Hi all and welcome to Wärtsilä Strategy Call. I'm Hanna-Maria Heikkinen. I'm in charge of Investor Relations. Today, our CEO, Håkan Agnevall, will discuss some of our long-term opportunities, and after Håkan's key messages, there is a possibility to ask questions. Håkan will show also two slides which are part of our roadshow material, which is always available on our IR cover page. My colleague Noora will send you the slides after the call. As a reminder, we will host a pre-silent call on September 30 together with our CFO, Arjen Berends. Let's leave the questions related to recent trading and detailed financials to that call. If you have a question, please use raise your hand functionality in Teams. In the case you don't have the possibility to use raise your hand functionality, you can also send an email to me. Håkan, please.
Yes, hello everybody. I see the questions are already starting to come in. That's good. It's going to be a good session today. I suggest that I give a couple of overview comments on where we are and how our strategic outlook, how we see the future. Then we go over to he Q&A. Marine and Energy, if we start on the marine side, we see a demand situation that is holding up rather well, I would say, both for new build and services. We do underline the continued message that our core segments continue to be strong. We all know the market volumes are coming down from the record 2024. Please note that even using Clarkson data, their forecast and outlook is continuous actually to be higher than the 10-year average. That is even more accentuated for the Wärtsilä core segments: Cruise, Ferries, Offshore, Special vessels, etc.
Still rather strong view of the market. I think our strategy in the space, our decarbonization journey, it's also adding or supporting the growth. We see MEPC 83. The vote is coming in October. If the industry will start to introduce carbon fees, that would be a landmark. The U.S. is opposed. We don't know how the vote will go. After the vote, there is also a process afterwards where the flag states also need to approve it. There is still an approval process. Nobody knows for sure how this will play out. I would say a vote for, I mean, an introduction of carbon fees on a global scale will certainly support the Wärtsilä strategy. In a situation where for some reason this will not go through, I think Wärtsilä is still rather well placed. Shipowners take a long-term perspective. A vessel will sail for 25, 30 years.
They will have to have a long-term perspective. Regardless, the whole topic of energy efficiency, fuel efficiency, and fuel flexibility will still be strong. Flexibility will be key because even with, I mean, there will continue to be uncertainty. From that perspective, a positive decision will certainly support Wärtsilä. A negative decision is not the end of Wärtsilä, rather the opposite. I think we have a very strong offering and we will also fare relatively well in that situation. For the world, it would be really good with a decision to move on with the IMO suggested framework. Core segments, decarb, it continues to play out. Services, we pointed that out last time, book-to-bill continues above one. We have a bit of a cyclicality on the retrofit side, but we also said this is a cyclical business. We still have a positive view of our strategy.
The whole focus on moving up the service value ladder continues to be our strategy. We do see it playing out. Our agreement business is growing and still with a good renewal rate above 90%, etc. It's the same narrative as before. It's working for us. It's really working. I will cover the U.S. tariffs later on. I do energy first, and then we can certainly talk about the U.S. tariff situation. If I go over to energy, it is the narrative of balancing. Now we have a new end baseload, and the new or somewhat new theme on baseload is, of course, data centers. I will come back and comment. If we start with balancing, we do see continued growth of renewables on a global scale, even in the U.S. I think the awareness of the need of balancing is growing in many parts of the world.
Certainly, when we see the market going forward, we see also a lot of interest for balancing. Various maturity in different markets, but still strong interest. If we talk about baseload, we have our traditional baseload, providing baseload for South America, Asia, etc. There are markets there. Clearly, there is the demand. Of course, it's project business, so it's a bit cyclical. It can shift from one quarter to the other. I think the underlying fundamentals are strong. Now, of course, we add the whole data center piece to this equation. Let's share some slides here. It's not a new message, but we can reiterate it. First of all, I get a lot of questions from analysts these days that, you know, Håkan, what's the size of the data center and how big is it going to be? I would say nobody knows.
There are a lot of reputable institutions and players that are trying to make estimates. When I look at those estimates, if you look at accumulated five-year new installations, those estimates vary between 20 GW and 100 GW. Quite a big span. To do predictions is very, very hard. However, the underlying demand is certainly there. There is a lot of activity. We can clearly see it. As you've seen from last quarter, we start now also to have our first orders in the U.S. We had orders before in Europe. Now we had our first U.S. orders. What is happening is, and this slide we have showed before, basically, if you go back a couple of years, and it's not so many years, you know, the typical data center need was 20- 100 MW.
If you are building a data center, you went to your utility and asked, you know, to have this power available. Then to back up the grid, basically, you bought some high-speed engines, not from Wärtsilä because we don't have high speed, to stabilize. In this, you could say the old business model, Wärtsilä, this was not in Wärtsilä's sweet spot because we do bigger engines for bigger power. Now what is happening, as the size of the data centers are growing, we are talking 5,300. I mean, we are talking all the way to 1 GW . We don't play there in the giga, but we certainly play in this, I would say, 20- 400 MW.
When the data centers have that demand and they go to the utility, they say, the utility says that, interesting business opportunity, but we need to run this through our process, planning process, because we will need to build additional infrastructure and generation, and that takes five to seven years. Now the data center developers are looking for, not everywhere, but in many cases, they're looking to generate their own power. The amount of power is coming into our sweet spot, so to say. This is what we start to see. This is, of course, a very interesting business opportunity for us. I think the technology that we're offering, we're always competing with the gas turbines. We have also in this area, we have some interesting advantages. First is our modularity.
As you know, or you might know, you build engine power plants of several engines, whereas you build a gas turbine plant with a few gas turbines. This modularity with several engines is actually help when you want to have an affordable, very high uptime reliability. Of course, our other argument of high fuel efficiency compared to the gas turbine still holds in this because it's going to be baseload. It's a lot of energy. We don't need water or very limited water compared to the gas turbine. Those advantages that we have been talking about before, they certainly are relevant in this segment as well. I think what is accentuated in this particular segment is our modular design. That's pretty interesting. I'm sure some of you want to ask or will ask, so I take that already. OK, you had your first order in the U.S.
Will you have more orders in the U.S. this year? As I said in the second quarter, we are working with many different opportunities. The opportunities are in different stages of maturity, as they always are in project business. Sometimes they go up and go down. We have more than a handful of interesting opportunities in different stages. We could have another order. I'm not saying we will, but we could have another U.S. data center order this year, clearly. That is a little bit on the energy side. To sum it up, the balancing narrative is still there. We still see it playing out. We have a baseload business. You saw the record decoder, these types of opportunities, mining opportunities in Pakistan, they are still there. Now we start to see the data center orders also coming in. That is good.
On energy services, it's a similar narrative like in Marine, moving up the service value ladder, moving customers more and more into agreements. It's playing out also with a very good renewal rate, about 90%, and still also with growth opportunities going forward. That's Marine and Energy portfolio. We continue the divestments. We closed on ANCS. You saw that. We have signed on marine electrical systems. We still have Water and Waste and Gas solutions, and that is ongoing. That is also playing out, as we've been talking about. Of course, tariffs and geopolitical tensions, clearly. We live in times of high uncertainty. We underline that in our guidance as well. There is a lot of uncertainty. I think underlying business is still developing. Demand side is developing in a good way.
Now, if we talk about the tariffs, I would say, I've been to China, I've been to Korea recently, talking to shipyards and customers. We know, we all know that U.S. part of global shipping, it's less than 15%. I think the industry is still moving ahead. There are still vessels being built in China and Korea, for sure. I would say that the impact so far is limited. Of course, if there is a global downturn of trade, the international trade, that will certainly affect the industry. So far, I would say limited impact. If we look at our Energy business, our thermal power plant business, yes, we are exporting the engines from Helsinki, from Vaasa in Finland, to the U.S. Yes, we are applying the relevant custom duties. There has been some dynamics on which rates and how you apply them, etc.
I would say that the impact of the most recent changes, our analysis so far is that the impact will be limited. We also underline, this is a very dynamic topic. I think the rules that are applied one day could definitely change the other day. So far, we still say limited impact on our thermal business, so to say. Finally, on energy storage, this is clearly impacted in a major way by the import duties, those in Asia. I would still say that the U.S. storage market is muted, which also leads to competition in other markets increasing. It is a very competitive space right now. It's a tough space. We still continue with the storage strategy that we have kind of communicated earlier this year, that is, it's still a focused growth in certain core markets. We are executing on that. It is a competitive space. There is no doubt about it. All right. A quick summary, exposé of the strategy and how we see the future, so to say. With that, let's open up for questions.
Thank you, Håkan. The first question comes from Daniela Costa. Please go ahead.
Hi. Good afternoon. Thank you so much for taking the question. I have a two-part question, but related to your commentary on data centers. We have seen, like the turbine companies, I would say, the last two years, they've had a lot of orders. Their backlogs are really long. I guess the first part of the question is, why do we think we're only now starting to see the growth for engines? Have these benefits that you talked about just been developed more recently, or the customers weren't aware? Maybe a bit of why are we only seeing this demand now? When we look at some of the turbine companies saying they're going to double capacity in two years and adding more, how does that interplay with your space? How should we be thinking about that?
I think many of the, I mean, you should ask our gas turbine competitors, but I think that some of the really big gas turbine orders have been for the, you know, we talk about the gigawatt projects. These are the segments that we are not playing. My understanding is also that, on top of the data center orders, they have also had quite a lot of big utility orders in many parts of the world, which are in this giga, in the higher power space where we are not playing. I think that's pretty straightforward. Now, when it comes to delivery times, etc., yes, the gas turbine guys are sold out. We are sold out for certain parts of our portfolio, but we can still deliver on 12 months on other parts of our portfolio. I won't go into the details of that for very obvious competitive reasons. I think in Vaasa, we have a good loading situation. For some part of our portfolio, we are sold out. For others, we can still deliver.
To this point, you said sort of the giga projects.
Daniela, it's a bit hard to hear you because I don't know if your microphone is strange or something here on our end.
Oh, OK. Y ou hearing me like this without microphone?
Yes, much better. Much better.
Yes, on this comment on the giga projects, does that mean the turbines are maybe more for hyperscaler type projects and you're maybe more exposed to the co-locate? How exactly can you help understand there?
If we talk about the hyperscalers, they do bigger data storage. I mean, super big giga data storage. They do the more mid-sized data storage. There are different types of data storage. Those data storage are acquired or built or contracted by the hyperscalers. It depends where they are located, the situation of the local power system. I think the key thing here to create business opportunity for us is that, of course, you have a grid situation where you cannot rely on the grid. You need to generate your own power. That's the starting point. If you have the super big, the giga, you will go with CCGTs, gas turbines. If you're in this segment, which I indicated, sweet spot 20 to 400 MW, then you would be looking at engines and gas turbines. I would say that we have some pretty interesting advantages there.
Thank you.
Thank you, Daniela. Next question comes from John B. Kim. Please go ahead.
Hi. Afternoon, everybody. Thanks for the opportunity. Continuing on the energy theme, can we talk a little bit about the shape of the market? Now, without holding to numbers or size, if you look at your business plan or early strategy here, over the next three to five years, how much would you expect from baseload applications? The examples you gave about smaller data centers running on a bank of engines versus backup. Thank you.
First, I'm going to, just for nomenclature, we don't call it backup. We call it balancing because backup is more the functionality that the high-speed engines are providing, i.e., you have a grid, and if the grid malfunctions, you have a backup high-speed engine. We are not in that. That's not our sweet spot. We talk about the balancing when you have renewable swinging and you need to balance those. It's a nomenclature thing. Coming back to you, now, what's your share of balancing and baseload? What's in your business plan? First of all, we don't approach our target setting like that. It swings a lot between years. I think last year, balancing was 70% approximately. This year, I think it's the balance. I mean, the split between balancing and baseload is almost the opposite. It swings a lot between different years.
OK, helpful. Outside of the U.S., which markets or regions are you most excited about over the next three to five years for energy?
I'm excited about the world because, as I said, you know, the U.S. is clearly a leader in, sorry, it's a clear leader in data centers. We do see other parts of the world also moving on the data centers. The U.S. is the biggest. Balancing power, the U.S. is also the biggest for us. We start to see this in Asia as well. Baseload, it's all over the place. South America, Brazil, maybe Argentina, Indonesia, Australia, etc., etc.
OK, thank you.
Thank you. Next question comes from Antti Kansanen. Please go ahead.
Yeah, hi, guys. Continuing with the same theme on the data center side, I wanted to ask a little bit about the aftermarket opportunities in these cases because, as you mentioned, Håkan, already, a lot of the demand is created by the fact that the guys can't access the grid. What do you think happens in five to seven to 10 years' time when they perhaps get the access and the grid operators have invested in their capacities? Are you still, or how do you calculate the lifetime profitability on those projects without knowing exactly the long-term capacity factors? Back to Daniela's point on the turbine makers being sold out, you're selling scarcity. Are you getting rewarded on the new build side? How should we think about these things?
Yeah. First of all, you should talk to the data center developers about their TCO model because they would know how they plan their investments. To your point, this is like a mining operation, no pun intended, where you run the power generation 24/7. Uptime reliability is key, key, key, key. I would say that we will have a very strong service business. I would also say, as you said, you should talk to the customers directly. I would say that we will have a strong service business beyond the five years because let's say in a scenario that the grid will come back, if I was a data center, I would start to export the power to the grid. You have it there. You have it on super energy efficient Wärtsilä engines, hopefully, if you're lucky, which has better energy efficiency than the gas turbines.
They are highly flexible, so you're also prepared when you want to integrate renewables because renewables, as we all know, they are good, but they're intermittent. You have the balancing story again. I think we will have a strong service business on the data center side, not only for five years. Coming back to your second part of your question, and I'm missing it. Can you, Antti, can you repeat it?
Antti, I think you are muted.
Can you hear me now?
Now we can hear you.
It was on the new build profitability. I mean, you're selling scarcity. A lot of the competitors are sold out.
I would say that, you know, of course, in this situation, you know, we have a good price realization. Of course, you know, there is a limit of how you can go because we have customers that are professional. You know, they will not pay too much, so to say. I would say that we have a good price realization.
All right. Maybe I'll try to ask a little bit of details on your already mentioned the project pipeline. They are on various degrees. If we just look at the news flow regarding the entire kind of data center AI stuff in the U.S., last night we saw Oracle popping up 30% on their numbers, which is very driven by that market. Are you seeing a sense of urgency increasing on the portfolio that you are talking about, that decisions are speeding up and they are more eager to get the project up and running?
I would say I see a high level of activity, clearly. I'm not sure if it has increased the last three or four months. I would say it continues to be high or very high. As I said, projects are in different stages of maturity. There are projects that come up. They pop down and go down again because people need to find the permits. You know, when you do develop big projects, sometimes you run into some blocks. It takes two, three months, and then you're through. From that perspective, the data center power business is like any other power business. There is a clear demand for power for data center, and there is a lot of activities, clearly. If I compare to one year ago, the derivative is clearly positive.
Maybe on the timeline regarding kind of executing those projects on these negotiations, is your delivery capability really the bottleneck? I mean, you said that maybe you're sold out in a lot of the applications for the next 12 months. Maybe the delivery time is 18, 24, whatever. Is that the bottleneck on those projects that you are negotiating? Is it any ways that those projects will be delivered in that type of a frame that it doesn't really matter?
No, I would say the bottleneck is more that our customers, that they need to get their, you know, because our customers are selling to the hyperscalers. We have us as an OEM, you have the power or data center power developer, and then you have the hyperscaler. Of course, you could say it's a middleman. I mean, I underplay that role, but there's in the middle between the OEM and hyperscaler. Of course, they need to contract with us, but they also need to contract with the hyperscaler. There is a lot of negotiations ongoing. They are building a power plant, so they need the environmental permits, the building permits, and all the other permits. This is never a linear process. This type of project development dynamic is what is really, you know, setting the pace, I would say.
All right. Thanks, Håkan.
Thank you, Antti. Next question comes from Sven Weier. Please go ahead.
Yeah, hi. Hi, Håkan. Hi, Hanna. Thanks for doing the call. A few questions from my side, please. I'll start with marine, and then I also have a question on data center. On the Marine side, just curious about cruise. We've now seen Clarkson data, three months without any cruise orders, at least in the current data. It's a lumpy market, we know that. I was just wondering what you see in the forward pipeline. Is this just a pause, the usual lumpiness? Do you see more projects? That's the first one. Thank you.
Yes, it's lumpiness. I think we see an underlying demand side that is strong. There is a limitation on yacht capacity. There are three big yachts that are the dominant in place. They are selling their slots. I see a strong underlying demand. I would talk more about lumpiness than anything else.
At the same time, offshore was pretty good. I mean, was that in your neighborhood for your type of offshore exposure? Or should we take comfort from the data?
Yeah, I think that, yeah, I mean, short-term data, as you know, Sven, I think you should take them. You need to discount them with a very high factor. We talked about that. You need to see the longer-term trends. I think for us, and I know you're seeing it, but in our quarterly report, we always show the volumes there. We show what the Clarkson forecast is and how it compares to the 10-year average. You have a top graph and a bottom graph on that slide. You see, we have done what's the Clarkson data for Wärtsilä's core segments. You see it's way above, even if Clarkson has revised down. The current levels that they are predicting are still way over the 10-year. I think this is what is important.
Yeah, agreed. The follow-up question I had on data centers is, again, you know, on the claim to say, look, on energy efficiency, we are relatively similar. There's not much difference to the turbines. Honestly, it's a little bit difficult, at least on my end, to always, you know, have the evidence, right, because different people say different things. I was just wondering, on the U.S. example, was that a project where you are, you know, competing against a turbine and you won the project because your solution is better? Or was that a project where they had a problem getting access to turbines because they sold out? Can you give us a little bit more insight into this very specific project?
There is, I would try to sketch something. I cannot go into the details because this was an anonymous customer, and you know, we have signed an NDA. I need to respect that. I would say we are always competing with gas turbines, even in data centers. The dynamics is very similar to what we have on balancing and baseload. It is about, you know, our fuel efficiency. It's about, you know, low water consumption. As I said, the additional element on data centers is our modularity. More engines than gas turbines has an impact on how you calculate the reliability. It's also clearly the CapEx, OpEx. Some people put more focus on CapEx, so sometimes that, you know, emphasizes the advantage of the gas turbines. We have, of course, delivery times as well, where, you know, some of the gas turbine manufacturers are sold out.
Not everybody is sold out, but some are. It's the same complexity, you know, in the customer's decision matrix like we have on balancing and baseload, maybe with the caveat there that, you know, the modularity is something that plays out extra favorable in this high uptime reliability application.
The projects that you have now in the pipeline in the U.S. that are not signed yet, it's not a case of the turbine guys can't deliver, so you're in the game. It's a level playing field.
It's a mixed event. There are those, clearly, but there are those opportunities where the customers really see the benefit of what I've just been talking about for engines, so to say.
Maybe the final one, if I may, is just on what you said on storage because you said, you know, things are getting a bit more competitive outside the U.S. I guess everything is still in line with the kind of strategy that you had. You told us that short-term margins would be a bit lower and in the long term, 3%- 5%, I think. It's still, everything is still very much consistent with that, I guess.
It is. I would add that competition has increased. There are two things we did. We didn't predict Liberation Day. We didn't predict that some of the battery cell providers are moving, integrating forward. They are starting to compete with us in a stronger way. That is a consequence maybe of the first one because when the U.S. market is muted, people are still trying to push their volume. That tends people to stop to think, you know, how do I compete in other markets? If I sum that up, the difference between when we launched our financial targets to where we are now is that competition has increased. We are still sticking to our strategy, for sure.
Understood. Thank you, Håkan.
Thank you, Sven. Next question comes from Max Yatesv. Please go ahead.
Thank you. Good morning, both. I had a couple of questions. Just on the tariffs on Section 232 and what came out over the summer, I guess one of the dynamics is components and now kind of products with high steel content are now captured under the tariffs. I guess the very simple question is, are the engines with potentially high steel content, are they now facing 50% tariffs as per the agreements? Or is it still sort of they're facing the tariff coming out of Europe? I guess that's a relatively straightforward question of what tariff are you trying to push to your customers?
The short answer is much closer to the 15% than to the 50%. This is rather complicated. It goes into the custom codes. It comes into, you know, material choices and weights, etc., etc. To answer your question, it's much closer to the 15% than to anything else.
OK. Maybe just the second question is how you're thinking about capacity in Vaasa because I guess if I look at most of the gas turbine companies in gigawatt terms, they're increasing by anywhere between kind of 30%- 50%. I guess two-part question. How able are you within the kind of brownfield confines to do that in Vaasa? At what point would you actually start to do that? Is that very much kind of moving up the plank?
We have ample space to expand in Vaasa. We are actually expanding as we speak. If you visit Vaasa, which I encourage you to do, and I know we will arrange things, you will see the digging machines and it's happening. In April, we announced that we are expanding R&D, but also manufacturing. We haven't said how much. I won't say how much. The gas turbine competitors have given you those numbers. We are expanding, and we have capacity. We have capability to expand further.
OK, very clear. Just thirdly, how quickly can you do that? If you were to add sort of 30% capacity, is that a sort of one-year, two-year, three-year project? Are there any constraints on the supply chain that would make that impossible?
When you expand capacity, and to your point, we are, you could say, our industrial system, we do final assembly. We do some in-house manufacturing of some critical components. To ramp up our capability is also very much related to ramping up supply chain. Can we do that? Yes, we can do that. How much time does it take to add capacities? There is no, how long is the road? It is a year, it is two years, it is three years. It is in that time domain. It is not five years. It is not one year.
OK. Just two more quick questions. If I look back at kind of when this business was at its peak in terms of what you were taking in order intake, I think you were somewhere in the region of kind of 3.5- 4 GW annually. Given the capacity that you have and the lead times, can you ever get back to that kind of stage? At what point does it just become ridiculous that you're taking orders that are so far in advance because of capacity? What is actually reasonable from what is physically possible and where the lead times are acceptable?
If you compare, you're right. I think if you look at the maximum gigawatt we ever delivered, it's probably somewhere in 2018. Since then, we have also, of course, closed manufacturing interests in Italy. Now when we talk European, we have our JVs in China. We have capacity there. If we talk about the energy space, those engines are delivered from Finland. Now, what's our capability in Vaasa and how can we expand? I am saying we have such a team there. We have such a strong support from the city, from the region, from the whole cluster. You know, the sky is the limit. We should do this in a wise way. As you know, we are well capitalized. We have the financial firepower to invest, and we are investing. We have the ecosystem there to support us. I think we have good suppliers also to support us. We will take our capacity to where we see the market going and what becomes attractive, so to say. As I said before, we are already expanding, and that is even public if you go back to the April announcement we made.
OK, just final one. Just to focus on the energy storage comments, because I guess you do have some visibility on what the sort of price per megawatt hour and the kind of pre-calculation margins on some of these projects. I guess for us, the other side, it's a very different outcome if this business does 2% margins or goes back into loss-making. I know you never like sort of commenting on margins. I mean, are we talking about, I mean, it sounds quite severe when you talk about it. Are we talking about this business going back loss-making when we think about the margins? Are you just saying, you know, we might drop out temporarily out of our 3%- 5% guidance range, but we can keep it profitable? Just any way to kind of quantify the severity of those comments.
I would say it's a very relevant question. I think it's very much related to how the U.S. market will evolve going forward. How that will go, I think nobody knows. I think competition is growing, that's for sure. How these dynamics will play out going forward is very unclear, I would say. I will leave it at that.
OK, very helpful. Thank you.
Thank you, Max. Next question comes from Akash Gupta. Please go ahead.
Yes, hi. Good afternoon, and thanks for your time. I got a few as well. Maybe starting first with marine. I think in the past down cycles, we have seen some cancellation, mostly at Asian shipyards, more so at Chinese shipyards than other Korean and Japanese ones. I mean, so far, we have seen a significant decline in contracting activity. Maybe if you can talk about, have we seen some cancellations already? How do we think about the risk of cancellation in this cycle? Linking this with the MEPC 83 vote that is coming next month, could that be a trigger for, let's say, a bit different behavior from your customers' customers?
Once again, I urge you, first of all, everybody's talking about the big decline. Yes, 2024 was a peak year, absolutely, and the volumes this year are much lower. Please look at what Clarkson is forecasting. You can challenge if Clarkson are right or wrong, but I think they are the most reputable institute in trying to predict the future in this industry. Their estimates for the coming three years are above the average 10-year demand. I don't recognize this, that there is a crisis, so to say. Yes, volumes have come down from 2024. That's a general comment. I would really encourage you also, please look, maybe I've done that already, but please look at this slide in our quarterly report. If you look at Wärtsilä's core segments, Clarkson is very optimistic about them, clearly about the 10-year rally. When it comes to cancellation related to Wärtsilä, no, we have not had any cancellations. Rather the opposite. We have new orders coming in.
Thank you. I have a couple of questions on the energy side. The first one is more technical. I think you have said that you're operating in a market for up to 400- 500 MW. I mean, you can stack these engines, and probably you can go even higher. Why can't you go even higher and have a setup that could be 700, 800 MW and could be more suitable for some of the hyperscaler customers' requirements? What are the technology challenges there? Is it something else?
No. I mean, it's very practical. There is no technical limit. You can build gigawatts of engine power plants. The thing is that it will be quite a few engines, and it will take quite a lot of space. Those are the constraints. It's always a balanced decision, you know, how much space you want to allocate. The other thing, if you intend to run a combined cycle, I mean, use the heat that you generate. Are you going to use this for district heating or for something else? That also lends into the equation, so to say. For practical purposes, it's similar. Why don't you build 1 GW with high-speed engines? You can do it, but it's going to be a hell of a lot of engines, and you're going to fill a lot of space with engines. Of course, they need to be maintained, etc., etc. That is very practical.
Maybe a follow-up to that. Is there anything that stops this combined cycle technology in terms of capturing the heat and having a steam engine to steam turbine to fire up with engines? I guess, let's say if you have 400, 500 MW setup, there will be a lot of heat. Can it be done theoretically? Are you looking at it to further improve maybe attractiveness of your offering?
We are doing, I mean, this is not, I mean, to put a steam turbine after the engines, you can do that. We're doing it in Mexico in some certain installation. It's about how much value, if you can use it, how you use it. It's part of the equation. Normally, if you're looking at the 1 GW installation, it's a better equation, energy equation for the CCGTs than it would be for engines with the steam turbines, so to say. Up to 400, not 500, but 400 MW, we are certainly competitive, not in all occasions, for all applications, but in many applications, we have a very interesting alternative.
Lastly, a follow-up on your earlier comment. I think you mentioned your customers are selling electricity to hyperscalers. Just to double-check, you have not been approached by hyperscalers directly. It's basically the customers that are building infrastructure for hyperscalers.
Right. I mean, we are not doing business directly with hyperscalers so far, I would say at least.
Thank you.
Thank you, Akash. The next question comes from Vivek Midha. Please go ahead.
Thank you very much, everyone. Good afternoon. Hope you can hear me well. I have a few more sort of follow-ups on some of the questions earlier. The first was on the capacity commentary. You mentioned the expansion of Vaasa, the Sustainable Technology Hub, and based on your release, the comment was that the completion is expected into 2028. I was wondering if you could maybe just give a little bit more color about the phasing of that. Should we expect that there's some incremental capacity coming from that expansion, bleeding in in 2026 and 2027 before the eventual completion? How should we think about that? Thank you.
I would say that, you know, when it comes to testing capacity, maybe there is something earlier. When it comes to manufacturing contribution, it's probably in 2028. Please also note the comment that I made that, yes, we are sold out. You could say second half of 2027 for certain of our engine offering. For others, we still have 12 months delivery time. Please don't ask me to elaborate on which engine. I don't want to give that to competition.
Understood. Just following up on that, I guess it's a tricky one to answer. I mean, testing sounds like it is one of the key bottlenecks. Is there any way you can quantify for us just how much of a bottleneck testing is at the moment, that testing capacity?
I don't have even a public API for it. Sorry.
Fair enough. There's another question around service. Clearly, you've had a strong upcycle now on the order intake in energy, and particularly now with, for example, data center, that feeds into your baseload service fleet. When can that start to feed into the service revenue growth? What's kind of the lead time between deliveries and when that becomes relevant for your service business?
Basically, you know, we take an order. I'm talking very generically now. There is a lot of, but just to give you because I think I understand the logic behind your question. We take an order. Let's say it takes 12- 18 months to build it. These days, we, not for all, but many times, we sign some type of service agreement, sometimes even operation and maintenance agreement, very early. This service agreement starts to kick in very closely to when we start operating the plant. We have a kind of an internal accounting guideline. Ayan has been talking about this in the past. It's nothing new. If we sign a 10-year service agreement, we only book the first booking for the first two years. We don't book, I mean, let's say we deliver after 12 months.
Let's assume that the service agreement starts immediately when we start to operate. The customer starts to operate the engine. Let's say that we have signed a 10-year maintenance agreement where we will book when we start operating. It's only two out of the 10 years.
Totally understood. My last question is just sort of summary thoughts around how you're thinking about service growth going forward because you've highlighted the book-to-bill. Maybe the Q2 order intake was not quite so strong. A lot of that was retrofits. Over the last few years, you've seen some very strong service growth. As we go forward, do you expect that to be sustained? Is it likely to go close to maybe your group targets? How should we think about normalized service growth going forward? Thank you.
I think we send very strong signals. Our service order intake continues to be well above book-to-bill. We also made this breakdown of the four disciplines within services because we clearly acknowledge that the retrofit is cyclical. I know in the past, maybe still, it causes some concerns among the analysts. I think we've been very consistent with our message. Services continue to grow. It will continue to grow. It's a major driver of both our growth and profitability. Even more importantly, it's a major driver of our customer happiness and satisfaction.
Understood. Thank you very much.
Thank you, Vivek. Next question comes from Tom Skogman. Please go ahead, Tom.
Good afternoon. Good afternoon, Håkan and Hanna-Maria. October 1, Europe is moving to a 15-minute pricing market in power. What will happen and what regions are included in this? I guess, can you draw some similarities for what has happened in certain states in the U.S. from this?
I don't know exactly which regions are involved. I can clearly say that the more granular, I mean, the shorter time for this resolution, the more value will be realized in balancing power. They will put even further focus on balancing power, and this is good for Wärtsilä. From that perspective, this will support Wärtsilä's business going forward. Just to put in references, and because you alluded to the U.S., people have moved from a couple of hours to one hour. In Japan, there is now one hour, and we see that change from a couple of hours to one hour. That is clearly putting the focus on the balancing power. Now when Europe is moving to 15 minutes, it will create additional focus. I would also say it's still early to say. I mean, we remember the Spanish incident, power outage earlier this year.
I think there are still ongoing root causes, and there are different hypotheses, but it's clearly so. I think that there are more and more stakeholders realizing that, you know, renewables are great, but we need more balancing power. Balancing power, yes, and also to strengthen the grid. Green is not black or white. There is no single simple solution. Moving to 15 minutes will support us. Maybe next quarter, no, and maybe not next year, but in five years, it will definitely support Wärtsilä business.
Will it trigger closure of some certain form of power generation?
I would say, coming back to the narrative, and Tom, you've been following us for long. Normally, you chase us, you know, when is the big revolution coming? We say that, you know, it's coming, but it's more linear. Yes, it's a story. When the need for balancing goes up with more renewables, you put the balancing in. You will start to phase out the big, often coal-fired power plants. This will happen.
All right. Moving over to data centers, to my understanding, the power demand goes up and down extremely quickly, even more quickly than, you know, your engines can, you know, ramp up and ramp down. How is this technologically done, basically? What is your responsibility here? Are the batteries in between, or who is responsible for no failures, basically, there?
Basically, this is where the system developers, our customers, they are the developed power systems for the hyperscalers. We talked about this three-stage, you could say, business system: OEMs, the data power center developers, and the hyperscalers. They design these power plants to reach the appropriate reliability. The reliability requirement from the hyperscalers is very, very high. How do they do that design? They introduce redundancy in the engines or gas turbines. You need less redundancy from a modular engine power plant. You can add some batteries also if you need it. We are not into that. We deliver the engines. Of course, there is a certain reliability requirement on each engine from our side. We need to meet that. The step change from that level up to a level where the hyperscaler wants to have, this is one of the major tasks and engineering and also risk management competencies of these power developers for the hyperscalers.
OK. Then.
Now in four minutes. One more question from you, please. It needs to be the last.
Yeah, the final question then is, when I just look at marine equipment and the different kind of segments, I can just notice that the merchant segment has been clearly larger the last three years than before that. Of course, they have had a good order cycle. I mean, is there something beyond this that you sell more products to the merchant segment than to other segments that you have increased compared to history?
I mean, you could say there is a little bit, you know, we sold more auxiliary engines, especially with the introduction of methanol. We talked about that before, that, you know, for the new fuels, we have a higher market share in main engines, but also in auxiliaries. That has fed in a little bit. I would also say cruise has been very muted for a number of years. Now it's starting to come back. If you look on the percentage, this means that percentage-wise, merchant has been strong. As I said, I'm optimistic about cruise going forward.
Thank you.
Thank you, Tom. We have two minutes. Nicholas Dunn, you can raise one question, but only one.
Thanks for fitting me in. Just one on the, I'm not sure if you looked at this CAT agreement with Hunt Energy. I think it was like a 1 GW over some undetermined period of time. Maybe you can talk through why, I guess, why they went with CAT as opposed to you. Did you compete in this? Would you be interested in signing a similar type of agreement with someone else where you agree to provide X amount over a few years?
I must say I pull a blank on the CAT agreement. I mean, Caterpillar , I guess. I don't know. I can't comment on that. I can comment on, you know, are we in our business thinking, do we want to sign up agreements, you know, for power for a long- time forward? We are a little bit hesitating on that. I mean, if you want to sign, you will have to pay the full amount. There are very few customers willing to do so. To charge like a 10% or 20% reservation fee, it normally doesn't work very well. That's why our general approach is that, you know, if you're willing to place a firm order 100%, we book you in. To pay 10% as a reservation, no.
I'm afraid that we need to close the call now. Thank you, Håkan. Thank you for all of the great questions. Pre-silent call will be on September 30. I hope you can enjoy September before that. Thank you.
Thank you very much. Thank you for today.
Thank you.