A decade ago, clean energy seemed like a pipe dream. Solar panels, windmills and electric cars were widely considered to be something for wealthy tree huggers to assuage their conscience, rather than components of a serious energy policy. Now, however, Morgan Stanley predicts that renewables will overtake fossil fuels by 2020.
The appeal of clean energy has gone way beyond climate change and the environment. It's now being increasingly driven by basic economics. Wind and solar energy have achieved grid parity in many places and, over the next decade, clean energy will become far cheaper than traditional sources.
The sticking point is energy storage. While today's dominant technology, lithium-ion, has made great strides, it is approaching theoretical limits. So we need to discover fundamentally new chemistries in order to continue to lower costs and increase the clean energy footprint. To get there, we will need to forge a new partnership between government and private industry.
The Race To Build A Battery That Can Power The 21st Century
The integrated circuit was invented independently by two Americans, Jack Kilby at Texas Instruments and Robert Noyce at Fairchild Semiconductor, in the late 1950s. Yet by the 1980s, the Japanese had begun to dominate the technology and the American semiconductor industry was in peril. Seeing the matter as critical to national security, the US government set up a public-private consortium, SEMATECH, to regain dominance.
In very much the same way, the lithium-ion battery was largely an American invention that became dominated by Asian competitors. Yet unlike computer chips, the industry was not deemed important enough to inspire a major effort and today firms like LG, Panasonic, Toshiba and Samsung are the top players.
And much like Moore's law, lithium-ion batteries are nearing their theoretical limits. While prices for battery packs have fallen from over $1000/kWh in 2010 to under $200 today, it is doubtful that they will ever get much under $100/kWh. That kind of progress will serve us well over the next decade, but to advance beyond that we need something more radical.
In 2012, the Department of Energy established the Joint Center for Energy Storage Research (JCESR), a partnership between the National Labs, academic institutions and private industry, to develop new prototype batteries within five years. Now, having accomplished that mission, it hopes to go even deeper and reimagine how new battery technologies are created.
Solving Fundamental Scientific Challenges
A battery is made up of three basic components: an anode, a cathode and an electrolyte. An improvement in any one of those components can raise the efficiency of the whole battery. So, for the last 30 years, scientists have been tweaking each part of the battery to make it more powerful and cost less.
Another important factor are manufacturing efficiencies. Just like any other product, as you increase scale you find ways to lower costs, either by increasing your ability to procure materials cheaply or simply by improving the process. That's a big reason why the price of batteries has come down so fast and we're seeing the market for electric cars and power storage expand.
However, there are limits to how much we can get out of the lithium-ion battery. Lithium can only hold so many atoms and eventually manufacturing improvements will level off as well. So we will need to come up with fundamentally new chemistries if we are to be able to improve the clean energy economy for decades to come. That's the essence of JCESR's mission
"What we found in the first five years is that the primary barriers to next generation batteries are scientific," George Crabtree, Director of JCESR told me. "That doesn't mean that there aren't serious engineering challenges as well, but that for all the chemistries we looked at, there was at least one material that doesn't exist yet. So a big part of the challenge going forward will be to discover those materials."
Using Bits To Discover Atoms
Traditionally developing materials was almost obscenely tedious and time consuming. To come up with something with the properties you were looking for, you had to test every conceivable option in the lab and, even then, there was no guarantee that you would find anything useful. More recently, however, scientists have begun using computers to automate much of the work. Essentially, they're using bits to drive atoms.
The early stages began around ten years ago, when the ability to simulations with high-performance supercomputers first became available. More recently, that ability has been augmented with machine learning algorithms that are able to identify significant patterns in enormous quantities of materials data. In the future, quantum computing may also play a role.
One of the major achievements of JCESR over the last five years has been to develop large materials genomes for the algorithms to analyze. In some cases, researchers have been able to predict materials that can lead to better batteries, but don't exist yet. For example, scientists at JCESR identified chromium-oxide as a promising candidate for candidate for magnesium-ion batteries, but it took months in the lab to figure out out to synthesize the compound.
Today, JCESR can be considered a real success. It has developed four prototypes for advanced battery chemistries and some of the technologies are already being commercialized by companies like Form Energy, Blue Current and Sepion Technologies. Yet serious challenges still remain and, in order to create a true clean energy revolution, JCESR is beginning to reimagine it's role.
Scaling To Impact
Identifying promising new battery chemistries is one thing, creating a true market impact is another. Going from milligrams in the lab to metric tons in the real world is no trivial matter and significant engineering and economic hurdles remain. To overcome these, we need to envision a new partnership between government, academia and the private sector
For example, prototypes for lithium-sulfur batteries already exist and perform very well, but can't maintain that performance across hundreds of battery cycles like conventional lithium-ion batteries can. So JCESR is shifting from helping to build a battery ecosystem to empowering one that increasingly exists.
"Traditionally, most people saw the national labs as a place where new discoveries could be spun out into new technologies," says JCESR's Crabtree. "Over the past 5 years, we've been able to flip model somewhat and companies are coming to us for help in getting answers to advance their own R&D."
Hopefully, five years from now instead of talking about merely extending conventional lithium-ion batteries, we'll have validated new chemistries that can power the next 30 or 40 years. That will be the point at which a real clean energy economy becomes truly viable.