The Future of Energy Storage: Five Key Insights on Battery Innovation and the Clean Energy Shift

Developments in batteries and other energy storage technology have accelerated to a seemingly head-spinning pace recently — even for the scientists, investors, and business leaders at the forefront of the industry. After all, just two decades ago, batteries were widely believed to be destined for use only in small objects like laptops and watches. But now, batteries have expanded dramatically in both size and relevance to assume a critical role in enabling society’s transition to clean energy.

At a recent gathering of global energy storage experts hosted by Columbia Business School, Dan Steingart, a professor of chemical metallurgy and chemical engineering at Columbia Engineering, recalled that just over two decades ago, his PhD project, to develop a lithium-ion battery that could power buses, was scrapped when the U.S. Department of Energy decreed that such batteries could never be safe enough for large vehicles. 

Steingart pointed out that the biases against the viability of these batteries extend even further back than that. In 1883, Thomas Edison stated that storage batteries could amount to no more than “a catchpenny, a sensation, a mechanism for swindling the public by stock companies.” That impression stuck around for far too long, he said.

“The battery industry, through brute force in some ways and through incredible scientific innovations in other ways, had a massive commercial legacy to push past to get to where it is today,” Steingart said.  

One means of achieving this feat has been in proving incorrect the once-sacrosanct beliefs about batteries’ safety limitations. A crucial factor motivating these safety improvements — and the broader focus on developing energy storage solutions more generally — has been the realization that energy storage is a necessary component in scaling up clean energy solutions to power society. Energy storage has the potential to abate up to 17 Gt of CO2 emissions by 2050 across several sectors, primarily by supporting the establishment of renewable power systems and by electrifying transport. 

The rapid scale-up of renewable energy solutions like solar and wind power will need storage solutions to keep pace with their growth. What’s more, the rapid growth in electric vehicle (EV) sales will similarly push massive demand for batteries, especially lithium-ion ones. 

According to workshop participant Shirley Meng, professor of molecular engineering at the University of Chicago Pritzker School of Molecular Engineering, the world’s current annual production of lithium-ion battery capacity stands at roughly 1 TWh. While that capacity is an achievement, she said, it represents only about 1% of the lithium-ion battery capacity the world will need to manage the transition to clean energy.  

Her message, and the message of several other participants in the daylong workshop, was that energy storage solutions will need to diversify in every sense: in production geographies, in technology types (and materials required), and beyond batteries. 

“Right now, Asia is leading in battery production, but I really think the U.S., the EU, the U.K. — everyone has a fair shot,” she said. “The question is, who wants to be part of this opportunity? Because energy storage is obviously going to be an important part of the future.” 

Five key points emerged from the experts’ wide-ranging discussion. Click on the links below to dive deeper into each. 

Key Point No. 1: There’s an EV battery tech race underway, and a combination of factors could influence which companies, geographies, and technologies pull out ahead.

Read more about Columbia Business School’s Climate Knowledge Initiative

According to Meng, the battery market for EVs is on track to grow fivefold over the decade ending in 2028, mushrooming from $17 billion in 2019 to about $95 billion. Most of that growth has happened, and will continue to happen, in lithium-ion batteries, which are the most prevalent choice for EVs, thanks to their high energy density and reliability. 

Meng pointed out that when her career began 20 years ago, Japan dominated lithium-ion battery production. But that geographic production picture has shifted radically over two decades, with China now commanding over 50% market share in the space and South Korea roughly 25% (for its part, Japan’s market share has fallen to 10%). One Chinese battery producer alone — CATL, founded in 2011 and a supplier of batteries to carmakers Tesla, BMW, and Volkswagen — represents 35% of the global EV battery market. 

Learn more about Columbia Business School’s Climate Knowledge Initiative

As a broad category, lithium-ion batteries actually contain a range of diverse chemistries. The most common of these chemistries up to this point has been NCM systems, which use nickel, cobalt, and manganese oxide cathodes, though fast-growing Chinese companies have pushed a recent shift to LFP systems, which use lithium iron phosphate as a cathode.

LFP chemistries are cheaper because the raw materials they require are less expensive than the nickel and cobalt used in NCM. This exemplifies a cost conundrum embedded in battery technology that researchers regularly confront — with some frustration, according to workshop participant Jeff Dahn, a professor in the Department of Physics and Atmospheric Science and the Department of Chemistry and head of a research group at Dalhousie University in Halifax, Nova Scotia.

Dahn, widely recognized as one of the pioneering developers of the lithium-ion battery, explained that in his lab research (funded by Tesla), he and his colleagues have subjected NCM and LFP batteries to repeated tests and have found NCM to be a fundamentally superior, longer-duration system than LFPs. 

And yet, he acknowledged, NCM technology falls short in the comparison in an important way: upfront cost. “But over the long haul, you will win with this,” he said. Dahn’s lab has tested NCM technologies — even some that dispense with cobalt, a rare mineral — that could have lifetimes upwards of 20 years. 

Adding another layer to the discussion, Alex Mass, managing director at Goldman Sachs, where he leads European climate investing, asked the question: “But have we over-optimized for lifetime?” In other words, by creating a better system at the expense of higher upfront costs, have we failed to address broader system-level challenges? “Because the point of failure won’t be chemistry at the cell level — it’ll be everything else in the system that breaks down,” Mass said. “This is part of why we’re seeing a shift to LFP systems, which offer a better upfront cost dynamic.” 

He also highlighted the persistent challenge of soft costs in energy storage. “While we’ve made remarkable progress at the cell level, the cost of installation lags behind the decline rate of hardware costs, creating a high ratio of soft costs that are the hardest to eliminate,” Mass explained. “To accelerate adoption, we need to get the rest of the system onto the same trajectory of innovation and cost reduction that batteries have been on for the last 20 years.”

The tension between upfront costs and long-term benefits underscores a critical challenge in the EV battery industry. Still, workshop participants agreed that the financeability of long-duration batteries — which admittedly come with a higher upfront price tag — remains a major barrier to widespread adoption.

Another market-imposed barrier to delivering the batteries the clean energy transition will require is that, as various manufacturers race to compete, they don’t share their learnings along the way. Meng wondered aloud, could that change?

“If we declared a national emergency for climate change so the whole country had to manufacture batteries, Tesla would have to share all of its know-how,” Meng said. “It happened in the Second World War — it’s possible — but that’s an extreme case. But it’s terrible that every company has to repeat everything Tesla has done since 2011, and it’s going to cost a lot of money.”

CBS’s Steingart pointed out that there is precedent for such a policy, from the U.S. semiconductor industry in the early 1980s, when the Reagan administration pushed to establish the first labs within universities that taught the basics of semiconductor manufacturing, which had earlier been the secret purview of private manufacturers. 

“I think it’s not just a matter of a wartime effort, but it’s a matter of saying, ‘Look, this is a commodity, and if we want to get better, we have to share knowledge and create a common baseline,’” Steingart said. “We’re not going to get to where we need to go if we keep this know-how contained.” 

Read more about Columbia Business School’s Climate Knowledge Initiative

And that rapid scale-up is important, Meng added, because the EV battery industry has other pressing improvements it will need to make in four areas over the next decade: (1) safety, particularly by developing lithium-ion solid-state batteries that pack more energy into less space and never catch fire; (2) energy, by continuing to lengthen battery lifetimes; (3) cost, in part by developing batteries that can be fully recycled, and (4) power, by developing batteries that can be fully charged in as little as five minutes. 

Importantly, the magic will be in the tech mix, Meng added: “We want to give the public multiple options so we don’t all go after one single resource. Otherwise, lithium could become like oil, in that people will compete for that resource.” 

Key Point No. 2: EV batteries will soon outlive their vehicles, which means the industry needs to figure out how to treat them as infrastructure rather than disposables.

One of the many ways in which EVs diverge from gas-powered vehicles is that an EV battery is likely to outlast its car, unlike an internal combustion engine. 

“It all comes down to the fact that no one is investing in batteries as infrastructure,” Meng said. “We really have to think of batteries as infrastructure, not disposables, to pay for the benefits of having something that lasts for decades.” She added that while she’s confident scientists can eventually deliver a battery with a 50-year lifetime, the question of whether it will actually attract financing and penetrate the market — and thereby manage to accelerate the energy transition — remains. 

In other words, these batteries should be treated as longer-term assets than today’s vehicles or their component parts, implying that they ought to be insurable for second use. But is this even possible? 

Dahn shared that in his renewable energy storage lab, Lukas Swan, a mechanical engineering researcher, has managed to collect four different battery packs, in various states of health, from retired or crashed EVs (all found on eBay). He then connected the batteries to one another and then to the grid — and he’s now storing energy to the grid with the reused EV batteries and delivering it back.

Learn more about Columbia Business School’s Climate Knowledge Initiative

“Now, it’s really possible to do this,” Dahn said. “But what pisses me off is that no one will ever do this — because no one will ever insure it.” 

Steingart replied that he’s currently at work on changing that, noting that he hopes to find a way to perform the appropriate measurements on used batteries so that they one day might, in fact, be vetted for performance and safety and therefore insurable.  

“I’m glad it’s pissing you off,” Steingart said, “because I know it’s driving research to give people the tools they need to potentially do this.” He added that to make used batteries insurable and therefore available to be connected to the grid (or placed in a new EV), “you have to have a much better estimation on the battery state of every cell in the pack. The question is, what level of granularity would you need to do that?”

Meng pointed out that if batteries can be freely reused with multiple applications, their economics suddenly become orders of magnitude more attractive: “If we can make batteries last 10 times longer, storage costs fall by a factor of 10. The way to achieve that is ultralong life.”  

Key Point No. 3: A successful energy transition employs EV batteries as utility storage.

When EVs are parked (which is how most cars spend the majority of their time), their energy remains stored, though it often could be better used as part of a distributed utility grid system. 

As utility grids add more renewable sources like solar and wind power, they need some help balancing their supply with demand. In the middle of the day, for example, solar energy is abundant, but peak demand usually happens at night. Too often, conventional energy sources are called in to smooth out the demand imbalance. 

Batteries can help store energy for when it’s needed by utility systems — and EV batteries could serve as a readily available and widely distributed source of this storage. In fact, a study by UK Power Networks found that integrating EV batteries into the grid could help reduce peak load by 10%, thereby delaying the need for grid infrastructure updates.

Several of the workshop participants agreed that vehicle-to-grid (V2G) uptake will be an integral component of shifting to a clean energy system, because of how it helps avoid the need to rebuild a new grid from scratch.

“I think EV-to-grid has to happen,” said Meng. “I mean, if it doesn’t happen, I don’t know how the energy transition can be successful because we would have to build so many extra batteries.” 

Steingart agreed, adding that he likens EV batteries’ relationship to the grid to hard drives’ relationship to the internet: Without the wide prevalence and interconnection of these technologies, decentralized and distributed systems of sharing couldn’t be born. 

Other participants noted that V2G connections — and the energy security and economic benefits they represent — could serve as a motivator for more consumers to opt for EVs. After all, an EV battery pack stores enough energy to power the average home for a day, which could serve as an important backup source in the event of a power outage. What’s more, EV owners in some cases have the option of getting paid for the energy they discharge back to the grid, especially during times of peak demand — and those incentives could be better packaged and advertised to make the opportunity clear.  

The promises of V2G point to a massive opportunity: Whatever company or entity manages to connect EV batteries and the grid in a profitable, scalable way might thereby help solve some of the challenges discussed in earlier points — namely, finding ways to extend batteries’ useful lifetimes and helping the grid adapt to today’s rapid additions in renewables capacity. 

How to get there? Some small-scale, emerging policies may point to a way forward. For example, New York City has mandated that all rideshare vehicles be zero-emission (or wheelchair accessible) by 2030, which is likely to push efforts to improve grid resiliency through faster charging and innovations that simplify and incentivize V2G uptake.

Key Point No. 4: Recycling batteries and mining for their raw materials present interrelated challenges — and opportunities.

Meng projects that a future version of the world that relies on clean energy will require between 200 TWh and 300 TWh of lithium-ion battery storage. That is an intimidating figure, she acknowledged, given that so far, the world’s battery industry has achieved only 1 TWh annual production of lithium-ion battery capacity. 

But Meng emphasized that there’s a more hopeful way to look at this capacity requirement: If the world can get to 20 TWh annual production of battery capacity and heavily integrate battery recycling into the mix, it might be possible to reach the same goal. 

Learn more about Columbia Business School’s Climate Knowledge Initiative

To get there, however, plenty of questions about battery production will need to be addressed.

The good news is that, in theory, the materials in batteries are nearly 100% recyclable. The tougher news is that it is currently cheaper to mine the raw materials for lithium-ion batteries than it is to recycle them. 

Vivas Kumar, CEO and co-founder of Mitra Chem, a startup that develops and commercializes iron-based cathode batteries, raised pressing questions about how the economics of a battery recycling industry could work.

“How do we reconcile a push for much longer lifetimes on the technologies with the fact that, for recycling to economically make sense, companies need a very large input feedstock so that operating costs can come down to parity with virgin materials?” Kumar said. “We’re not seeing that happening, because the impetus is to push for much longer lifetimes.” 

Steingart suggested that the solution may lie in the mining majors’ entrance into the recycling business. He emphasized that if batteries are going to be produced at the scale required, certain raw materials will be more in demand than ever before. Depending on the battery technologies that gain traction, he added, it’s possible that society “will have to extract more copper in the next 15 years than we’ve done in the last 3,000 years.” Steingart noted that both mining and recycling will be necessary (and recycling won’t be sufficient on its own) to support the energy transition’s storage needs. 

For this reason, Steingart believes, mining majors will need to expand in the coming decades from primary expertise in raw material extraction to recycling as well. 

Meng agreed: “Recycling and mining go hand in hand,” she said. “If you want to achieve true circularity, you have to think about the process starting from the moment the atoms are taken from the earth and consider how they can perpetually cycle through the industry.” 

Key Point No. 5: AI will both spur the need for new energy storage solutions and help devise new solutions. 

Workshop participant Paul Jacob is CEO of Rye Development, which helps develop utility-scale energy storage projects, with a particular focus on pumped storage hydropower. He shared that as he travels the country and meets with representatives from utilities, he’s increasingly hearing that they’re receiving interconnection requests that exceed the size of their entire utility system because of new AI data centers. Since many of the tech companies behind the new data centers have published net-zero goals and want to be associated with renewables, Jacob said, this type of spiking demand could drive interest in the sort of long-duration storage his company develops. “That could be the key to driving this,” he added.  

Still, the grid in the United States presents a tangle of challenges and necessary updates that will need to be addressed before it can truly sustain a clean energy system. Ken Caldeira, Senior Staff Scientist (Emeritus) at the Carnegie Institution for Science's Department of Global Ecology, Senior Scientist at Gates Ventures, and Visiting Scholar at the Stanford University Doerr School of Sustainability, highlighted the disconnect between how utilities model demand resources and how grid operators actually dispatch them. “For example, pumped storage hydro is often used as a peaking resource or to provide inertia during ramping periods, cycling multiple times a day,” he explained. “Yet, utility models typically treat it as a simple two- to four-hour demand resource. This disconnect makes it difficult to plan for the integration of renewables and long-duration storage.” 

Caldeira said we need better technology and policy to modernize the grid and make it capable of handling the demands of a clean energy future. But perhaps developments in AI can help there too, some workshop participants suggested. 

Kumar argued that the impact of autonomous vehicles on battery technology hasn’t yet been fully appreciated. If a fleet of robotaxis, for example, are on a city’s roads constantly, “your fleet starts to look like it’s managing the grid,” — and maybe a smart grid of the future, which can monitor and manage flows of electricity, can finally start to emerge. 

Learn more about Columbia Business School’s Climate Knowledge Initiative