Battery Gold Rush

The US is expected to install 24,000 MW of new utility-scale battery capacity this year, which is a 60% increase over installations last year. This is almost double the global growth rate. FEOC, tariffs and local concerns about fire risk do not seem to be holding batteries back. What is the key to separating spreadsheet optimism from what is actually bankable?

Five battery company CEOs talked about these are other topics at our 35th energy finance conference outside San Francisco in mid-June. The five are Brian Hayes, CEO of Key Capture Energy, Bailey McCallum, CEO of Goshe Energy Storage, Julian Nebreda, CEO of Fluence, Chris Taylor, CEO of GridStor, and Andrew Waranch, CEO of Spearmint Energy. The moderator is Caileen Kateri (Kat) Gamache with Norton Rose Fulbright in Houston.

Rapid Growth

MS. GAMACHE: Last year, the main question for the battery storage panel was whether storage could survive FEOC and blistering tariffs. A year later, the question is whether we are in the midst of a battery gold rush. Chris Taylor and Andrew Waranch, tell us about your past year and whether you are screaming "Eureka!"?

MR. TAYLOR: There has been quite a change in the past year. We have seen the headwinds from the Trump administration moderate. There are still headwinds, but the tailwinds of exponential growth in demand and the need for capacity that can be deployed quickly are more than offsetting the federal policy headwinds.

We have grown our company by about a third in terms of headcount in the past year. We will more than double our operating projects this year and expect to do that again next year, and we are seeing a lot of opportunities in the M&A market. We just closed our fifth acquisition in the last year and a half, and we have about 3,000 MW of projects in advanced stages or under construction. We see a strong growth trajectory for the next few years, driven by the fact that storage is the leading source of near-term capacity.

MR. WARANCH: I agree. I describe us as post-hype. In 2020, batteries were the hot new thing. You had a lot of time, effort and money thrown at them, which was great, but projects take three to 10 years to get through the queue and there was regulatory uncertainty.

As we get into 2026 and look ahead to 2027, we see an enormous number of projects coming out of the interconnection queue. The sizzle has passed. We are in the real steak. There is going to be a solid construction boom for the next decade.

MR. NEBREDA: I agree. I wouldn't call it a gold rush in the sense that we are looking for a mine. We already have the mine. It has taken a tremendous amount of work to change our supply chains and bring them to the US to be able to deliver batteries that pass muster under FEOC and to get a new product ready that can adapt to data centers and the speed and other things that they need. We found the mine. We built it. I think it is ready now to start producing.

MS. GAMACHE: Brian Hayes and Bailey McCallum, welcome to the club. How are you feeling about the industry outlook?

MR. HAYES: Very good. I have been at Key Capture for two and a half years. Until about a year ago, my kids would ask me how things are going, perhaps out of nervousness.

Things have changed in the last year in terms of demand. Utilities are recognizing how battery energy storage fits from an accredited capacity standpoint. That alone has unlocked tremendous opportunities.

Time frames have also changed. On the supply side, the battery manufacturers have responded to FEOC concerns and have been good at working through them.

MR. MCCALLUM: Any question about a gold rush suggests there may be an element of irrational exuberance. Are we in a battery gold rush where we are overbuilding? I don't think we are because of the escalating demand for capacity. We are the maker of picks and shovels and supplier to that gold rush for capacity. We are in a really strong spot as an industry, bringing a lot of projects out of the queue. I expect to see a further maturing in a significant way over the next year and a half.

MR. TAYLOR: Many of us are experiencing a shift away from heavy reliance on merchant revenues from arbitrage and ancillary services, which was the dominant revenue source in the early days in both ERCOT and to a lesser extent CAISO, to long-term contracted revenue, whether that is through investment-grade utility offtake or well-capitalized hyperscalers. That creates a much more stable foundation for long-term growth and investment than the more market-facing merchant model.

MR. HAYES: I view storage as a service for the grid. If storage continues to try to go after arbitrage, we will kill ourselves. Storage is needed for the grid to operate reliably.

MR. WARANCH: I gave a lecture for more than 25 years that power is the most unique commodity because you have to produce and consume the same amount at the same time. What is lost in some of the minutiae is that the entire power trading industry changed with storage. We used to be all about probability of $5,000 spikes and running out of power. You don't have that in corn because you have silos. You don't have that in metals because you have warehouses. Power never had that.

We are in the midst of a transformation across every ISO, every utility and every corner of America. Storage is needed everywhere, and it will require at a least a decade-long buildout, regardless of everything else that is happening.

Customers

MS. GAMACHE: Brian Hayes, are your counterparties the grid?

MR. HAYES: The utilities are usually the contracting entities, but it depends on the project. There are certain state programs, like New York which has an index storage credit, where the state is effectively sponsoring the project. Illinois is coming out with its own index storage credit.

Data centers come in with large load requests. They go to the utilities, and then the utilities effectively act as brokers.

MS. GAMACHE: Bailey McCallum, with whom are you contracting?

MR. MCCALLUM: The same general parties. We started as a business focused on the arbitrage model, but I don't think that is held against us in this market in the way it might have been in renewables in the past. We have projects that are both in operation and under construction that are able to play in that market. Conversations about such projects are able to move much more quickly than others.

MR. NEBREDA: The biggest recent change for us is we are getting direct demand from hyperscalers and developers for behind-the-meter storage. Two quarters ago, we had very little in that pipeline. On our last earnings call, we reported 37,000 MW of battery orders and that pipeline is growing quickly.

MS. GAMACHE: By show of hands, how many of you are working with hyperscalers? [Pause] All of you. I have been surprised to see hyperscalers turning to gas where I would expect them to use storage.

MR. NEBREDA: The ones that are using gas are not yet well educated about storage. If they are looking for capacity, there is no cheaper, faster and better source of capacity. If they are looking for electrons, that is a different story because we do not produce electrons. We store them, manage them, provide them, shape them and make them do what they need to do.

Part of the big demand increase is due to data center developers realizing what they really need is capacity.

MS. GAMACHE: Are you focused on regions that have capacity markets?

MR. NEBREDA: Our customers may need capacity because the local utility is asking for capacity. They need the ability to disconnect from the grid. They can get the electrons from renewables or other sources. There are different reasons why they may need capacity and not electrons.

MR. TAYLOR: The first speaker today had a slide showing the impacts on gas turbines that can be harmful to the turbines if you don't have something like a battery attached to address the extreme oscillations. It was an AI training model. That creates opportunities for companies like ours. There is a pretty short list of people who have gas turbines ready, that have gas supply, that have permits, and are also experts in battery energy storage systems. Raise your hand if you are such a company. [Pause] We have one back there. Companies that are well-capitalized and well-positioned to deliver on the gas component are going to look to partner with people like us who can deliver the battery storage component.

MR. WARANCH: Trying not to get too deep into it, the voltage ride-through issues that ERCOT has pointed out are real, and we are going to solve that problem. The frequency response issue and other ancillary service issues are real. They are just coming to the forefront now.

Data Centers

MR. NEBREDA: There will not be a data center built in the US without battery storage. How many batteries they need is a different story.

MR. TAYLOR: There has to be a more robust partnership with the hyperscalers. They are going to have to disclose more about what is behind the curtain for us to deliver solutions that will work behind the meter. I worked at a hyperscaler for seven years. They do keep that stuff very close to the vest, even for those who work there.

MS. GAMACHE: Connect some dots for me. Julian Nebreda said every data center will need battery storage. Yet, as we heard from the last panel, perhaps 40% of data center projects projected to be completed this year will suffer delays or be cancelled. How do you deal with that risk when working on a project with a data center developer?

MR. WARANCH: If you have 370,000 MW of data centers coming to ERCOT and 40% are canceled, we will be fine.

MS. GAMACHE: Okay, but what are you putting in your contracts with data center developers to protect yourselves? Are you working with creditworthy data centers? What does the credit look like?

MR. TAYLOR: You will have contractual provisions for dry-hole risk if you are the developer and you are advancing capital to support a project whose only reason for existing is because of the offtake. The hyperscaler has to have some skin in the game alongside you. This cannot be a one-way street. It has to have the capital, the wherewithal and the willingness to do that.

That is a viable approach with the well-capitalized players. There is enough demand from the top of that stack that I do not see much reason to dip below the top of the stack in terms of the best capitalized creditworthy counterparties. That does leave a lot of the market unaddressed. Maybe other people are chasing those opportunities.

MR. MCCALLUM: We are finding that if Google or another hyperscaler is not the name on the term sheet, it is probably backing the group that is there and willing to provide maybe not a 20-year backstop but some period of time that is enough to get your capital returned and have a satisfactory project.

MR. NEBREDA: A significant number of developers are developing data centers with the view that they will sell them. They are the ones that will have problems. That is where most of the cancelations will occur.

MR. WARANCH: The type of letter-of-credit backstop support required is evolving rapidly, and it will be probably be different again six months from now. We spoke to a hyperscaler recently to whom we said we need a $100 million backstop LC to do what it wants. No problem, it said.

MS. GAMACHE: We are seeing collateral covering only about a year or two of expected revenues in contracts that are under negotiation currently. I am concerned about what will happen when those projects go to market for financing, even if a Google or similar entity is on the other side. Another challenge is data centers usually only want to sign contracts with five-year terms. How are battery storage projects with such contracts getting financed?

MR. TAYLOR: I don't think five years is financeable. You need a longer term. The big hyperscalers have dozens if not hundreds of people negotiating PPAs every day. They know what is financeable in my experience.

They understand the difference between using private equity on balance sheet to build something versus a blended cost of capital and what that does to their pricing. They would prefer not to pay the higher price.

MR. HAYES: We usually see tolling agreements in the 15-year-plus range. The interesting part is some hyperscalers need the electricity so badly that they are willing to pay enough for a short-term bridge deal that it covers the battery cost.

MS. GAMACHE: Are the hyperscalers doing tolling agreements?

MR. HAYES: Last year, it would have been a contract solely for capacity, but in the last six months, they have been more open to tolling agreements, especially in areas where a toll works for accredited capacity and they know they need it.

MR. WARANCH: They are a lot more focused on optimization services about balancing power from the grid, on-site generation and load because they don't have that skill. They want someone else to fix it for them.

Biggest Risk

MS. GAMACHE: We are in a world with lots of change-in-law risk. What is your biggest concern today: FEOC, change in tax credits, tariffs, or something else?

MR. TAYLOR: Interconnection is a way bigger fear of mine than change in law. We see the more sophisticated utilities that have significant procurement experience baking in some pretty reasonable two-sided protections for change-in-law risk that have been approved by public utility commissions. We also see less sophisticated buyers that frankly decline to share in any such risk. Maybe they will find someone who will do that, but it won't be us.

The interconnection backlog that other panels have talked about extensively is hands down the biggest challenge for any of us in this room, no matter what technology we are trying to deploy.

MR. NEBREDA: The rules are going to get more stringent on allowing equipment, control, software and cyber security from countries that the US government does not trust.

We are integrating AI into the way we work. As that technology plays a more important role in ensuring that the lights remain on, the security establishment is going to be a lot more worried about involvement by bad actors.

We expect to see a lot more trade barriers and higher tariffs over time. We are working full-time to build out our supply chain in the US all the way down to the mining of minerals.

MR. WARANCH: If we really want the load growth that we all think is coming, the NIMBYism and the community pushback, a lot of it based on ignorance, can be brutal and can delay projects that you think are nearing the goal line. Community pushback and NIMBYism are my biggest fears.

MR. HAYES: Developing storage projects has become tougher because of fires and unpopular data centers. Storage developers can separate themselves from data centers on the ground game and how they do permitting.

MR. TAYLOR: There are simple, viable solutions to the legitimate concerns about cybersecurity. It is all about the energy management system. It is not the cells. It is not the hardware. Having been to China and visited some of the leading manufacturers there, I don't think we are going to catch up to them anytime soon on cell production. I would take the long side of any bet on that.

MR. NEBREDA: I will take the other side.

MR. TAYLOR: But I think domestic EMS solutions are available, and they could be better than what we are getting from China.

MR. MCCALLUM: NIMBYism is also my biggest concern. The driver of demand for our industry is the growth of large data centers. If that growth were to slow dramatically due to community opposition and regulatory barriers, that would be challenging, not just for us but also for the economy as a whole.

MS. GAMACHE: Andrew Waranch, I understand Spearmint started -- or is backing -- a lobbying group. Is the group focused on NIMBYism or on government policy more broadly?

MR. WARANCH: Alongside a lot of people in this room, we helped create a group that is now known as Storage Powering America, which was really influential over the last couple years in working on some of the statutory language for investment tax credits and general education about battery storage. We had another group in Texas called the Lone Star Energy Storage Alliance. The focus is mostly on educating legislatures about the benefits of batteries.

I never thought that would be our responsibility to do. There are a dozen or so companies that were willing to join together to back the effort. There have already been tangible results from the education effort and advocacy.

MR. TAYLOR: We have been partners in that effort with Andrew's company, and I agree with what he just said. State level advocacy is important. We saved the day in Texas by defeating proposals in the state legislature basically to legislate us out of existence. I flew down to Austin last year. We just passed a constructive safety bill in Oklahoma. I used to be a state house lobbyist before I became a business guy. Your chances of winning at the state level are infinitely better than at the federal level, especially if you have your act together and are well organized. It is not beating your head against the wall like dealing with Congress. The cost is dramatically lower, like orders of magnitude lower.

FEOC

MS. GAMACHE: Chris Taylor, you mentioned going to China. Andrew Waranch, you talked about going to Thailand. What diligence are you doing to get comfortable with FEOC?

MR. TAYLOR: We are buying domestically manufactured batteries.

MR. WARANCH: Where are the minerals mined and refined?

MR. MCCALLUM: I'm relying on Julian.

MR. NEBREDA: Exactly.

MS. GAMACHE: Julian, are you representing that there are no FEOC issues with the batteries that you supply?

MR. NEBREDA: We are confident in what we are doing. The law is clear enough on what you can and cannot do. As long as you do not try to be too clever and rely on good lawyers like we have here, you will be fine. We are confident about what we are offering.

MR. WARANCH: What scares me the most is we were at an American Clean Power Association meeting a few weeks ago and met with a bunch of cell manufacturers. Each said it is FEOC compliant, but none of them is. You have as much as 90% of the minerals coming from China. Not everyone can be compliant if 90% of the inputs are coming from China, but a lot of people claim to be. Verification is thorny. I got in my email box this morning emails from multiple accounting firms verifying claims. There is only so much a US developer can do. You can hire people in China. You can look at satellite imagery. It is really tough.

MR. HAYES: The supply base has narrowed, at least for us. In the recent past, there were probably 20 to 30 suppliers that you could try to visit in China and elsewhere. Now there are about five that we would actually consider.

It is all about understanding the FEOC risk and then putting the onus on the supplier to explain how it will be addressed. We also look for suppliers that have the ability to manufacture in places like the US and Thailand and not just in China to give us a way to mitigate the effects of sudden changes in tariffs.

MR. TAYLOR: I second what Brian just said. That's pretty much our approach as well. We want agreements with manufacturers that have global reach and that are standing up domestic manufacturing as well. The policies are going to keep shifting. You need to be nimble and be able to respond to changes.

MS. GAMACHE: To what extent are offtakers willing to take tariff and other change-in-law risks?

MR. HAYES: It depends. Some offtakers will take tariff or certain other types of risks. We would love for them to take lithium risks. They will not.

MR. WARANCH: Some of the risk-sharing provisions are very prescriptive.

MS. GAMACHE: My understanding is that lenders will take zero change-in-law risk. I think interconnection costs go into that bucket as well. True?

MR. TAYLOR: Lenders have so far been less flexible than offtakers.

Long Duration

MS. GAMACHE: We are nearing the end. Let's ask the audience whether it has any questions.

MR. GREEN: Bill Green, managing partner of Climate Adaptive Infrastructure. In the face of all of the optimism, how do you think about what, from the outside, appear to be fairly credible advances in longer-duration energy storage. How will such storage affect the economics of projects you are completing today? Do data centers want eight to 10 hours of storage or much longer duration?

MR. NEBREDA: I think of long-duration storage like a form of fuel tank. Our best products today are not fuel tanks. They do not do the work of a fuel tank. They are electronics connected to a chemistry that can send and collect electrons very quickly from the grid. They can do so efficiently with very little conversion loss in energy. That allows you to move the peak time of use. It allows you to manage voltage.

The fuel tank technologies that are coming are great. They can store energy for 100 hours, but they lose 30% of the energy, and they cannot respond quickly. They need to charge slowly and discharge slowly. They play a different game. We are technology agnostic as a company. We would like to start offering a long-duration storage alternative, but it is a different business model.

Data centers need the rapid response times that the current products offer. It is not clear yet whether they will need long-duration storage. The grid can serve the same purpose without having to put storage equipment at the demand source.

MR. WARANCH: I used to be opposed to long-duration storage because I didn't see the business model. In many ways, I still don't. I remember being up here a few years ago talking about flow batteries. If we stop building wind as a country, off-peak prices will increase. If you look at where ERCOT had a real problem this winter, it was on a cold Tuesday night when the wind didn't blow and all the batteries had dispatched in the evening peak and they were empty through the overnight valley, and they were not there in the morning to dispatch for the morning ramp.

If we do not build any more wind and we build peakers instead of combined-cycle gas plants, that overnight valley becomes very shallow. It is hard to recharge a battery economically and be ready for that 6 a.m. to 8 a.m. peak. The lack of wind is what will push wintertime battery need toward longer duration, but it is artificial because of government-imposed barriers to wind.

MR. TAYLOR: In cases where the need is for up to eight hours of storage, lithium-ion batteries are competitive. I don't see much competition or need for other technologies in the up-to-eight-hour storage market. When you get out beyond eight hours, then lithium-ion becomes less competitive, but I am not convinced how many gigawatt hours of more-than-eight-hour storage are truly needed. That use case is still small compared to the up-to-eight hours. It exists for sure and it will grow over time, but it is small today.

MR. MCCALLUM: I agree with Julian. Long-duration storage is a different business model. We have some offtakers say to us that they like the fact that we can respond quickly and provide everything that a fast-responding battery does, but they also want to explore different, longer-duration technologies. There is an opportunity to work together as partners because that long-duration storage still needs a site. It needs an interconnect. It needs everything that we already have available. We see some of that as well.