Between Two Turbines
The world of energy and the technology shaping it.
Powered by NAES.
Between Two Turbines
Powering the AI Era - Alan Bull
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Between Two Turbines — #11 | Alan Bull
In this episode of Between Two Turbines, Vice President of O&M Services Alan Bull returns to join host Josh Tate for a look at the forces reshaping the power industry. As AI accelerates demand for electricity and data centers expand at an unprecedented pace, Alan explains how utilities, independent power producers, and technology companies are adapting to meet the moment.
The conversation explores the rise of co-located generation, why some AI companies are investing directly in energy infrastructure, the evolving role of renewables, and how emerging technologies like small modular reactors (SMRs) could help meet future demand. Alan also shares his perspective on where the industry is headed and the challenges operators will face as the grid enters a new era of growth.
Powered by NAES.
You'll see these companies just going to where gas is stranded. We have the same thing up in the up in the north slope in Alaska. Um we take oil, there's an oil pipeline to get that out, but there's tons and tons of methane uh uh in those same caverns, and there's no way to get it to market.
SPEAKER_03Right.
SPEAKER_01So maybe we build data centers up on the north slope of Alaska.
SPEAKER_03I mean, it makes a lot of sense. Those the data centers really can exist anywhere.
SPEAKER_01Right. You need fiber, you need good fiber and the gas. You know, it's like the old um logging towns or things like that, where they they would build it where the logs were. Sure. They'd just build the town and whatever infrastructure they needed right there.
SPEAKER_03Just build it where the energy's at. Yeah. The world of energy and the technology shaping it. I'm Josh Tate, and this is Between Two Turbines.
SPEAKER_01Powered by means.
SPEAKER_03Alan, welcome back. Glad to be back. Yeah. So last time we talked was a bit of a history of the grid, and um, you know, I enjoyed that conversation. I think what I want to do uh more regularly is just kind of an industry update.
SPEAKER_01Okay.
SPEAKER_03You've always been my go-to guy when it comes to what's happening, you know, in the industry, what's happening uh in the news. And and so yeah, I mean, to start out, I mean, the big story of the last year and and maybe more is AI and data centers. Um what's what's the scoop there? What's happening with data centers?
SPEAKER_01Like you said, it's it's really uh changing rapidly and it's really been growing fast um over the last two years, really, and then accelerated more in the last year. Um and the data center space is unique because the hyperscalers like the Metas and the Googles and the Microsofts of the world aren't used to having to wait for anything.
SPEAKER_03Of course.
SPEAKER_01And so um they have money to spend. They got money. They got money, so money usually means that you know you you can you can make things happen.
SPEAKER_03Um don't have time to wait for transmission lines for renewables and all that kind of stuff. So they need something always on right now.
SPEAKER_01Yeah. And so one of the big challenges that that they're running into now is that the cues uh in the various markets in which data centers uh are trying to get built are long. You know, three, five, seven years. Uh you mentioned transmission lines, those take even longer, a decade plus uh has been our more recent experience.
SPEAKER_03Aaron Powell, some of them are what, a hundred years old now? Yeah, yeah, exactly. Or even more.
SPEAKER_01Yeah, that we'll need to replace. And so these data center developers um or you know, the hyperscalers that are that are gonna purchase the data centers once they're developed are trying to find other avenues to get their data centers built. Typical construction timeline for the data center itself is about 18 months. So they want to make sure that everything can can happen within that that time frame. And so the shift has been now to go kind of off the grid to what's called behind the meter. And so many, I mean it's it's it's national news, not just not just in kind of the market space in which we sit in, but kind of all around, is that these data centers are are they're trying to get them built everywhere. Yeah. Um, and particularly if there's a place that has um available gas and maybe long-term available transmission lines, power transmission lines. And so NASE has been approached by dozens of new players in the space, which has been interesting for us to try to vet. Yeah, you know, are these you know there's folks like Chevron that clearly know what they're doing, yeah. They're real uh companies that are gonna get something built. There's other developers, yeah, exactly. And then there's other developers that you're like, you know, is this two guys in a truck or is this, you know, yeah, uh real project with real backing with real money? Right. Um and so all this behind the meter uh uh construction and development has really been a big change for the industry as a whole. And the backlash um that has been bubbling up has been around multiple things. One is the data centers that were connected to the grid. Smaller ones as well. Smaller ones have have been perceived to drive up the price of electricity for the common user, you, me, and everybody else. The um also, you know, they're they're buying up land that could be used for other things in in certain parts, particularly in um Virginia, where there's more data centers than any other part of the world. Um they're they're going home by home by home, offering houses, you know, three, five, tenx what the value of the home is to buy it, so they can get enough physical land to build a data center. So that's clearly changing what a neighborhood is like, uh, particularly for people that maybe retired there or moved there for the schools. Um now there's a huge data center in their backyard, or they just have to sell out and move somewhere else.
SPEAKER_03And yeah, you know, everything that comes with that.
SPEAKER_01Um some of them in Virginia are even just uh uh you know suburban communities that are that are that are you know commute into DC or something like that. And so even in those places now, they're the the data center developers are offering a tremendous amount of money for for property.
SPEAKER_03Yeah. And recently uh Seattle kind of voted a memorandum to kind of hold back more data centers in here. Any thoughts on that?
SPEAKER_01I don't think it's a good idea, I guess just to be blunt about it. I think that the the proposed data centers here were very small. Um, and the ones that we already have in the city of Seattle are very small. Yeah. And we're LAN constrained here anyway. Sure. So the chance of a a one gigawatt or something that really affects the average user is. Yeah, it it's pretty small. And so I think we want to be willing to listen to as a as a city, as a community, to listen to what the data center developers are offering. Sure. Um, and not just say, no, we're gonna have a moratorium for two years. You know, because I I just think those data center developers will just move on. Yeah. The city will lose any opportunity associated with that.
SPEAKER_03I I think just the AI, the brand of it right now and and the public sphere is uh it's not positive, obviously.
SPEAKER_00Yep.
SPEAKER_03Um, but the demand is, you know, it's beyond. I think they said 2026 the uh natural gas is now 71% more is planned than in 2025. So I mean it's it's growing faster and and faster than we can even get turbines out there to to combust um and and create that steam. So I mean the demand is great, obviously. Uh we live in a country where um a lot of people don't want it in their backyard. Yep. Um, and I can understand that. And I think I would imagine the politicians, maybe their position is, well, the public's not ready for it. Um, so um, but at the same time, these companies have the money, they're gonna figure out where to put it somewhere, they're gonna put it somewhere, and yeah, and development's gonna happen. And it's uh sort of trying to be pragmatic with that, obviously.
SPEAKER_01I think if if if we take a bigger picture, longer term view, um, construction of of this volume of generation will be good for everybody. Um, you know, it it it may in the near term uh you know create problems with with land use and and and the the the price of power uh for your home. But I think long term when we have a more robust grid, that'll eventually drive down pricing. Um and we saw that in the big boom in the in the late 90s, early 2000s as well.
SPEAKER_03Um So you see it as an investment.
SPEAKER_01I see it as an investment, and I know that will be painful probably five, maybe even ten years, but but long term over time, you know, and the grid is something that has got to be viable for hundreds more years, really. Um I think long term it'll it'll benefit the community you know, the North America as a whole.
SPEAKER_03Yeah, and our part to play is really just you know, we're not owning any of these facilities. We just want to make sure they're run safely and uh efficiently. Yep, you know.
SPEAKER_01And reliability. That's the reliability for data centers is the most important uh metric uh or KPI out there. And so, you know, traditionally we've we've as an industry have kind of strived for you know 90 98 to 99% reliability. These data centers are talking about 99.9999, you know, or you know, four nines, five nines as as they call it. And so one thing that does is change the design philosophy because in order to have that level of reliability, you have to have more redundancy. Okay. Uh the other thing that it changes uh for us as operators is the number of people uh at the site and the amount of spare parts that you have at the site to make sure you can make repairs as fast as possible. Yeah. You know, five nines reliability means you have something like 15 minutes of downtime a year. Uh that's very challenging. Yeah, 15 minutes. Yeah.
SPEAKER_03Yeah, that's seems extreme. Yeah. And, you know, at this point, like you said there, the money is not the issue. Right. The issue is how do we make this as reliable as humanly possible? And good thing we're pretty good at reliability. Yeah. Good, good record there.
SPEAKER_01Yeah, we we do. And and you mentioned um, you know, the the availability of of turbines. Um, it's not just turbines, you know, the the turbines are are one of the large components, but generators are becoming challenging to to get a hold of. Uh generator step-up transformers are getting hard to get a hold of. Um, so it's just, you know, every large piece of equipment that takes significant uh manufacturing know-how is is becoming harder and harder to to to source. Yeah. And so we see these developers going out and getting you know all kinds of equipment that you would never really want in a power plant.
SPEAKER_03Yeah, you know. And stuff that may not be as efficient as the most modern, right. Right. But we just need it. Yep. Uh there, or at least the demand is there for it. Right. Yeah. Exactly.
SPEAKER_01Exactly.
SPEAKER_03Yeah. I wonder the nuclear SMRs, what what's the scoop on that? Where are we at with that?
SPEAKER_01I think last time we talked, we talked a little bit about you know, some of these SMRs have been upscaled in size. They've gone from 50 megawatts to 100 megawatts to now like 350 megawatts. I saw an SMR. Yeah, is it really an SMR anymore at that size? Um, at least that's that's domestically how we're how we're looking at it. And small is relative. Yeah. And I think that that um because of that shift over time, and I think a lot of that shift is is due to the regulatory burden of still of getting these licensed, um they're not as fast as as I think many of us had hoped. Many of us in the industry had hoped they would be to come online. And so there's a there's a couple, I think there's one at uh Idaho National Labs that's being installed. There's one up in uh near Toronto at a large nuclear plant that they're installing an SMR on. So progress is being made, but it's it's relatively slow. Um I think I I read in the news that uh there there has been an SMR that in China that has been uh recently commissioned and put online. Uh that may be the first one. Yeah, they beat us. Yeah, exactly. Exactly.
SPEAKER_03Yeah, it's hard to keep up when um you know they can really make sweeping regulatory changes. Um we don't really have that same advantage. Yeah.
SPEAKER_01Uh and they've been building, you know, 20 plus nuclear large-scale nukes a year for the last 40 years. Yeah, they've been doing everything. Yeah, exactly.
SPEAKER_03I think most probably most of almost every category.
SPEAKER_01Yeah, renewables, gas, and gas, coal, the whole thing. Right. They're doing it all um because they need it all.
SPEAKER_03They're mega dam and yeah. So I mean it's some of it's pretty impressive stuff. Yeah, yeah, exactly. I mean, what's what's the future there for us? Do you think NACE will be involved with SMRs? I know um I'd imagine it's gonna be one of the things because we kind of have to throw everything at it.
SPEAKER_01Yeah, so I do think that that NACE eventually will be involved in SMRs. I I see this rollout being uh probably first with utilities um and and the utilities uh placing them at sites where there's already nuclear generation, as I mentioned, the one up in in Toronto or near Toronto where they're doing that. But I think once it becomes uh more commercially viable and you get private equity money uh in there, like our our uh most of our customer base, um then you'll see opportunities for NACE. But I mean it's at least five years, maybe ten years down the road, at least in my opinion.
SPEAKER_03So kind of midterm.
SPEAKER_01Yeah.
SPEAKER_03Yeah. And a lot of these hyperscalers or or these uh tech companies really, um they've made an agreement that on these larger facilities, I think above a hundred megawatts, uh, that they will kind of do the funding. They're not gonna pull from the grid on these. Um it's not legislation, maybe it will be in the future. Right. Um what are your thoughts on that? Is that optics? Is that um, you know, or is it just the reality of the market right now?
SPEAKER_01I mean, I think there's a little bit of uh call it politics, if you will. But but I do think that that realistically they have the money and they need to change the perception. And so this is a good way to start that process. You know, not not only will your rates not go up, but but once we get these all connected to the grid, it's it's it's likely that your rates will go down or at least go up more slowly uh than they have historically. Right. And so I think and and again, these hyperscalers, uh the Amazons, the Microsoft, et cetera, they have the money. So why not use that in a positive way, uh, you know, as as perceived by by kind of the average Joe, yeah, so to speak.
SPEAKER_03Yeah, it's interesting watching even just the Dow or any of the markets. It seems like it doesn't really matter what news we get, it just keeps going up.
SPEAKER_01Right. Exactly. And it's like it there's on the back of AI, I mean, primarily. But exactly.
SPEAKER_03It's it it's kind of a a long uh running room to kind of make some big decisions right now, at least the way I perceive it, is what we invest in now really has implications the next 20, 50 years. Yeah.
SPEAKER_01Um yeah, it was amazing to see the um the the SpaceX IPO and kind of you know, a lot of pundits were saying it's overpriced, but in fact it went up, it almost doubled, now it's come down since then. But uh you know, that's an investment in AI. That's not an investment really in in SpaceX as it sits today. Yeah. Um so there's still a lot of positive energy behind that.
SPEAKER_03Definitely. Have the anthropic IPO coming. Yep. I think uh opening i sometimes behind them. Yep, exactly. I mean, these are gonna be huge. Um, I mean, I use Claude regularly with my work and understanding some of its ins and outs. It I mean, it's uh it's such an impressive tool. Yes. Um and I mean there are times we're like, sorry, you've used too much, and it it's sort of a a direct reflection on how much energy we actually have going to the data centers right now. It's that's only gonna scale up, you know, as as jobs are built around AI and and these infish efficiencies are are um sort of become the norm. Um I kind of wonder what roles will will come up. And you know, part of my hope is that that will translate to each individual uh worker actually becomes more valuable to companies. Maybe they can start investing in things like hey, marketing, you know, marketing guy. Maybe more companies think, hey, one or two people could do the job of a team. Yeah. You know, yeah. Um I think that's been the experience with our team at least. Uh try to try to do that.
SPEAKER_01Yeah, the the productivity improvements, right? Yeah. Yeah. So which is economist's favorite term, right? You get more productivity from the same amount of work. Exactly. That's how you grow an economy.
SPEAKER_03Right. And so I know there is a lot of, hey, we're doing layoffs, it's because of AI. Sometimes I wonder if that's uh a little bit of lip service. Um, that it if it is that really the fundamental, or is it just that wasn't a project you want to keep investing in? Or because I don't know, we're getting a tool that's making workers more powerful. Why would you want less workers? Right. But that's the optimist in me, I think.
SPEAKER_01Yeah. I mean, I, you know, anecdotal stories from friends, you know, we obviously here in Seattle where there's a lot of folks in the uh in the tech industry, and they mostly play on the latter part of what you're saying there around, hey, yeah, it's you know, they're saying it's AI, but we just had too many people you know to do extra.
SPEAKER_03There's a lot of hiring during COVID. Yeah. I think some of that is still has maybe that'll be uh less of a an excuse in the future, but yeah, definitely obviously you know had an impact.
SPEAKER_01Yeah, yeah, yeah, exactly.
SPEAKER_03So we've seen these natural gas facilities or or plans proposals uh a lot more coming in 2026. Yeah. Um I think renewables have kind of flatlined a bit. Um what's your take on on the renewables market and where it's kind of headed next? Obviously, uh like we've said multiple times, we need everything we can get. Um what role do you see that kind of developing into?
SPEAKER_01Yeah, so well, it's true if you look at kind of year over year growth, um, renewables are are relatively flat, particularly compared to gas. Um, but it's still the largest uh overall uh uh new generation to be added to the grid over the next three years at least. So easy to add. Yeah, it's so easy to add, and um it's uh more cost competitive uh than depending on where exactly, but it can be much more cost competitive even than than natural gas. Yeah.
SPEAKER_03Still a good investment in a lot of places, yeah.
SPEAKER_01Exactly. So I I don't really while gas is growing, I don't see that that renewables is gonna slow down. I think it's gonna, you know, continue to be uh continue to grow um uh as it has been. Yeah um and and so over time a larger percentage of our overall generation mix I think will be renewables. Yeah. Um and I and I think it's it's you know, like we've talked, I think both the couple other times we've talked, I just think it's good to have it all.
SPEAKER_03Absolutely.
SPEAKER_01Um and so that's what I'm a big fan of.
SPEAKER_03Yeah, that flexibility.
SPEAKER_01Yep. The hyperscalers want renewables in their portfolio, but they recognize you can't get 99.99% reliability when only when the sun shines or only when the wind blows. Yeah, absolutely.
SPEAKER_03I mean you have a stretch of a week without sun, right. Uh which you know, up here that can happen. Right. Obviously, not the best spot to do uh solar. Right. But um, but that you know, how do we get the energy from Nevada up to you know, up to Seattle or um, you know, to the east coast where a lot of these these places are being put in.
SPEAKER_01Yep.
SPEAKER_03Um Texas as well. I mean, they're getting a lot of things.
SPEAKER_01Yeah, Texas is unbelievable. I think the I think the last statistic I read is there's over over 350 gigs of proposed large loads. They don't they don't designate whether it's an industrial facility or a data center on the Texas grid. And the Texas grid is only about 90 gigawatts today. So it's it's astronomical. It's astronomical. Um and so you're gonna need that much you know, not all that'll get built, clearly, but but you you'd need some, you know, a doubling of of ERCOT's grid size, maybe to support what does end up getting built.
SPEAKER_03Right. But data centers can kind of be placed anywhere, and I imagine that gives them a little bit of an advantage, uh depending on where some of this energy is actually being derived from.
SPEAKER_01We are close to signing our first data center project in Canada.
SPEAKER_03Uh huh.
SPEAKER_01And so um, you know, while while north of the border hasn't seen the same volume of desire to to build data centers, um that's starting to change because uh particularly in in uh kind of north central uh Alberta, they there's a glut of gas that that is that right now can't really get to market. Um there's just not enough uh pipeline transmission to get that gas to market. But you build a data center right there, you got the gas and it's cheap because it's stranded. Um that makes a lot of sense. Yeah. And so well, that's one thing that that's a shift since kind of the last couple of times that we talked where hey, maybe, maybe you'll see these companies just going to where gas is stranded. We have the same thing up in the up in the north slope uh in Alaska. Um we take oil, there's an oil pipeline to get that out, but there's tons and tons of methane uh uh in those same caverns, and there's no way to get it to market.
SPEAKER_03Right.
SPEAKER_01So maybe we build data centers up on the north slope of Alaska.
SPEAKER_03I mean, it makes a lot of sense. Those the data centers really can exist anywhere.
SPEAKER_01Right. You need fiber, you need good fiber and the gas. But you know, anyway, so that's it's an interesting slight shift that we're seeing is go to where the gas is. Yeah, you know.
SPEAKER_03Yeah, yeah. That I mean, it is interesting. You a lot of these facilities you think have to be Somewhat centrally located to get transmitted, but it's a different game altogether.
SPEAKER_01Yeah, when it's behind the meter, you know, it's like the old um logging towns or things like that where they they would build it where the logs were. Sure. They just build the town and whatever infrastructure they needed right there.
SPEAKER_03Just build it where the energy's at. Yeah. Yeah. So wherever we're fracking or so we talked about the the SpaceX IPO. Um, you know, I don't know that we're gonna get to Mars anytime soon, but uh some of the plans are pretty interesting that that uh they've talked about putting uh these solar panels and data centers out in into space because then it would be always on. Right. It's not like the sun stops shining. It's just uh you know the retail uh rotation of the earth. Right. Um as long as you believe it's a globe, I suppose. But uh but yeah, I mean what's your take on that? Do you think that's feasible?
SPEAKER_01I think long-term feasible. I I do. I think that um it's a really interesting concept. You know, the as the things like Starlink um and Amazon has their own, I think it's called Leo, maybe I can't remember now. They're trying to launch their own satellite uh uh uh network. They've shown you know less latency um and and high speed connection. I think there's more and more viability to having you know not no no wire plugged in and and you know kind of beam the things back and forth between the the data center in space and what we're doing here on Earth.
SPEAKER_03And on Starlink in particular, I mean it's it's amazing how many rural communities are or really anywhere around the world that has completely transformed. They've gotten online now. Yeah, um, I think about some of these countries that are more developing, they didn't have to do the 2G, 3G, they're going straight to LTE or landline, right? Can you imagine landline? Yeah, exactly. It's like they've kind of got to skip a few steps. Right. And some of that infrastructure is easier to install than you know the hard work we kind of put in earlier on. Yeah. So yeah, I mean with Starlink, it's it's interesting. I have family members, they they found that it was better speed, better rate, and then than uh what was already hardwired there. Interesting. And uh I'm like, man, that is it's pretty impressive.
SPEAKER_01Yeah.
SPEAKER_03Um who would have thought sending so many satellites into space would be so lucrative.
SPEAKER_01Right, right. So yeah, my my uh my father-in-law, like 25 years ago, he he uh lived out in the country and um he had satellite internet because it was the only thing available. Um, and it was kind of through Dish Network or some offshoot of the Dish Network or something like that. And it was horrendous. I mean, it was it was slow, the latency was terrible. You know, you click a link and it would take yeah 10 seconds to like pop over to that that link and then load it. Yep. And it's amazing to see what's what's happened here in the last 10 years, really. Yeah. Um, just absolutely phenomenal. The other interesting thing that I read about the uh the data centers in space is that because you don't have to worry about the cost of cooling, which is a huge cost for it for a data center.
SPEAKER_03It's pretty cold in space.
SPEAKER_01It's pretty cold up there. Projections are that within roughly five years it'll be equal cost to build a land-based data center and one in space. Now, there's still some challenges with can you keep the speed as high and and all that kind of stuff, but that it was an interesting analysis that that I read about that.
SPEAKER_03Possible definitely bodes well for the future of humanity, I hope. Exactly. You know, we'll be uh soon uh have some kind of a base on the moon and that's right, uh data centers out there probably too. Yeah, yeah. So uh maybe Mars. I don't know about my lifetime. I suppose it's possible late, late, and yeah, perhaps, yeah. We'll see. Uh that'll that'll it could be pretty rigorous. But um, yeah, I mean it's it's interesting. A lot of things changing. It's almost hard to plan because uh, you know, we're getting all of these innovations and realizing we have solutions to specific problems that actually change the outlook of how we should strategically approach this stuff. Yeah. So speaking uh about the grid and uh some solutions, you know, especially with renewables, uh, a big problem is what do you do with the energy when it's not in use or when it's not needed, I should say. Uh we talked with Jim uh recently and uh Jim McConville, and he uh is over uh the ACES project. Um so I mean, what what is the the point of hydrogen? I remember Toyota being really into hydrogen-powered cars that never seemed to take off. I mean, what's the deal?
SPEAKER_01Yeah, I mean the the nice thing about hydrogen, like you said, is you can store it. And and the location of this this ACES project is above uh a huge cavern where you can store hundreds of days worth of hydrogen and then burn that in a in a natural natural gas type plant. Sure. Um it's a hydrogen burning plant, but the technology, the turbine technology is roughly the same. Yeah. Um so that's where it kind of enters into the energy mix. I don't know much about why the the car uh hydrogen-powered cars never really worked. But I do know that that one of the challenges that that we've had with hydrogen on the power generation side is that it's very, very costly.
SPEAKER_03Yeah. To convert it, you lose what half of the energy. I think that really is the the culprit behind the um, you know, energy within cars is just wasn't economical enough. Yeah. Um so I mean, with this project, I mean it's cool. It's cool technology. I I know where um our bread and butter is some of these combined cycle and that type of thing, but it is cool to see uh how wide some of our expertise is and what this hydrogen plant is. Is it the biggest in the world or is it?
SPEAKER_01I believe it will be, yeah, once it's fully operational. Yeah, that's pretty cool. Yeah, exactly. Exactly, exactly.
SPEAKER_03So is the idea essentially the power that doesn't get used, we're even if we lose some of it, you know, in that conversion process, at least quite a lot of it's saved, and it's saved very stably for, you know, long ago.
SPEAKER_01Exactly it, yeah. And then we can just, you know, instead of burning natural gas, we'll burn this hydrogen which is green. Yeah. Um, and so and and there's the uh the combustion process itself for hydrogen versus natural gas, you're just not emitting the same type of or or volume of just water, right? Basically.
SPEAKER_03Yeah. Yeah. Well, Alan, appreciate your time. And uh next time we're looking at an update, I'll be sure to uh reach out and and hope to have another great discussion like we just did.
SPEAKER_01Anytime, you know where to get me.