about a variety of technologies and companies that will benefit from explosive growth in the sector. In late December, I started to explain why I believe the energy storage sector needs to take baby steps before it can run and ended up doing a cost-benefit analysis of electric vehicles. Today I’m going to take another step back and try to concentrate on broader issues that will make the sixth industrial revolution fundamentally different from each of the five industrial revolutions that came before.
Since the early ‘70s, the dominant business mega-trend has been the age of information and telecommunications, the fifth industrial revolution. During the ‘60s, my father invested in computer mainframe leasing and did well. I studied Fortran at Arizona State University in 1974 and used punch cards for data input. I bought my first computer in 1986 and thought direct modem communication at 1,200 bps was heaven. I established our firm’s first website in 1995. Today our IT network has terabytes of storage, we have more phone lines than people and we seamlessly represent clients in Europe, North America and Asia. Even though the calendar tells me that it took over 40 years to develop the IT systems and infrastructure we rely on today, I immediately assume that the commercialization path for new IT products will be measured in months, rather than years and decades.
That fundamental assumption was valid during the late stages of the fifth revolution but it is useless for the early stages of the sixth. The issue we need to understand as investors is how the sixth revolution is likely to unfold.
Mark Twain said, “History doesn’t repeat itself, but it does rhyme.” If history teaches anything, it teaches that the cleantech revolution cannot emerge fully formed in the twinkling of an eye. Like any other infant, cleantech must first learn to crawl; then learn to stand; then learn to walk and finally learn to run. Unfortunately, we’ve gotten spoiled. We’ve forgotten the early days of the IT revolution when progress was slow and painful. We’ve also forgotten the virtue of patience. Investors who blithely assume that the rates of progress and growth in cleantech will mirror the rates we have come to expect from IT are in for a grave disappointment. Investors who remember the words of Dorothy who said, "Toto, I've a feeling we’re not in Kansas anymore,” have an opportunity to become wealthy. However doing so will require patience, a global focus and common sense.
I first mentioned a thematic report on cleantech from Merrill Lynch strategist Steven Milunovich in a Seeking Alpha article that discussed the savage beating energy storage companies suffered in last fall’s market meltdown. I haven’t quoted the Milunovich report extensively, but the following paragraph merits special attention:
On the positive side, cleantech markets dwarf IT to the tune of two orders of magnitude. Unlike tech names, cleantech companies often don’t need huge unit growth to succeed – modest improvements mean more. New IT vendors often face a hurdle of a 5-10x improvement over incumbent technology to succeed while in cleantech doing the same amount of work with reduced CO2 emissions might be enough. Moreover, we think the application of the VC model to energy could result in an acceleration of results. It’s entirely possible that real change could be achieved in the next 5-10 years, change that would take decades in the existing energy markets.
The implications are staggering! Mr. Milunovich starts out by saying we can value the entire IT sector and add two zeros. He then confirms that baby steps matter and modest differences in price and performance will be critical factors; which only makes sense when you realize that behind the noble talk of a greener planet, the driving force for the cleantech revolution is economic – minimizing waste while getting more useful energy for less money. He finishes by suggesting that the cleantech revolution may advance much faster than its predecessors. These are very exciting conclusions, but they leave no room for doubt that the sixth revolution will be very different from the fifth.
The first four industrial revolutions were primarily North American and European affairs. The fifth included Japan as an invited guest and other Asian countries have certainly been having a good time since they crashed the party. As the sixth revolution matures, North America and Europe might be little more than footnotes. The word is already out and there are 6 billion people who are working diligently to earn the lifestyles and comforts that 500 million of us already enjoy. The trick will be finding a way to help raise the standard of living in developing economies without crushing the standard of living in developed economies. For that to happen without catastrophic conflict or horrific environmental consequences, the world must find relevant scale solutions for persistent shortages of water, food, energy and every commodity you can imagine. The sixth revolution is not going to be pleasant and we will likely be plagued by rising prices, commodity shortages and intense global competition. But with 6 billion new consumers striving to modestly improve their lives, the power of the sixth revolution will be an order of magnitude greater than anything the world has ever seen. The Asian giant is not only awake; he’s hungry.
The fifth revolution was largely a revolution in physics. Electronic circuits became smaller and more powerful with each new generation of products and researchers found amazing ways to use precision manufacturing to slash raw material inputs while improving product performance. It was difficult work for manufacturers that resulted in huge benefits for consumers. We’ve already seen many of the same dynamics at work in the development of smart grids, solar panels, wind turbines, tidal power systems and flywheel energy storage. Overall, there is good reason to expect that future advances in power generation and distribution will be far more impressive than past advances. But advanced power generation and distribution technologies are only part of the solution. The rest is energy storage.
When we get down to basics in battery technology, the ugly truth is that chemistry is far less flexible than physics. Every element on the periodic table has a fixed electrochemical potential and most battery chemistries are rapidly approaching maximum theoretical efficiency. Accordingly, future gains in the ability of a given battery chemistry to store energy will be measured in single digit increments at best. To make matters worse, about 75% of the cost of a typical battery goes for raw materials and every time you reduce the amount of active materials that go into a battery, you reduce its storage capacity proportionally. That means the performance gains and cost reductions we’ve come to expect in IT are extremely unlikely in batteries. If anything, demand from 6 billion new consumers will increase battery costs, not reduce them. With due respect for the last 40 years, I believe that people who expect battery costs to decline significantly from “economies of scale” are ignoring the Asian Elephant in the living room.
This leaves cleantech facing an immense challenge. Batteries are a dreadful way to store energy. They’re big, they’re heavy, they’re expensive and they’re easily damaged by careless users; but they’re the only choice we have for many applications. Pumped hydro, compressed air and bulk thermal storage systems may prove very cost effective when it comes to storing hundreds of megawatt-hours (mWh) of electricity, but they’re useless when it comes to storing a few kilowatt-hours (kWh). Flywheels and supercapacitors work great if you need power for 60 seconds, but they can’t deliver stable power for more than a few minutes. We’ve all heard about the wonders of hydrogen fuel cells, but storing and using hydrogen gas is no easier or cheaper than storing and using compressed natural gas. At least for now, batteries are the only game in town and choosing a battery is a lot like choosing a congressman; you carefully weigh the positive and negative points, and then vote for the best bad choice.
I’ve been studying energy storage and batteries for over five years and still find technology comparisons confusing. My first level of confusion comes from the fact manufacturers don’t adhere to uniform standards when publishing performance data. A second level of confusion comes from the fact that performance data is invariably expressed in metric measurements that many find difficult. To help reduce the confusion, I decided to spend some time working on a table that would answer two simple questions I believe every battery buyer should ask.
- “If I want to store a kWh of electricity, what will the battery weigh and how much space will it take?”
- “If I want to store a kWh of electricity, what will the battery cost and what will my average battery cost per charge-discharge cycle be over the expected service life?”
The following table isn’t perfect and I’ll undoubtedly draw comments that dispute the source data, but I believe the table is accurate enough to provide a general overview of how the numbers compare for 1 kWh of storage capacity.
Most of the basic performance data for the table comes from an educational website on battery technology but I did incorporate some information from other sources when I needed to fill gaps. I then added cost data from a recent report published by Sandia National Laboratories and performed the necessary conversions and calculations.
The first three performance metrics in the table, weight, volume and price, are pretty straightforward but the other two, cycle-life and user cost per cycle, are incredibly mushy because they assume that a battery owner will use every charge-discharge cycle he buys, which is usually not the case. Unfortunately, understanding cycle-life and cost per cycle issues is critical to a well-informed choice between competing battery technologies. So I’ll drill a bit deeper.
To analyze cycle-life, engineers install a battery in a test rack, hook it up to sophisticated electronic equipment and then repetitively charge and discharge the battery until it loses 20% of its rated storage capacity. After repeating the test with a statistically valid sample, the average becomes the reported cycle-life. A big problem with cycle-life testing is that human beings are less consistent than testing equipment when it comes to using energy and following instructions. A bigger problem is that cycle-life estimates assume that each cycle will discharge the battery down to its recommended limit and then recharge it properly, which doesn’t often happen in real life. To further complicate matters, many battery chemistries including lead-acid and Ni-MH last longer at modest depths of discharge than they do in deep cycle testing, which is why a 400-cycle NiMH battery can have a 5-year useful life in an HEV.
I can’t speak for others, but my experience with the lead-acid battery in my car and the Ni-MH batteries in my house phones has been better than the table would suggest while my experience with the Li-ion batteries in my cell phone and laptop has been worse. Just for kicks, I want to take a highly unscientific straw poll and ask each of my readers to jot off a quick comment to this article that ranks their personal battery replacement history on a scale of 1 to 5. To keep the comments consistent and simplify my tabulation work, I’d like everyone to tell me how many years they average between battery replacements using the following format: Lead-acid 3, NiMH 5, Li-ion 2. (My personal ranking) With a little luck the straw poll results may give me something to write about next week.
As I’ve learned more about energy storage, I’ve come to view cycle life claims as providing a useful but unreliable indication of potential battery life. Lab tests are fine but I prefer road tests. I’m also skeptical about cycle-life claims from development stage companies that are not actually manufacturing a commercial product or pre-commercial prototype. The reasons are simple. First, you can’t draw a statistically valid test population until you have a product population to draw the sample from. Second, laboratory prototypes are unreliable indicators of product performance because a PhD working in a well-equipped laboratory can always generate test results that are vastly superior to the best results one can expect from a factory staffed by skilled and semi-skilled manufacturing workers. Third, life-cycle tests on individual cells are not a good indicator of how a multi-cell battery pack will perform because system complexity increases at astounding rates as the number of cells increases from tens to hundreds to thousands.
The answer to the most important question, “What will my average battery cost per charge-discharge cycle be over the expected service life?” is perhaps the most difficult because it is impossible to provide an answer unless you can describe what you plan to do with the battery and how long you plan to own it. First, you need to match your cycling demands with the cycling potential of the battery because regardless of the technology choice, using only a half or a third of the potential cycles will double or triple your effective cost per cycle. Once you have a firm grip on what your needs are going to be, it’s a relatively simple matter to go through the type of cost benefit analysis I presented in an earlier Seeking Alpha article, Alternative Energy Storage Needs To Take Baby Steps Before It Can Run.
It is my fervent hope that some creative soul will eventually come up with a brilliant and cost-effective way to store small amounts of energy in a portable form that will make batteries obsolete. The guy who invented petroleum did one heck of a job; he just didn’t make enough of it. For the time being, however, we’ll all be forced to make the best bad choice. I for one refuse to pay a premium for that dubious privilege.
I’m a firm believer in the upside potential of companies like Exide (XIDE) Enersys (ENS) C&D Technologies (CHP) and Axion Power International (AXPW.OB) because they make inexpensive products that can satisfy the energy storage needs of most users and applications. I can’t and won’t denigrate the technical performance of exotic Li-ion chemistries (at least until my straw poll results are in). Toshiba (TOSBF.PK) manufactures a fine Li-titanate battery and development stage competitors like Altair Nanotechnologies (ALTI) and Ener1 (HEV) make impressive performance claims. The same goes for the Li-phosphate batteries manufactured by A123 Systems (IPO Pending), China BAK (CBAK) and Valence Technology (VLNC). My only issue with these companies and their existing and proposed products is simple economics. Expensive battery technologies do not work in applications where the user’s goal is to save money. Likewise, I’m unimpressed by vague happy talk about future economies of scale when I know that 6 billion new consumers will drive a seismic shift in global demand for all commodities and products.
Increased performance combined with lower end user cost was a reasonable assumption during the fifth revolution. But we’re not in Kansas anymore and the rules of the game have changed. As we embark on the sixth revolution, the watchwords for investors are thrift, patience, focus and common sense. We have a choice to either embrace the change and profit handsomely or fight the change to our detriment. But the change will come either way.
Disclosure: Author holds a large long position in Axion Power International, recently bought small long positions in Exide and Enersys and may make other energy storage investments in the future.