The data center power crunch: “Physics is going to win” — Robert Gaudette, President & CEO, NRG

Let's Talk Energy — Rob Gaudette, CEO, NRG Energy Show: Let's Talk Energy (Rystad Energy) Host: Noah Brenner (NB) Guest: Rob Gaudette (RG), President and CEO, NRG Energy Recorded: Sidelines of Rystad Energy's 2026 Americas Summit, Houston, Texas Runtime: approx. 31 minutes Timestamps: every 4 minutes Speaker labels: full names in the introduction, initials thereafter 0:00 Rob Gaudette (cold open): Ultimately, over time, you need to connect all of that load together. The reason you do that is that when you create a network of generation and load, you can absorb the shocks and the bumps in the road a lot better than you could by having a handful of islands trying to do it on their own. [Music] Noah Brenner: This is Let's Talk Energy, your go-to podcast for smart energy insights. I'm Noah Brenner, joining you today from the sidelines of Rystad Energy's 2026 Americas Summit here in Houston, Texas. US power demand has been essentially flat for decades, but recently trends like re-industrialization, electrification and the rise of AI-driven data centers have caused projections for US electricity needs to soar. Rystad Energy sees US demand growing at about 2.5% annually — up from less than 1% historically. This is straining grids, the market and the pocketbooks of US consumers, and it's putting power producers at the center of some of this year's hottest election issues. To give us an inside look at how these trends are playing out in real time, I'm joined today by Rob Gaudette, CEO of NRG. Rob, welcome to the program. Rob Gaudette: Thanks for having me. Really appreciate the time. NB: All right, let's talk energy. Now, NRG is not the same as a regulated utility, and I don't know that all of our listeners necessarily understand the differences between those two types of companies. I was wondering if you could place NRG within the landscape of power producers — and while you're at it, give us a glimpse into your journey to the CEO's office. RG: Oh wow, there's a lot there. Let's talk about where NRG is first. When you think about the energy industry, especially from a power perspective, you've got two kinds of companies. You've got regulated utilities — those are folks who have rate bases, and how they make money is determined by their PUCs. They're long-standing, and they basically have a monopoly. Then there's the competitive energy space around power. In that case, those companies have to compete for their customers every day. You tend to see more innovation, you tend to see a lot more competition. NRG is in that space. We are a competitive energy company. So I have to win my customer every day. I have to fight to find something new every day. And we bear all of the risk of what we do. For me — and we'll talk about it in a second — that's the exciting part about NRG: we are always working to better serve our customers, and we always have to work to be a bigger part of our community. So we work really hard to get there. As to how I got here: I'm from Houston. I was in the military before business school, and then I started on the wholesale side. I was a trader in both power and natural gas, then moved up through the organization for 10 or so years. Then I made the move to retail and started running businesses that serve business customers. So I really got to understand what a customer needs and what innovation can do for them. I got the opportunity to take what I learned on the wholesale side and apply it on the retail side, and five years later I'm sitting here. I have the opportunity to lead 17,000 employees, and we lead across multiple spectrums as a company. One more difference is that we're not just one piece of the equation. We have consumers, so we have homes as customers. We have businesses as customers. We have our smart home business — home automation and security. And then we have the wholesale side, the side that I started in: big power plants, 25,000 megawatts. 4:00 RG: So at NRG it's 8 million customers, 17,000 employees and 25 gigawatts of power generation. It's an exciting day every day. NB: We're going to come back and talk more about competition — I can see you light up as you're describing the role of competition in power and power markets. But first I want to talk about NRG's "bring your own power" initiative. This has been a major focus for you going forward. BYOP — how is it different from the approaches of some of your competitors? And what is your argument that a BYOP approach is the best approach as the nation looks to provide the power that's needed for everything from electrification to data centers to just general economic growth? RG: Electricity is the fabric that society is built on top of. You need electricity for every business, every home. It's important, and it's become much more important over the last few years with re-manufacturing, EVs and data centers. To your question on what BYOP is: when you think about how you would supply a customer, one option is you just plug them into the grid. That works, and it's worked historically over time, but it creates strain on the system. It ultimately raises prices over time, and it can create disparities around who's using power and who's paying for what. The second option is you just plug a big load into the back of a power plant. Some people have done that. You've effectively retired a power plant by doing that, and that creates problems around price and can create actual reliability problems on the system. And then you have our approach, which is bring your own power. What that says is: if you're a really big load — a very large industrial or data center load — you need to be a good grid citizen. So you bring your own generation. The product that we have, they're bringing state-of-the-art CCGT — combined cycle gas turbine — which provides enough power to power their facility and contributes to the grid. It enables our customers to walk into a community and, in addition to creating a tax base and a large load and an employer, not do damage to their power system. You're not creating problems around reliability. With our product, BYOP, we've been thinking about that the whole time. Remember, I said 8 million customers — I think about their needs too. As we designed this product over the last couple of years, it became abundantly clear that these large loads want to, and need to, be positive grid citizens and positive contributors to their communities. And we've got a great opportunity with what we're doing to help them do that. NB: I want to talk more about that idea of being a positive grid citizen, and about the impacts of these large loads on those 8 million customers you have. I'm not going to single out individual states — states are taking different approaches, and they're going to do what's best for their constituents. But I was honestly floored by some of the comments from Texas Governor Abbott. I'll just quote: "They basically dug their own grave for the problem that's been caused for them, and that's why they got the backlash they deserve." Those are pretty harsh words coming from a very business-friendly governor. How do you respond to that criticism of the data center industry, and by extension the people who are going to be helping them bring additional power? RG: Rather than how we got there, I focus on how we move forward, how you get beyond it. We have always focused on what the right way to do this was, and the product that we have happens to be that. 8:00 RG: So we're focused on helping our customers, or potential customers — the data center guys — think it through. Look, these politicians and elected officials are protecting their states. I empathize with their needs. Their voters are my customers. I get it. The product that we have, the thing that we're offering you — you can contribute from an affordability perspective. You can protect reliability. Our design is smart from a water perspective, which is important in some states like Texas. And by the way, we've been in these communities for decades. We have employees who live here. We know what a community needs and what engagement looks like. That's why you should partner with us, NRG. That's a message that resonates with our customers, and it was one we had two years before it became a thing. So we find ourselves in a really good spot, where we have the ability to serve these big customers while also protecting and making sure the communities we're in thrive. NB: Now I want to talk about how that BYOP model fits into your broader strategy and the direction you want to take the company. Your first BYOP project — I believe it's 1.4 gigawatts? RG: It's 1.2 gigawatts. NB: 1.2 gigawatts. And you've estimated that by 2030 — or if everything stays roughly as is, 2029 — RG: End of 2029. NB: — that's going to account for, I think I said, $500 million in annual earnings and $375 million in free cash flow. I think that's about 10% of your total free cash flow. And if you take that out to the 5.4 gigawatts that you say you've got in an achievable pipeline, it's a pretty big chunk of your business. How do you think about that much exposure to an industry that's pretty new? And is there a plan B if it turns out these data centers need less power than we think they do? RG: There's a lot in there, so let's take it step by step. My job as CEO, and ours as a company, is to make sure we're investing the capital from our shareholders the smartest way we can, and we're trying to create long-term cash flow. We're going to create more cash flow, but we're also trying to get long-term cash flow. We've been very careful about the product that we're creating. This is something that's got a 40-year life. We've got the design down so that it makes a lot of sense, and we're only going to do the deals with investment-grade counterparties with contracts. So I protect our shareholders from that perspective. The thing I want to make sure you walk away with, though, is that that's not the only thing in our business. We continue to develop products around our consumer business. We continue to serve lots of other commercial and industrial customers — that's our bread and butter. And we're one of the largest owners of generation on the continent. All of those things combined get us to one place: we are going to serve our customers and the communities that we're in to make sure everybody can thrive, and we're going to do that in the right way, while protecting the interest and the value that I've been entrusted with and the company's been entrusted with. NB: I want to talk about some of that phenomenal growth we're seeing in power demand. It's certainly not the same as we've seen over the last 20 years. By your own analysis, here in Texas, ERCOT peak load is projected to increase by about 40 gigawatts — RG: Mm-hmm. NB: — between now and 2030, a pretty big amount, with limited firm supply additions. And at the same time, PJM — the power system that serves where I'm from up in Pennsylvania, as well as parts of New Jersey and a large chunk of the eastern US — 12:00 NB: — is also facing a similar setup. Is this a demand super-cycle, or is some of this load perhaps more speculative? Are we going to see this massive wave of growth that people are saying? RG: I would say yes — it is a demand super-cycle, there's no question about it. I've been in the business for 25 years. Over that time period, growth was about 1% on a good day, at least for the first 20 years. Then you saw Texas kind of explode as industry moved down here, and then you've seen it across the country. More electrification has occurred. Manufacturing's coming back. You have EV mandates in states. Homes use more power than they used to. All of those things already create growth in the sector that's kind of unheard of, at least not in my time period. And then you add data centers. So I would say we have a demand super-cycle — I wouldn't question that at all. When you see some of the demand forecasts for data centers, though — in ERCOT, there's a number out there that says 500 gigawatts. That's on a system where 91 is the most recent peak. That's not going to happen. Physics is going to win. When you think more rationally, 40 gigawatts is probably the right amount. And in a system where you're talking 40 gigawatts on 91, that's a big growth period — but we have a lot of supply already in Texas and we can meet a lot of that. It also goes to BYOP. You already have these trends happening and demand moving up in an industry that's used to 1% demand growth. BYOP helps keep things under control, and our product is actually 1.2 gigawatts of supply for every 1 gigawatt of load. So you are contributing to the market and making it better. NB: Can the power industry as a whole keep up with that? If we're really getting 40 gigawatts of growth here in Texas, can the industry keep up — across engineering, turbine manufacturing, availability, construction, permitting? Do you have enough electricians? Where do you see the biggest constraint? As we talk about the limits to this growth, is the constraint we should be looking at actually the industry's ability to build? RG: I would say every constraint you listed is real. There is a turbine constraint. There are high-load switchgear problems. There are definitely labor constraints — EPCs, the guys who build things, are completely booked, and getting enough electricians and welders and all of those things are going to be challenges. Will we find a way? We always seem to find a way. Constraint usually drives the thing that I love about competitive markets, and that is innovation. You're going to need every megawatt we've got, every megawatt we can build. But we're also entering into a world where load management — helping customers use less electricity when it's needed — is also a big and important part of the solution. And I'm proud to say that NRG is a part of that too. NB: I want to come back to load management in a little more detail in just a second. But as we're talking about this wave of behind-the-meter generation that's being built today — 16:00 NB: — these are assets with 20-year lives, and the grid environment could look quite different as we move through. We're kind of at the beginning of this super-cycle we've talked about. Do you see this push for behind-the-meter power as more of a bridge to grid connection, or is it really a shift in how we think — as a country, as an industry — about how to serve these types of facilities that require a lot of power? RG: I'm going to have a view that's different from others. NB: Good, means we've got something to talk about. RG: Exactly. So you call it behind the meter. I call it islanded. NB: Okay. RG: Basically, load that isn't connected to the grid and has its own power source — and we've got a design that can do that too. But ultimately that load and that generation source need to get connected to the grid. Ultimately, over time, you need to connect all of that load together. The reason you do that is that when you create a network of generation and load, you can absorb the shocks and the bumps in the road a lot better than you could by having a handful of islands trying to do it on their own. You actually strengthen the grid more generally, and you protect that customer and that setup better as you get connected to the grid. So our strategy is to create a product where you can be an island if you need to, while always looking for the opportunity to ultimately get connected to the grid and provide a service to the grid, while also extracting the value of being on the grid. I think everybody gets connected eventually. NB: Okay. And under that scenario, do these assets then participate in wholesale markets and things like that? Is that eventually how the industry shapes up? RG: If you think about it over time: when you start with an island, that generation has to run all the time. It's running to power that data center or that large load. Once you connect to the grid, then you're making economic decisions. There could be a very windy day in Texas, or a very sunny day in Texas, with lots of renewable power and low power prices. That data center should benefit from that. That customer should benefit from that. And in that case, maybe that CCGT is not needed at that time. So the stuff that gets built on the island needs to be flexible — it needs to be able to turn off and on. Ultimately you're solving a better financial and physical equation if you get connected to the grid, for the customer and for the market. And you also get the oomph of that generation: if there's no wind and no solar, the data center can still run and not create damage or an impact on the market, because you've got that big generator next to it. NB: I said at the top of the discussion that we were going to come back to this idea of competition in the power sector. There's a school of thought that says deregulated power markets are broken — that they're not performing in the way they need to in order to give companies like yours the signals that are needed to provide the power that it seems like we're all going to be demanding. The concern has been particularly big in PJM, like I said, where I'm from, where it fell short of its reliability requirement for a couple of straight auctions. Are competitive power markets working as they should? And if they're not, is it a structural problem — that they're not fit for the environment we're moving into — or do we just need to tweak them at the margin? RG: I'm completely biased — I'm a huge competitive markets supporter. The reason is because they always put the risk to the right people. 20:00 RG: As a company in a competitive market, I make a decision and I wear that risk. In a regulated market, somebody makes a decision, it's a bad decision, and the rate base wears that risk. So I believe that risk transfer is an important part that nobody talks about in the competitive markets. The second piece is that in competitive markets you can move much quicker on things that are more innovative. Things like load management, demand response, peak shaving — all of those came out of competitive markets. They're developed and processed and built in competitive markets, and then sometimes they translate over to the regulated side. When you ask whether competitive markets can meet the needs, or do what we have to do, I firmly believe they can. The regulatory structures, depending on what state and what market you're in, are going to drive people to the right decisions. Here in Texas, we have a very competitive market and no capacity price, but we also have mandates and statements from regulators and politicians that say if you want to come, you need to be a good grid citizen. Well, in my market, I know exactly how to do that. And I know that with our product up in PJM you have a three-year view on capacity. But what they're seeing is, you know what, Rob can't build on a three-year, one-price signal. So they're creating things like the 15-year auction. Those kinds of solutions get developed, and then you can put my capital at risk — not the rate base, not you in Pennsylvania. We put our capital at risk to go build things and solve the problem. Looking at one year and evaluating whether or not competitive markets work isn't fair. Let's look over the last 25 years and look at how those markets have scraped out efficiency and kept prices low for two decades compared to where they would have gone over time. Then let's assess what makes the most sense, and who really bears the bad-decision risk — because I think that is a key attribute that a lot of people don't think about. Now, I am different from others because I'm on both sides of that equation. I have the customer who doesn't like the prices they're seeing, and then I have the generation side. So our perspective has always been that we need a system that works. We need market rules that work for both parties. Let innovation and competition drive — and that's been successful for a quarter of a century. NB: So innovate on the model; don't throw the baby out with the bathwater, if we're going to use a generalization there. But I want to come back to this idea of customers. I was watching — you were on a podcast earlier this year, the Electric Power Supply Association podcast — and you said something I thought was really interesting. You said what we can't have is an affordability crisis. Unfortunately, I think many people might say we already have an affordability crisis. But at the same time, we can't build additional capacity to meet the demand that we have if it doesn't make sense financially. So can the US power sector grow its generation sufficiently at prices lower than they are today? RG: They can if they're given longer-term certainty. If you want to build a power plant and you're only going to get one year of certainty, the price goes way up. But if you said 15 years, or 20 years, now that investor, that company, has a profile they can work against, and prices go down. If you create market signals that are not just price but also term, then people can go build things. That's the "build the big steel" — getting real power plants built. The thing I would tell you, though, is that the price signals today — especially where you live, up in PJM — help drive innovation around things like load management. They start inspiring commercial and industrial customers to — 24:00 RG: — do demand response again. When the price of capacity is north of $300, it's much more interesting to think about how I operate my manufacturing facility than when the capacity price is worth $30. So the best answer for high prices is high prices, because it drives innovation, it drives investment, and ultimately it drives behavior that pushes prices back down. That's the beauty of a competitive market. NB: You've mentioned demand-side management a couple of times and I want to dive into that in a little more detail. For those who aren't necessarily familiar with the concept, it's basically: use less power at times of high cost and you get a break on your bill. Where are you as a company, and where are we as a country, in applying demand-side management and actually seeing the full potential of all the different ways we can do that? RG: As a company, we're the largest demand response provider in the country. NB: Yep. RG: Through acquisition and otherwise, we have about 7,000 megawatts of DR — demand response. I have 25,000 megawatts of gen; I've got 7,000 megawatts of the opposite. We developed a product called VPP that we tested down here in Texas. It had really good uptake — 10 times what we thought was going to happen — and now we're moving that into the east. That's working with consumers to adjust their thermostat — not turn everything off for five hours, but adjust their thermostat — to help them save money and ultimately provide relief to the market. The work we've done shows they can save 200 bucks a year. For some people, that's a lot. And then on the third piece, from our perspective, we're finding ways to marry up people who need capacity — think hyperscalers — with their ability to invest, to create opportunities and programs so that consumers and business customers can save money. So I believe we are moving forward on all fronts around demand response and load management as a company. As a country, there are places where it's being used and used well, and there are places that could use some work. Over time, we as a country are going to need every megawatt that we can build, every megawatt that exists today regardless of technology, and every megawatt that we can get to be managed — because that helps the customer, that ultimately helps the market, and by doing that you actually help the country. So I think we've got work to go. I think there are lots of really interesting tech solutions being built over time, but we've got way more runway in front of us on what that looks like. I think it's a real opportunity for the industry itself. NB: And as you point out, if we're able to bring demand-side management to these large loads, that just increases the size of the prize that we might be able to achieve. RG: Absolutely. NB: NRG's recent growth strategy has been really natural gas focused. As load growth accelerates and existing assets age, how are you thinking about the role of low-carbon technologies — solar, wind, geothermal, people are talking about SMRs these days? How are you thinking about their ability to meet the demand we're seeing from these large loads? And if you could weigh in: the UN, I think, had said something like every large-load data center should be powered by 100% renewable electricity. How does that goal look from where you sit? RG: Look, we have been proponents of renewable energy over time, in the history I've been at NRG. We have about 2,000 megawatts of solar and wind under contract, so we're believers in renewables. We are not the big bad — 28:00 RG: — we don't dislike any of that stuff. We are also in contact with all of those technologies. Assume that the leaders in geothermal, the leaders in SMR tech, batteries — we're talking to them, and we're talking to them on a contractual basis, not necessarily "I'm going to go build something." But I will take offtake, because I believe they're important parts of the portfolio. Every component in the system plays a role, but no component can play it by itself. SMRs — whenever you think those are going to be commercial — but let's just talk about nuclear generally: carbon-free energy, great baseload energy, not super flexible, but they have a role to play. Batteries are great, but they don't actually produce energy; they just move energy between time. Wind and solar are zero-cost resources, as long as the wind's blowing and the sun is out. All of them play a role. When you look at our portfolio and what we've done, we've invested in flexible natural gas-fired generation, because it's the thing that can fill in all the gaps and run on its own for a very long period of time if necessary. I'm trying to build a portfolio of generation that serves my customers, and by proxy you end up with the portfolio that we've got. I like what it is. I came from that side, and I think natural gas is going to be an extremely important part of everything we do as a country for decades to come. NB: In our earlier conversation you termed it "spinning steel," which is very apt. RG: That's a physics thing — it's important to have inertia in the system. It was very underappreciated, I think, until perhaps not too long ago. NB: Well, Rob, we've talked about competition, we've talked about demand-side management, we've talked about your BYOP model and how that could alleviate some of the concerns being felt about the growth of these large loads across the country. Is there anything we haven't talked about? Is there anything we should be asking? RG: I would say that I get out of bed every day knowing that I've got incredible people working on the challenges that are in front of us. The challenges in our industry are opportunities in our industry too, and they're pretty important. We are not a sideshow in this economy. We are the foundation the economy is built on. I have faith and I have the belief that there are incredibly intelligent people working to solve these problems every day. And the customers that I meet and the customers that we work with, I know are working with us to solve those same problems. I'm excited. The demand cycle in front of us is the real deal, but we've got the engineering, we've got the intelligence, we've got the really bright minds working on it. This industry is fun to be in again, and it's a great place to work. NB: Well, Rob, I really enjoyed our conversation, and I want to thank you for coming down here to join us on the sidelines of the Rystad Energy Americas Summit. RG: Happy to do it. Thanks for having me. [Music] NB: Thanks for listening to Let's Talk Energy. This podcast is a production of Rystad Energy, and today's episode is produced by Rystad and Lerner Productions. Check out the show notes for further analysis on the topics we've discussed today, and find us on social media — we're @RystadEnergy on all your major platforms. While you're there, give us a like, hit that subscribe button, and please leave us a review. You can also keep up to date with us on our website. If you'd like to comment on today's episode, or maybe you've got an idea for our next one, reach out to us directly: podcast@rystadenergy.com. And finally, don't forget to keep an eye on your feed for the next episode of Let's Talk Energy.