In this week’s episode, host Kristin Hayes talks with Daniel Poneman, a senior fellow at the Council on Foreign Relations and former Deputy Secretary of the US Department of Energy, about the role of nuclear energy in meeting increasing demands for electricity. As the use of artificial intelligence grows, so does demand for electricity, raising questions about which energy sources can provide reliable, clean, consistent power. Poneman discusses whether nuclear energy is a viable option, how the safety and performance of nuclear technology have evolved, and why some retired nuclear power plants are being revived—including the Three Mile Island nuclear plant in Pennsylvania, where a reactor meltdown caused by equipment malfunctions occurred 46 years ago as of last Friday. Poneman also describes what challenges the nuclear energy industry is facing, barriers to wider adoption of nuclear energy, and how public perception of nuclear energy has shifted over time.
Listen to the Podcast
Audio edited by Rosario Añon Suarez
Notable Quotes
- The energy density of nuclear power could help meet rising energy demand with a relatively small footprint: “One uranium fuel pellet is equivalent to 17,000 cubic feet of natural gas, 149 gallons of oil, or a ton of coal—in a fuel pellet the size of your thumb. People also don’t take into account that land is scarce. For a traditional reactor—not one of these fancy, new, advanced, small modular reactors, but a traditional 1,000-megawatt or one-gigawatt reactor—you need land the size of Central Park. For 1,000 megawatts of solar, you would cover Brooklyn. For 1,000 megawatts of wind, you’d cover all five boroughs [of New York City]. That is not a very practical solution in many parts of the world.” (8:19)
- The cost of scaling up nuclear energy is a challenge: “There’s no question that the nuclear industry has to perform better. But if you look at a big nuclear power plant, and it costs, say, $5,000 or $6,000 per installed kilowatt, and a natural gas combined cycle costs $1,000 per installed kilowatt, and a nuclear power plant takes at least five years to build, and a combined-cycle gas plant takes a year and a half or less to build … Put yourself in the shoes of a CEO, a CFO, or a board of directors of an investor-owned utility. That’s a hard case to make.” (11:10)
- Reviving old nuclear reactors can help meet energy demand and climate goals right now: “We need every jot and tittle of reliable power that we can get … [Reviving] any of these [existing nuclear reactors] is going to be faster and cheaper than starting from scratch with a greenfield and a piece of paper.” (29:08)
Top of the Stack
- Double Jeopardy: Combating Nuclear Terror and Climate Change by Daniel Poneman
- Washington: A Life by Ron Chernow
- Team of Rivals by Doris Kearns Goodwin
- Speed of Heat by Jeff “Skunk” Baxter
- Henry M. Paulson Jr.’s writings about biodiversity
The Full Transcript
Kristin Hayes: Hello, and welcome to Resources Radio, a weekly podcast from Resources for the Future. I'm your host, Kristin Hayes.
Today's episode is focused on nuclear energy. In particular, it's about whether rising demand for electricity—driven to a significant degree by artificial intelligence, or AI—is leading to a new nuclear renaissance. Joining me for this discussion is Daniel Poneman. Dan is one of those people whose bio, honestly, I really can't do justice to in the short space that I allot for introductions. But I'm going to choose a few highlights.
Right now, Dan is a senior fellow at the Council on Foreign Relations and a senior fellow at the Belfer Center for Science and International Affairs at Harvard's Kennedy School of Government. From 2015 to 2023, he served as president and CEO of Centrus Energy Corporation. And before that, he spent over five years as Deputy Secretary of the US Department of Energy, including a brief stint as acting secretary. He also served for six years on the National Security Council under Presidents George W. Bush and Bill Clinton. Dan has tremendous knowledge of the nuclear industry from both the technology and the security perspectives, and I'm really excited to have him on the show today. Stay with us.
Hi Dan, welcome to Resources Radio.
Daniel Poneman: It’s great to be with you.
Kristin Hayes: Thank you again for making the trek over to Resources for the Future on a very rainy day here. It’s so nice to see you in person.
I introduced you a little bit, but I'd like to invite you to introduce yourself a little bit more. I would also welcome some thoughts on how you ended up focusing on nuclear energy in your career.
Daniel Poneman: Well, like most things, it was serendipity. I grew up in Toledo, Ohio. I used to watch Lloyd Bridges on Sea Hunt as a kid, so I decided to be an oceanographer. Then I got to college and it turns out there's a lot of science and math. Meanwhile, I took an international affairs course that got me excited. So, I took a summer internship for my home state’s senator, in Ohio, John Glenn.
Kristin Hayes: I've heard of that guy.
Daniel Poneman: After three weeks working in the mailroom signing constituent mail with an auto pen, I asked for a research project. They said, "Sure, on what?" And I said, "Anything." And they said, "Well, why don't you look at the NPT?" And I said, "I'm on it." Now, this is 1975. There is no Google. There is no internet. It took me all afternoon in the library to find out NPT stood for Nuclear Non-Proliferation Treaty. I became very excited about this topic, and I still am.
Kristin Hayes: That's amazing. I'm always tickled by how many people launched their career based on one good professor or one good course they had at some point in their life. It really sets people on a course, but it's great that you're still enthusiastic about it.
I love to start conversations with some baseline information, so I'm going to ask you for that. As our listeners will have likely heard, we're in a new area of expected growth in electricity demand, driven in no small part by data centers needed to feed AI. That's the combination that we're talking about today. Let me start by asking you, What do we know about where that growth in electricity demand is going to happen and how quickly it will happen?
Daniel Poneman: It is going to happen in many, many places. So far, 40 percent of these data centers are in the United States, but they’re being built in Asia, in Europe, and people are looking at building them in the Middle East. It’s not only a question of where this demand will grow. The question of by how much is also very front and center. Just a couple of years ago, 21 gigawatts of power were consumed by data centers in the United States. It's now projected by McKinsey & Company that over 70 gigawatts will be consumed by 2030. This is a staggering number. Just to put this in perspective, 50 gigawatts is like 50 San Franciscos. One gigawatt is enough for San Francisco. So, the need for prodigious amounts of new power generation is just off the charts.
Kristin Hayes: It’s unlike anything that I've seen in my lifetime. So, there’s a good reason why people are taking this quite seriously—this question of where we are actually going to get this power.
I have another stage-setting question. This one's about the nuclear power industry, particularly in the United States. Can you remind us what percentage of US electricity currently comes from nuclear power and give a little bit of a historical perspective of how that has evolved over the past several decades?
Daniel Poneman: It has been quite steady. The amount of electricity provided by nuclear, traditionally, has been about 20 percent. Now, it's just under 20 percent. Even though, as a percentage of what they call the “installed base,” it's gone down, relatively speaking, as solar and wind have come onto the grid. But because nuclear power runs 24/7, its contribution is outsized. It produces more actual electricity than the mere nameplate of the facility would suggest. So, it's been 19 percent for a number of years. But even though nuclear power is 20 percent of our grid in terms of power generation, it’s about half of our carbon-free power generation. So, it's a very significant contribution.
Of course, now, with the AI revolution—and I should note it's not only AI—after decades of basically flat demand, we have this vertiginous growth driven mainly by AI, but also by the electrification of transportation. We all know about electric vehicles and industrial processes. Companies have been looking to find non-carbon-emitting ways to make steel, like Nucor steel, which has been looking at nuclear. Or to produce petrochemicals, like Dow Chemical. So, you not only have this direct increase in electricity demand for power generation, but you also have electricity demand to support other applications that had earlier been provided by fossil fuels.
Kristin Hayes: I'm going to ask you if you know one more statistic off the top of your head, which is what percentage of our clean baseload power comes from nuclear. I imagine it's even higher than that half, because it's really just nuclear and hydropower that would provide clean baseload power in this country.
Daniel Poneman: In terms of baseload, yes, because wind and solar are intermittent. Traditionally, I think hydropower (if memory serves) is about 7 percent of baseload power. That would make, between nuclear and hydropower, about 26 percent of the installed base carbon-free power.
Kristin Hayes: It's really important. If you're looking for clean and, as you mentioned, 24/7 power, nuclear plays an outsized role.
Let's dive into some of the connections between these two themes. There have been a couple of announcements in the last several years between technology companies and existing or new nuclear facilities. They've signed agreements for power. In your view, what makes nuclear energy a good choice? We've already started down this path, but what makes nuclear energy a good choice for meeting this growing electricity demand from data centers and perhaps the other sources that you mentioned, too?
Daniel Poneman: I'm going to cheat a little bit here, Kristin.
Kristin Hayes: Okay.
Daniel Poneman: It's okay if I cite my source, which is Bill Gates. In his book, How to Avoid a Climate Disaster, Bill Gates came up with the one-sentence case for nuclear power: “It is the only carbon-free energy source that can reliably deliver power, day and night, through every season, almost anywhere on earth, that has been proven to work on a large scale.” As they say in Spain, punto final, end of story.
For many years, while the data centers were starting to show these vertiginous growth demands, they were very focused on getting to all renewable energy by 2030. I would scratch my head. I said, "How do you do that?" And I said, "I think it's rather quaint. You're not going to operate those data centers at night." Not really. So, it was inevitable, in my view, that they came to the conclusion that they're going to need nuclear, as well.
There’s one other thing that is worth saying: In terms of power density, we're on a crowded planet. One uranium fuel pellet is equivalent to 17,000 cubic feet of natural gas, 149 gallons of oil, or a ton of coal—in a fuel pellet the size of your thumb. People also don't take into account that land is scarce. For a traditional reactor—not one of these fancy, new, advanced, small modular reactors, but a traditional 1,000-megawatt or one-gigawatt reactor—you need land the size of Central Park. For 1,000 megawatts of solar, you would cover Brooklyn. For 1,000 megawatts of wind, you'd cover all five boroughs. That is not a very practical solution in many parts of the world.
Also, if you care about biodiversity—if you've read the writings of people like Hank Paulson, the former secretary of the US Department of the Treasury—biodiversity really values not using too much land that is needed to sustain biodiversity. So, there's a lot of good arguments for nuclear power, and that's why you see this strong attraction to it now.
Kristin Hayes: I hear you. I want to talk about some of the headwinds that may have been holding it back. I think it has faced a number of headwinds for years: cost competition from a variety of other sources; difficulty in citing these new facilities, even if they are, ultimately, fairly compact; economic pressure from lower-cost natural gas; and, even though it might be a thumb-sized piece of waste, it's still a kind of waste that has people concerned. So, we're talking about additional demand and all the values that nuclear can bring, but have any of those headwinds fundamentally changed in this conversation, as well?
Daniel Poneman: It's a little hard to calibrate them in the gestalt, overall. So, maybe if we could unpack it a little bit, why don't we take them one at a time?
In terms of the classic “NIMBY,” or “Not In My Backyard,” phenomenon, I think that, to use a German phrase, the zeitgeist, the vibe around that has changed a lot. Overall, the statistics show that in the United States, 70 percent of people support nuclear energy. You find more support in communities that already host nuclear power plants. That’s not terribly surprising, because they're there, they see that they're not dangerous, and they provide jobs and economic support for the community.
Kristin Hayes: High-paying, quality jobs.
Daniel Poneman: The one demographic, interestingly, perhaps, that is negative on nuclear remains the over-60 age demographic. Idiosyncratically, not scientifically, but in my conversations in Europe, and in Germany, in particular, I also found that the younger generations are much more open to nuclear energy.
Kristin Hayes: Interesting.
Daniel Poneman: Cost is a big deal, and there are many, many drivers of cost. There's no question that the nuclear industry has to perform better. But if you look at a big nuclear power plant, and it costs, say, $5,000 or $6,000 per installed kilowatt, and a natural gas combined cycle costs $1,000 per installed kilowatt, and a nuclear power plant takes at least five years to build, and a combined-cycle gas plant takes a year and a half or less to build … Put yourself in the shoes of a CEO, a CFO, or a board of directors of an investor-owned utility. That's a hard case to make. There are ways to address that, and we need to. Later, we’ll get to how we overcome some of these things.
You mentioned waste. Again, a lot of times people don't really have a proper sense of the dimension of these things. If you took all of the waste generated since the dawn of time in the United States, it would fill one football field, seven yards deep. In terms of volume, it's not that much. We know, technically, how to take care of that. You can put the spent fuel (or used fuel, as we sometimes call it) in concrete casting and canisters for hundreds of years. If you need to contain them for longer, you can put them in geologic deposits that are extremely stable, that have been stable for millennia. For example, there are these salt formations where you take spent fuel and put it in a thing called the Waste Isolation Pilot Plant, or WIPP, in New Mexico. There, you can put the fuel into a borehole, and then the salt will seal up around it.
The problem has been political. In the United States, we picked the state that didn't have very many electoral college votes and said, "Let's stick it there." I'm being a little bit glib, but that’s kind of what it was, right? The Blue Ribbon Commission on America’s Nuclear Future that began in 2010 in the United States studied programs around the world and found that, if you start with the community on a consent-based approach, in which you're educating people all along the way, every step of the way, you end up with, as you did in both Finland and Sweden, two communities who are fighting over the opportunity to host a repository. So, from a climate change perspective, I view the next 50 years as critical. We're seeing state-sized pieces of Antarctica float off into the ocean. We're losing ice at an incredible and alarming rate. And, with nuclear, we're talking about a problem that is hundreds, or perhaps thousands, of years off. And we know very well, from a technical perspective, how to control that.
A lot of times, frankly, I think people trade on the fears and the anxiety about radiation as an invisible force. But we live with radiation. Every time you eat a banana, every time you get on an airplane, you're getting doses of radiation. That fear factor is something that has to be addressed up front. We have so many fears, going back to movies, with Godzilla and things like that. But radiation is a part of daily living, and we have to overcome that fear factor. Then, some of these other things will fall away. But even if we fixed all of the political and emotional aspects of these concerns, we’re still going to have to tackle this economic challenge.
Kristin Hayes: For sure. Later we will get to the policy conversations around incentives and how people are looking to make the economics work. But before we do that, I'm really intrigued by your comments on the age divide, because that does seem to fit into the narrative of being more concerned if you've lived through some of these accidents that were extremely high profile, quite scary, and did lead to deaths and long-term consequences for communities. Although, interestingly, one of those very plants is one of the ones that's coming back to life, which we'll talk about in a second. I could see that if that is a stronger memory for you, you might have a different view moving forward. But also, today's nuclear energy is not the nuclear energy of the 1960s, either. So, there seems to be an education piece there about what's actually changed. I wonder if you have any reflections on how nuclear itself is different than it was in those days.
Daniel Poneman: Let me take two pieces of that, because the first thing you mentioned is really quite interesting. You talk about the demographic differences in responses to nuclear. If you ask people of my generation what they think about the term “TMI,” they'll immediately talk about Three Mile Island and where they were when they heard what happened. If you ask people of your generation, if I may be so presumptuous, they'll say, “too much information.”
Kristin Hayes: Exactly. That's what it means to me.
Daniel Poneman: No one died from radiation at Three Mile Island. At Fukushima, 19,000 lives were tragically lost. Of those, so far as I know, there was one radiation fatality, and that was from a first responder who died about eight years later from cancer. Those 19,000 lives were lost, by and large, from the tsunami and the associated earthquake, which was a millennial event. And actually, sadly, some number of people died in a botched evacuation. If you look at it in terms of deaths per terawatt-hour, the amount of deaths caused by nuclear is way, way, way lower, down there with wind and solar, as opposed to the millions of people who die from carbon dioxide around the world. That's something we need to understand.
Now, Chernobyl was different. You had a lousy reactor design that, without getting into the physics of it, got hotter and hotter, and it had none of these so-called “passive safety features” to cool the core. Then, they had a botched response to it and a cover-up. So, around 4,000 people died. But 40,000 people die in the United States driving cars every year. We don't ban driving. We invent seat belts, collapsible columns, mirrors on both sides, and airbags. We have to respond. We live in a world full of risk.
Now, that being said, this is not to gainsay that we need nuclear to be as safe as possible. We need everything to be as safe as possible. After these various accidents, a number of passive features of safety were invented so that you did not need, for example, electric-powered pumps to pump cooling water through a core that was being uncovered in a loss-of-coolant accident. You would allow a pool of water to be above the core and, therefore, gravity would draw it down, for example. This whole new generation of reactors, sometimes called Generation IV reactors, sometimes use very different physical principles that are inherently safe. For example, let’s say you use metal to cool a very hot core instead of water. The melting point of metal is much, much higher than water. So, the problem of water evaporation uncovering the core and leading the rods to melt down physically cannot happen.
The reactors that operate with some of the metals are like liquid sodium. For example, TerraPower or Oklo use those kinds of metals. They can also operate at ambient pressures, which is also a safety feature. So, there's a lot of features that are inherent to these new technologies that are helpful. That's not to say that the existing generation of reactors is in any way unsafe, because, as I said, if you just measure it clinically, in deaths per terawatt-hours, it's minuscule compared to what happens with hydrocarbons.
Kristin Hayes: I'm glad that we're talking about the new features that have been added here.
I do want to go back to the agreements that we talked about that have been signed between these technology companies and nuclear facilities. Some of those agreements have been with traditional, large-scale nuclear plants—ones that we'd be familiar with, like Three Mile Island. It seems like others are more focused on a new wave of what we sometimes call “advanced nuclear” and in particular, small modular reactors or even fusion. That’s the magical word. What would you identify as the pros and cons of working with those more traditional plants, versus with some of these newer technologies?
Daniel Poneman: I think you have to go back, Kristin, to the first part of our conversation. What is the demand signal that you are responding to? In the Conference of Parties, the big global environmental conferences, over 30 governments have committed to more than triple their nuclear capacity by 2050. In the United States, that would mean adding 200 gigawatts. If you define a small modular reactor as capping out at about 300 megawatts, which is about a third of a gigawatt, that's a lot of reactors. So, my hunch is that you're going to need and see a lot more of the larger, existing Generation III or Generation III+ reactors.
What are the trade-offs between those and the small modular reactors? Well, you're going to have to divide (and I'll come to this in a moment) the small modular reactors into two categories. But the first point is that with the existing generation of reactors, of which we now have 94 operating, there's a very well-established regulatory regime. There's a very well-established supply chain. There are very well-established fuel markets to sustain them. There are very well-established engineering firms and firms that handle the outages. It's all very familiar. The downside is very big ones have tended to be bespoke and individually tailored to the site. Because they're so big, you have to do a very detailed environmental and seismic analysis on them. And, like anything you buy that's bespoke, it’s more expensive than something that you buy off the shelf. So, there's a trade-off there.
Now, you can't just have a generic comparison to large and small reactors, because the smalls come in two broad categories. One broad category is ones that are, basically, smaller versions of the large ones. They're cooled by light water, and the light water cools the core. It moderates the neutrons, so they're more likely to collide and cause a fission that will create steam and electricity and so forth.
The more advanced generation, which we started to talk about a moment ago— the so-called Generation IV reactors—don't use light water anymore. They will use liquid metal. They could use high-temperature gas to cool the core, or they could have the fuel circulating in molten salt. The good news about those reactors is they have some of these intrinsic safety features that I mentioned. They also have certain performance characteristics. For example, the outlet temperature of the steam from a high-temperature, gas-cooled reactor is so high that it can not only generate electricity, but also support industrial processes.
That's why, for example, X-energy has signed a deal with Dow Chemical to build a plant in Texas to support their petrochemical production of products. The good news about those is they have those kinds of advantages. The bad news is they're new, and like anything new, you've got a lot of technical risk, first-mover risk, and regulatory risk. There isn't an existing supply chain to the same degree. Most of these new Generation IV reactors require a new, special kind of fuel that's enriched in the Uranium-235 isotope not to 4 or 5 percent, which is what the 94 existing reactors in the United States use, but up to 19.75 percent. It's kind of technical, but if you think about it more broadly, if you get the enrichment of the Uranium-235 up to 90 percent, it can actually produce a nuclear weapon or a very small reactor that can fit on a submarine.
If you want to have the enhanced performance of that higher power density but not get into all the regulatory overlay of getting too close to weapons grade, you go to just below 20 percent, which is the legal limit. And so, many of these advanced reactors—for example, TerraPower, which is the other large reactor being supported as a demonstration by the Department of Energy, and X-energy—need what's called high-assay low-enriched uranium. That's been a real choke point. The company I used to lead, Centrus, is now making that stuff. But this was the first new, US technology-enrichment line to start production since 1954, and it would need to be scaled up a lot to be able to satisfy the demand that we expect to see.
Kristin Hayes: I'm really curious how some of those features that you just mentioned also play into our economics conversation. I would imagine some of those things might be reduced costs over the comparison between these large scale plants, and, historically, some of them might be more expensive. So, we'll come back to that.
I don't want to miss the opportunity to ask you about Three Mile Island—what might've been considered a cautionary tale from a different perspective that's coming back to life right now. It’s one of two specific examples that I think, it could be argued, were declared dead and are now being resurrected. We have Three Mile Island in Pennsylvania and the V. C. Summer plant expansion led by Santee Cooper in South Carolina. These are infamous for different reasons, but they are both being rejuvenated right now. How are each of those projects being slated for a second life? What lessons do you think those two, in particular, might offer for the nuclear industry moving forward?
Daniel Poneman: Well, I think the revival of those two projects really speaks to the broad issue we talked about up front, which is the voracious appetite for power—I’m talking about AI and data centers here. We need everything you can throw, including the kitchen sink. Everyone knows it's ever-so-much easier to either upgrade an existing reactor, if it's not too late (and often, unfortunately, it is too late), restart a closed-down one, or do anything other than start with a green field. That's the first way to think about it. It's inevitable. And by the way, I'm not talking about something exclusively nuclear. They're going to need every jot and tittle. They're going to be looking to soak up every electron, and you're going to get into issues of data centers having a little bit of an argument with ratepayers and communities as to who gets the electrons preferentially.
There are different cases, obviously. The Three Mile Island unit that's now under contract with Constellation and Microsoft, with all the power being taken up by demand that's already been booked, was a perfectly well-operating reactor. It was not the Three Mile Island unit; it was the twin of the one that had the episode in 1979. The sad aspect is that we were up to 112 reactors at one point, and even within my relatively recent memory, we had 104. We're now down to 94. Unfortunately, what happens is, pretty soon after the thing is closed down, you do things that make it very hard and very expensive to restart. If there were more of these reactors that could be brought back to life, I think you would find that. But there may be, one hopes, a couple more. But I think the message out of Three Mile Island’s restart is that, if it's not too late, these things can be salvageable.
There's been a relatively recent debate about this in Germany, as I'm sure you recall. There was a big argument, first, after Fukushima. They decided to kill everything that was on a schedule. In my view, it was a terrible decision, which has had a terrible impact, including on health, in Europe. We won't get into that. But there were three reactors left. And at the time of the invasion of Ukraine, somebody said, “Well, maybe we shouldn't close those last three.” They decided to do it anyway, and now, they just had elections in Germany, and there's been some discussion about whether they should try to restart those. I don't think it's trending in a good way, but that's being looked at. Many of the countries in Europe that had been virulently anti-nuclear have now either stopped, paused, or reversed their earlier decisions to shut down reactors, like in Sweden and Belgium. So, I think you're going to see that.
V. C. Summer is a different case. After Three Mile Island, there was a 30-year hiatus in building new reactors. What do you suppose happens when you don't do something for 30 years? You get rusty. You lose your supply chain, you lose your talent pool. Full disclosure, I was Deputy Secretary of Energy and chairing the Credit Review Board when we approved loan guarantees that enabled the Vogtle 3 and 4 reactors to get built, and the V. C. Summer projects were going to be built at the same time in the same wave. They ran into the same problems that you probably read about with Vogtle 3 and 4. Delays …
Kristin Hayes: Cost overruns.
Daniel Poneman: The project was mothballed. And so, I think the message is that we need every jot and tittle of reliable power that we can get. I think they're taking a look and seeing if that one is going to be available. Any of these is going to be faster and cheaper than starting from scratch with a green field and a piece of paper.
Kristin Hayes: That makes a lot of sense.
All right, let's talk about the policy side. We're heading toward the end of our conversation. I don't want to lose the opportunity to ask you about that.
I have a two-part question for you. My understanding is that there have been a number of state and federal subsidies that have allowed nuclear to live on past what it might have done strictly based on economics. Those were in service of trying to meet clean energy goals. What does it mean for the broader grid and for those states if nuclear power plants are partnering with data center companies, like you were mentioning? If they're giving this power specifically to a customer, what does that mean for broader clean energy goals if it's not going into the grid more holistically?
Daniel Poneman: Well, I hate to quibble with a premise, but there is no energy forum that I know of that has not benefited from massive government subsidies. Do you remember the oil depletion allowance? Master limited partnerships? Certainly, oil and gas have had prodigious federal support. You would not have a solar industry without the investment tax credit. You would not have a wind industry without the production tax credit. In my humble opinion, the smart way to do this is to put a price on carbon, because the biggest problem that nuclear has had is that its virtues have not been rewarded by the market. Those virtues include clean energy and running 24/7. How many people would've died in Minnesota in the polar vortex of 2014 when the coal and gas plants shut down if nuclear didn't keep humming along? That's been the challenge in terms of nuclear power. So, making the production tax credit or the zero-emissions credit available to nuclear is a surrogate for a smarter policy, which would be just to make a very clean, level playing field by putting a price on carbon.
Now, to the other part of the question—it's really a game changer what has happened here with AI. You have a convergence of two things, which together, I think, create a new dawn. First, we have this vertiginous increase in electricity demand. But here's the second thing: We talked about the Vogtle plant, which turned out to be very expensive—around $35 billion, $14 billion over budget. That's a big chunk of any balance sheet of an investor-owned utility. To go to a CEO or a CFO or a board of directors and ask them to tie up that much of their balance sheet on a very long-term project is a very big ask. But if you have two things that you didn't have before—this huge demand in electricity, and now the hyperscalers, each of whom have a market cap of around $1 trillion or higher—it's a game changer.
Kristin Hayes: And a hyperscaler is, in this case, one of those very large-scale data centers—is that right?
Daniel Poneman: Yes, like Amazon Web Services, Google, and Microsoft. If you just put it in context, if you took all of the market capitalizations of all of the investor-owned utilities in the United States, it's about $1 trillion, and each of the hyperscalers is $1 trillion. So, it's a game changer. And so now, you have both the demand signal that was missing and the balance sheet that was missing to possibly make this thing work.
Kristin Hayes: I feel like I could ask you many more questions about this, but I do recognize that we're at the end of our time, and I really don't want to miss the chance to have a robust Top of the Stack conversation. I wish our listeners could have heard our pre-recording Top of the Stack conversation. I want to end by asking you to recommend good content, either on this topic or, more generally, that you might want to recommend to our listeners.
Daniel Poneman: Well, I never missed an opportunity to make a plug. Even though it came out just before COVID, my book, Double Jeopardy: Combating Nuclear Terror and Climate Change, is still a good read, I think, and highly relevant to today's conversation.
More historical books that I recommend, because I think they're still incredibly relevant, are Ron Chernow's biography of George Washington and Doris Kearns Goodwin’s book, Team of Rivals: The Political Genius of Abraham Lincoln.
This goes back to our pre-taping conversation, since I'm now allowed to recommend music. There's a new CD out called Speed of Heat by the rock-and-roll icon and two-time Rock Hall of Fame inductee Jeff “Skunk” Baxter. Some of your listeners may know that Skunk Baxter is not only a genius as a musician, but he's also a genuine nuclear and national security expert. So, if you want to both enjoy good music and get some of the brainwaves from somebody who has thought deeply about nuclear energy and national security, you can't do better than going to Skunk Baxter and Speed of Heat.
Kristin Hayes: Amazing. Since you're not going to brag on your own behalf, I'll brag on your behalf and note that you actually play music with this particular gentleman, as well. So, there's a strong legacy of the music-nuclear energy connection here that we're celebrating at this moment. That's a great recommendation.
Thank you again for coming in. It's been a pleasure. I really enjoyed it.
Daniel Poneman: Thank you. I've enjoyed it.
Kristin Hayes: You’ve been listening to Resources Radio, a podcast from Resources for the Future (RFF). If you have a minute, we’d really appreciate you leaving us a rating or comment on your podcast platform of choice. Also, feel free to send us your suggestions for future episodes.
This podcast is made possible with the generous financial support of our listeners. You can help us continue producing these kinds of discussions on the topics that you care about by making a donation to Resources for the Future online at rff.org/donate.
RFF is an independent, nonprofit research institution in Washington, DC. Our mission is to improve our environmental, energy, and natural resource decisions through impartial economic research and policy engagement. The views expressed on this podcast are solely those of the podcast guests and may differ from those of RFF experts, its officers, or its directors. RFF does not take positions on specific legislative proposals.
Resources Radio is produced by Elizabeth Wason, with music by Daniel Raimi. Join us next week for another episode.