Hi all. I'm Hanna-Maria Heikkinen, I'm Leading Investor Relations Wärtsilä. Well, warm welcome to strategy call with Wärtsilä CEO, Håkan Agnevall. The purpose of this call is to provide equal Q&A opportunity with Håkan, and there will be no material new information and no slides on this call. The purpose of this call is to focus on our long-term business opportunities.
As a reminder, we will host a pre-silent call on the October 1st, meaning next week, together with our CFO, Arjen Berends. So let's leave the questions related to detailed financials, that call. On next week's pre-silent call, we will also discuss the details of our energy storage joint venture. So let's leave all the questions related to energy storage joint venture also to next week's call. Håkan can discuss.
Yes, and warm welcome, everybody. It's time for the CEO call again. If we continue the practice, I try to give you an overview of where we're heading and how we see the market evolving, and then we open up for Q&A. If I start overall, I think we continue to see a very strong demand on the markets. If I start with energy and then go to marine. On the energy side, yes, its data center growth continues in a good way, I would say. High level of activities. We have a dynamic portfolio of opportunities in various maturity stages, but it's evolving.
There are new orders coming in, and if and when we agree with the customers, we will of course communicate those orders. But as you know, there is always a time lag between or normally there is a time lag between when we sign and when orders are announced. But I would say that the data center demand is holding up. I like to highlight, once again, it's not only about data centers in Wärtsilä energy. Balancing power are also very strong. U.S., absolutely, midsection of the U.S. Australia coming very high level of activities, a lot of tenders coming out. Starts to come in certain countries in Europe as well.
So Finland is one example, but there are more examples to come. So the balancing narrative, it's really supporting us. On the traditional baseload side, also good levels of activities in our traditional markets, Southeast Asia, Latin America, et c. So many drivers coming together. It's the general electrification of industries. It is need for more cooling in the hotel worlds, air conditioning, it's data centers, it is the aging energy infrastructure, certainly in the U.S., but to some extent in Europe as well, and it's the balancing power. It's all those opportunities together that makes us optimistic.
This is also why, as you know, we have been investing for the future, expanding our capacity. Then on the marine side, also demand is holding up in a good way. You probably saw the latest Clarksons data, which kind of, it's an external proof point of the story that we are conveying, and that is our core segments are performing very strong. We are clearly not in tankers. Tankers was up a lot in the Clarksons status, and that's definitely not our core segment. But cruise, LNG will come probably in 2027.
Offshore, even containers is hot, so to say. I think there are good opportunities going forward. I think Hormuz is having a little bit of an impact on our service business. Not huge, but there is a little bit of slowdown because there are one type of category of customers, the fuel prices are up, and then they tend to postpone certain maintenance forward because you have a budget and, if you spend the budget on fuel, you prioritize that for a while. It works for a while, but not for the long run. There are also some of our customers in container side, they have very favorable rates, so they postpone maintenance for those reasons. So the business are not lost.
Definitely it will come, but there is a little bit of slowdown on that. But in the end of marine, it is holding up. On the service side in general, you saw Q2. For marine and energy combined, order backlog was up double digits. Marine was a bit flat if you look organic, but that was based on a high comparable quarter last year. As I said, there is a little bit of the Hormuz impact in that as well. So overall, our growth narrative on the energy side is also continue to hold up going forward. What is new, so to say, and what we have made public. We have the cooperation agreement with Tokyo Gas.
These are not orders, just to highlight that. But we have a longstanding relationship for several decades with Tokyo Gas, and we have supplied engine power plants for them in Japan. They are now identifying data center opportunities in Japan, and we will work very closely with them. It is a fantastic partner for us to have long-term relationship and very well-established in Japan. It also underlines a little bit what we have been saying before that, yes, there is a lot of growth in the U.S., but there will be a growth in several regions going forward, and I think this is one of the concrete proof points for that.
So that is one thing. I know also that on the data center side, there has been a lot of discussion about how will the off-grid segment grow going forward, and there are different kinds of statistics. Everything from 30% even to 50% of the new build for data centers in the U.S. will be off-grid or behind the meter. I do not think nobody knows the exact, but it is going to be a significant growth avenue. Also, this whole notion that six, seven, eight years later, there is a grid connection coming in. What does that mean for Wärtsilä? You can go back to our data center call earlier this year.
This is going to be good and this will be good. Why I say that with a certain surety is that when I talk to some of our data center customers, they are envisioning this, and then they will use the power for, or they will actually add a couple of modules to the engine power plants, and they become an actor on the grid. The key thing to be a competitive actor on the grid, you need fuel efficiency and you need flexibility. As you know, these are some of our strengths of our technology. So we clearly see that.
Also the other third element that we identified with the increasing share of renewables. I mean, the hyperscalers, they for sure, they would like to go green and renewables will come. It is a very affordable source of energy, and then you can use the engine power plants for balancing. So in all of these scenarios, I think we are having a strong technology.
We also come back to the Q2 notion of we are building an order backlog with good profitability. We have good opportunities for price realization, and we disclosed earlier, in Q2, that the new build order backlog in energy. If you look at the margin there, it is up 500 basis points compared to when we started in 2025. So it gives you a certain indication. I think that is a short summary, and let us open up for questions.
Thank you, Håkan. To the Q&A now. If you have a question, please use raise your hand functionality in Teams. If you cannot use the Teams functionality, please send an email to me. The first question comes from Sven Weier. Please go ahead.
Yeah. Hi, Håkan. Hi, Hanna-Maria. Thanks for the call. Few questions if I may. The first one, Håkan, is on the Texas update we saw this week that permissions are put on hold. Was just wondering how that impacts your order pipeline. With a longer-term view, we have also seen that water is going to be much more important in the decision-making going forward. So could that actually end up being a positive outcome for you? That is the first one.
If I start with that, because it is a very dynamic environment, that is why I make the caveat. My understanding of the regulatory or the announcement that was made by the governor, that was related to if you wanted to hook up to ERCOT, to the power system. There is this one-year consideration period, or I do not know the right formal term. It does not affect off-grid. This is our major market in Texas on the data center side. We are on the utility side, but there we are on ERCOT, but more for balancing power. So when it comes to data centers, it is off-grid in Texas, so it does not have a major impact so far, at least. What was the second part of your question, Sven?
It's just on the water consumption, because I think that was now featuring more prominently.
No, and that is a big advantage for ourselves, clearly, because we have a closed-loop cooling system, and you need to top it up a little bit. But we consume thousands of times less water than the gas turbine competition. As this becomes more into consideration, it's an advantage for our technology, clearly.
Another question I had was just because you said with a view to data centers, there are new orders coming in, but you haven't announced one. Should we take that as a sign that you booked orders but haven't announced them yet? Is this completely fair?
Yeah, I would put it like that. We continue to take in orders, and like always, like as many times, there are delays between us signing orders and announcing them. Overall, the order intake side on the data center side continues to look good.
Understood. Thank you, Håkan.
Thank you, Sven. The next question comes from Antti Kansanen. Please go ahead.
All right. Thank you. A couple of questions from me as well. I start with the cooperation with Tokyo Gas that you, Håkan, highlighted. You said that you have worked with them for a number of years already. What is really different with this cooperation? Would you be able to talk any kind of the market opportunities or the Japanese data center power situation in general? Are they suffering from the same issues with the grid? Are they getting baseload off-grid solutions from you, or how does it look like?
Basically, our long-term relationship with Tokyo Gas is to provide baseload, but actually balancing power. If you look at the most recent orders, the last, let's say five years. We have a relationship with them for, I think, 30 years. In those, if you go about 20, 30 years, it was smaller power plants for baseload in different parts of the Japanese energy system.
Yeah.
The last five years, it has been balancing power plants. The latest one is Sodegaura, and it is a 100 MW, 100 MW, 120 MW balancing power plant in the Japanese system. Why Japan? Because they have actually created a capacity market. They have also decreased the granularity where you evaluate the power swing. I think they have a five minute time consideration when they see how the power is moving. That is why we are so established with Tokyo Gas in the Japanese power mix. What is new here is that Tokyo Gas, and you could talk to them, they see the demand for data centers in Japan.
There will be data centers coming in, so there will be need for power. I think there will be a mix between both on-grid power and off-grid power, so to say. Here we need to see how it evolves. I think people are getting formulated. I think for me, what I am exciting about, Tokyo Gas is a very well-reputed player in the Japanese system, and we have worked with them for decades.
When they do something, they do it seriously. I cannot go into details of how many megawatts et c or gigawatts it will be. That is too early to say. There is certainly potential in the Japanese market. Of course, Japan is also. We also know the general narrative in the world is national sovereignty. Of course, Japan wants to have data centers in their own soil.
Okay. The second question was more broadly on the timing and deliveries on the data center side. I guess you are almost or fully sold out for 2028 and filling up 2029 right now. Could you talk a little bit about what the clients are asking? Is there fear that at some point when you are filling up 2029, the delivery schedule is too far out and that might cause some type of flutter? How do you stack up versus your engine competitors in terms of delivery schedules?
First of all, I can confirm that we are sold out for 2028, and we are selling 2029. Started to sell 2030 now as we speak. Yes, if somebody can bring a very short delivery time in 12 months, they certainly have a competitive advantage. I cannot comment competition so much. I can just comment on our own pipeline and what we see. We see a trend that continues. Right now, as we are moving towards the end of this year, and now we start to sell end of 2029, beginning of 2030.
There is a window moving forward. You cannot sell in. We have not been able to sell in 2034 yet. My key point here is that this moving window, let's say, you talk about three years delivery time or however you want to position it continues to move forward in time, so to say.
I mean, it's not an issue if you are talking about, let's say, ability to deliver 2030, 2031, 2032. There are still other constraints on the projects, so are kind of still well in line with where the customers are.
Today, we are not selling for 2031. We are selling for 2029 and 2030. My key point when we talked six, seven months earlier here, is that then we were selling end of 2028, beginning of 2029. Now six months later. That's why I said this window keeps on-
Yeah.
It's not stopping. It's not that all of a sudden, "Oh, we cannot sell anything because we have too long delivery time." No, that's not the case. The market is digesting the delivery time, and we continue to have good strong order intake.
The last question was on the service contracts for these data center power plants. At what point of time in the delivery schedule or the startup schedule would you expect to get those service agreements, if you expect to get them? Is it before the warranty period or is there a two-year warranty already baked in, so after the startup you will start to then discuss on the longer term? How should we think about phasing of those announcements?
I think that there is not one story fits all customer. But if I talk about the clear majority of our customers, it will be very early. We are in discussions on service contract with certain number of customers that have placed the new build orders that we have announced. These service contract, they are in the very close future, so to say. It's not going to be that this will wait for two years or something. It's going to happen. But right now, I think many of our customers, they are so busy signing new build contracts. So that's why they haven't signed immediate service contracts. But there are clearly negotiations ongoing, so there will be orders. Absolutely.
All right. Thank you.
Thank you, Antti. The next question comes from Uma Samlin. Please go ahead.
Hi, good afternoon. Thank you very much for taking my question. Thank you for hosting this call. My first question is, you mentioned starting to see also very strong demand on balancing power in the U.S. Just wondering, is that also data center related? How do you see a market for balancing power for the data center space? If so, how big is that market do you anticipate that to be?
The general narrative on balancing, especially in the midsection of the U.S., is not new. It's been with us for a couple of years already. I just want to underline that continues, and it continues to be a very interesting market. The major driver for that market is cheap energy. At the end of the day, wind and solar provides very competitive power, but you need to balance it. That proposition, balanced renewables, so you combine, let's say wind farm with some batteries, with thermal balancing. If you look at the LCOE, the levelized cost of electricity of that, you can say system, it is very competitive.
That's the key driver. Now, on top of that comes now the data centers. Of course, we know there is a lot of data centers in Texas and that is on top. But I would say in data centers, and you know it's everything about delivery time. Right now, in general, there is not so much focus on renewables. But as I said, over time, clearly, the hyperscalers would like to have renewable power. But right now it's the rush to power. The balancing narrative so far from our side has been more the traditional balancing, the traditional utility balancing power narrative.
Yeah. That's super clear. Thank you. My second question is on pricing. You mentioned during the Q2 results that you saw 500 basis points of margin increase. How can we think about that going forward? Do you see continued increase there? Have you started to see customers pushing back on prices? Just any update on that would be great. Thank you.
I have been also, I think, I hope, consistent in this message. We have good opportunities for price realization, but at the end of the day, the business cases need to make sense for our customers, and it is always a balance. We have had good price realization, we have had happy customers that are excited about the projects that can get the businesses together. I continue to see good opportunities for price realization for us.
That is super helpful. Thank you.
Thank you, Uma. The next question comes from Anders Idborg. Please go ahead.
Thank you. Good afternoon. Just wanted to ask a bit about, it was good to see the divestments of the final portfolio businesses come through in June. I was just wondering if you could open up a bit more about the structure of those deals. I saw in terms of cash flow, there was a negative in Q2. Have you provided for contingencies in existing projects? Can we be certain that there are not other project risks remaining in the sold assets? Just wanted that clarification.
No, I think without going into all the super details, then I will have to refer you to Arjen next week. Sorry about that. But overall, to what I understand is the core of your questions, will there be negative surprises from those divestments that we have made? The reason to no, they are closed.
Okay. That's good to hear. The other question, because it's early in the ramp-up of Vaasa, and of course, as we have looked at the delivery schedule, the idea is that it will accelerate in 2028 rather than in 2027. So in terms of phasing of the cost of the ramp-up, do you think that will be matched pretty much with deliveries, or do you see a period where costs increase before those deliveries take place?
No. So basically, just to clarify, then come to your core of your question. So yeah, because I know that there's a lot of interest. So just the data that we have provided. So 2025, we were running at 75% of technical capacity. This year, we will be running at 100% of technical capacity. I'm talking about Vaasa now. Then we will, 2027, primarily running on 100%. 2028, we have kicking in the first up 30%, 35%, then beginning of 2029, the second 30%, 35%. So you could say that's the ramp-up, if you know. Then when it comes to the investments, et c, we capitalize some. It's not going to hit the cost in one major blow, so to say. So you could say it's spread out over the revenues and over time.
Okay. Good to hear. Okay. Thank you.
Thank you, Anders. The next question comes from Sven Weier. Please go ahead, Sven.
Yeah. Thanks for taking the follow-up question. Was just coming back, Håkan, to your comments on the marine service side, where you talked about a bit of a softening. Was just wondering, in the first half, marine services was down 4% on retrofits. Is this now more going into the spare parts bit, and is that a new phenomenon that you've seen in Q3?
Yeah. Q2, you have to point it out. I think we were 1% organic in Marine. Then we were down on, if you consider the FX effect, et c. But it was a little bit slower growth, and we highlighted that one thing was the major thing there was actually a high comparable quarter last year. Then we also said, and as I said, now Hormuz have had some impact in postponed maintenance. You already started to see that in Q2, and I think we will see this postponement for a while, because you can postpone certain maintenance, but then you need to do it. You will probably see this in this quarter as well.
But it is not like a sudden thing that happens in Q3. It is more continuation of what you saw in Q2 already.
Right. I think, I hope it's understandable logic when some of our customers, because for instance, the container rates, they are very high now. Then of course they take opportunity to really operate and make a lot of money. Then they postpone the maintenance. Then you have other customers that where the fuel and increased cost of the fuel, it's a significant impact on their operating budget, so to say. Then they reallocate the budget and put more money on fuel and the maintenance, and they have to postpone the maintenance. Then of course, well knowing that they're taking additional risk by not maintaining and at a certain stage they will need to maintain because otherwise there might be serious things happening.
Yeah. Makes sense. Thank you, Håkan.
Thank you, Sven. Next question comes from Antti Kansanen. Please go ahead.
Can you hear me?
Yeah, please.
No, I guess it's Yeah. I wanted to come back to the capacity expansion plans on a longer term. Obviously there's a bit more concerns now that everybody is building up over capacity in the market as the short term kind of a data center demand is driving a lot of excitement, and you are doubling your delivery capabilities and your competitors are doing the same. How would you kind of address these concerns over the long term? If you think about kind of the magnitude that you decided to do in Vaasa and also with your suppliers, how much are you doing yourselves?
How much are you adding to suppliers? When you kind of made those plans, how did you look at the long term kind of prospects, especially on the energy side, beyond the data center side? How much was driven by all of these factors that you are already discussing regarding electrification and coolings and things like that? How are you kind of addressing the concerns that the industry is over-investing right now?
Yeah. First of all, if we look on the demand side, and I try to highlight this is a broader demand side than just data center. Clearly data centers are contributing, but the balancing power narrative is really strong. What I'm excited and encouraged with it's in the U.S., but now it's coming also in other countries. In Australia, now for the first time, there are engine tenders coming out for balancing power. There's an acceleration of the shift from coal to renewables in Australia, and it's in several parts of the country. Australia have been one of the pioneers in batteries, and that is good.
But they have also understand that it happens to be the Wärtsilä message as well, that you cannot only do balancing with batteries. You need thermal also. So now as a result of this journey, there are concrete tenders coming out for engines, not source gas turbines, for engines. That is new. That's a very concrete. As I said, a couple of countries in Europe it starts to happen. I take Finland as an example, which is fairly new. We delivered our first balancing power plant in Finland ever.
We have just delivered it and it's operating. I think the operating experience there is triggering a lot of excitement and people are making money, etc . Now we see there are also new tender opportunities coming up, very concrete. So there are those kind of proof points on the whole balancing narrative. Then on top, as I said, we have our traditional markets like Southeast Asia, electricity demand is growing, electrification of industries, et c.
Quite frankly, our expansion is clearly not only driven by data centers in the U.S. Clearly not, because then you would probably have a different kind of risk, reward assessment, so to say. It's driven by this broader narrative. We do think, we have visibility. We are sold out for 2028. I think we have fairly okay visibility for 2030. So it's really what's going to happen until 2030. So what's really going to happen in 2031- 2035 and beyond. We think that the fundamental trends will still be there. That's a I would say that's a fairly unique perspective that you can have this long term.
Now on the supply side, if we go there, clearly, gas turbines are ramping up, we are ramping up. We saw the HiMSEN announcement. We've seen the MaK announcement by Cat, et c. Now, so that would be capacity coming in and let's see when demands meets capacity, right? That we can have a long discussion. I don't think anybody knows. Now, what makes me excited about the future, then I'm talking beyond 2030, is a couple of things. One is this notion I've been talking about before, and I say this a little bit colored by the U.S. situation.
When customers have tried the new candy, medium speed, not high speed, medium speed reciprocating engines with high fuel efficiency and high flexibility. They like that engine, and they will come back. So I think that structurally, this is a great opportunity for us to grow our market share, because more end customers will be acquainted and familiar with our technology. That's a great opportunity for us because the fundamentals are there. The fuel efficiency and the flexibility.
That's why I see that as a I like to keep the gas turbine guys in the lake, and we will continue to do so. Because we have in our sweet spot, when we talk about it, the 20 MW- 600 MW, 700 MW, 800 MW , we have a very competitive technology. Yes, combined cycle gas turbines, we will not be able to compete with them, but the industrial and the aeroderivatives, ladies and gentlemen, we have a winning technology. That's one thing. The other thing, then on top of that, I talk fuel efficiency, but now we have water coming in as an increasing issue. We have thermal derating in a hotter world, et c. So there are a number of those parameters. It's nothing new. Then, of course, you could say, "Okay, Håkan.
We believe that reciprocating medium speed engines actually has a vital role to play in the mix going forward. What about your medium speed reciprocating engine competitors? Then I say, "We are a technology leader. We have the most energy efficient medium speed engines with the best flexibility, and I dare say we have the best service network in the world. We have been in the U.S. for decades. We have been in the U.S. for more than half a century."
That's fairly unique among our other competitors. There are a couple of those strengths. Let's see how, Antti. We will follow our, hopefully, the coming years, and we will see how this plays out. I think we have, and that's why I like this. We have fundamental, concrete, competitive advantage. If we play this right, and we grow in a sensible way, in a robust way, where we deliver, we serve our customers with uptime, reliability, and fuel efficiency. We have something really good here.
I want to follow up on one thing. Let's assume that you have the best medium speed recip technology there for fuel efficiency and uptime and all of that, and you have clients who agree. If we talk about balancing in Australia or in the U.S. or base load in Southeast Asia, what are those guys' investment horizons or delivery times currently?
There would be somebody who is maybe not as good as you, but close, but they can deliver six months, nine months earlier. Does it matter for those guys, or is it only the D.C. guys where the time to power is so critical? How do you balance on being on board with those maybe more traditional guys who follow their own schedules versus maybe taking those kind of high margin D.C. deals right now?
I think, a customer, regardless if they're D.C., if they're utility, they will always appreciate a shorter delivery time, compared to what it is right now, clearly. I think there is not one answer to your question. It's like a balanced decision for customers where they take, "Do I take a penalty in the life cycle economy of the plant because I will pay more fuel?
How do I weigh that towards having a delivery earlier?" Right now, our medium speed reciprocating, I'm not aware that they have significantly better delivery time than us. Let's say if they would, there is not one answer to that. It's how different customer evaluate life cycle economics versus time. On the data center side, that valuation is clearly skewed to delivery time. On the utility side, I would say it's more balanced.
Okay. Thank you.
Also, I got some question on this about HiMSEN, and of course you should ask them. My understanding that they're going to put up a house in a year, one year and a half. But it's going to take them to 2030 to ramp up the capacity. So yeah, let's see. But you should ask them.
Will do. Thank you.
Thank you, Antti. The next question comes from Adrian Hill. Please go ahead, Adrian.
Yes. Hi, everyone. Now you can see me even. Just a question on the ramp-up that you're doing at your Vaasa site. I was just wondering if you can elaborate a little bit on the flexibility that you have at this site over the next couple of years, and how you do it in terms of, will you have some flexibility in terms of temporary staff on lines that you can put on, put off, depending on the demand and how it evolves over the next couple of years?
Which is actually a question contributing to what has been asked before about if the demand at the longer end is getting a bit more difficult. That's the first one. The second one is c oming back to Tokyo Gas, obviously, they must be aware of the scarcity of your production slots. You're saying it's not directly coming, any orders or so, but what would be the next step, how to evolve this partnership, then eventually leading to orders? What would be your wish list on how that progresses? That's it. Thank you.
Well, so if I start on the flexibility, and this also comes back to answer this earlier question on competitiveness and how do we build that. So it ties to that. Because we have an industry system where we build a lot on the suppliers. So you could say it's a little bit automotive style, but it, of course, with much less volume. We do R&D, we do engineering, and we do final assembly and testing, but we buy a lot from our supply chain. And of course, we do certain in-house machining, but a lot is done on the supply chain.
And of course, that gives a certain flexibility because you share the volume swings with the supply chain. The other area where we are clearly and that's nothing unique. With the exception now of HiMSEN, nobody has been building new factories. If you look at the gas turbine players, and also if you look at the other reciprocating players, they have all been extending or rejuvenating or revitalizing their existing or former facilities. And our strategy is the same.
Because that is one way to deal with, to create flexibility. Because obviously, if you set up a new factory, you introduce new fixed costs, so it increase your general fixed cost levels, so to say, in a different way than you scale up an existing plant. And using consultants or temp labor, it's part of our normal mix, so to say. Like it would be in automotive, whatever. You have a certain share. We have always had that, and it's one way to deal with flexibility. Then I will also highlight, clearly, we have a very good, strong relationship with our unions. Scandinavian unions, I commend them for their maturity.
Of course, you always have the battle of terms, so to say. But overall, when it comes to structural measures, I think if you follow statistics, you have the environment. I think we have a very favorable union. Because we have unions that think about and just clearly realize, how do we secure the most number of jobs over time? And then sometimes you need to take structural measures. So I am very happy of that, and I think it is a little bit different maybe from some other parts of Europe. Or I know it is different from some other parts of Europe. So actually that is part of the whole equation. Then Tokyo Gas, coming to that.
Now, let us see how fast it will turn out to orders. I think you should ask Tokyo Gas. But they are seriously looking into building power for data centers. I have said before, and I hold to that statement, that how do we allocate our capacity? Because in this very favorable market situation, it is a little bit how you allocate.
And we said that we do not put all our eggs in the data center basket, and we have also said we want to diversify geographical presence. We have strategic customers both in marine and energy, and we will honor those relations. And Tokyo Gas, I would say, it is one very strategic customer for us, where we have decades long, three decades long relationship. So clearly when we talk about where do we put what, Tokyo Gas is high on the list.
But I took it right that you are not working with any kind of slot reservations, and that is also a topic for Tokyo Gas. They won't get a reservation in that sense, right?
We are not working with any customers on slot reservation. I think we have been consistent with that, and we continue to be consistent with that.
Thank you, Håkan.
Thank you, Håkan. The next question comes from William Mackey. Please go ahead.
Yeah. Good afternoon. Thank you. A couple of follow-ups, really. The first one relates to the emissions and permitting backdrop for the U.S. power market for data centers. We've seen tightening, I think, at the state air rules level, and maybe loosening at some of the federal levels with particular technologies. I wonder, and there's also a debate when we look at prime power and the number of hours that prime power is expected to run, that some of that creates delays to permitting.
There's clearly no problem in demand, but I think when you build your order book and when you look at your order book for data centers in the U.S.A., how do you think about timelines and the risk of slippage? And to what extent could permitting or emissions-related slippage impact your profile around that business?
I would say, we don't take risk on our customers' project execution. And included in that bucket of our customers' project risk is they are responsible for the permitting. We don't take risk on that because we sell our engines. You need to sign up for an order. If a customer, in our case, and sorry to be a bit black and white, but here, if a customer gets problems with the permitting, they still need to buy the engines. And if they need to terminate the contract, they need to terminate the contract, and there are provisions for termination, and there are cost compensation, et c.
It doesn't have direct impact. The other thing that we've been trying to do is, and that is also putting our eggs in different baskets. On the data center customer side, there is also a vast array of different type of customers with different level of experience. Everything from the very experienced to the newcomers. We try to work with some of the well-established players that have experience and also, in many cases, have global ambitions or actually global footprint, so to say. So customers running into permit problems, it is their problems. We have, of course, obligations on our side, delivery, time and quality, et c, but permitting is not in our scope.
That is very clear. Coming back to an earlier question about attachment of service. To what extent do your customers agree in principle to attach a service contract to your engine sale? To what extent do they have an alternative? What I mean is, actually the service contract will come, is it just a matter of time or do they have an alternative and is there a risk your attachment rates fall below 100%?
No. So over time, basically, for the vast majority of services, we are the only ones that can provide it because we are the only ones that really know in-depth technology. Let us look at our service offerings. I want to give you a little bit more granular answer, expand here. But the general answer is the high level of stickiness. We talk about our service value ladder. If I sum up our strategy, it is moving up the service value ladder. There are four steps in this service value ladder. The first step is the transactional spare parts business.
The second step is the agreement business. We have a fairly broad array of different type of service agreements. We adapt them to customer need. The third step is the retrofit project. So these are upgrading existing engines to new fuels or making them more energy efficient. The fourth step is the performance-based. This is where we make long-term agreements, 5- 10 years, and we have bonus, malus clauses, et c.
Moving up the service value ladder is that we want to move all customers up this ladder because there is more stickiness, there is more revenue, and it is a way to grow our business. Now coming to your question. We will always have the spare part business, clearly. But that is not sufficient for us and normally not for the customers as well, because they need more help. So on the data center, there is certainly going to be agreements. They are going to be with us. Now, some customers want full operation and maintenance agreement. Then we even operate the plant for them.
But some customers say, "No, I'm going to use somebody else to operate the plant, but I still need Wärtsilä to service and maintain the technology," so to say. Since this is very early stage, I would envision that we will have also data centers that will go to the performance-based contract. So answering your question, there's a high level of stickiness. On the first step, there is nobody else that can do the job. On the agreement side, if you look at operation and maintenance, yes, you can contract to somebody else. But the technique, more regular service and agreements, basically it's Wärtsilä because we have the technical knowledge. On the performance base, that will only be Wärtsilä.
Great. Thank you.
Thank you. The next question comes from Anders Idborg. Please go ahead.
Yeah. Thanks. Thank you. A follow-up. On the marine side, I agree. Looking at the contracting, it has increased further. So normally when we look at this, it should lead to sequentially higher equipment orders for marine. But given the tightness in the engine market, is there anything that is different now? Are customers placing engine orders earlier perhaps to secure a slot? How are the lead times working out in regards to marine customers here?
It is in line with what we have said before, and that is also on the marine side, the lead times are getting longer and longer because maybe not so far for engine, but more for slots with the shipyards. Especially if you go into cruise, you are way into the 2030s if you want to order a new cruise vessel. Then, of course, once you ordered it, you want to look up the equipment and it is very important equipment, so early. So also on the marine side, the lead times are getting longer.
Yeah. Are customers ordering engines earlier than usually or?
Yeah. Earlier, clearly than-
In relation to delivery date of the vessel, obviously. Yeah.
Correct.
Okay. Thank you.
The next question comes from Tom Skogman. Please go ahead, Tom.
Yes. Hello, Håkan and Hanna-Maria.
Hello.
You seem very positive about the marine prospects, and you do not really have as overheated orders as in the energy side. I just wonder what would it take for you to upgrade the demand outlook for marine? We already know that vessel orders are strong. Shipyards are expanding faster than expected. What is the kind of missing piece to say that demand really will improve for you in marine?
I think there, and we kind of alluded to it. We are already on all-time high level. When we have issued our demand guidance and we guided similar to your point, both in marine and energy, we also said clearly, you need to see that we are running at all-time high level. For our order intake, of course, Wärtsilä is running at 100%, but it would take us some time to ramp it up, and we are already sold out for 2028. If I had more capacity than we are currently having and have announced, I could sell it.
Then another question about AI in service. I've noticed some engineering companies are now more outspoken about how they aim to use AI in pricing and resource utilization in service. Can you open up what you are doing and when we could start to see some benefits from that?
Tom, I appreciate you a lot. Now I'm going to make a little joke here, so forgive me. You can see it already in our annual report for 2025. It's there, concrete. It's nothing new. There is so much going on the AI side already. For us, not being arrogant, but this is nothing new. When we talk about AI in Wärtsilä, we've been talking about two focus areas. One is how we use AI as a tool in our toolbox for continuous improvement. This is happening all over the company. We are not unique here, I would argue. Any industrial business, they are trying to do.
There is very concrete. It used to take us a day to compile documentation for the classification societies that needs to approve certain things. Now with the help of AI, we do it in minutes. These are very concrete examples. Here I would argue, the most important to be successful in AI, in continuous improvement is the continuous improvement culture. This is something we are working on in Wärtsilä and evolving.
But the AI tool itself, it's not the latest Claude version or whatever. It's the mundane AI, so to say, but very valuable. It does create value. There is no doubt. The second focus area that we have been talking about when it comes to what we are doing in AI in Wärtsilä is exactly what you mentioned, Tom, and that is how the strategic bets that we are making to support our service business, because service is more than half of Wärtsilä revenues, as you know, and it's a good source of customer interaction and profitability also.
What we are doing there, what we are concretely doing there is that we are creating an edge platform and a computer platform, several stages, and a data architecture, so you can have data flowing, usable data flowing in a good way because we are in remote areas or we are at sea, et c. There's a lot of investment. You're talking hundreds of millions of euro going into that. You have the WISE program in Finland, it's EUR 200 million over five years. So investing in this edge platform data architecture, then combining that with the knowledge of our people.
We have so many knowledgeable people. So they can use the right tool and have the right data-driven support. What do we provide? Uptime, reliability, and reduction of fuel consumption and reduction of emissions. This is kind of customer proposition, this is what we are integrating in our service business. This is one of the fundamental pillars for moving up the service value ladder that we talked about earlier. So it's very well integrated into our strategy. It's ongoing. We have a lot of potential here, but we have started clearly.
But I guess there must be a lot of potential in kind of trying with pricing and pricing power in certain products and areas and stuff like that as well. Are you doing that already in a sophisticated way, or will it be very different a couple of years from today?
I cannot comment on the level of sophistication, but we are clearly using, you could say, data-driven pricing, leveraging a lot of data that we have. I'm sure we can evolve it going forward. But as I said, I'm really excited because we are really, of course, we are providing power to AI, so we benefit from that and we are excited about that. But I am, and we are really excited about the potential of AI to drive efficiency and develop the customer offering and making sure that we get paid with pricing, et c. There is a lot of potential. What I'm happy with is that there is curiosity in Wärtsilä. We want to adapt.
There is much more curiosity than negativism, because we also know in the current overall discussion in society, there is a lot of concern and rightful concerns about AI and what will it do and will people lose their jobs and could it even kill humanity? All those big questions, but from how AI is perceived in Wärtsilä, it is a positive. We see it as an opportunity, and the concrete proof points are the good things that there are so many good things going on in continuous improvement. Then it is our hundreds of millions of euro investment into AI for our service business.
There, evolving the expert insight platform and the customer offering. The performance-based contracts, you know them, we are evolving them. There are more customers coming in. I think we have good proof points, and there is a lot of more opportunity to come, and we are investing. It is one of the areas where we are really investing. Like we are investing in engines and the fuel efficiency and fuel flexibility. AI is a second major area where Wärtsilä is investing.
Then finally on-
Sorry, and of course, not to develop large language models. That is clearly out of our competence and scope. How do we apply those to create value for our customers in our two industries? On the second, sorry, and then the second area is how to leverage AI for service business. We want to be best in our industries, not in the world. We will never be able to do that. But we want to be leading in our industries. Like we are leading in fuel flexibility and fuel efficiency in our industries. We want to be leading in how we apply AI in our service business, in our industries.
Then finally on Brazil. There was 90 GW of power that companies committed to installing before 2031, whereof 10 GW is gas. To my understanding, this is balancing power. So it is either air derivatives or your solution. You have only announced one order so far. So what am I missing here? They should be in a brutal hurry, or is it so that there will not be any sanctions if you are late or what happens?
I think we have some orders, but also to be frank, I do not think there will be so much more coming because we are sold out. We have sold our capacity. We could sell more in Brazil. But if there are some newer auctions coming out, we cannot meet the delivery times because we are already sold out. To your point, could we have sold more in Brazil? Yes. But we did not have the capacity.
Now the last question comes from Sara Jamil. Please go ahead, Sara.
Hi, Maria. Can you hear me?
Yes.
Thanks for including me in. Just a follow-up on your various comments about running at all-time high demand and capacity. Would you hypothetically have the actual physical ability to increase capacity further than you have already announced? For example, do you have the physical space in Vaasa and would the supply chain physically be able to do that? If so, what would you need to see in order to justify it?
The answer to the first question is yes, we could expand further. Second, what do we need to see? I am sure you noticed that we have had a staggered approach on making announcement of new capacity investments. That is one of the de-risking strategies from our side that we form a certain view, and then we invest, and then we go, and we get even more confident, and then we invest even more. What is the key decision criteria is that we, for further investment, is that we continue to see this trend extending even further out in the future.
Now I am talking well beyond 2030, as I highlighted before. Then we can certainly take more steps. Yes, there is more supply. We talked about that coming in. Of course, we need to consider that. We need to be disciplined. However, I am still coming back on this. Our technology, the medium-sized reciprocating technology, has in our sweet spot, very competitive features. That sweet spot in terms of the overall market is growing. That is good. Then within the medium-speed reciprocating engine, population of suppliers, we are a technology leader, have very strong service offerings. We look probably more on the demand side than the supply side.
Thank you. I am afraid that we are running out of time. As a reminder, we are hosting pre-silent call together with our CFO, Arjen Berends, next week on October 1st, and then please remember to register to our CMD. There are only a couple of seats left for the face-to-face event. Luckily, there is always plenty of seats on the online event. Thank you.
Thank you, everybody.