DVL Power Hour Podcast
DVL Power Hour Podcast
Delivering Data Center Power: How Utilities Enable Uptime
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
For this episode of the DVL Power Hour, we’ll welcome Stephen Stegall, Senior Project Manager at Georgia Transmission, whose project portfolio includes large and complex initiatives dedicated to data center interconnections. Drawing on his experiences, he’ll be walking us through how utilities scope, engineer, procure, construct, and commission the substations and transmission assets required to bring reliable power to these high-value sites. He’ll provide a grounded look at what it truly takes to deliver megawatts on demand.
Join us to gain a valuable perspective on how thoughtful collaboration and realistic planning translate into faster, smoother power delivery for mission critical operations. You’ll also get an understanding of:
•The full-lifecycle process of energy interconnection,
•Regulatory and environmental considerations,
•Pressures behind maintaining schedule deliverables,
•Importance of coordination between engineering teams,
•And more!
Thank you for downloading DVL's Beyond the Product Podcast. This podcast was originally released as a live webinar and has since been edited to fit this format. To view the original webinar and reference PowerPoints live, please visit DVLnet.com slash resources slash webinars. Have a question or want to submit feedback, please email us at marketing at dvlnet.com. Thank you. We hope you enjoy the podcast.
SPEAKER_03Welcome to the first DVL Power Hour of 2026. My name is Robert Leek. I'm the director of marketing here at DVL. We're going to have a great conversation about what it takes to deliver power to data centers. Over the past decade, you know, data centers have quietly undergone one of the most dramatic transformations of any sector in our economy. You know, what began as an IT function, simply supporting email, storage, enterprise applications, data centers have become a form of critical infrastructure that now rivals transportation, manufacturing, and honestly the energy systems from both the scale and importance perspective. Data centers today sit at the center of most everything that defines our modern life, whether that's artificial intelligence, healthcare systems, financial markets, logistics, government, and honestly, even national security. Everything is going through data centers, quite simply. But the workloads may be digital, but the constraints are very physical. Data center growth is outpacing the global GDP by a wide margin. The most recent numbers that I've seen really come down to about the GDP of 3% and data centers at 7 to 10% going through 2033, if I'm not mistaken. It's a huge demand of electricity from data centers. And it's growing several times faster than overall power consumption in general. And many regions, the question is no longer how fast can we build, but how fast can we deliver reliable power. This is truly elevated, you know, power planning from a late-stage engineering detail for data centers to now a board-level discussion. Utilities, regulators, developers, and infrastructure providers are now being asked to work together earlier in the process than ever before, often years before a shovel hits the ground. And that's where real world experience matters. And that's why I'm so excited about today's DVL power hour. You know, today's conversation is about what it actually takes to deliver power to modern data centers, navigating things like grid constraints, large load interconnections, permitting timelines. It's everything, right? You know, and so our goal today isn't to talk about theories, to talk about what's happening on the ground, on the front lines, what's happening, what's working, what's not, um, and how infrastructure decisions are working and being made from whether it's the data center side, the utility side, or really a collaborative side. And I would like to now welcome, you know, our guest, my guest today, Stephen Stigal, uh from Georgia Transmission. So, Stephen, hopefully you're on the line without the technical issues I've had, and uh you can turn that bad boy on. Stephen, how are you, sir?
SPEAKER_01I'm great. How are you? Can you hear me?
SPEAKER_03I am well. I am just to kind of catch everybody up. You know, you I'm here in Denver and I'm in DVL. We have a lot of territories across the country, and if we had a map of them, you would not see Georgia on that map. We've got some, I've got some great friends in Georgia. Um, but we actually met here in Denver at a at a power event back in the fourth quarter of 2025. And that power event was kind of an all-encompassing thing, but data centers really stole the show, I think. And um, with that, I was happy to you agreed to come be a part of the DVL Power Hour. And maybe you could start by just introducing our audience a little bit about you and your background, if you'd be so kind.
SPEAKER_01Yeah, certainly. Uh thanks for having me on. Uh, excited to be here and share my opinions. You will find I have lots of them. Uh, hopefully, you're I can I can entertain your guest as well as provide some useful information. Uh, so as you stated, I I've been in the electric utility industry for 20 years. I've worked for both IOU and now cooperative uh business models, and then I've also worked on uh the nuclear generation side in the last 12 to 13 years in the transmission side of the business. So I've gotten to do a wide variety of things working with uh solar interconnections, uh large manufacturing now data centers. I've done portfolio management, project management. Currently, I'm a project manager for all uh transmission facilities and assets that are needed to energize data centers with related to the cooperatives. So that's uh that's the oversimplification of 20 years. Um so hopefully that uh provides some useful background.
SPEAKER_03I you know it absolutely does. So, first off, 20 years in the energy sector is pretty amazing in and of itself. Um, and I don't want to give away too much of our conversation, but I know you've been doing data center specific stuff for a couple of years now. And could you give us maybe a little bit of a highlight about maybe some of your early learnings as you transition from your previous role into this kind of project management focused on data centers?
SPEAKER_01Yeah, so I I think the biggest adjustment would be the dealing with the fact that as large campuses, right? So if you go to any transmission utility, be it IOU, cooperative, municipality, any of them, and you just say, Hey, I need a single 60 megawatt MVA transformer. We're gonna serve these two or three buildings, manufacturing, commercial, whatever, uh, that's very simple. Majority of your power is gonna come from off-site. We're gonna go to the nearest point of the property, build a substation, maybe even handle it with distribution feeders off-site. That's pretty much the history of serving what we used to think of as large loads. But when you have these campuses that come in and they're a thousand acres, and you're gonna bring in multiple transmission sources from different directions, um, it gets it gets to be a little bit different, right? Because now I'm gonna impact their site design, I'm gonna impact their utility layouts. We have rules and requirements related to our transmission easements. So all of a sudden, it becomes very much a DOT type environment where you can't be within 10 feet of this, and I've got to go over here, and your building needs to slide back, and your protective fence needs to move. And they're going, wait a minute, you know, we if we'd done all that, we wouldn't have put the substation where we did. I didn't tell you to put the substation there. So it it begins this long journey, and it usually takes about a year, depending on the size of the campus, to really nail down where were our facilities go and where were their buildings need to shift and their roads, um, their protection schemes, be it a large wall, large fences. Um, yeah, I would say that's the biggest difference that we've learned in the last couple of years is how do we deal with their campuses?
SPEAKER_03That's you know, I think we're gonna get into that a little bit. I'm sure. So thank you for for teasing us uh a little bit with those. And and it's interesting, you mentioned just fence lines. Um, I was at it recently at an iMasons event, and uh one of the speakers was from that kind of super duper Oracle centric um campus down in Abilene, Texas. And I forget how many gigawatts that they said they're running through there, but the the one thing that truly stood out was seven miles of fence line around the perimeter of this data center campus that you just mentioned. I think that really ties into both the scale of land that we're now dealing with, but also, you know, a lot of different variables that that we don't typically think of. So with that, let's dive in. You mentioned the size of the loads, right? And how it's a little bit different than your typical call it large load or C and I applications. Once we start talking about, you know, you mentioned the dealing with a campus was one of the biggest things that kind of that changed and you know, kind of in that capacity. But once you start dealing with an actual hyperscaler or colo operator and they request a new interconnection, what is that very first step that uh that call it Georgia Transmission or XL Energy or maybe one of the co-ops that I know that you deal with quite often? What is the first step once they get that request for interconnection that that that the utility goes through?
SPEAKER_01So I think almost all of us now have a large load application. I wouldn't call it an interconnection at this point, because to us a load is very different than the FERC interconnection process where you have third-party generators who want to connect to a to a system. Yeah, I would say it's a large load application, and the very simple simple part is where is your site located and how much capacity do you want? What is your load ramp? And do you have any other special requirements? Redundancy, um, do you know what your high side voltage needs to be, your low side voltage needs to be? So we ask those basic parameters, and then we'll do a geographical study of the transmission network in that area, how much capacity is available now. If you need 300 megawatts in 18 months, that's a that's a different answer. But if you need 300 megawatts over the course of 10 years, that's a lot more digestible. And so it does two things. It helps us figure out how we would develop what we call a method of service for your campus. And then the other part is it helps the EMC or an IOU uh figure out a power procurement plan. That's not an official term, I'm just calling it that.
SPEAKER_03I kind of like it, it sounds very official, especially if you put procurement into anything. I feel like it's a procurement plan, that's definitely very official. How uh how long does the um I think the application process is pretty easy? You fill in a few simple fields, it sounds like, but on the back end of that, that study and kind of looking at the land and kind of scope depending upon, I'm sure, the load ramp that you mentioned on the application. But for you to come back and I guess I'm curious, one go through that process of receiving the application of whatever after it, how long does that typically typically take?
SPEAKER_01From what I understand with our current internal processes, it's at least a three-month study. Three month study. Three month terms, yeah. That's the that's the bare bones. It's it's more of an easy either A, we can serve it in a reasonable amount of time, or it's five to ten years, right?
SPEAKER_03Okay, so that was gonna be my the kind of my next question. So after three months, there's a response given, and it's kind of the simplest of responses either yes or no, but the no may have some other stuff. But it's I I don't want to get too far. How can you share a little bit uh with your experience as far as what the yes-no percentages have been over call it the last year and a half or so?
SPEAKER_01I'd say 50 percent yes. Okay, that's fair.
SPEAKER_03I hey I'm okay with 50-50 because you know it's middle of the road a little bit, but I think it also speaks to some of the challenges that that's right.
SPEAKER_01Um the queue's long I don't see I don't see everything that comes in on our system, right? And I know that the the IOU in our state, they they had a six-month wait to even have a conversation to even get in the queue. Wow. So yeah, that their side was uh is a lot more extensive. And earlier in 25, I think it was, uh the IOU told all of the data centers, look, unless you're one of the big ones, the really, really, really big ones, go talk to your local cooperative. And so that began the cooperative side all of a sudden in the state of Georgia at least, had to start developing formal processes that we never had to worry about before. And so now we're in a dated, and I know that we've got it, you know, we have 20 or 30 NQ to study. It's it's not a small number.
SPEAKER_03I think that number's one gonna obviously get bigger, right? But before we get too far, I know you keep mentioning IOU, and I know I owe you 10 bucks from that drink that from that show and whatever. But for those who what does IOU mean in in your world specifically?
SPEAKER_01Yeah, so it's an investor-owned utility. So next there, uh Southern Company, Dominion, uh, these larger utilities that are pretty much vertically integrated.
SPEAKER_03They have generates primary energy providers in some of those instances, right? Yeah, that's right. And then that's on the flip side of the co-ops, those little smaller, a little bit more rural areas where they are have a lot of different types of strategies, right? Um, as it relates to how they're generating power for their region. Is that a fair statement? Yes, that is. What so how does the world of the co-ops differ from the IOUs as it relates to being able to? I I mean, obviously they're maybe a little bit more behind on processes and handling things like data center requests and things of that nature, but just overall handling of large capacity requests, how do they go about fulfilling those? If it's any different, I would say, than the IOU side of why.
SPEAKER_01Yes, I don't want to oversimplify for your listeners because they probably some of them I'm sure know more about the cooperative level.
SPEAKER_03If they're listening and they want more questions, ping that question in there and we'll have follow-ups for you. That's fair.
SPEAKER_01So in the state of Georgia, though, the cooperative network is one of the largest in the country, if not the largest in the country. And so the the cooperatives in Georgia, they own a GT. So they own generation assets through Oglethorpe Power and then transmission assets through uh Georgia Transmission, my company. So I think their response and ability to deal with large loads is a little bit different than other cooperatives across the country who don't have GT assets. So for us in the state of Georgia, we're the only state that has an integrated transmission system as well. And so what that means is I can call up Georgia Power, the IOU in the state, and I can say, hey, we want to put a 500 megawatt turbine, you know, turbine asset on this line that goes through this property that we uh we have right to all of a sudden. And so we can go through that through the interconnection process and not pay the tariffs and all the other things that um IPPs have to do. So it's all to say is the co-ops in Georgia can handle it just like an IOU. I think the challenge is do they have the firm, that's what I'm learning, the difference between firm capacity and non-firm capacity. And so, from what I understand, without giving too much away, they they broker, right? So there is a schedule member group that really helps them make sure that they're in compliance with how much power they're wielding to the state of Georgia. But each cooperative has their own ability to procure power, and they have their own ability to procure new generation if they wanted to do so, at least in the state of Georgia.
SPEAKER_03So let's maybe move into the projects a little bit as as we kind of dive into the weeds. And Lord knows no one wants to talk about you know schedule risks, but they're out there. And um I know you've dealt a lot with your responsibilities in general around scoping, estimating, citing, environmental compliance, procurement, construction, commissioning. What's where in these worlds of high capacity data center projects are you seeing the most scheduled risk surface?
SPEAKER_01So I would say that the projects that we're seeing the most risk on are the ones who pick territories where there is no current infrastructure, right? So there's several data centers across the state, not just on on the cooperative side, but also on the other side, that picked an area that requires, say, a 230 high side voltage, and they picked an area that doesn't have any 230 high side voltage available, no transmission lines in the area. So and they did that because the land was cheap, or the um the local and tax incentives or whatever the local development authorities were able able and willing to give, that drove their decision to pick that site. And so once they picked it, then they went to the utilities and said, Hey, we need power.
SPEAKER_03Oh, so they did no real background on power availability, they were just looking at the price per acre, probably, and like, okay, hey, this is this is in our budget. Now let's let's energize this place.
SPEAKER_01That's right. And typically, you know, if you look at you know our state in particular, the the cheap land is going to be where there's less infrastructure. Sure, sure. And I know the water was also a consideration, too. What counties and cities were willing to provide good deals on water. I think that was just an equal part. Um, so those two things, the tax incentives, or maybe three tax incentives, land, water seemed to be the determining factor of whether or not they were going to put a an option on that specific piece of dirt, which again is part of your application for us to look at it. You got to know where you're going. We don't you don't want to pay the money as a data center for us to study it, and that might be really where you're going. So they tend to do that due diligence first to pick the site, secure whatever rights they need to for the site, and then they have the utilities study it. The electric utilities study it. Yeah.
SPEAKER_03That's, I mean, it's I don't want to say data center companies are kind of painting themselves into a corner, but obviously they've made that investment. Just curious when if if you can share anything about that particular project or generalities around that type of a situation. Reminds me of like the triangle. I remember like uh, you know, for a triangle of business. You can have it good, cheap, or fast. You just only get to pick two, right? I feel like this and so they picked cheap and fast, maybe, or whatever. But yeah, how how much type of investment was it going to require to energize? I'm gonna mess up the 230 reference that you meant a moment ago, but the type of interconnection or or how much does it would it have is it going to cost that company to build out the infrastructure to to energize their data center at that address?
SPEAKER_01So I would say that it the risk, the the the issue is not so much cost. It's not that they're gonna have a higher cost because the way we work in Georgia is you largely pay for what's on site. You know, if you want eight transformers or ten or three or two, you want two substations, you want redundant lines, you know. So there is limits to what we charge the data centers, if you will, to build infrastructure in the state of Georgia, at least before all the new latest pushback came through. We might see some rule changes going forward, at least in Georgia. But as far as the as far as the cost goes, it affects their schedule more. I would say that's the more important piece because if I have to build a one mile transmission line to get to your site or 10 mile transmission line, that's a very different level of risk to your data center. And then if I if you have a large low data center and you're in the hundreds of megawatts and you need multiple sources from the east, west, north, and south, that's a lot of land and easement right, land rights and easements to go get to get the lines to you. So it's just a schedule risk.
SPEAKER_03A schedule risk that might not have an uh an end date. If I I mean what if what you're posing, right, is not really customer challenges. Again, it comes back to infrastructure challenges and coming to partners like Georgia Transmission to make sure that you can get 10 megawatts into the facility or whatever you're trying to do. And so you so okay, this is interesting. So, does the data center not have to be financially responsible for building out that infrastructure somehow they just say, hey, I want to address this, I need you to energize it, and that you pay for it, it's just a matter of how long it's gonna take you to do it.
SPEAKER_01Yes, essentially.
SPEAKER_03Well, that's my big first big epiphany because I just I assume that your customers would be somehow at least off offering some level of investment to whether it speed up the speed up the schedule or to maybe you're looking for some level of collaboration on on cost management. But okay. So really it's just a matter of your schedule is your really is your biggest thing, along with a lot of question marks.
SPEAKER_01If you have an open checkbook, then yes, schedule is your best biggest risk. But in the state of Georgia, two of us utilities have the ability to do eminent domain from a transmission perspective. So with that ability to do that, we don't necessarily want a data center, let's pick a big one, right? AWS, um, Google, just pick big people, right? We don't want them out there buying property for us, buying rights for us. So we do kind of push them to the side a little bit and say, hey, we'll give you a commitment on timeline, but it's going to be our way. We're going to use our processes so that we are following the state law. We don't want you risking our ability to have imminent domain. And so, because we have 500 assets we build, we do 115 assets. We don't just serve data centers and we can't just put all that risk on you. We have a business model outside of data centers that's actually a lot bigger. So neither one of us in the state of Georgia want to lose that ability. So in that case, it goes back to schedule, not cost, because the data center, a data center, could give us a billion dollars, and I we don't really want to do that. We don't want to start that precedent.
SPEAKER_03Well, this actually kind of leads into I think our next question really nicely. I mean, you mentioned like from a cost perspective, I s really about once it gets on, right? And I think that's data centers, as you just mentioned, right? They found some land, they just wanted to get some power on to it, and the infrastructure is not there. And that's a nice example that maybe we can just use as an example. You talked about transmission lines, you talked about, you know, generation points, you talked about, I'm sure there's some repeater. What has to, let's just say there was a 10 mile gap between infrastructure and an address, or maybe even something even more manageable, like three. Five miles. I'm sure this is the same stuff, but what has to be built? How does that process go around? You mentioned laws, you mentioned kind of land issues that you guys deal with. I mean, you've got private ownership, public land ownership, I'm sure. What are really, what really let's just say everybody has green lit this project, okay? And now the data center is waiting for power and you're working with one of the co-ops or whomever. What then happens to then deliver the power from wherever the co-op or even Georgia Power currently has it and get it to that now remote address?
SPEAKER_01Yeah, sure. So we have a whole writing, uh routing and siting process, right? And so that's where we evaluate existing corridors. Is there a gas line easement that already runs that direction? Uh, what's between the local source, the nearest source, and where we're headed? Is there environmental features we need to avoid? Is there navigable waterways? Is there protected weather, protected historic districts, right? Archaeology studies. We don't want to run over top of cemeteries or Indian mounds or any other protected thing that you could come up with. Yeah, I saw the movie Poltergeist. I don't think we want to do that either. No, you don't want to do that. Yeah, you don't want your lights flickering, do that. Um, but but in all reality, in that siding and um and routing process, we look at all of the topography changes and then we come up with a a reasonably achievable route that doesn't include that that tries to avoid how much cost we're going to incur. So so as an example, if we had to build extremely tall structures that were completely oversized with 50-foot foundations at each one, I mean, there are cost limitations that we have because for again, for us, following state law, we've got we can't just overburden our base with the cost of a new facility, a new asset. And of course, our base is the cooperative network. Georgia Powers base is their uh every person who has a meter spinning on Georgia Powers system, you know, they're they're a ratepayer. And so so for both of us, there are limitations, but again, that goes back to what was the original method of service. And we do put a lot of thought in that window that we're looking at this project. We say, oh, we've got to come from a line over there for six miles, a line over here, but it's shorter over here, but there's a river we have to cross. That's extra permitting, extra environmental reviews. Um, the structures are probably gonna have to be taller. I learned a thing that there's a sailboat requirement on certain on even private lakes. If there's a private lake and it's allowed to have a sailboat, our wire has to go from 27 feet off the water to 55 feet off the water. Oh my gosh. I just learned that the other day.
SPEAKER_03And you got a sailboat on your wallet, even.
SPEAKER_01I know I do. I should yeah, I should have known that. Shouldn't I?
SPEAKER_03I mean, but again, these are like I mean, this is really intriguing stuff that I think the the normal layman who's not in this world of of utilities just doesn't consider the the the bylaws and the the limitations or or or whatever the case is.
SPEAKER_01That's right. And so and then once we pick a route, right? Then then what I've learned is you you have all kinds of things. If heaven forbid, one of the parcels that's impacted by your route have a trust, and that trust is 25 family members from 200 years ago, and one of them lives in France, and one of them may or may not have died. Um it's it's wild because once you get into those trusts, everybody has to sign off on it. And if you can't find somebody living in France to sign off on it, then it doesn't happen. Yeah, then it doesn't happen. You gotta wait, wait more time, more schedule risk for the data center. Not that a data center can control every variable, but uh I think certainly in in this round, when I say this round, the projects that are under construction in the state of Georgia for both uh Georgia Power side and and our side, they certainly didn't consider how we could get electricity to that site. There they had other things that were priorities, and then once they got into the risk, and so we try to hedge our bets. We try, we try to guess where that line's gonna end up, and and we try to say, well, that's a three-year project, that's a four-year project. And so we're we're making assumptions before we even have authorization to move forward. So we try to hedge, but there's nothing, but we caveat everything. If we end up in legal processes with the trust with a family member in France, that three-year commitment is gone.
SPEAKER_03Well, I and granted, there's always going to be these curveballs, like you pointed out. And again, I I get I don't think we just naturally think of, so thank you. But on a normal issue-free, call it routing and siting process. Yeah, how long do you think that normally takes? On average, and obviously all the other caveats I know we're dealing with. But just on average for a data center type project like this.
SPEAKER_01Yeah, I would say six six to eight months. That's about what it takes to do routing, siting, get the route approved, and then go public with it, right? So six to eight months to go public. So then after that, once you have announced to the public that this is what we're doing, then we go into what we call phase two due diligence. But that's where we send environmental people, archaeological teams to go evaluate that route and make sure that there's nothing there, like I don't know, protected frogs or protected tortoises or whatever you could think of. Um, so so then we do phase two due diligence. We'll also take soil bornes if the property owner allows us in different places just to validate that it's good dirt and make sure that because again, it's part of the evaluation of can we build this transmission line where we want to, where we think we should. Um, but we've certainly had to make changes because the dirt wasn't any good. We found out that maybe that area was contaminated 50 years ago by the previous property owner, and so now we have to move that segment of line to a different side of the property or even a different route altogether. Um, and then once you phase two due diligence can take anywhere from three months on the low end to six on the high. And so now I'm almost a year into the project and I still haven't ordered poles, and so the challenge that we have today is lead times on steel poles. So the other impact that data centers have brought to the transmission network is the size of wire that's needed and how many conductors per phase that they need. The bigger the conductor, and the more conductors you need per phase, the bigger the structures are. So we're getting into territory where these structures are the size of lines that used to only go to generating facilities, but now they're we're putting them up everywhere, everywhere there's a data center. We call it bundled conductor, it's going up on top of these poles and it's causing these poles to be 130 feet, 140 foot, versus what we used to be able to get away with at 105 concrete pole. Now we need 140-foot steel, and those steel poles need seven, eight feet diameters, which has a bigger base, a concrete foundation at the bottom. So these designs are getting harder and they're getting bigger. And because they tend to pick sites that are cheap and land, usually the terrain's terrible, which leads to saying when you're going, I I'd lived in Georgia for a while.
SPEAKER_03The mountains of northern Georgia can be a great hike, but probably a horrible place to put a power pole.
SPEAKER_01That's right. And these pier foundations, it's hard to get a concrete truck up certain slopes. So now I'm incurring more cost just to get a concrete up the side of the hill, um, to pour the foundation for the 140-foot tall structure. So now all of my cost and schedule are going up.
SPEAKER_03This just uh opened up so many doors, quite honestly, Stephen. So, first off, you mentioned to some degree supply chains are are an issue, right? Um, what call it steel, call it everything else for the conductors and just general supply chain challenges. I know whether it's the data center world or probably many industries around the world, we've we felt those, especially back during the days of COVID, but gotten a little better. I'm sure they get a little tighter when we get into different styles. Um, but curious. So you mentioned that the the cables are certainly more robust, more like power generation cables, et cetera. That's having an impact on your pole sizes. Um, from and this is just my curiosity. Why do the poles have to be taller because the wires are bigger and obviously supporting more energy rolling through them, I assume. Is that just to more energy popping off and setting forest fires? What's why the why the required call almost a 30% increase in pole size?
SPEAKER_01So we have the NESC code that requires um you know minimum distance between the conductor and say roads or the conductor and the ground. So just from an NESC standpoint, we're having to go bigger. The configurations are the what part of what's causing it, right? So if you go near generating facilities, you tend to see the big lattice steel towers that have the conductors horizontal. Well, where these data centers are going, that people want us on the roadside as much as possible, the local communities. And nobody wants to give us an easement to have the conductors flat. So we roll them vertically like you would typically see on a distribution circuit. So when you ask me to have generation-sized transmission lines vertical, because that takes up less easement along the roads, the structures get 30 feet taller.
SPEAKER_03I see. So it's almost you're compensating from that horizontal to vertical flip to some degree. Yes. And I know we're talking about like the remote areas there. Um, let's maybe reel it a little back in a little bit closer to where where people are having dinner and stuff today. And I think the answer is obviously as early as possible, right? Especially once if you're acquiring new land and you're figuring it out. And the question I want to talk about is how early should a data center provider engage a utility, right, with information that can provide an accurate estimate. But right, if we're talking about, you know, a brand new greenfield build and all that stuff, they need to figure out final stuff, right? But let's just say a data center company is exploring somewhere, call it maybe not Metro Atlanta, I know it's pretty populous, but maybe somewhere else in the spans of Georgia, maybe Savannah, maybe they're a golf fan, want to build out in Augusta or something. What how can what do they need to do if they say they have three or four addresses that they're looking at across the state? When could they come to the utility or how when would be the right early stage to say, hey, I need 10, 50 megawatts of power, whatever the number is, at this facility? Can you tell me if this is possible and I can go invest my money and buy this place?
SPEAKER_01Well, unfortunately, because the queue's so long, now you have to be have to have good information and be a real player before you even come to the table. Three years ago, we would take prospect prospective, prospective phone calls and and and say, Yeah, well, if you're just thinking about it, here's some thoughts. Here's a recent study we did that we used six months ago. Yeah, this this place in Augusta, that could work. So we'd give you a very general answer on prospective, but we don't do that anymore. I mean, the line's too long. So you don't come, yeah. If you don't come to us with a solid, uh at least a land option, an option on the property, we're not gonna study it because there's too many people in line who want to study. Yeah, who have options, who have, in some cases, real customers with in some cases firm power already in place.
SPEAKER_03Well, I the most thing that's most surprising to me is the fact that what a short timeline in three years can do. Because I think it's uh it's you picked it, it's great that you picked that, right? Because call it three years ago is when the AI bubble was just starting to float up and become a thing, maybe even two years ago. But three years ago, we you most I'm sure utilities, we kind of go through that kind of just nice process of exploring something and whatever. But now that we've evolved to a place where there just isn't enough time in the day and there's too many projects on the table, you got you all, the utilities, are requiring just a little bit more solidified information before even getting into call it a partnership with it with a data center company, I guess. That's right. Wow, yes.
SPEAKER_01So, I mean, I would say the fact we uh I'll say we're working with one data center company now and they're legit, they're big, they know what they're doing, and so this is almost like planning out a subdivision. So I did distribution design when I was in college and just largely putting in underground cable for distribution substation, or not distribution, but distribution level circuits for residential areas. And it was simple. I mean, you learned it, you rinse and repeat. That's kind of how this company is. They they prioritize the water and the power before they put a land option out. They use existing relationships, they kind of already know what power is available, and they have so many studies done from the last couple of years that they they know where they're doing, and so we have a timeline. We already we've already procured long lead equipment. It's the the transmission lines right next to the property. I mean, that one's super easy. They get all the megawatts they want on the schedule they want, no risk. They're very mature data center, they have mature processes, and they're not guessing, they know what they're doing.
SPEAKER_03I think that's what I was about to say is once you go through this a few times, I'm sure you really start understanding whether it's on the utility side or the data center operator side, where your challenges are and the pitfalls you're you're you're trying to avoid.
SPEAKER_01On the transmission line, I kind of got distracted there, rabbit trailing on the different topics. Um but on transmission line timelines, it's it's at least a year before I go public and can order poles. And then from the time it takes to get poles, if it's if it's smaller wire, I can do concrete poles, if it's steel poles, that's with the the supply chain limitations that we mentioned. You know, you're talking six to eight, sometimes ten months to get all the poles you need. So, I mean, we're telling prospective data centers for a transmission line two to three years. If, like the one we just discussed, the mature company that's used to this uh already within a three-year period, they've gotten really good in the in our state, they they picked a transmission line right next to the on the property, actually. They won't they don't have to wait three years for a transmission line, but a data center is trying to figure it out, uh, and you pick a spot not near current infrastructure, minimum three years.
SPEAKER_03So a lot of these conversations are coming around this idea of capacity constraints, right? Yeah, yep. You can't have 10 megawatt or you can't. What I guess is the source? And again, now we're talking not in the middle of you know the boonies, we're talking back in Augusta or somewhere that's got some some power. What usually drives those constraints? What's the barrier? Is it the generation, the transmissions, substations, or people even, I guess? But when somebody actually has an address and they go through the process and they you say no, you can't have that, why what's the most typical answer about why they can't have as much power as they're wanting?
SPEAKER_01So typically it's not that they can't have it at all. It's just you can have it in 10 years, you can have it in five years, right? So it just depends on what it is. If you want to put a I'm gonna make a absurd number, but some of the things you see in Nevada, Texas, even um what are the six six thousand megawatts? Let's say you want 6,000 megawatts in Augusta, it's just time. You're just not gonna get it in a year, even if you put yourself next to a transmission line or several transmission lines, you're just not gonna get it. So, what we're learning is is it's generation capacity. It's it then if there is enough generation within your RTO to handle it and take credit for that you can buy, then it's can you actually get it to that local area through transmission network upgrades? Because if your firm power is located, I'm just gonna make up a crazy example. I'm not an expert in this, but I've I've learned enough or asked enough questions, learn a little bit. But if your farm power is an IPP in North Carolina, it may or may not be pointed towards Augusta, Georgia, right? So if you got a 6,000 megawatt campus, you still got to get it there. And so the local grid near Augusta may not be able to support 6,000 megawatts. So how the question is is how long is it going to take a company, George Power or my company, to upgrade all of the transmission facilities to be able to even import, I'm gonna use the word import, 6,000 megawatts to your facility. I mean that that's the issue, that's one of the other issues when you talk about constraints and what are data centers doing to us. You know, historically, from a man, you know, manufacturing was the largest concentration of power of load, would say load, uh, for the history of electricity. And so the largest manufacturer, you will, in the state of Georgia, is an LNG gas plant. And they're only a little over a hundred megawatts, and they're consistent, they run 115, I think, 120. And it's a gas facility. Being from so I'm from Augusta, Georgia originally, and for the longest time there was a mile in Augusta that had Procter and Gamble, International Paper. They called it the miracle mall for Georgia Power when I used to work there. And so back when manufacturing was a thing in the US, uh I forgot how many, it's like 25% of Georgia Power's income was from that one mile. And actually, the um one of the plants was Augusta newsprint. They used to make all the newspaper for people to print on. They used to make all of it, right? And so they ran three shifts a day, they had big boilers, right? So I think 25% of the income at Georgia Power at one point was just from that one mile in Augusta. Well, even then, I think the megawatts was only 150 at best for the one mile. And now I got data centers that are taking up farms and the easy ones are 150 megawatts. I mean, the scale is just incredible. That's why your local transmission grids, regional transmission grids, just aren't built for a thousand megawatt campus. And so the question is, is not can I get it to you? It's it's how long can you wait for me to get it to you? You give me 10 years, I can do just about anything.
SPEAKER_03Sure, sure, sure. And and I think to your point, the the answer is not what is the single thing that is at the source of a capacity constraint, it is literally the entire infrastructure.
SPEAKER_01That's right.
SPEAKER_03That has to be either upgraded or augmented or or something.
SPEAKER_01Um and one thing I did want to touch on too is cost, right? We we keep seeming to ignore cost, right? Um but uh in my opinion, the way I understand the business models, right? And I know there's a lot of push, public push, and really make sure that data centers are paying their way. But uh what I'll say is is they are paying their way, right? If your power companies do it right, data centers can absolutely put downward pressure on rates. I mean, you don't you don't necessarily want your local data center to make prices directly cheaper for your residential customers, but if they can lower your fixed cost, if they can improve your reliability, they can pay for network improvements to where so my house is what we call radio fit. And so when a tree limb hits one section of that multi-mile line, I'm out until Georgia Power gets out there and fixes it. But if all of a sudden some big important customer lived near me that didn't live near me before, and more power is brought to the area, maybe I'm on a what we call a looped system and they can switch me remotely, and so I go from being out for days to being out for 20 minutes.
SPEAKER_03Well, hell, I think we just solved the PR problem that data centers keep having with the communities. But I I'm not even kidding, Stephen, because I you it everybody who's listening to this, I'm sure is aware of I'll say Northern Virginia is a great example, has been very public in a lot of the news over the last 12 months or more. Community pushback, picket lines, town hall meetings that are against data centers coming in. I think most of that isn't due to misconceptions based on what maybe that facility is or is not going to do. But if you could clear those up with a cherry on top with reducing rates or making your own homes and your residences more viable or more disaster recovery just naturally, that's a that's a hell of a thing to consider as somebody who, you know, just living in a you know in a cul-de-sac next door to a data center and just knowing that maybe I could have a lot less power outages, especially if something that they considered uh experienced in the past.
SPEAKER_01Sure. And and we try, we do sell, but you know, most power companies are not in the business of marketing how good this is for you, right? I mean, uh I would say traditionally utilities take the approach of we have a cut, we have a load to serve, right? That's I've read my entire career is we're there to provide power. You know, we we always we the power companies always try to make the argument that we're another member of the community and we have a new member of the community to serve. Yes, we try to communicate that, but we're not going to go on a marketing ploy uh for everybody to understand just how great as much as it is yours for sure.
SPEAKER_03But have a little bit of an aside, but have you experienced that type of community pushback at all, whether it's data center-centric or maybe something else, but that type of hey, stay out of our neighborhood?
SPEAKER_01So I used to work in nuclear.
SPEAKER_03Oh that was gonna be that was gonna be how we wrap this up and and tease the next time I invite you on, Stephen. But no, please.
SPEAKER_01Yeah, so I used to work in nuclear. I was a part of the young professionals organization that would used to be called N A Y G. And I think they've since got an a much more marketable name. Um, but part of our job was to do outreach. And I was involved when Vogel 3 and 4 was uh going through the licensing process before it actually got built over near Augusta, Georgia. And um, it was tremendous. And what we found in those public meetings was misinformation, right? Fake news, all the taglines we see today. Uh, we didn't know we didn't know what to call it back then, but you know, people would walk in and say, You're you know, you're gonna kill us by building new reactors, you're gonna make this Chernobyl, you're gonna make this what we've seen in terrible movies about the big plume and a nuclear bomb go off. And all the reality is none of that's true. I mean, even from a nuclear standpoint, yes, there's an exclusion zone that the NRC requires the nuclear plants to have, absolutely. But are you gonna have a nuclear bomb in your backyard? No. Could radiation get out? Potentially, in the worst case scenarios, yes. And so just getting the public to understand who has preconceived notions already of that this is dangerous, I don't like it, it's not much you can do. So the same thing with data centers, right? So there's a I saw an article over uh you know, one of the data centers an hour outside of Atlanta. They said that they're affecting our groundwater. I'm not even sure how that's possible. I've been out to the site, right? I'm not a high, I'm not a hydraulic engineer at all. I don't understand that stuff. I'm not a civil engineer. I ask questions when it's important, but I'm just not sure how that specific property owner could be affected by the groundwater. The data center doesn't draw groundwater. Yes, they developed the site, but they didn't take out any springs, they didn't tap the aquifers. They I'm not sure why her groundwater is going down. It may be because of the thousands of homes that have been built around her. But she made the news and she got everybody's attention in Metro Atlanta for how this data center is ruining her groundwater.
SPEAKER_03I think we're all making notes about to go Google certain trying to find these headlines, honestly. But uh I bet, I bet. Okay, so from that type of community concern, right? We kind of speaking of data center and you mentioned your history with nuclear. Pretty sure there's a Venn diagram happening right now that we're gonna see in the future that's just gonna get bigger and bigger than Vinny, if you will. Are you? I mean, maybe Georgia Power, if you can speak to that without breaking any kind of, you know, whatever. But is Georgia using nuclear power today? Just curious. And subsequently, do you see as we can get the public to realize we're not gonna be creating mushroom clouds with these things, nuclear as a potential option, whether it be small reactors at closer to the site or maybe even as power generation from the utility side? Call it in the next five years.
SPEAKER_01You're not gonna see nuclear reactors hit the ground in the next five years, right? There's a licensing process that's painful. We saw that with Vogel three and four. Vogel site is license was licensed a long time ago in the 70s for five nuclear reactors, and they stopped at two. Then they came back and built you know three and four 30 years later. So currently there's six reactors operating in the state of Georgia. Yes. Right. That's that's public information, that's not any secrets. The problem with nuclear is it's it's just not economical to start, right? Any utility that looks at the cost of a gas plant, even if it's baseload or a peaking gas plant, the cost is just not even close. And the worst part is the NRC really hasn't proved approved, not that to my knowledge, and I try to follow it pretty closely, they haven't approved any of these small module reactors for commercial use yet. Yes, they're being built on DOE facilities, but what they're not really communicating is how by building them as test reactors on DOE facilities, they're bypassing a lot of the processes that uh that a Duke or Nextera would have to go do to build. Or let's say, let's say Amazon, right? We see Facebook, we see Meta, we see Amazon making these deals with nuclear power plants to either keep them running or to turn them back on. You know, as long as those kind of capital investments keep coming, I think nuclear will be a part of the distill a CEO's uh phrase, all the arrows in the quiver. I do think that we reached a point though from traditional power generations and power transmission and distribution, yes, the data centers, if this thing is going to keep going, like you mentioned earlier, if this is truly the next infrastructure, it's time to start getting creative. If every community needs a data center, just because of the things that we do, if Amazon has to be in every community to have a have their own servers, have their own data centers, if the local hospital has to have a data center to operate going forward, then yes, we're gonna have to get creative. Nuclear has to be a part of it. The problem is between licensing and cost, it'll be a part of it, but it's not gonna be the easy button. Yeah, but I mean, what I think they ought to do is that they should have done 10 years ago. Now that you have complete drawings from the AP 1000, and this is not me, you know, trying to put a um put a boon out there for Westinghouse or whoever owns Westinghouse these days, but I mean, one of the biggest issues was the detail drawings were not done when it was licensed, just the shell and the major components were philosophized on what it would look like. So that was one of the issues we had as you went through construction, and I was on the Vocal 3 and 4 site for two years in the middle of all this, and so the NRC was approving the design as it was being designed from a detail perspective. And so now that you have a complete set of drawings for anybody who's able, copy and paste. Let's go, AP 1000s, because you'll get better at it. And we see that with any industry that you copy and paste, you get better, the supply chains get better, which drives the cost down. But I don't know how much capital inertia you would need to actually get that ball rolling to then see the economies of scale kick in. Um, but I'm a huge advocate of nuclear. I would love for it to be a major player. I I just don't know how you get past the licensing um people's fears and and and the political games that can be played with it. I don't know how you get past that to get the economies of scale.
SPEAKER_03Well, I think to your point, time and standards um will probably have a big, big thing. Time just because over time we're gonna continue to run out of available power. And I know that right now it seems like delivering power, it sounds like is a three to five year plus project, depending upon what part of the boonies you're trying to build a data center. From what I mean, to your point, you mentioned Westinghouse. I think GE is in the world of trying to commercialize nuclear in some capacity. Numbers I've heard toss about is like five billion dollars. That's a crazy number. And to your point, I don't know how much of an investment you have to, you know, 20%, you know, to get to put a down payment down. I don't know. But 70 to 80 years worth of energy on the back end of that. It's and that I I I just feel like again, to your point, three years ago, you'd entertained a data center conversation over coffee. In such a short period of time, you've had to change your processes, create a queue, and now create a pretty significant timeline of wait. We're just gonna feel these things over the next shorter and shorter periods of time. So, Steven, this was a great way to, like I said, wrap up this conversation. We're here at the tail end of the DVL Power Hour. Want to be respectful of your time and certainly our audience, but I'm not even kidding when I want to come back and talk to you again. Um, and if it's only about two years spent on the site of that nuclear development, maybe that's something. But it's been a joy, it's been a pleasure. So much wonderful. So any last thoughts from the utility side for our data center friends?
SPEAKER_01Yeah, I would just say like it's a it's a growing industry. I I've enjoyed learning a lot about the data centers. It's been fun to see how the data center, the what what is the definition of a data center is can continuing to, I would say, to tier, right? So now you've got artificial intelligence only sites, you've got cloud server sites, hyperscalers, mom and pops are now getting into it. I call them mom and pops. The smaller companies are getting engaged. It's interesting. I'm interested to see where it actually goes. Is this sustainable or are we in a dot-com bubble? I don't know where we are with this. It feels like we're in an arms race and they're just buying up capacity to be able to use long-term. And I think that's gonna be the trigger. Are it is is is Google, are the big ones Google, Meta, are they really gonna use 6,000 megawatts of power in five, 10 years? I don't, I don't know. That that's a lot.
SPEAKER_03That is a lot, but uh, I think the more and more that we as a society around the world continue to become reliant on things like AI and high performance computing, and it's gonna be wild. But uh, I'm glad you're now my new friend in this world, Stephen. Um, thank you again for for participating on today's DVL Power Hour, and um thank you to all our guests. Thank you, and I do hope to have you on again sometime. But until then, y'all have a great one. Appreciate it.
SPEAKER_01Yeah, thanks everybody. Enjoying it.