2026.09.29 | TB | Power | Aeroderivative Turbines Expert: GEV & Woodward

Client-led • 14 Sep 2026

Companies: WWD US, GE US, SIE DE, CAT US, BKR US, FTAI US

Expert Former Senior Role, Global Services Aeroderivatives at GE Vernova Inc

Agenda

• Over 10 years’ experience in the turbomachinery and heavy power generation equipment sector, with deep expertise in reciprocating engines, gas turbines and specialised fuel control systems
• Well-placed to evaluate power generation market dynamics, capacity expansion feasibility and emerging demand drivers such as behind-the-meter power for data centre infrastructure
• Knowledgeable on supplier selection strategies, component outsourcing and structural differences in original equipment vs aftermarket margin profiles across major OEMs, including GE Vernova
• Well-versed in sub-tier supplier relationships, particularly regarding long-term service agreements, component pricing power and aftermarket service channels involving major system providers such as Woodward

Questions

00:03:24] We are interested in learning about component suppliers for engines and turbines, particularly those specialising in fuel systems such as Woodward. We want to better understand industry dynamics, aftermarket operations, supplier selection processes and contractual agreements within Caterpillar. Who are the top suppliers with the deepest relationships and most invaluable IP for Caterpillar Solar Turbines and reciprocating engines?

Expert [00:05:18]: I don't know much about the recips side of the house. On the Solar side, basically, the IP to a big extent is owned by Solar. Now having said that, there are a lot of manufacturing processes that are owned by the supplier. I'm thinking of single crystal blades, for example, the geometry of the blade and the number of cooling holes and those kinds of things is owned by Solar, but how you actually manufacture the blade so that the grain structure is correct is owned by the supplier. Howmet is a big one, just picking on blades for now. Generally, I would say the engine components that go inside of an engine blades, shafts, disks, those kinds of components tend to be higher IP, higher reliance on suppliers, a lot of just single suppliers for those. As you work your way from the inside out, then you find that there is flexibility in having a multisource solution. Engine typically tends to be one, but then you get into things like fasteners, for example, you can have multiple suppliers for those or you get into package-level components, a lot more optionality there as well.

Client [00:06:54]: Did Solar Turbines work with Woodward as a supplier?

Expert [00:07:01]: No, not that I'm aware of. We didn't.

Client [00:07:04]: Whom do we use for fuel system components, fuel metering, actuation and similar parts?

Expert [00:07:13]: Injectors were all in-house, built in-house. All of the servers and injectors and all of that was all done in-house. Then for actuations and fuel metering, I think a company used to be called PECC, Precision Engine Controls. I think they were bought out by, gosh, Sundstrand, I think, or United Technologies, I think owns them right now.

00:07:52] You mentioned that injectors were produced in-house. How do we evaluate which components to produce in-house vs outsourcing? How does that process look?

Expert [00:08:05]: It wasn't an annual review. It was, I would say, a big discussion, big point of debate whenever a new product or a modification to a product that we did that we did the sort of make vs buy decision. Some of it had to do with it was a proprietary design that didn't want anybody else to have and so then that in-house or did we already have capacity available and we could use it, increase the utilisation in the factory, but generally, I think it was all about if it was something that could outsource, could switch from one supplier to another would be something that would decide to do that. At the end of the day, it was a big decision around cost and the IP side of it that made that decision. Once the decision was made, I would say, generally, we're stuck with that for a long time. Maybe every few years, we would kind of revisit, "Oh, we need to start outsourcing a part," because it's gone too expensive to build it in-house or we need to insource something because our supply chain isn't as robust and the delivery times aren't good or whatever it may be. It was a constant review, but something had to have been changed either on the cost side or on the delivery side or the quality side to go through that cycle of make-buy decision again.

00:09:53] What is the length of that cycle? Are we tied to our supplier list for every turbine generation and were there regulatory barriers or primarily inertia preventing transitions?

Expert [00:10:09]: No. Yes, there was no regulation. It's a good word, inertia to a big extent because it takes a lot of effort to… It's not so difficult to outsource. It's very difficult to insource. If you decide to bring a component in, do you have the right machines? Do you have the right people with the right training, the right tools, whatever to insource something. Then there's always this ramp-up. You don't want to cut the supplier off immediately. let's say, the suppliers was financially unstable, we would start insourcing some percentage of that just in case something happened with the supplier. That was another decision criteria of why insource vs outsource.

00:11:05] How were original equipment and aftermarket operations typically handled for Solar Turbines? Was it structured similar to an aerospace razor-and-blade model, where original equipment is sold at a relatively low price and margins are generated from the aftermarket?

Expert [00:11:30]: Not so much.

Client [00:12:45]: What was Solar's general margin philosophy regarding original equipment vs the aftermarket?

Expert [00:12:53]: Ballparking, if you got in the high teens, that was a good threshold to be at or higher was the expectation on the new unit in the power generation side. On average, I would say that would be the kind of numbers we would look at, we would see.

Client [00:13:23]: What was the approach on the aftermarket side?

Expert [00:13:30]: A lot more. It's kind of dependent on what you were doing. If it was an overhaul, it was probably twice that still. It depended on what you were doing on the aftermarket, what services you were providing.

00:13:50] Did Solar typically handle its own aftermarket servicing or did any suppliers handle their own aftermarket services?

Expert [00:14:01]: Generally, I think I would say a good general rule was that if it left the Solar factory, from the Solar factory. If it was a component that was on the package or it was shipped with the package, generally, Solar would do the servicing for that. If it was a bought-out item that got delivered in the field that was most likely commissioned by the supplier, then usually, they would handle the aftermarket. If it was, I don't know, coolers for oil, lube oil cooler would be even a wrong answer, but generally, the servicing would be done by that supplier. I don't know, we'll use the fuel metering valve as an example. Fuel metering valve gets installed on the package. If something went wrong with the gas turbine package, Solar would generally be the first person that would get called and then Solar would handle the logistics of providing a replacement valve, shipping the old valve back to the subsupplier to get repaired and then either sold back to the customer or put back into inventory.

00:20:40] Woodward is a major supplier to Vernova. Could you expand on that dynamic, particularly any details regarding Woodward specifically?

Expert [00:21:05]: What's different about Vernova was that you may or may not know this, but the engine itself, the aeroderivative engine, the design is, I would say, owned and the manufacturing of the engine is done at GE Aerospace. GE Aerospace provides an LM2500 engine tested to Vernova or Baker Hughes and it is the same with the other products. A lot of the components that are specced and installed are done by Aerospace. I don't know exactly, to be honest with you. For Vernova to go and say, "Oh, I don't want to use Woodward, I want to use something else" is going to be very difficult back to its already been done. In that case, it's a little bit different in that GE Aerospace has a direct relationship with Woodward for the new components that go into that engine. Once that engine is sold to, let's say, GE Vernova and the aftermarket needs a replacement valve or they need a replacement component, that doesn't get consolidated like at Solar's case, that's a completely different contract relationship, everything demand stream that comes in case of an injector or a fuel valve or something like that and (talking over each other). Baker Hughes, exactly the same thing happens at Baker Hughes. They have a completely different contract relationship buyer supply chain going with Woodward. Whereas Solar consolidates all of that demand into one buyer, one mechanism, in case of the aeroderivative engines that come out of GE Aerospace, it's three different companies that interact with Baker Hughes [sic] for basically the same product.

Client [00:25:36]: Vernova recently pushed to sell its own services, achieving a high attach rate for H-class services. Does this mean Vernova deploys its own field technicians and facilities or is there another arrangement?

Expert [00:25:57]: Yes, that's been around for many years. I would say that 30 years ago wasn't like that, but in the last easily 10, 15, 20 years, there's, in the whole industry, a shift towards retaining your own services, field techs, repair shops, so on and so forth.

00:26:22] Woodward recently noted on an earnings call an intention to reclaim aftermarket servicing previously delegated to engine and turbine OEMs. Does Woodward have the contractual right to deploy field technicians and set up local shops? Would that infringe on existing OEM relationships?

Expert [00:26:55]: I don't know which specific products they're talking because they provide injectors and provide fuel metering valves and they provide control systems. I don't know what else they do, but on the aeroderivative side of the house, if the injector was kind of dependent. If the injector is dirty, you could put it through a sonic bath and kind of clean it. That really doesn't require Woodward. You just put it through a little machine and it cleans the thing and you put it back and it's fine. If it's damaged, that's something that Woodward has to do. GE, on the aeroderivative side, we didn't have the know-how or the technology to take the injector apart and fix it. Same with the fuel valve. Those two would always go back to the Woodward shops, would get repaired and then they would get sent back into Vernova who would then send it to the customer or keep it in inventory or whatever. The third thing was on the control system, that's sort of the same thing.

If it's a minor tuning that they had to do, usually, the engineers that were already there, the field technicians were there, it didn't really make a lot of sense to hire or call out a Woodward technician for a minor tuning or a set point change, but if it was more involved, then Woodward technicians will get called out. For example, we would call out a Woodward field tech to come in and help with the commissioning process. At least on their side, I don't know what other services they would want to reclaim except if they want to get involved with the minor tunings and stuff, but then you run into a cost issue. If you have five minutes' worth of work, you have to mobilise and demobilise the technician. That doesn't sound very efficient to me.

00:29:22] Historically, its aftermarket activity was focused on spare parts, where Vernova managed repairs and sourced replacements from Woodward. Does Woodward now intend to directly perform field replacements to capture higher-margin opportunities?

Expert [00:29:51]: Yes, they could absolutely do that. That's the other thing that's a little bit different between Solar and at least on the aeroderivatives side is that GE Aeroderivatives. The number of suppliers in the ecosystem in the aftermarket was just a lot. The aftermarket business couldn't really efficiently have a warehousing strategy or supply chain strategy to be able to maintain all of these components. These suppliers oftentimes went direct to the customer. In the case of a Woodward, they could go directly to Woodward and say, "Look, instead of going through GE to buy your fuel valve or repair your fuel valve, just send it to me directly," and they probably can do that.

00:31:14] Regarding long-term agreements, Woodward typically enters contracts with OEMs. Is that how you understand the relationship and what are the standard terms, price escalators and scope inclusions regarding new builds vs the aftermarket?

Expert [00:31:37]: I couldn't answer that question. I don't know.

00:32:52] Regarding power generation and the demand driven by AI build-outs, load growth was stagnant for a long time but there is now significant demand for both grid and behind-the-meter power. Is it realistic for engine and turbine original equipment manufacturers to double or triple capacity over the next few years or are there upstream bottlenecks?

Expert [00:33:32]: Yes. I'm a little bit of a pessimist on this point. It's interesting I think if I read between the tea leaves, they're kind of thinking similarly. On the heavy-duty side, there's definitely an increase in capacity, an active drive to increase production volume because I think that demand is going to continue to grow in my head. On the medium-sized machines, and that's Siemens with their SGT-800. They've said they're going to double their volume of an SGT-800, so a very specific product and we kind of know how many they were doing annually, they're going to double that. That product is a really good fit for baseload operations. If a data centre needs 200-300MW of power, an SGT-800 is a really good fit for something like that. If you look at the smaller-sized machines, Vernova has not made any statement on increasing capacity on their aeroderivatives. and neither has Baker Hughes. They're the big players in the smaller side of the medium gas turbine side. Solar Turbines has said they're going to double their output on a gigawatt basis and that's true. I think they are, but that was part of their Titan 350 programme anyway.

They were introducing that product for the oil and gas market. The AI boom slows down, they're going to refocus that product back to the oil and gas. Outside of that, nobody else is on the smaller or medium-sized machines, nobody else has made a statement of doubling capacity,(talking over each other). I think on the engines, the reciprocating engines, they top out at about, I don't know, 10-15MW. I think Wärtsilä has made some comments about increasing size. I'm not too familiar with the recips size too much either. I think Caterpillar has also made some increases, but again, from a megawatt perspective, that doesn't really add up to a lot. It's the turbines that drive the gigawatts.

Yes, when you're talking about the AI side of the house, when you're talking about each facility is 200MW, 300MW, 400MW and now the latest designs are 1GW in size, reciprocating engines are [sic] going to solve that problem. Recip engines are going to be 10MW, so much smaller than what's actually needed.

Client [00:37:09]: Do you believe turbines will present the main bottleneck?

Expert [00:37:14]: Yes, it's the turbines that are going to satisfy. Two things are going to happen with data centres. The ones that are going to be these large 1GW size data centres, they're going to buy their own combined cycle power plants. They get dedicated power plants that are going to just power them. Then the ones that are 200MW or 300MW in size, in the next, call it, 4-5 years, they're forced to have a behind-the-meter solution. They're going to have to have their own gas turbines producing power within the confines of the data centre, but in parallel, every single utility has put a large number of orders for new combined cycle power plants. Once those come online which is starting at 2030, and they're going to start getting added to the grid,. 200-300MW power plants for these size machines are going to slow down significantly, I think, starting 2030-31.

00:38:33] Caterpillar appears optimistic about the demand for reciprocating engines in behind-the-meter applications. Given the size limitations of these engines, is this adoption primarily driven by speed-to-power requirements?

Expert [00:38:47]: I think so.

Client [00:38:48]: It happily picks up early recips.

Expert [00:38:51]: Exactly. I think that's what's going on right now. What you're finding is whoever can provide the kilowatts or the megawatts from a delivery perspective, will do very well in the next 3-4 years. I think that's what Cat is banking on. Hey, so what, they have to have 30 of these machines there. It's bad design, but they're willing to pay for it, and we can deliver in a year. The more optimum design would have been five gas turbines, but that's not going to be available for five years. Let's just give them 30 reciprocating engines earlier. I think that's the thought process.

00:39:32] Caterpillar is discussing tripling capacity on the reciprocating engine side. Is that a realistic target or an overestimation?

Expert [00:39:46]: I think Caterpillar has a lot similar to Siemens that have a good control over their supply chain on the gas turbine side. I think Cat, they're pretty vertically integrated on their reciprocating engines. I think they have a much better chance of being able to up the volume quickly vs a turbine manufacturer. The turbine manufacturers are so dependent on subsuppliers and those subsuppliers are very few. When you talk about a blade manufacturer, you got Howmet, maybe a couple of other ones that supply all gas turbines, heavy-duty, flight engines, gas, industrials, whatever. In order for you to be able to increase volume on a gas turbine, you have to have every one of your suppliers agree to give you more parts. That's hard to do when you're talking about 300-400 suppliers.

Client [00:41:17]: Which is interesting. It seems that industry participants appear to be preparing for a substantial increase in production.

Expert [00:41:26]: Yes.

Client [00:41:28]: Could you discuss markets outside of power generation?

Expert [00:41:33]: I was going to say that's interesting because like I said, they're probably making the inquiries now with their suppliers. If they can get 80-90% of them to shake their head, yes, we can do it, then they can make the announcement at the turbine side, because right now, I think they're all worried if they make the announcement, they'll sound like FTAI. They said they're going to do 100 and they're going to do six.

Client [00:41:57]: Outside of power generation, what are the emerging growth areas if data centre demand declines? Where could excess capacity be redirected?

Expert [00:42:20]: I would say the demand and supply capacity was quite even until this AI data centre started. Before that, the typical places where gas turbines were used were either in power generation, sort of a big umbrella power generation, which included utilities. On the bigger size machines, 60MW machines, you would sell blocks for utilities, both large utilities as well as in small municipalities. You would sell them into industrial applications, factories, hospitals, shopping malls, whatever, you would sell them into peaking power plants. I would say those are the big three pieces on the power generation side. Anything else? I think that's about it. Then the other big kahuna that uses a lot of gas turbines is oil and gas. Then in oil and gas, you use them across the entire chain from upstream, midstream and downstream. You use them both in compression, in pump applications, in power generation in oil and gas, so everything. I would say, probably in the small and mid-sized gas turbines, it's about a 50/50 split. 50% of volume used to go into power generation. The other 50% would go into oil and gas.

00:43:53] Woodward categorises its end markets as roughly one-third power generation, 40% marine and transportation and the remainder O&G [oil and gas]. It is surprising to hear that O&G accounts for 50%, as it appeared to be a smaller segment for them.

Expert [00:44:19]: It sounds like Woodward also provides a lot of components to reciprocating engines.

Client [00:44:25]: They likely allocate to reciprocating engines.

Expert [00:44:29]: Yes, yes. That's a big volume. Yes, there's a lot more recips that get sold than gas turbines. I don't know what the number, I would suspect because there's so many of them out there. You got Cummins, you got Caterpillar, you got whoever and they do high volume.

Client [00:45:29]: Are there any other relevant areas we have not yet covered?

Expert [00:45:35]: With the PECC design, we only had the one valve, one metering valve. On the Woodward side, the design is different. You have multiples of them. It's not apples-to-apples, so I don't want to talk about that. Anything else? I think I mentioned to you that at least my pessimistic view of these mid-sized data centres, I think where the volume is going to go on that one. There's definitely a lot of folks that are trying to get into this market.