Energy storage as an industry is no longer a curiosity or mere academic pursuit. With its ability to support the weaknesses of intermittent resources such as wind and solar, low-cost storage is the catalyst to a carbon-free world. But there are doubters who claim costs can’t drop without deep science and huge breakthroughs. For those doubters, let’s take a trip down the memory lane of a similar disruption we just witnessed in the solar industry.
In 2003, when I entered the solar industry, a solar panel cost $4.50 per watt, and entire installations were not uncommon at $10 per watt. By virtually all measures, these were not economically competitive systems. As entrepreneurs back then, when we went to pitch VCs and talk to the media about the coming solar revolution, we would often refer to a $1-per-watt number. This was described as a distant aspiration, likely to take 20 years. Even from the most sympathetic of VCs or journalists, we were met with a “yeah, right” response. Few stopped to ask details about whether this was distributed or central system reference, or even whether we were talking about just the panel or an entire installation.
This $1-per-watt target was an extraordinary goal -- we were talking about an order of magnitude reduction in the cost of installation of solar. The last reduction of that scale in solar took nearly 30 years. Yet, here we are today, seeing large-scale projects being built with costs either at or just hovering above the one-dollar mark, fully installed. Let’s take a moment to soak in that milestone, and acknowledge that even the most passionate evangelists among us doubted we would get here this quickly.
While there are many people to thank for this transformation, and many areas of cost reduction to credit, there is one component which stands out: crystalline silicon. The radical reduction of the cost of solar grade silicon over the last 15 years has been the single largest component driving the cost curve of solar.
All the while, CEOs, pundits and investors continued to believe that new chemistries and technologies would leap-frog polycrystalline cells and beat the old standard to the dollar-per-watt goal. I was one of those entrepreneurs, focused on building a tracking concentrator based around gallium-arsenide cells. When silicon was $300 per kilogram (or even when it was half that), our company's unique offering had terrific economics.
Of course, in hindsight, we know how that story ends. We were wrong. The cost of a kilogram of silicon indeed dropped below $200, then $100, and seems to have finally settled below $20. This, combined with incredible learning curves throughout the production and installation supply chains, is what has brought us to our magical number of one dollar per installed watt. This also means that the tracking technology, along with various flavors of thin film companies (Solyndra, MiaSole, Nanosolar, etc.) simply couldn’t compete. The staggering dominance of this singular old-school technology cannot be overstated. Billions of dollars of venture and other investment were wiped out by the strength of silicon’s ability to drive down costs. With the rare exceptions of First Solar and Solar Frontier, no other companies created material scale without crystalline silicon as part of their formula.
For the coming energy storage revolution, there is much to be learned from this solar history. Energy storage as an industry is about to undergo the same radical scaling exercise we just witnessed with solar. In this scaling, you can also expect to see some of the very same trends and results.
A central theme, of course, is cost reduction, and the challenge of differing technologies to get there. Out in front is the current industry standard of lithium-ion. While there are many flavors of lithium-ion, they (like silicon before it) share a common chemical core. If energy storage is the new solar, lithium-ion is the new crystalline silicon. There are and will continue to be many challengers to Li-ion’s dominance, but just as with silicon in solar, these chemistries are likely to fail to become a winning solution. There may be a few breakouts for specific applications (flow batteries are a leading contender), but these will be rare exceptions.
Lithium-ion and its derivatives are simply too large in scale at this point, with too many uses across too many industries to be disrupted. If the technology is already used in cell phones, computers, hybrid vehicles, etc., how could an entirely new technology scale to match it? Massive Korean (LG Chem) and Japanese (Panasonic) companies (along with Tesla here in the U.S.) are building gigafactories to fill the Li-ion needs. China (Foxconn, Huawei, BYD) is just getting started. Just as with our $1-per-watt-peak goal in solar, we will find that we can quickly approach the magic $100-per-kilowatt-hour target in storage (the price at which mass adoption becomes economic).
Nothing seems to focus the mind these days quite like a tweet from Elon Musk. OK, maybe a Hollywood-style announcement from the same Mr. Musk (reality’s very own Tony Stark) gets us even more engaged. And these days, both his tweets and his announcements are laser-focused around energy storage. The hype around this space is starting to reach levels we in the space-formerly-known-as-cleantech haven’t seen since the days of Nanosolar, Solyndra and other solar-coaster headlines from a decade ago. We’ve seen this movie before, and while the ending was terrific, there was also tons of carnage. Let’s learn from these misdirected efforts to avoid the same disasters this time around.
A more productive space to investigate and invest in, as was the case with solar, will be the business engineering, rather than the chemical engineering, of storage. Companies like Stem and Enphase are chemistry-agnostic, and will thrive based on all cost reductions, not on any specific winning hardware.
Investors, entrepreneurs and newly minted graduates should all look carefully at the many claims being bandied about in energy storage technology business plans. Flow batteries, compressed air storage, and more exotic storage chemistries (liquid metal, etc.) will mostly go the way of tracking concentrators, CIGS and nanomaterials. For the sake of the storage effort as well as the entire industry, let’s learn from history and improve our hit rate this time around.
We are only now, after decades of hard work, beginning to declare victory in solar cost reduction. The last big push, from 2004-2014, was what edged solar over the line. Certainly, when it comes to storage, we have a long way to go -- perhaps another decade in this field as well. I hope we can look back at the recent history of solar, extract some lessons, and see if we can’t speed up the storage revolution just a little bit.
Andrew Beebe is a managing director at Obvious Ventures.