Loading summary
A
Foreign.
B
Welcome to the sarawik Podcast, where the world's energy leaders and innovators share insights on the future of energy, technology and climate. I am Atul Arya, chief energy strategist at S and P Global. In each episode, we dive into the critical issues and bold ideas shaping our energy future. So let's get started. Hi everyone. Welcome back to saraweek Podcast. I am Atul Arya, your host and joining me today is Bill Newsom, President and CEO of Mitsubishi Power Americas. Bill, welcome to Sarah Week podcast.
A
Thank you. I'm really excited to be here and the market's really crazy right now, so I'm looking forward to our dialogue.
B
Excellent. Thank you. I do hope you're getting some sleep. Bill. So let's start with what's going on, setting the scene and the speed of change. So give us your perspective on sort of the gas turbine growth for Mitsubishi Power and how does it compare to the global market? You are a big player, I think one of the top companies in the gas turbine market, correct?
A
Yes. I'd like to start with a perspective based upon McCoy data. That's public information out there. And I'm going to take it from a position of north and South America gas turbines sold over the last five years and give you a little data. And also heavy duty gas turbines. So that's 100 megawatts and above the market growth over the last five years. Let's touch it by saying in 2020 in north and South America, the entire market was 6.1 gigawatts sold for heavy duty gas turbines in north and South America. Now in 2025, entire market for the same sector, north and South America and heavy duty gas turbines was 40 gigawatts. That's a 6.5x growth in five years. And effectively the major growth is the last two years. So in 2024 is about 16 gigs and in 202540 gigs. And let me also state that our Mitsubishi power gas turbine sales followed by basically what that percentage growth was year over year. And now I want to compare that to the global market in 2020. The global market for all gas turbines sold of all sizes in 2020 was 40 gigawatts. This past year in 2025, the global market for all gas turbines was 100 gigawatts. So let's digest those numbers for a minute. In America's north and South America in 25 we had 40 gigs and that's the same as the total global market five years ago. So unbelievable growth that we've seen in the last five years. I look at this growth where we're at in three phases. Phase one is the last 18 to 24 months where as an OEM, we've realized how to spell hyperscalers and recognizing this is a real demand that we must service. And then we work to commercialize a big slug of work. So phase two is really looking at how do we execute on this big slug of work and recognizing, okay, is this growth continuing? Phase three is understanding the future and we need to automate, we must innovate and we must invest in order to strike a balance between sustainable business growth in the short term and then long term support for demand that's needed.
B
So let me just stay on this for a minute more. This was a massive growth for the industry and for you as mhi. So did you increase capacity? Are you increasing capacity as we speak in North America or beyond?
A
Yes, the demand right now is frothy and driving really longer build cycles. And so two things. One, the demand is more than all OEMs can service right now. And I've been in the industry over 30 years and I've never seen it quite like it is right now. The fact that maybe a few years ago, before this 2024, 2025 growth, we were seeing lead times for gas turbines about 20 months or less. And now we're seeing lead times be at about 36 to 40 plus months lead time. And that's because all of these hyperscalers are lining up to get a manufacturing slot. So you can't just flip a switch and increase capacity. It does take time to invest and grow. As I'm sure you're aware, both in our facilities and the supply chain and the Labor Force, all OEMs are in a mode where we're selling manufacturing slots right now. We haven't done that in more than two decades. So that's a big change. And so Mitsubishi Power and Mitsubishi Heavy Industries, we've committed, we've come out recently and committed to over 30% growth in our capacity. We're actually ramping up now to do more than that and also reduce our cycle time. So can we bring turbines to market even quicker? Yes, we're investing in the entire value chain. We're investing new facilities, existing facilities, training our people, doing all of that. And we're working to scale our operations to meet the demand responsibly and sustainably.
B
Very interesting how the market dynamics as you've described and shifted and how as Mitsubishi Power are responding. Let's step back for a little bit and look at the power Demand growth. And you already said what I want to say. The market is a bit frothy. We see lots of big hockey stick growth. What is your view on the power demand growth for the US we look
A
at this enormous demand growth in front of us and three key drivers that we're seeing. AI and data centers are about 45% of the big growth right now. And then the other 55% is reshoring of manufacturing and then electrification. We've seen over 300 new manufacturing facilities announced. As we ramp up and scale up our factories, we need more electricity as well. If you look at the last two decades in the United States, our electricity demand growth year over year has been effectively flat less than 1%. And if you look across the United States right now, there's an inflection point in 2025 where we're looking at 3% per year growth in the entire United States, an average. And in pockets in the Southeast where we're building a lot of these data centers, we're seeing maybe 8 to 10% growth in those pockets compared to what it was 20 years ago. So the question is really how long do we see this continued growth in the United States? And based on contracts we're signing and the demand to service those, we're confident we're going to continue to see this growth for at a minimum, five to 10 years. And we're hyper focused on hiring, training and developing our team to support this growth. We are focusing on new facilities and expanding our facilities. And we recognize we must ramp up to meet this demand in the market. Also, as we build these new facilities, our fleet size is going to double over time and we're going to need to be able to service out in the market. So we're planning and focusing on that. Our big focus is on making sure we execute with excellence and deliver for our customers.
B
One thing I want to go back to very quickly. You know, you mentioned the 45% hyperscaler, 55%, everything else. I think that's a really important point because I think the public perception, if you're not an expert, is all driven by hyperscalers. That's not really true. Correct.
A
You're exactly right. And we're seeing new manufacturing facilities. We've got a huge facility in Savannah, half a million square feet. There's a Hyundai has a big facility. They open there for EV cars. We're seeing LNG terminals in Texas. We're supporting Entergy as they build for industrial customers. So it's not only AI, and you're exactly right. Some of the mantra out in the marketplace is win the AI data center race, and that's the only thing. And also worried about cost. In line with that, there's definitely a manufacturing industrial growth reshoring here in the United States, making sure we have security of this supply chain in the United States. And that's a big push.
B
One of the interesting things I observed, Bill, is the different time frames in the different sectors work. Right. Hyperscalers were just by the name hyperscaler at Hyperspeed. And when we think about power companies, we think about them as stable. I don't want to say boring, but really stable. And there you are in the middle of this group, power companies and hyperscalers, and being pulled by everybody. So how is this working? Give us a little bit of an insight into the conversations you're having with these very different industries.
A
You're exactly right. And you touched on a very impressive dynamic in the market. And really the key is speed to market and surety. Those are key factors today that we're all focused on in the marketplace. Hyperscalers want to win that AI race, and they have this ferocious need for electricity to power these data centers. They're building these facilities, and we need to continue to build power plants to support them. When you look at the market, as you touched on, there's about three to four major hyperscalers, a couple trillion dollar market caps. You have about three to four OEMs, and then in the middle, you have hundreds of developers and utilities that are all working on real projects. The question is, what makes a real project? We're over here, the OEMs are over here saying, okay, look, what makes a real project is you got to have a gas turbine to power it, and then you've got to have an offtake agreement with a hyperscaler to fund it. At Mitsubishi, Power America and mhi, our role is to work with our partners so that we can scale up and meet the demands of the markets. We're hiring two people a day, and we're growing by 12% this year so that we can meet the demands and scale up our business. We also are focusing on standardization. This helps us to not have to recreate the wheel. We can cookie cutter these designs and not have to recreate it each time. It helps us get it out quicker and it helps our EPC construction partners so that they can build the same thing again and again and we could replicate it. Otherwise, we have no chance. And helping these hyperscalers win this AI
B
race, are you seeing the impact of that Sort of standardization in your speed already. You mentioned that earlier also that you're trying to reduce the sort of the cycle time, if you will.
A
Absolutely. Let me just give you an example. I was going through quarterly project reviews last month and we have a contract requirement to provide 30 drawings to our EPC partner. And we provided over 300 drawings. EPC partner since. Thank you so much. We're really appreciative that you're giving us more information earlier so that we can get ahead of schedules. We can continue to duplicate and standardize. We can get ahead. We can help them get ahead. We've got another project where it's even before all the big slug of AI projects are going in the ground. And we're seeing some of those projects stub their toe by poor or lack of abundance of labor in the field as well as late designs and things like that are causing some delays in schedules. So if we can get ahead, we can help our partners be successful. And across the board, we all want to be successful together.
B
Is the power industry from. You've been working with the industry for a long time. As I said, is it changing now? How is it evolving as they see this kind of rapid change? Working with you, working with hyperscalers?
A
Yes, I'm really seeing a huge shift in the power industry today from what it was previously where we were focusing on lowering emissions to now we're focusing on extreme demand growth and electrons as quickly as possible. Power. Now this is somewhat changed. What's changed is pragmatism and the AI market drivers and reshoring and electrification. These all need to be serviced by baseload generation. Renewables can't service 24, 7 through 65, so we need to bring baseload generation as quickly as possible. We're still focused on clean, reliable and cost effective generation. Absolutely. And you know what's interesting is each state is racing to attract economic growth in their state and that's being driven by being able to build power plants in their area. And that's something that we haven't seen in many years. We're seeing the US power generation industry drive economic growth across the United States and it's impressive. We're also utilizing hydrogen and gas turbines to continue that decarbonization goal. We're excited about reaching FID on Broad Wing project, which later this year will reach fid and that'll be the first large scale gas project with carbon capture on the back end. We also are excited about our Asus Delta project up in Delta that should go commercial this year. As well, which will utilize green hydrogen. So we have not gone away from those goals of decarbonization. The goals of providing clean, cost effective combined cycle natural gas is the answer to solving the challenge with a path to clean.
B
As you said, the 2040 target, you haven't changed, you're still committed to that. It may require a slightly different path to get there as you're responding to the demands of your customer.
A
100% we're committed to that. We've got to do it in 2040, net zero by a company and develop the products and the solutions so our customers can still meet their goals in 2050. So pragmatism has set in and we must build electrons now. And the hyperscalers have a social commitment to decarbonize. And so we're bringing to them the carbon capture and the solutions to be able to allow these growth and baseload generation to meet the demand needs today.
B
I'm also seeing that even the oil and gas companies are talking about the baseload ccs with power. You've seen big companies here in the US saying we will build a gas power station, but with CC has, which is a good thing to know, isn't it?
A
It is. And I think there's a huge focus on costs of electricity and getting those electrons in the ground today. And in a market where demand is outpacing supply, there is inflationary cost that drives for that. And I think utilities and regulated markets are able to really defer and align that cost to hyperscalers and contractually have them pay that additional cost because they're the ones demanding it and not the ratepayers and customers in their prospective states. And I think that's really important to recognize in this big growth up that we have as well as hyperscalers are going to hold us accountable to meet the decarbonization.
B
Do you see opportunities for bit of a better alignment? You have power companies, you have equipment providers, hyperscalers. And then we haven't talked much about the policymakers and you're already a leader to the cost issue. And I've also heard and read that the hyperscalers are very sensitive and they want to help defray any kind of cost increases. But do you see some more opportunities for everybody to work together and address the speed issue on one hand, but also the cost issue on the other hand? Cost, particularly for the consumers?
A
Yes. I think to meet the demand in the market right now, we must have alignment and trust to work quickly. Minimizing the uncertainty and the risk is how we're going to be successful and driving those stronger partnerships and collaboration with transparency across, as you mentioned, the hyperscalers all the way through utilities and down to OEMs and EPCs. So every day I wake up trying to discern how big and how long this market continues and understanding what are the real project opportunities. The relationships and deep trust are critical to complete these projects on schedule, on budget and safely. When we talk about minimizing some of the risk and uncertainty, having some understanding of where policy is going to go with the United States is important as well. So that we can invest in long term facilities and know that the market's still going to be here. It's not just here today. And in five years the market changes, or four years after this administration changes, the market totally changes. And so I have confidence that we're across the spectrum, we still want to win the AI race and we're going to continue to build manufacturing and industrial growth in the United States. And in doing so, I have confidence that the policy will help support that.
B
And also, as you alluded, it is impacting the GDP growth, right? It's helping the US still have a positive, pretty strong GDP growth last year and our own S and P forecast for this year, also strong and to a large extent on the basis of all these big billions of dollars of investment, isn't it?
A
Yeah, it's amazing the numbers that are going to be invested in the United States and the GDP growth. You touched on that. I was reading an article recently and there was a comment on a power plant next to a mega data center. And the cost of the power plant with carbon capture, it's a couple of gigawatts was about 10 to 14 billion dollars. The cost for the data center is about 45 billion dollars. And then the cost for the chips is 100 to 120 billion dollars. These numbers are just staggering. When you and our traditional businesses as we focused on plant by plant, when you talk about these numbers, it's staggering.
B
It's mind boggling. By the way, I do have a question for our audience. I actually grew up in power business. My father built power stations in India. I'm just curious right now, what is the biggest single power station you're supplying turbines to? How many gigawatts is that? Is it more than a gigawatt?
A
I was at a groundbreaking recently and this is all public. You know, I'll just tell you how impactful it was for me. It's in Louisiana and I got to the site for the groundbreaking and as we're driving along the site, we're driving for five miles to get to the back of where the power plant's going to be. And this power plant is right behind one of the world's largest data center facilities. This data center facility is for Meta. It is five miles by one and a quarter mile wide. And the perimeter is like 13.6 miles around the perimeter. So again, the size of these facilities is unbelievable. A lot of these facilities you start with a one on one. So we're starting with two one on ones and then potentially adding two more one on ones to one on ones is a little above a gigawatt and then two more would be above 2 gigawatts. So we're seeing one on ones to be a good selection because we can standardize on that gives you the flexibility to duplicate and replicate.
B
Wow. Very impressive. Congratulations. A couple of other questions before we wrap. We haven't talked about nuclear yet and of course Mitsubishi mhi. You are a big investor and developer of nuclear technology. Tell us about your thinking on nuclear.
A
Absolutely. On mhi we have a large nuclear division that is growing significantly due to bringing back PWRs in Japan after the tsunami. And we have long standing history with that technology. We are looking at large scale new nuclear and Japan over time and we're looking at SMRs as well. For nuclear I think there are many considerations, regulation, developing and certainly the cost to build large scale nuclear makes economic sense. If we could manage the cost to build the last plant built was took 15 years and about three times the budget. That's just not sustainable with the risk of cost overruns for these big projects like that, a single utility would be very hard to manage building these large nuclear projects. I think for large scale nuclear in the United States you really need the government to stand behind those and make backstop the risk of construction. I think the operation risk is not a big challenge. It's really that cost overruns and front end and getting the NRC challenges through. And so if we can shrink that down or minimize the risk, I think that's the big challenge. But SMRs, the small margin reactors, it's a great concept. I'm a nuclear proponent, I've worked on nuclear projects before. I just think the economics are challenged. At $20,000 a kw it's hard to make that dog hunt. And I know some will get built. I think they're going to be test facilities. But when you're looking at $20,000 a kw and the first plant hasn't been built, I think that's challenging and we get to nth of a cost will we get there? It's over the years in the 70s, we never really got there in the nuclear market. So I do think it's challenge. I think we will build nuke. It's just going to take time, 10, 15 years from now before you're going to see any of that really happen. So I'm really bullish on gas. Now we know how to build that, we can manage the risk, we can scale up and then we can clean it up. So I think we have the solutions for that right away just for the
B
benefit of our audience. When you said pwr, what we mean is pressurize water reactors of the people who are not into the nukes. So Bill, this has been a great conversation. But before we wrap, you're coming to Sarah Week in about a few weeks. What are you looking forward to at Sarah Week this year?
A
Every year I'm excited about going to Sarah Week and so every day my challenge is I wake up and I'm working to try to understand how long, how big and what are the key levers to be mindful in this current market situation. And one of the things that I'm uncertain about is the business model for hyperscalers and data centers. How does all that work and how sustainable is it over time? Those are questions in my mind that every day I wake up to try to understand. So I'm really looking forward to having some deep conversations with folks from all different perspectives of the industry. That's why I enjoy being part of SARAH Week because it's a forum for many different players to come together, debate, discuss, challenge in front of each other. This all helps me personally discern whether we're executing our right strategies or if we need to look at things from a different perspective. So I really think the self awareness is extremely important. Being self aware about where is this market going, how are we able to participate in this market is critical to our long term sustainable business. And I reflect and I talk to my team all the time about the moment we're in and the ability to serve our customers. The impact and difference we can make in everyone's lives. This is an incredible opportunity and we're blessed and excited to be part of it.
B
Yeah, future generations are an empower generation, just to be very clear. Bill Newson, thank you very much for joining us from Mitsubishi Americas, Mission Power America. It was a great view. I think one of the things we are looking forward to is having you other companies from across the various players we talked about and come together and say, let's be sure that something is going to be delivered and will create value for all of us. So that's we are looking forward to at SARAH Week and having you and your colleagues at sarawik. Thanks again.
A
Thank you. I really appreciate it.
B
Thank you for joining us on the sarawik Podcast to stay connected with the ideas driving change across energy and technology. Subscribe, share and rate this episode. It helps us get the word out. Let's continue having impactful conversations. I'm Atul Arya until the next time,
A
the saraweek Podcast with Atul Arya is brought to you by Sarah Week, the
B
world's premier energy conference.
A
Be part of the conversations moving the
B
world of Energy Forward, March 23, 27, 2026 in Houston. Discover more@saraweek.com.
Date: February 24, 2026
Host: Atul Arya (Chief Energy Strategist, S&P Global)
Guest: Bill Newsom (President & CEO, Mitsubishi Power Americas)
This episode features a dynamic conversation between Atul Arya and Bill Newsom, exploring the extraordinary surge in power demand across the Americas, the pivotal role of Mitsubishi Power in scaling up infrastructure, and the unprecedented impact of hyperscalers (large-scale data center operators) and reshoring of manufacturing on the energy landscape. Newsom articulates the urgent need for speed, collaboration, and both pragmatic and innovative approaches as the industry contends with extreme growth, supply chain constraints, and decarbonization imperatives.
Industry dynamics:
Standardization & replication is critical:
Balancing emission goals with rapid build-out:
Cost Allocation:
Bill Newsom paints a vivid picture of an energy sector stretched to new limits, balancing hyperspeed requirements, massive investments, and the imperative to decarbonize. Mitsubishi Power’s response is practical and ambitious: innovation in scale and speed, robust collaboration, and a clear-eyed focus on meeting extraordinary power demand with both traditional and emerging technologies. The episode offers indispensable insights for anyone tracking the intersection of energy, technology, and global economic development.