Announcer: Welcome to “Not Your Father’s Data Center” podcast brought to you by Compass Datacenters. We build for what’s next. Now here’s your host, Raymond Hawkins.
Raymond: So thank you for listening in to another edition of Not Your Father’s Datacenter. Today we are joined by Matt Crozat, the senior director of strategy and policy at the Nuclear Energy Institute. Thanks for joining us, Matt.
Matt: Thanks for having me, Raymond.
Raymond: So thank you for listening in. Today, Matt and I are recording on July 30th, 2020. And our world is still struggling with the global pandemic, COVID-19, what it means and what is social distancing and all things that many of us probably never thought we’d be talking about. Today, we’re gonna talk about energy in the data center business, certainly one of the hot topics. When we think about sustainability, we talk about water and we talk about energy and we talk about PUE and we talk about things that help the facility operate at an optimal level but we often don’t talk about where that energy comes from and how to think about clean energy other than just renewables and thought we would have Matt join us today to give us some insight into how nuclear could help answer some of our clean energy challenges as data centers continue to use more and more power. Matt, how’s that sound for a charter?
Matt: That sounds like an excellent conversation.
Raymond: Matt, as we get rolling here, I’d love for the folks to understand a little bit about you, you personally. As we talk about here on “Not Your Father’s Datacenter” is that it’s a really, just a conversation over dinner, you and I wrapping up, talking about politics and religion and our families and a little bit about business and love to get to know you a little bit better before we get into the business side of things. So, Matt, tell us a bit of your life story?
Matt: So I was an Air Force brat. My father was in the Air Force for 30 years. We moved around a bit growing up. My background is I got to this line of work through social sciences, political science, economics, and I knew nothing about energy especially nothing about nuclear energy. But while I was in the workforce trying to use these economic tools, I learned to make better sense of public policy, I ran into these nuclear issues and they became fascinating. What I really wanted to do was work on interesting problems that mattered and realized very quickly that nuclear energy was full of them. And so in the 2000s, I was working on trying to figure out how on earth we were going to be able to expand nuclear energy in the United States factor of two, a factor of three to meet what looked to me like a coming challenge for human climate change. And I went into the Department of Energy, the end of the Bush administration, trying to work on that problem.
And I stayed there for nine years but during that time, a lot of the world changed around us. We had the massive global recession in 2008, 2009, that changed a lot of the attention around energy, and that was followed by seeing a real change in how natural gas was produced in the United States and that changed a lot of the economics. So I went from trying to figure out how we were going to expand nuclear power a lot to then trying to figure out well, geez, if I’ve got plants that are well run but are at risk of closing early because of economics, how can I help stop that from happening? And that’s how I landed at the Nuclear Energy Institute to help work on the policy design to ensure that these carbon-free nuclear plants aren’t lost but we’re trying to figure out how exactly we’re going to charter a path forward on climate change issues.
Raymond: I liked the way you said that, Matt. “I wanted to work on problems that mattered.” I think lots of us look for, “Hey, how do I do something that purpose? What’s my purpose? How do I make an impact? How do I do things that influence my fellow man and the planet positively?” And so problems that matter, I like that. So, Matt, you and I share the army brat, the military brat. My father, like yours, was in the Air Force for 21 years and we moved all over. What was your move all over experience being in the Air Force? So what parts of our great country did you get to be stationed in with your parents growing up?
Matt: I barely remember California but we were primarily in way upstate New York, almost to Montreal. And my dad flew a fairly obscure plane so we kind of ended up staying there much more than most experiences. I had a little bit of time in Omaha, Nebraska. My parents went on to North Dakota and Georgia after that but by then I was out of the house in college.
Raymond: You got to avoid the North Dakota’s tour, that’s what I call there.
Matt: I did. I did.
Raymond: Now you’re in the DC area today, is that where the NEI is? Is that the right way to refer to it? NEI?
Matt: NEI, that’s correct. Yes.
Raymond: Is that an acronym you guys use? Okay. Yeah. Sure.
Matt: It is. We’re based in Washington, DC. We’re the trade association for the nuclear industry. We try to help the nuclear industry speak with a common voice wherever we can. This involves having a membership of all kinds of companies that are involved with the nuclear industry from those that own and operate large nuclear power plants to those who design them, build them, service them, companies that provide legal services to academics who work on research for the next generation of reactors and technologies. So we have a very broad membership base and NEI’s mission is to promote the use and growth of nuclear energy through efficient operations and effective policy and that’s where a lot of my day job is was try to find, well what are those policies that can help nuclear contribute to our energy future along with the other technologies we’ll need to be both secure and carbon-free.
Raymond: Got you. So if I think of the Nuclear Energy Institute, just like when I see those commercials for the “Beef It’s What’s For Dinner” or the “Cotton Growers of America” it’s the same concept. It’s just for the nuclear energy industry. Is that a fair analogy, Matt?
Matt: It is a fair analogy though we tend to not have national television campaigns. Those tend to be pretty costly. We tend to focus much more specifically. So we tend to be in the background but still trying to find ways to provide an understanding of what nuclear energy is and is not, and educating people about what the technology can do. That is a large part of what we hope to accomplish.
Raymond: So I’m not gonna see you in a national TV commercial on the news tonight is what you’re telling me?
Matt: It would be a shock to me as well.
Raymond: We’d both be surprised by it. Okay. Very good, Matt. All good. But similar things, just like there’s the producers, the analogy is still the same, right? You represent the whole industry, right? All of the member institutions? Is that a right way to say it or member companies?
Matt: They’re companies.
Raymond: All are part of the institute. Yeah. Got it. Okay. Good. Good. Well, that’s helpful. All right. So we’ve got a little bit of understanding about you, Matt. Thank you for that a little bit of understanding about the Nuclear Energy Institute. Let’s tie this a little bit here on the front end into the data center conversation and then let’s really get into the nuclear side of the business.
So as we think about energy in the data center, what often gets talked about the most is or let me rephrase, when we think about sustainability, the number one thing people talk about is how much power gets used and talk about the efficiency of that power. So how good a job is the data center at efficiently using electricity to power the computers and then using electricity to reject the heat from those computers.
And we recognize that our industry is a measure of PUE, power utilization is a big deal for us. When we think about where that power comes from, a lot of our big customers are very concerned about the sustainability of that energy. And when I think about all of that energy coming into the data center space today on a macro level, could you tell me where that energy comes from? And let’s just think about it from a U.S. footprint standpoint, if I look at the grid today, the grid is powered from a percentage perspective, how’s the grid get powered today from different sources?
Matt: Sure. So today about 38% of electricity and by today, I mean, actually I’m using last year’s numbers, I’m referring to, about 38% of the electricity came from natural gas which is a fossil fuel that is extracted largely from shale in the Northeast but also is a co-product with oil production in places like Texas and Oklahoma and that’s been very cheap recently. So 38% is a growth for natural gas over the last 20 years. About 24% of electricity comes from coal production still. Next, behind that, is nuclear. Nuclear provides 20% of the U.S.’s electricity with wind and hydro about 7% each and then solar about 2%.
Raymond: Wow. I did not realize. So the first two, kind of both fall in the fossil fuel category. And then third on the list is nuclear. I didn’t realize it was that high.
Matt: Yeah. Nuclear provides 55% of the carbon-free electricity generation in the United States.
Raymond: Wow.
Matt: Wind has been growing strong. So is solar. But still nuclear has been the backbone of our carbon-free electricity generation for decades now.
Raymond: I like that phrase “carbon-free energy production.” When I think of that, I think most people think of carbon-free or sustainable energy or renewable energy. They tend to think wind or solar. Hydro falls in that category and nuclear both as well, don’t they? Is that an accurate statement, Matt?
Matt: It is because in all those technologies there’s no combustion of fuel that would release carbon dioxide. And so there’s no emissions that affect the climate that come from the operating of these facilities.
Raymond: Got you. So today about 38% natural gas, 23%-24% coal. So that’s 60% to 63% around there. I’m using round numbers. And so two-thirds basically fossil fuel and a third the renewables with nuclear being a part of that one-third of the production?
Matt: That’s about right. Yep.
Raymond: Okay. Got it. Well, I’m a child…I was born in the ’60s and grew up in the Cold War and certainly lived through the news cycle around Three Mile Island. We can take a few minutes to talk about that but there’s I think a bit of a stigma, a bit of, you know, elephant in the room thought of, “Hey, is there safety issues around nuclear?” Can you take a few minutes and just talk about the history of nuclear energy? I love it from a sustainability standpoint, right? We’re not burning fuel, we’re not putting carbon in the atmosphere, but can you talk about it from a safety perspective, Matt?
Matt: Sure. Happy to do so. So we think of the history the technology, of course, was developed largely in World War II to support the weapons program what became the Manhattan Project but nuclear fission was developed as part of that technology exploration. The neutron itself had only been discovered in the 1930s. And so the idea that you could use a neutron to split apart an atom and release energy was still a very new concept.
After the war, there was the realization that this energy especially it’s so dense and so powerful, you could use that to create civilian electricity and that’s what we pursued through the Atomic Energy Commission at the time. We had a commercial program that began in the late 1950s where we’re actually producing electricity from fission, uranium, and pretty soon nuclear power. You know, in those early days, there was a real optimism and justifiably so about how this could be developed. But as this technology began to expand, you know, we began to see more and more plants built. By modern standards, I won’t say that they were being operated as efficiently as we’ve come to really understand now, and I’ll come back to that notion of power utilization in a moment because that’s interesting and it’s the story, but in 1979, we did have an accident at Three Mile Island.
One of the two units there had a partial meltdown when a valve got stuck open and the plant operator didn’t know that it was actually open. And so some of the fuel melted in the reactor and that created a great deal of alarm understandably. Often what’s not appreciated is there were no deaths associated with the Three Mile Island incident, there were no injuries even, but that was a pretty significant wake up for the industry and that set off a path to really emphasize safe operations in making sure that these kinds of challenges and threats would not be a feature going forward.
We set up something called the Institute for Nuclear Power Operations and this was one where all the CEOs of the companies that own nuclear plants were brought together and measured on their performance on how well they operated their plants. And they would use peer pressure to basically…it forced the norms that, “No, you’re not being effective enough. We all need to do this together.” And that really began to change how companies thought about operating these plants. Let me get back to power utilization because not only have we operated them much more safely as a result, we’ve also operated them much more efficiently that once we began to really focus on operating these plants well it became something of an economic opportunity also.
Early 1980s we probably ran our nuclear plants about 60% of the hours during the year. We call that a capacity factor. That’s gonna be focused on a lot. By the 1990s that had gone to about 90% of the time. So we were getting 50% more energy out of these same capital assets than we had been in the early days. Last year, we actually operated our fleet over 93% of the hours available in the year and that, I think, speaks to the way in which the incident at Three Mile Island really forced a better focus on how these plants are run both in terms of the safety and efficiency of the operations. And I do see parallels to the story that you talked about with the data centers in trying to maximize the efficiency of the energy being used in those facilities.
Raymond: So I want to make sure I’ve got this right, Matt. So you guys refer to your industry, the nuclear industry refers to it as capacity factor. So let’s imagine my nuclear power plant and I don’t know what the right metric is but it can produce 100%. And previously you were running them at about 60% of that number and now they’re running it over 90%. And so you’re getting more production out of the facility year-round and you’re increasing that capacity factor, so it’s running more efficiently. Did I grab that right?
Matt: You did. And what really happened is there’s a real focus on making sure you were running as often as you could.
Raymond: Got you.
Matt: And so right now when the plants are offline, it’s predominantly from pre-planned outages in the low demand portions of the year so you can refuel because nuclear plants don’t need a constant source of fuel. About every 18 to 24 months you come in with a few truckloads of fuel, you remove about a third of what’s in the reactor and replace it with the new stuff and you’re good for another year and a half to two years. And so when we’re offline for those refuelings, you do some of the maintenance too but that used to be a process that could take months. Now it’s down to a process that takes a couple of few weeks. And that I think speaks to the recognition that these are very valuable machines and you want to be using them and you want to use efficiently as you can.
Raymond: Like any other plant and equipment, the more it runs, the better the return on the investment in the plant equipment. It’s the same concept, right, Matt?
Matt: It is. And also what we found is by focusing on keeping the plant running and running well, it also reinforces the need to have safe and digital and oversight in the operations, and that’s what we’ve seen in the data over the last 20 years.
Raymond: So we measure PUE to data center and it’s stated as a ratio 1:2, 1:3. And then we talk about data centers in megawatts, kilowatts, and megawatts. What’s the right metric thinking about a nuclear power plant? What do you guys measure from an output perspective?
Matt: So we measure megawatt-hours for the most part.
Raymond: Megawatt-hours.
Matt: Millions of megawatt-hours. So right now the nuclear fleet has a capacity of about, I want to say 96,000 megawatts. So if all of my plants are operating full out, I’m gonna be producing 96,000 megawatts every minute of that day. And so we tend to calculate them in hours. So last year, the nuclear industry produced over 800 million megawatt-hours of carbon-free electricity.
Raymond: 800 million megawatt-hours just in ’19?
Matt: Yes, sir. That’s right.
Raymond: Okay.
Matt: Which again goes back to earlier, that’s what 20% of electricity in the United States works out to be.
Raymond: Got you. And if they were all running flat out, we’d be producing 96,000 megawatts across the fleet? I like that word fleet.
Matt: Yes.
Raymond: Is that right?
Matt: That’s about right. And, you know, where I think it’s important from our point of view is that when I look at how the nuclear industry and the nuclear fleet operates, it operates 24 hours a day, 7 days a week, that because they’ve done that refueling once every other year in a spring or a fall, once the plant is up and running, it runs 24 hours a day. And for this reason, when I think about the sustainability argument you brought up earlier, I think we’ve got a real opportunity to have a pairing with some of the other technologies that are less capable of running on those cycles but instead have their own benefits like solar and wind that are clearly a growth portion of the grid and trying to ensure that nuclear can compliment those technologies as we expand is in part the appreciation of the last few years in the sustainability conversation.
Raymond: So, Matt, I liked you use the carbon-free word. If I called nuclear energy green would somebody throw a shoe at me? Would they say, I don’t understand it properly because when I think about green, I’m going, “Hey, just don’t put carbon emissions in the atmosphere.” Is it safe to say that or am I gonna get looked at cross-eyed by somebody?
Matt: Oh, I can’t speak for everybody but clean, green, carbon-free, I think they’re all trying to get the same basic idea, which is I need to be really cognizant of how I’m producing my energy if I want to make sure that the environment I want to preserve is going to be part of my future.
Raymond: So, okay. If I use the word clean instead of green, is it a fair understanding that the 96,000 megawatts produced by the nuclear fleet today here in the U.S. is the only 24/7, 365 green source that I can get today. Is that a fair way to think about it?
Matt: I think there’s a small slice of the grid for a geothermal production that will make a similar argument and hydro as well could be more 24/7 if the seasonal rains wherever your locality tends to support it. That being said, nobody has the capacity factor as high as nuclear, I think, and that I think makes the [inaudible 00:19:31.931].
Raymond: Matt, I appreciate you bringing up hydro. My thought on hydro is oftentimes you may have a hydro plant where it’s utilizing the water flow but that water flow is controlled by environmental forces and then there’s other times where hydro is just pumping, I mean I’m going to be crude in my description, they’re pumping water up a hill and then letting it run down a hill. So the pumps have to run, right? Is that a fair understanding of the two different ways hydro facilities produce electricity?
Matt: Well, I think it is but let me take a slightly different framing. So you have the hydro facilities that are designed to generate a lot of electricity now. So sometimes you think of the classic large reservoir with the big wall of concrete like the Hoover Dam that’s kind of a canonical hydro. And honestly, most of that hydro that I mentioned earlier falls in that category as well. These are facilities we’ve had for quite a while, and they’re still operational.
We’ve seen environmental concerns about what that does to aquatic life on those rivers. And so we haven’t built many of those for quite a while. We’ve seen much more attention to things like run of the river where you just have a turbine in a river that can use that flowing water to help produce electricity not the scale of the Hoover dams, but at least a way to take advantage of some of the opportunities that certain places will afford. The part you get at though is something that’s really important which is how do I think about storing electricity? This is something that’s really a challenge because, in order to run the electricity grid, I need to make sure that everyone who’s trying to use power is being [inaudible 00:21:09.303] to someone who’s producing it at that spot at that time. And that’s a bit of a challenge. I don’t have a warehouse I can use to draw on reserves or store something if I had a little extra. I have to do a kind of a real-time dance.
And we’ve not had very many technologies that are good at providing that buffer and that storage. And that pump hydro example that you described is actually one of the very few ways that we’ve done this to date of saying, “Well, if I do have extra electricity maybe I can use it to pump it up a hill. And when I need it later, run it back down through a turbine to release a little bit extra.” And that’s where a lot of attention is going now as we think about things like a lot more wind and a lot more solar. Well, if not just those hydro facilities, are there other ways they can store electricity as well using batteries or similar technologies? And nuclear is part of the solution that says, “Well, because it can run 24/7 as you look for these additional storage opportunities, you don’t need quite as much of them. I can provide a balance to just what’s gonna take to scale up to really decarbonize electricity sectors.”
Raymond: So Matt, I hadn’t even thought about the storage aspect of it. That’s just a great thought. Am I right? We warehouse lots of things, but we don’t really warehouse electrons particularly effectively, right? I mean, battery storage but those aren’t not on a massive scale. So this production meeting demand inline concept is really what our grid and our sustainability of our grids are really all about, isn’t it? Let’s match production to demand.
Matt: It is. And I would think from the point of view of data centers they’re supposed to be a pretty constant demand for power over the course of a day. I would be surprised if they were dramatic spikes of when I need to run the cooling systems and the like and so having that 24/7, 365 backbone to compliment some of the other sources I think provides a nice fit for the kinds of profile of demand use for data centers and the profile of production for nuclear plants.
Raymond: Yeah, you’re exactly right there, Matt. Computers like the same amount of power whether to run electrons through the computer to do its work or to run through the air conditioner to keep the computer cool. It runs the same 24/7, 365, whether the sun is up, the sun is down, whether it’s summertime outside or whether it’s winter. Now when the outside air temperatures change, we can be a little more efficient about what we have to do to reject the heat in the data center but the demand for power in the data center never goes away. It is 24/7. And thinking about the notion that your side of the industry, the generation side of the industry is trying to match production with demand at all times because there’s not an effective storage answer is just a part of the business I never thought about. So great insight there and a way to think about, “Hey, we got to make electrons as we burn them because they don’t sit on shelves very effectively.” Fascinating.
Matt: No, and this is whereas people have tried to really do the analysis to look at [inaudible 00:24:13.450] and trying to hit these vicious climate targets what does the system need to look like? How do I produce electricity? What technology combination is going to get me to a working system? There’s been a growing appreciation for how nuclear fits as part of that portfolio alongside wind and solar. And even as we try to develop these energy storage technologies like batteries, doing that on a scale that can completely allow us to rely on things like wind and solar, that’s still a big, big, big challenge and ways off.
And so what we’re seeing is an appreciation that as you do the analysis that looking at this hour by hour matching of demand and supply, there’s a really strong role for how nuclear energy fits as part of that energy picture. And that’s where in the last few years we’ve seen a much greater appreciation from analysts and academics of the role nuclear is playing now in providing carbon-free electricity and what we probably need going forward. And I think that’s been a heartening development as we begin to move beyond a relatively simple way of thinking about a small set of technologies, trying to solve the whole problem and trying to expand that toolkit so that we can bring as much of our capability to bear while also reducing carbon emissions and providing the kind of stable power that your facilities will need.
Raymond: Thank you again for joining us. We’re gonna pause here in our conversation with Matt Crozat and we’ll come back on the next episode of “Not Your Father’s Datacenter” and continue the conversation with Matt about how nuclear energy and the Nuclear Energy Institute and the customers it supports can help contribute to the energy needs of the data centers that are fueling and leading our digital revolution. Thank you for joining us and we look forward to talking to you again on the next episode.
Thank you for listening. I hope you enjoyed this edition of “Not Your Father’s Data Center.” I’ll ask that you please join us again next time for a very informal chat around the businesses, the people, and the technology driving the data industry today. If you have any questions or comments about this episode or if you’d like to make suggestions about future topics, we’d love to discuss it. You can email me at our rhawkins, that’s R-H-A-W-K-I-N-S @compassdatacenters.com or you can reach me on our Twitter feed @CompassDCS. That’s Compass, C-O-M-P-A-S-S-D-C-S. Thank you again for listening to “Not Your Father’s Data Center” brought to you by Compass Datacenters.
