← 返回

2026.10.06 | TB | Energy | USA Compression Capacity & Project Economics

原标题：USA Compression – Compression Capacity, Supply Constraints & Project Economics

#energy

Third Bridge · 2026-10-06

Analyst-led • 29 Sep 2026

Companies: USAC US, AROC US

Kirk Schafer (KS) Director, Project Management at Kodiak Gas Services LLC

Curren Larson (CL) — Third Bridge Sector Analyst

Agenda

Key Insights

Introduction

CL [00:00:02]: Welcome to Third Bridge's Interview entitled USA Compression – Compression Capacity, Supply Constraints & Project Economics. I'm Curren Larson, and I'll be facilitating today's Interview with Mr Kirk Schafer, Director of Project Management at Kodiak Gas Services.

Kirk, before we start today's Interview, please state I agree or I disagree to the following statement: You understand the definition of material non-public information and agree not to disclose any such information, or any other information which is confidential, during this Interview.

KS [00:00:33]: I agree.

CL [00:00:36]: Would you mind briefly introducing yourself before we jump in?

KS [00:00:43]: Director of Project Management for Kodiak Gas currently. I've spent the last 12-15, somewhere in there, years in the industrial, specifically heavy industrial EPC space between the power, downstream and midstream and refining industries, specifically on the contractor side, was kind of in and around the compression space secondhand since I was in high school due to some personal relationships with some guys that were very influential with the way that the market is shaped today and did some work as a third-party contractor for Kodiak Gas back in 2015-16. Then obviously came on board just over five years ago to run turnkey projects out of their special projects group directly, but really understand the full scale of the project side of things and have been in and around and obviously more so over the last five years, pretty heavily involved in the compression side of things.

Questions

00:02:18] How do you see demand for reciprocating compression evolving over the next 3-5 years? What gives you confidence that the current supply-demand imbalance can be durable rather than temporary?

KS [00:02:48]: The market generally sees it as holding to what it currently is for a number of reasons. One is the well-known supply chain constraints, specifically around engines. Two is the amount of demand coming in from the major gas shale plays around the US and that demand just steadily projecting to go up and to the right, both due to various domestic demand, legacy demand as well as the added demand around the data centre boom, but also export demand. Again, it's pretty general consensus that that is going to continue to hold. Then obviously, on the constraints around supply chain, while there is a little bit of relief potentially in sight through primarily Caterpillar, we're still years out from my understanding and everything that folks are communicating and discussing within the industry from there being real meaningful relief from that end of things.

Also, if you're asking me, honestly, I don't know how much motivation the folks that's supplying the critical components have to really drastically relieve the supply constraints that they have in terms of manufacturing, whether that's right or wrong or whatnot is, I think, another discussion, but you've always had constraints around that that have popped up over time and not to downplay the efforts that they are taking, but they always seem to find their way back into a constraint at some point or another. With everything, it always seems to get bigger and bigger with the scale and whatnot.

00:05:13] How has the conversation changed with E&P [exploration and production] and midstream operators, now that lead times for horsepower units are stretched towards 200-plus weeks?

KS [00:05:46]: A lot more upfront planning that folks are looking instead of 12-24 months out in terms of horsepower needs, obviously, they're looking significantly further out. In some cases, that's leading to commitments. In others, it's just leading to indication of possible commitments. I think what it is doing is, in some cases, leading folks to look at alternative solutions to help relieve some of the issues associated with the horsepower lead times, whether that's through different technologies or optimisation of existing horsepower or what have you. It's a mixed bag on how people are approaching it. Again, some commitments that are being made that far out. From what I've always heard and seen, they try to limit that, committing super far in advance unless it's a horsepower that the end user is directly purchasing. If you're talking from a lease standpoint, they tend to try to thread the needle, so to say. It is also not uncommon for a lot of your leased compression across the market to be placed on contract long before the major components are hitting your packaging shops where they get put together and turned into a compressor.

CL [00:07:36]: You mentioned alternative solutions. What would those look like today? Do you view any other technologies as a threat to reciprocating compression?

KS [00:07:46]: Electric-driven compression turbines, although turbines are hard to deploy at the scale that your recip compressions are definitely in a lease fleet capacity. They're not as mobile historically. I have heard of some folks trying to figure that out and come up with a workaround there. Historically, the lead times and the sensitivity of the turbines is typically worse than your recips, which is why the lease market and a lot of more of these shorter-term assets tend to go to recips where it's your main pipeline, something that's going to have some staying power tends to utilise turbine-driven compression. What I can say confidently is there is a lot of optimisation discussions that seem to be going on, folks that may have too much horsepower sitting in a given location and looking at opportunities to redeploy that horsepower on another one of their locations where they need horsepower and take on some smaller horsepower because there does seem to be a decent amount of movement and availability generally of smaller horsepower. When you start talking about Cat 3500 series, 1875 horsepower series and above, 1500 horsepower even and above, that's where you start to see the pretty hefty constraints in terms of supply.

CL [00:09:35]: Is optimising horsepower a solution that a company such as USA Compression or Archrock is offering, or is that typically third-party?

KS [00:09:44]: Generally, that would be something that a company like that would offer. I can't speak directly if either of those offer it. Typically, especially if it means the ability to deploy idle horsepower that might be smaller that would otherwise just be sitting on the bench by just shuffling a deck of cards that from a business standpoint, that makes good business sense. I know that they and companies like them will do that at times.

00:10:31] Where is electric-driven compression's pricing relative to reciprocating compression?

KS [00:10:48]: That's a nuanced question to answer. It all depends on at what scale and size you intend on deploying it and where you intend on deploying it.

CL [00:11:06]: At what scale does reciprocating compression make more or less economic sense than electric-driven?

KS [00:11:11]: You can actually make a very practical argument for either in the right case. Usually, your constraint around the electric-driven compression is going to be the availability of power, whether that's generated or land-based power, where if you were to go set in just about every case that I've been exposed to, if you were going to go set an electric-driven compression, vs a single electric drive compressor package vs a single gas-driven recip package in the same horsepower deployment, same application, everything, the gas driven is going to be cheaper over the long term because of the amount… The lack of scale rather associated with the electric compression. However, as you start to deploy multiples simultaneously on the same site, there's some technologies that you can do on the electric-driven side that can help bring that cost down. I can tell you that the maintenance side of things from a PM standpoint tends to be a little bit easier on the electric-driven side if you can get the people, which is its own challenge. You're not having to do your typical engine PMs and overhauls and all that type of stuff. It's a little bit more straightforward on the engine or on the electric motor drives themselves.

00:12:50] I believe USA Compression's capacity is booked out in 2026, 50% booked out in 2027 and extended into 2028. To what extent are customers willing to commit to binding take-or-pay contracts two years ahead of delivery, given the current equipment constraints? Is that becoming more frequent? If so, what contractual protections are typically built into those agreements?

KS [00:13:35]: It's a mixed bag. Usually, your lease fleets will have some pretty strict cancellation language in there to protect the business itself. I don't want to go and commit millions of dollars to acquiring and packaging and prepping to deploy this thing to a customer and have them cancel and not have a little bit of recourse. There's certainly some. I will say that in previous stints where demand has been super high relative to the supply, my experience and what I've always heard is that most of those fleets are pretty willing to be forgiving, especially if it's a big client, to let folks out of an agreement before the unit gets deployed because the odds are is that you've got that contracted currently at a lower rate than you could contract it out at today. You may have one guy that's saying, "I don't need this unit right now," but I've probably got five or six that are beating down my door looking for that exact same unit that you don't see that cliff really drop off.

If it does drop off, that's where you've got the contractual protection cancellations and things of that nature in place that allow you to recoup a little bit. They can put it on standby or whatever the case may be, and that's what a lot of clients will end up doing. Maybe your mom-and-pops, you just don't see as many of those. Right now, we're going through a big series, it feels like, of M&A where a lot of your smaller outfits are getting scooped up and getting absorbed into your bigger E&Ps and mid-streams, where they have a little bit bigger appetite to sit on or relocate where that compression is ultimately going to end up going.

CL [00:15:48]: Let's say you had a cancellation today. How long would you expect it to take to redeploy that capacity? Is it pretty much the next day, given how tight supply is?

KS [00:16:04]: I've heard of guys that have redeployed cancelled units while they're on the call being notified of the cancellation. It's pretty wild right now. I wasn't as heavy into it during the COVID timeframe, but it feels like the way that they've described that at times when the supply got super, super constrained around them.

00:16:35] With demand currently outpacing supply, how much pricing power do tier 1 compression providers have on contract renewals vs entirely new deployments? Is it actually a discount?

KS [00:17:06]: You tend to see a little bit of a discount at times on older packages. It depends on the age of it. It's not drastic, but that's an aspect of it that I'm not super read up on or involved in. What I will tell you is that it is more economical even if you're paying a one-to-one same rate for a newly deployed package vs a currently deployed package. If I can leave that unit there and renew that contract and let it keep working, it's going to be a heck of a lot cheaper for me to do that than it would be for me to pay both ends having to demob the existing unit and then pay to have a new one remobed in at what inevitably would be a higher rate to begin with.

CL [00:17:57]: What's the cost for USA Compression on a contract renewal? Is it still having to go in and do a bit of maintenance?

KS [00:18:09]: It depends on the age and state of the unit. If it's a leased unit, you get the fleet owner, the fleet provider is going through their PM schedule and doing all their things. There may be some things that have to be done. For the most part, that gets built into the overall cost of just doing business on a month in, month out basis. Whether you've got to swing an engine or do some reconfiguring or whatever the case may be, that doesn't typically, unless it's something that's driven by the customer operating outside of the parameters that were agreed to in the contract, that's typically not something that the customer is going to feel.

00:19:02] We've seen USA Compression's service revenue jump over the last couple of quarters. Is that because there are a lot more contract renewals, so more maintenance is necessary?

KS [00:19:30]: I think it's probably more so caused by you're, what, five years out from a heavy chunk of new horsepower deploying out from the Big Three, so to say. You're just at that time of when you've got a lot of horsepower that's due to have some PMs done, some pretty heavy maintenance schedules on it. The lease compression market has been growing like crazy over the last handful of years. It would make sense without having even looked at the rate and schedule in which horsepower had been deployed over the last x number of years, it would make sense that service costs will continue to go up within those lease fleets. The more horsepower you have and the longer that it's been deployed, the higher your service costs are going to be.

00:20:32] Where do you think scale and fleet standardisation actually create an advantage in the current market? Are the larger providers commanding a premium relative to smaller local competitors, or are their earnings just stronger, given scale?

KS [00:21:14]: In terms of standardisation, that was kind of what, call it, mid- to late '90s when this market really started to grow and show up on the landscape en masse. That was the big thing that took your early players that ultimately morphed into a part of USA, took them from a small group to what USA in part is today. It's being able to have the overwhelming majority of my fleet using a very narrow subset of different parts and pieces because they are being driven by very similar, if not the exact same components within the skid. It makes training a lot simpler. It makes inventory control a heck of a lot more streamlined. You know that for the overwhelming majority of the big fleets out there, it's Cat-driven and Ariel compressed. If I know that I'm going to be using a lot of Cat parts across my very large fleet, it allows me to go in and get certain price breaks that where if I'm spreading that across multiple different engine providers, I won't necessarily get those types of price breaks. I think really where the standardisation shows up is in the training and the standardisation really of the training of your workforce.

I know that if I go to work on a 3516 package unit in Ohio, and I go and I work on one that's in west Texas, New Mexico, while there's going to be some subtle differences, whether that's in cylinder sizes, it may have a hot start in Ohio vs probably wouldn't have one depending on where it is in Texas and whatnot. Generally, the guts of that unit are going to be the same even over time. While, yes, there's been a fair number of changes to the major components. They're not so drastic that it's not like going and trying to write a paragraph with your left hand all of a sudden, you can typically pick up and figure your way around it. It makes moving people around and makes training people. It makes refresh training a lot easier. You're not having to go ground up because of some new technology that's come out. We're just doing incremental training at that point in time.

You would ask about the big fleets and their ability to impose their pricing across the market and whatnot. It depends on who you ask and it depends on which way the wind is blowing. It's like anything, if you're the new blood into a market and you've got a little bit of money behind you, am I going to go give away some work early on to build the name and put some folks to work and show that we're the best at what we do and then gradually try to work my rates up? Absolutely. I think it also works from the big boys. There are times where you see the big boys maybe not undercut the market, but go really, really competitive because they can and they want to make sure that they retain the market share and/or incrementally grab more market share. Then there's also plenty of instances where they know what their business is worth and/or they know what it's going to do potentially incrementally to their staffing needs, and we know staffing constraints exist right now and whatnot. They feel like the service and the stability and the experience that comes with being one of the big fleets is worth the money that they're asking for. A lot of these customers will pay it. There's something to be said about having a four million horsepower fleet behind you rather than a 200,000 horsepower fleet behind you.

CL [00:26:03]: How much of a premium do you think larger providers such as USA Compression can realistically command over smaller, local players before a customer has no choice but to leave?

KS [00:26:37]: That's hard to say because I'm not in a lot of those conversations.

00:26:45] If a customer urgently needs capacity but can't wait for a reciprocating compressor, how viable is a turbine as an alternative in practice? Could you elaborate on the key economic and operational trade-offs that the customer would have to make?

KS [00:27:20]: Turbines, I've not run across a practical application for a skid-mounted turbine, which is what, by and large, your lease fleets are employing. They're skid-mounted so that they can be moved around and redeployed when they're done at whatever location they're at today. Turbines are much more difficult to do so. The cost of them is a heck of a lot higher. There's a lot more precision that goes into setting up a turbine vs a recip compressor. Every application and every turbine that I've been a part of installing or I have looked at installing, it's a much more long-term asset than it is a short-term asset. That's why you see them on big pipelines, big transmission lines, plants, refineries, places like that, that aren't going anywhere anytime soon.

00:28:30] Is there any operational difference we should understand between installation and service in the Permian vs in the Northeast and the Rockies?

KS [00:28:56]: Operationally, you're dealing with the cold vs down in Texas, New Mexico, it doesn't get as cold so that you don't have freezing issues as much, at least associated with cold and whatnot. Labour market is a little bit different. You get further up to the Northeast and you're having to be mindful of union interference and things of that nature. Beyond that, operationally, it's hard for me to say, at least in terms of compressor mechanics. In terms of installation, your installation costs are typically a little bit higher in the Rockies or in the Northeast, again, due to weather, you got a lot more winterisation and/or you're doing enclosures or whatnot. Your heat tracing and insulating your liquid lines, you're putting hot starts on it to help your unit at start-up. Just the cost of setting that unit down is a lot more, and therefore, your rates tend to be a little bit higher from what I've seen beyond the winterisation and some of the considerations that you've got to make in terms of the foundation that you put on.

You get into the Rockies and you get into the Northeast, the soil is not as good as it is out in west Texas, New Mexico, where it's super flat and you can build a west Texas caliche pad and set that skid directly down on it, whereas I've not done any work in the Rockies myself. I've looked at some work and it can be hit or miss depending on what the soil conditions are, and it certainly is hit or miss when you start getting up into the Northeast, the soil tends to be garbage up there. You see a lot of units going on either helical piles or full-blown concrete slabs, which is, I would say, uncommon for rented compression in the really Oklahoma, Texas, New Mexico and even parts of Louisiana for those to go on anything, but we call them a chat box, but a gravel pad effectively.

CL [00:31:36]: Is it fair to assume a bit of margin expansion for USA Compression just from having more revenue tied to the Permian and the Gulf Coast than previously, relative to when it had more in the Northeast and the Rockies?

KS [00:31:51]: Yes.

00:31:54] Could you walk us through the specific bottleneck at Caterpillar's manufacturing level? Why is it so difficult to expand production at the 3600 engine block? What would need to happen before we can expect meaningful relief in lead times?

KS [00:32:15]: There's two things at play. One is the 3600 block for both the engines that are used in compression vs the engines that are used in the power gen world come off the same line. Naturally, when you've got a spike in demand on both sides of it, it just exponentially makes your issue worse. My understanding is that they are in the process of splitting that so that they come off of separate lines. Then in general, your 3600 block line, Cat only has, I think it's two machines that are able to mill out the block that the 3600 engine is made out of. They predate you and I both. I think one was from the 40s and the others from the 50s or something like that. One of them has to be fixed. It's limping along or whatnot. My understanding is that they are about to start or they're in the process of actually putting in a new one, which is a very long process in and of itself.

As soon as that gets put in, I can't speak to what any uptick in productivity relative to the existing machines would be, but you'd like to think new technology would allow it to move a little quicker. My understanding is also that once they get that new machine up and running, they intend to take one of the existing machines down for some pretty major repairs. I'm not super familiar with the overall timelines of those other than to say that it's pretty far out there. Then you couple that with the overarching just general demand. It's making everything really, really slow to get through the Cat shop, and it's making for really good business for Caterpillar.

CL [00:34:29]: Is pricing different across power gen vs compression for the same block? Why separate?

KS [00:34:42]: I think it's just purely to relieve the bottlenecks. I think there's enough demand on both sides for them to justify it. In terms of pricing difference, I couldn't speak to that. I'm not involved in that side.

00:35:07] Aside from the engine block itself, where are you seeing the most significant secondary bottlenecks in packaging a compressor unit today?

KS [00:35:15]: From my understanding, Ariel is in a fairly good place. That's not a massive bottleneck right now. Shop space tends to be fairly volatile. The shops that are not solely dedicated to packaging compression are busy with other non-compression work, especially for your big packagers or your guys that are on some of these start-ups that have been in the space for a long time, they tend to earmark some room for them. I do know that those shops are pretty busy right now, and there's none that I've heard of that are adding to their capacity. There doesn't seem to be a lot of relief coming down the road, which doesn't seem to be directly impacting because you can't get engines to save your life. If engines were to magically appear, it's not something that you can put on the floor tomorrow by any means. Then the other side of it is I do know that there are some shops that are primarily dedicated and historically dedicated to just packaging compression of various shapes and sizes that are looking to shift and diversify their product line to include some power generation. When you start talking about that in terms of bottlenecks for packaging compression, that is certainly, I would think, play a factor in it long term if they truly do start adding that and taking up floor space.

00:37:05] How long does the packaging side take today? What's the lead time for a unit to be packaged and made available?

KS [00:37:24]: The packaging itself, once the major components get there is not very long. It's really, really hard to say because the engine and the compressor are the two primary components that drive that, and they're ordered so far in advance. I think once you get your main components, you've already got the majority of your bottles, you got the majority of your piping done, a lot of your electrical… It's not super long once you get all your components. Beyond the engine and the compressor, your panel is probably your next long lead thing. There's some speciality items that some folks will put on that can trickle out there, but nothing near to the degree of the engine and the compressor. Once you get all the stuff in, I think the last lead time or turnaround time I heard was in the 20-25-week range. That's not one that I'm involved in on a regular basis, and everybody is so zeroed in on the lead time of engines that that's what we're talking about in terms of packaging durations.

00:39:01] USA Compression acquired J-W Power in January 2026, effectively giving it in-house packaging capacity up to 125,000 horsepower annually. Do you see that as much of a competitive advantage when, again, the constraint is largely just the engine block?

KS [00:39:19]: The only capacity in which I'm real familiar with J-W's packaging is around their EOR units, which they do some, but not a massive amount. My understanding is a lot of what they would push through to also would end up being smaller horsepower. It certainly doesn't hurt if you can run the size units through there, assuming you can get the components. If you can run the size units through there that support the incremental growth that you have planned, then in theory, you can drive down your cost or control that cost and also accelerate or decelerate so that you can have units as just in time as you need, so that you have stuff not sitting idle, at least in terms of new units. It all depends on what J-W is capable of pushing through, what in-house designs they've got that are at the ready or if you've got to flesh any out, do you need any new equipment? It's been years since I've been to J-W shop. What state is the shop in? Those would all have to be questions that you'd have to answer before you could really state that you've got a meaningful competitive advantage. I can't imagine that it is a detriment to your ability to grow.

CL [00:41:13]: How much margin does a company lose by having to go third-party with packaging?

KS [00:41:24]: Packaging in general is a single-digit margin business and a low single-digit margin business at best despite what a lot of people would say. You're certainly saving something off a USD 3m piece of equipment. If I can save 5%, I'll save 5%. That's not nothing. Short of playing some money games on how you look at internal resource costs, you're not getting rich and completely beating down the market. I also was never led to believe that J-W was overly cheap in terms of their packaging cost. Again, my experience with them was primarily with EOR compression, which is very, very specialised, very, very high pressure rated units for a very niche application that I would think, naturally be a little bit higher margin than your just everyday run-of-the-mill fleet unit.

00:42:37] How are you thinking about the industry's engine backlog today? Are there any early warning signs that the backlog is being inflated by double ordering, speculative orders or projects that ultimately may not materialise? If not, what early indicators would you be watching for?

KS [00:43:17]: There's definitely some speculative orders. There's definitely some stuff on order for projects that will never happen. I don't know about double ordering as much. It's a big industry, so you never say never. I do know that there's folks out there that they're trying to get their name on components and units and whatnot to make sure that they're not last one to the cookie jar, so to say. That's where your cancellation and stuff like that starts to come into play. Again, there's such a long line from everything that I'm being told, there's such a long line of suitors for units that if one were to get cancelled. It's a matter of hours and not weeks or days even in a lot of cases, it seems like to get those re-contracted.

CL [00:44:25]: Is there a way to know if people are starting to double order just to ensure that they can get engines as quickly as possible? Is it too opaque to tell?

KS [00:44:48]: It's going to be interesting to see if, I think strictly or specifically around some of the data centre stuff and the gas demand associated with some of it. You talk about double ordering and speculative ordering and stuff like that. You're starting to hear about a lot of that in that space, where companies and/or projects are either being put on hold or you've got some of these smaller companies that just aren't going to make it because they'll run out of funding or whatever the case may be. It'd be interesting over the next few years to see how much of that gets drawn back and/or dies just because they can't get permitting or can't get land or both. As of right now, assuming that the demand for compression generally remains what it is, I would imagine it would be the smaller groups if they've got the money that would fill any vacuum that the larger ones would potentially create by wanting to hold some of their timelines. Most of your big fleets have to be fairly strategic with the rate in which they are growing on a YoY basis because they're reporting that to the public markets.

To deviate from that is there's a lot of steps to go about doing that without creating a lot of hoopla and uncertainty in the marketplace. Never say never, but it would not shock me to see if that were to happen, some of your big boys say, hold generally to the timelines that they were either being quoted now or somewhere in between where they're quoted now and where they were originally quoted. Anything that remains, they're either pushing out or releasing altogether, and those would be the ones that would be interesting to see if they get scooped up by some of these up-and-coming groups that I wouldn't say don't care about money, but they're in growth mode.

00:47:12] If you were to see engine lead times suddenly normalise, what would be the immediate operational and pricing implications for major contract compression providers?

KS [00:47:35]: Operational, that's hard for me to say. Pricing, I don't know how directly that would follow through to the end user, especially with the contracts and whatnot in place. Waukesha for that matter, they're not known for cutting their price right before delivery, especially if they've already advertised an increase, and that increase is going to ultimately get passed to some degree down to the end user. I don't know that pricing would massively change, but theoretically, it could, especially if you see a bunch of engines that are currently spoken for that get released and re-contracted through a different fleet that theoretically could affect the price. That would be hard to say. That would be interesting to watch though.

00:48:48] How are you thinking about the inflationary environment today on the construction cost side? To what extent can USA Compression pass increases in labour, lubricant oil, parts and other operating costs through to customers or hedge them through contract structure?

KS [00:49:13]: There's a lot of different mechanisms that you can use. The biggest one that's thrown around is some pricing index that's more on your operations side of things. In terms of your upfront, just generally, if your responsibility extends beyond the compressor itself, you start talking about the build-out of full-fledged facilities, stations, what have you. I am seeing a lot more cost-plus structure contracts being thrown around, some of which end up dying on the vine and some of which go through. There's a ton of them on the data centre side of things. Obviously, that's the electric side, but will at times touch the compression world. They almost seem to be primarily doing cost-plus structured construction contracts. In terms of compressor facilities or compressor adjacent type facilities, amines and cryos and things of that nature, I don't know how deeply USA is getting into that, but the market as a whole is you take your best guess at what labour is going to look like one year from now and you try to manage to it. I will tell you that contingency buckets from everybody that I'm talking to on that space seem to be growing quite a bit because of that if you are in a lump sum type of structure.

CL [00:50:46]: Are operators doing anything in particular to combat lubricant oil cost increases? To what extent does that actually impact margins?

KS [00:51:04]: Massively. That's where you try to account for it in your pricing index levers inside of your contract. Sometimes that works, sometimes it doesn't. Then it goes back to just your traditional economic measures, buying at bulk through single users and things of that nature. There's some operational things that you can do to be a little bit more efficient with your lube oil to do things to prevent leaks and things of that nature. You're also that much more important on staying up on your PMs and whatnot to make sure that your natural either burn off or on-skid leaking of any kind of oil is kept to an absolute minimum to try to keep as much of that liquid gold working the way that it's supposed to.

00:51:58] How much of a premium would you assign to today's labour and lubricant costs relative to 2025's, given inflationary increases?

KS [00:52:16]: I can't really speak to lubricant because I'm very hit or miss when it comes to being involved in having to buy lube oils and such in my side of the business. In terms of labour, it depends on where you're working and what type of labour you're looking at. Mechanic labour has gone up. It's hard for me to say how much, but it absolutely continues to go up and the quality of that labour is pretty green. That's where the advantage of being one of these larger fleets, I think, comes into play because your training programmes are a little bit more well established and ready to take on some of your less experienced techs and whatnot. When you start talking about electrical, it is a brave new world, the absolute definition of the Wild Wild West with everything that's going on with data centres. The costs around electrical and controls techs, it's hard for me to put an actual percentage on it because it feels like it changes legitimately every single day.

The change hurts, but it's also the quality continues to go down because you have such an influx of new labour coming into the market. If for whatever reason, you or your contractors are on the lower end of the pay scale in the industry, you're getting the low end of the craft and then retaining that craft who will leave for an extra USD 0.50 per hour is very, very difficult as well. It's a weird time on the labour side of things.

00:54:15] Are there any other increasing input costs that are surprising you right now?

KS [00:54:23]: In terms of facility costs, cable tray is a weird one that I didn't expect. I'm seeing it on some jobs that I've been a part of recently that have a bigger cable tray scope than I've been a part of in a while. Where a compressor facility may have USD 100,000-200,000 worth of tray that's got to get bought and installed. You've got these data centres that are placing in these end users that are owning these big data centres that are placing USD 5m-10m orders of tray and it's just completely screwing the market when it comes that they're buying it in bulk so that they have it and then they're distributing it out to their various projects as needed. That was absolutely one that took me by surprise.

00:55:31] USA Compression is expecting enhanced telemetry and remote monitoring to improve route management and predictive maintenance, driving margin improvement by 2027. How much operational efficiency or margin improvement do you think can realistically be generated across the legacy fleet?

KS [00:56:00]: I think there's a break-over point in terms of the age of the unit where it becomes actual value. All of that feels so dang new that I have, maybe not doubts, but I'm a little hesitant to spike the football that it's going to be the cure-all in terms of extending out PMs and things of that nature and preventing failures and whatnot. I certainly think it will help, and I don't think that it's going to be all for naught. You got to pay to stand that stuff up. My thinking about it has been that in the immediate future, I don't know how much uptick in margin you're going to see because it takes a while for these new technologies to develop the trends and start to meaningfully be beneficial but also accurate on a consistent basis as to what the machine is telling them. From an operational standpoint, you got to get your guys up to speed on (a) interpreting what the data is telling you and (b) responding to it in a way that you can validate is just in time. That's what this is all about.

We're talking about trying to stretch these things that said that they had an X hour interval to a Y hour interval. Is what the data is telling you, is that correct? Is it way too early? Is it too late? I think that's going to take some time to flesh out. In any case, on the back end of this, you've got computers, you've got people, you've got infrastructure, you've got all sorts of costs that you're having to spend upfront to stand all this stuff up that has got to get paid up. Now what that total burden is at somebody like USA, I can't really say, but it's not nothing. There's upkeep and maintenance and upgrades and things like that that are going to go along with it. Do I think that it will have some benefit? Yes. Do I think it's going to show up as a substantial uptick in margin in 2027? Probably not. I think I'd be hard-pressed to say that it's going to be zero. I think you get two, three, four years and more down the road. I think that's where the meaningful changes are going to start to show up in my opinion.

00:58:55] What do you see as the biggest swing factor that could change the outlook for USA Compression's margins and returns over the next 2-3 years?

KS [00:59:14]: I really think that labour, it's hard to say that it's being downplayed because it is being talked about a lot. You've got some very influential people who have branched off and started up new companies, and they are doing some very interesting things and they're guys that have done it at scale very successfully previously. There's no reason to believe that they wouldn't do that. What that's going to do aside from market share is it's going to pull people from the pool that are available for your bigger fleets for a variety of reasons and being able to backfill with relatively unexperienced craft mechanics at whatever type and flavour or whatnot and be able to do so timely and cost effectively, I think, is going to be very, very interesting to see how those guys adjust over the long term.

Then you also are not just navigating some of these new start-ups, you're also navigating a lot of folks who are leaving the industry altogether to go try to latch on to this tech boom, to this AI data centre boom, whether that's electric techs that are making lateral moves in terms of the trade or guys that are going and learning a new side of the business altogether. I've seen a mix of that. Again, adapting to that, developing a really strong pipeline of new talent, getting that new talent trained and retaining them. I think that's going to be a major, major factor moving forward. I know that we're seeing it just from a construction standpoint in that end of the industry. I see no reason why operationally, it wouldn't be very similar.

CL [01:01:27]: Certainly, something to keep an eye on going forwards. I think that's all we have time for today. I just want to say thank you for joining me and sharing your insights on USA Compression and the broader market.

Transcription ends at 01:01:37 of the recorded material