Power & Sites · United States

Data Center Developers Face Tighter Margins as Construction Costs Rise

Credited to Data Center Knowledge · datacenterknowledge.com · 17h ago

Photograph · Data Center Knowledge

JLL says strong AI demand is not enough to protect projects lacking power certainty, anchor tenants, and credible delivery timelines.

Data center developers may be able to pass higher construction costs on to tenants, but only when projects have power access, committed customers, and a credible path to delivery, according to JLL.

In its Global Construction Costs Index 2026, the real estate services firm said annual construction cost escalation of 3% to 6% has become the new baseline, up from the 2% to 3% range common before 2020. JLL also forecasts another 3.1% to 4.2% increase in global tender prices in 2027, including 3.2% to 4.2% in North America. Those forecasts apply to construction broadly rather than the data center sector alone.

The JLL report notes that rising rents and asset values have helped data center projects absorb some of that inflation. However, Matthew Giddings, executive director of technical services with JLL’s Data Centers and Critical Environments team, said stronger market fundamentals do not automatically protect individual projects.

Related: Vertical Data Centers Face Cost-Proximity Trade-Offs

“The economics break down when costs outpace contracted revenue or power delays postpone income while financing costs continue to accumulate,” Giddings told Data Center Knowledge. “Higher asset values alone cannot close that gap.”

That risk is most acute for speculative campuses without anchor tenants, projects with uncertain power delivery, and developments tied to fixed-rent agreements that leave little room to absorb construction escalation. Older facilities being repositioned for AI workloads may also face difficult economics if upgrades become too extensive.

“Strong demand supports the market, but it does not make every project viable,” Giddings said.

Developers should reassess projects when rising capital costs erode expected returns, utility schedules slip beyond committed occupancy dates, or tenant demand no longer supports planned capacity, he said. Reduced power availability and delays in anchor-tenant commitments are also signs that a project may need to be reworked.

One way to manage that risk is through phased construction. By aligning spending with confirmed electricity supply and customer demand, developers can limit the amount of capital they commit before fundamentals are in place.

Skilled Trades Constrain Delivery

Labor constraints are adding to the pressure. Giddings said the electrical and mechanical trades are creating the greatest cost and schedule challenges, particularly electricians, high-voltage specialists, pipefitters, and advanced cooling technicians.

“Controls, commissioning, and startup teams have also become critical bottlenecks, as their availability often determines when a completed facility can be placed into service,” he noted.

Related: How to Secure Data Centers During the Construction Phase

Competition for utility and substation construction crews is also intensifying, adding risks that extend beyond the data center site itself. To reduce exposure, Giddings recommended that developers bring in trade partners early and test project schedules against real workforce availability rather than assumed labor capacity.

He also pointed to standardized designs, off-site prefabrication, and phased construction as ways to reduce peak labor requirements. In remote markets, developers may need to offer housing, transportation, and travel incentives to attract workers. In the longer term, investment in trade schools, apprenticeships, and union training facilities could help support future construction phases while expanding local employment opportunities, he said.

Buy Early Without Freezing the Design

Procurement strategy is becoming just as important. Long lead times for critical equipment have pushed many developers to place orders earlier, but doing so can create problems if tenant requirements shift, particularly as AI infrastructure continues to evolve.

“The key is to secure manufacturing capacity early without freezing every design decision,” Giddings said.

Related: Virginia Tops AI Data Center Readiness Ranking as Smaller Markets Advance

He said developers should reserve production slots for transformers, switchgear, generators, and cooling equipment while negotiating flexibility to modify specifications before fabrication begins. Orders should advance in line with validated tenant requirements and clear decision deadlines.

That approach is especially important as operators try to account for shifting rack densities and growing interest in liquid cooling. Modular power and cooling systems, built-in expansion capacity, and design provisions for liquid cooling can help facilities adapt without undermining the overall project.

“Early procurement protects the schedule, while flexible design protects the long-term value of the asset,” Giddings said.

Retrofitting older facilities can offer another path to faster delivery and lower upfront costs, particularly when a site already has available power, sufficient structural capacity, and infrastructure that can be reused. However, those advantages can disappear quickly if a building requires major mechanical upgrades, higher electrical distribution capacity, heavier floor loading, or large-scale liquid-cooling retrofits.

“The economics ultimately depend on how much existing infrastructure can be retained while still meeting current and future performance requirements,” Giddings said.

About the Author

Nathan Eddy covers data center trends and technologies across multiple industries. A graduate of Northwestern University’s Medill School of Journalism, he is also a documentary filmmaker specializing in architecture and urban planning. He currently lives in Berlin, Germany.

Want more Data Center Knowledge stories in your Google search results?

Editor's Choice

Industry Voices

Join the entire data center ecosystem September 21 – 23, 2026, at the Gaylord Texan Resort & Convention Center in Dallas, TX.

Power & SitesUnited States

The Energy & Sustainability Supplement

Credited to Data Center Dynamics · datacenterdynamics.com17h ago

Useful

As explosive AI demand threatens to break power grids worldwide, traditional data centers have hit a wall. Hemmed in by severe land shortages, water scarcity, crippling grid congestion, and mounting regulatory scrutiny, the industry is being forced to completely rethink how and where compute is built.

To keep pace without compromising environmental goals, operators are embracing radical infrastructure shifts. Our latest Energy & Sustainability Supplement explores the pioneering technologies and frameworks redefining sustainable compute:

Original · Data Center Dynamics

Front Method

Power & Sites · United States

Can the fossil gas supply handle the AI boom?

Credited to Latitude Media · latitudemedia.com · 17h ago

As data center-driven load growth continues to rise, hyperscalers are turning to fossil gas – both in front of and behind the meter — to secure power. But whether the gas pipeline system has the bandwidth to support this surging demand remains far from certain.

In this episode, Shayle sits down with Eric McGuire, director of gas and LNG trading analytics at Wood Mackenzie, to unpack whether gas is in the midst of a capacity crunch of its own. Shayle and Eric consider how data center demand has collided with massive LNG exports in the United States, why regional pricing spreads are informing data center siting, and how winter weather creates extreme daily price volatility. They also discuss:

Credits: Hosted by Shayle Kann. Produced and edited by Max Savage Levenson. Original music and engineering by Sean Marquand. Stephen Lacey is our executive editor.

Catalyst is brought to you by GridBeyond. GridBeyond is the world’s leading technology platform for managing distributed and flexible energy resources. Learn more by visiting GridBeyond.com.

Catalyst is brought to you by Antenna Group, the strategic communications and marketing partner behind the biggest names in energy, climate, and infrastructure. For three decades, Antenna has helped breakout companies amplify their stories, build reputation, differentiate from the competition, define new categories and accelerate growth. Learn more at antennagroup.com.

Catalyst is brought to you by EnergyHub. Peak season puts every grid to the test — and the utilities that pass are the ones that built flexible capacity before they needed it. EnergyHub works with more than 230 utilities to coordinate over 2.6 million DERs and more than 3.6 gigawatts of dispatchable flexibility through a single platform designed to perform when it counts most. See what that looks like at EnergyHub.com.

Shayle Kann: I’m Shayle Kann. I invest in early-stage companies at Energy Impact Partners. Welcome to Catalyst.

So we spend so much time here—and I spend so much time in my daily life—thinking about the constraints on the electricity grid that are being caused by all this load growth thanks to data centers, etc., etc. You’ve heard this a hundred times.

The direction of travel right now for many, many of these data centers, of course, is more natural gas generation. A lot of it in front of the meter being built by utilities or through utility tariffs, but increasingly also a bunch of it being built behind the meter to feed those data centers who are trying to connect directly into natural gas pipelines to deliver them the kind of reliability that they are used to or at least something approaching it.

And I realized that one thing that I haven’t really spent a lot of time thinking about is, we know the constraints on the power side and why it’s so hard to interconnect into the transmission system or even the distribution system in power, is that also coming in the world of natural gas, or is it easier there?

In natural gas in general, historically it has been easier to build new pipelines. It is also easier to ramp production. So there’s a number of reasons why natural gas shouldn’t experience quite the same constraints that power has.

But there’s also a confounding factor in natural gas, which is that there’s an even bigger relatively new demand driver, which is LNG in the United States, which is much, much bigger than demand is expected to be for data centers, at least for the next few years.

So that’s an additional lever on prices. It’s an additional constraint on supply. And so I wanted to see whether we should be talking about natural gas constraints coming up in the same way that we talk about electricity constraints.

For this one, I brought on Eric McGuire. He’s the Director of Natural Gas and LNG Trading Analytics at Wood Mac. He’s coming up next.

Eric McGuire: Thanks for having me, Shayle.

Shayle Kann: Thanks for doing this. I want to start with a basic question. I know exactly what it’s like if you are trying to build a new generator on the power system or a new load on the power system and you want to inject into or pull from the electric grid. I know how that works, the interconnection process. Let’s focus on the load side of this. If you’re trying to pull electricity, I know what that looks like. I actually don’t really know what it looks like if you want to do the equivalent thing in the natural gas system. If you were trying to tap into a natural gas pipeline, what does that process look like today? Like how is that market structured?

Eric McGuire: It depends on whether or not you already have the interconnect. And so the first thing, if you don’t have an interconnect, is you need to talk to a pipeline about building a header system to tie into their interstate pipeline. If it already exists and you can already tap into that, then there’s a whole process in terms of bidding on capacity, making sure that you do have the capacity you need to get the gas from the supply source to where you need it. And there are mechanisms for that for taking out long-term capacity on a pipeline.

Shayle Kann: Does it look similar to electricity interconnection in the sense that like, okay, if I want to interconnect to the grid, I need to go file an interconnection request with the utility in that case, they’re going to run it through this queue with these study processes to determine whether there is capacity on the grid, and eventually they’ll tell me, one, whether there’s capacity, two, when they can deliver to me, and like, three, what upgrades I need to pay for in order to interconnect? Is it similar to that, or does it work differently?

Eric McGuire: It is similar to that, but it depends on what scale you’re talking about. And so if you just are tying into a header system, that’s a much smaller thing. But if you’re trying to pull a very large volume across an area that there isn’t the capacity to move that kind of volume, then you’re going to start dealing with much more regulatory and environmental scrutiny.

Shayle Kann: Let’s presume you’re a data center. So let’s say you are, you know, hundreds of megawatts and you are putting enough capacity behind the—enough natural gas capacity behind the meter to fully meet your load for some period of time. So I would presume that is enough that you’re on the big side and you’re pulling just from the header.

Eric McGuire: Yeah. I mean, so we’ve seen that already with the LNG market, right? The LNG market has grown exponentially over the last several years, and we’ve seen, even particularly just in Texas, there have been a large number of intrastate pipelines that have come online flowing from Western Texas into the Gulf from the Permian to bring that supply across to where it’s needed. And so with every single one of those projects—and we can talk some about how long it takes from brownfield to greenfield and what kind is possible—it can take anywhere from 8 months to 6 years depending on what you want to get done.

Shayle Kann: 8 months to 6 years. I mean, that range is actually short when I think of it in an electricity interconnection context.

Shayle Kann: Which I guess is the point, right? Like if it’s—that’s why people are doing it, because it’s faster.

Eric McGuire: I think it’s faster, it faces less regulatory scrutiny, you’re not pulling from the grid. And so even if you look at the moratorium in Texas right now on data centers, it’s the in-front-of-the-meter data centers that there’s a moratorium on, but the behind-the-meter data centers are still progressing.

Shayle Kann: This is going to be a location-specific answer, I’m sure, but at the high level, are we in a period of capacity constraints in the natural gas pipeline system? Is it right now the case that it is getting harder and harder to find new places to interconnect quickly, or is there still enough excess capacity that we’re not approaching that?

Eric McGuire: I think we’ve been in a point of constraint for some time. I mean, the natural gas market has almost doubled in the last 10 years. And you look at how much production is up, you look at how much LNG demand is up. But the question has been, how can you get enough supply to the LNG facilities? Now we’re facing the question of, how can you get the supply to the data centers as well? And so between both LNG and data centers, we’re expecting a huge amount of demand growth over the next several years, and the question is, how do we keep up with that? Now, there’s lots of pipeline expansions that have been proposed, but then we have to sort through those and determine what are realistic, how long to get to market, and at some point, do you have to start rethinking about, you know, or rethinking where you’re putting these data centers or where you’re putting this demand to remove some of those constraints?

Shayle Kann: Before we, I guess, move on from how the contracts are structured, is it generally—if I’m, okay, again, I’m a data center and I’m going to build behind-the-meter turbines or natural gas engines or whatever it is, and I’m going to tap into a header natural gas pipeline—is my contract generally structured as like, “Here is my expected annual capacity that I need and I’m guaranteed that amount”? Are there—do you get curtailed ever in weather events? How does that work?

Eric McGuire: So let’s take the big example. So if you need a very large amount of supply from a pipeline, then you’re probably looking at a large expansion to a pipeline or a new pipeline altogether. If you’re doing that, then the pipeline’s going to ask you to be a backbone shipper and to essentially commit to 10-plus years of capacity that you need. If you do that, then there’s a few different ways you structure it, but essentially you’ll pay a reservation fee that is a use-it-or-lose-it. You know, you pay that fee whether or not you flow the gas, the pipeline knows that it’s going to get its revenue.

Then I guess that’s the first question, and so how do you structure that contract? If it’s an already established pipeline, then your contract might be a 3-year term instead of a 10-year term. But if you’re building a data center, I would want to secure that supply for as long as I can, and so I’d be looking at that longer 10-year term kind of contract.

Then there’s a question about priority. So when you take out capacity on a pipeline, usually when you have that use-it-or-lose-it contract, you’re paying that reservation fee, and that gives you the highest ranking to be able to flow on the pipeline. So there’s firm transportation, there’s interruptible transportation, there’s all these different ranks. Even within firm, you could be primary within path, meaning that the path that you paid the pipeline for, you’re flowing from point A to point B just like you said you would. If you decide to flow and take receipt at a different location, then that drops your priority a bit, and then lowers you in terms of the rankings for cuts.

But overall, so there’s this large kind of ranking structure that prioritizes how gas flows. And that’s pretty standard, and that’s not a big risk during normal flow conditions. However, you mentioned cold weather. When major events happen, let’s say there’s a force majeure event—right now there’s a major force majeure event in the Northeast going on—and the pipeline loses half of its capacity, well, in that case, if you’re flowing interruptible, you’re probably going to get 100% cut. And even the people with firm capacity are going to get cut because the pipeline isn’t running at max capacity.

And so that’s in a force majeure event. Then you have cold weather events. Now, cold weather events won’t necessarily impact pipeline capacity, but what happens in cold weather events that people forget about is we get freeze-offs. And so when it gets really cold, particularly in the Permian or in the South, you get condensate at the wellheads freeze off and you lose production. So during Winter Storm Fern this past winter, we had almost 18 Bcf a day of freeze-offs across the entire US. And think that you’re in a market where you only have 110 Bcf of production. So this is a huge amount of supply that got pulled off of the grid.

So we think about cold events, we generally think, what is the demand impact and is there going to be this demand-on-demand competition? But we also have to realize, when we’re thinking about security of supply, that during these cold weather events, the market loses a huge amount of supply, too. And so I think Winter Storm Fern in January, off the top of my head, I think it was a 350 Bcf event. 120 Bcf of that was supply loss, and then what’s that call it, 220 Bcf was demand gain.

Shayle Kann: I don’t know if you have these numbers off the top of your head, but I’m curious given those couple of reasons why you might lose supply if you’re on the load side, like on average, what is the uptime? What’s the reliability if you are a consumer? The reason I ask is because, again, in the data center context, right, the general expectation is three 9s or five 9s of reliability, and to get there historically, it’s like electricity plus backup power. The electricity grid, you know, has its outages too, but is generally pretty good—you don’t get that much of an outage. In a natural gas pipeline, do you get more or less than electricity?

Eric McGuire: I mean, in general, pipeline reliability is really, really good. Now, you’ll find different pipelines and different compressor stations are better than others. But my first job was bidding on pipeline capacity as a junior trader, and I would then assess that downtime pipeline by pipeline when I’d bid on it. And then when I bid on the capacity, I would then take that into account in assuming a certain number of days down.

The problem is when you’re dealing with something this size, when you’re talking about data centers and reliability, even if you’re only down for 3 days in the entire year, I mean, that could be a pretty big impact, and you then have to make sure either you have to think about that in terms of the compute that you’re then leasing out or that you’re selling, or you need to think about alternative energy sources.

Shayle Kann: Yeah, that’s where I think, you know, I’ve heard talk of a bunch of the new data centers that are doing behind-the-meter bridge power with natural gas, not also installing backup generators, like diesel generators, and they must be willing to withstand—I mean, you know, 3 days a year would be roughly 99% with no additional 9s, which is definitely not the historical norm for how data centers look.

You mentioned LNG a little while ago. I’m curious the extent to which we are starting to see real competition for supply between LNG and large loads.

Eric McGuire: I mean, I think that’s definitely what’s happening. I think LNG, right, is—we’re exporting I think somewhere around 19 Bcf a day right now, which is absolutely enormous. Again, in the context of the natural gas market, I’m going to roughly round it down to 110 Bcf a day, you have—let’s round LNG up—20 Bcf a day of LNG demand, right? That’s absolutely enormous.

We’re looking at some of the analysis we’ve done recently of up to 5 Bcf a day of behind-the-meter gas demand coming up over the next 5 years. So it’s not nearly as big as LNG, but what you have to remember with both of these is these are baseload demand, right? They’re not going to swing a whole lot, at least right now based on prices, kind of they’re all in the money. They’re not going to swing a whole lot with temperatures.

So a lot of the demand in the natural gas markets historically has been tied to temperature. So when it’s cold in the winter, you get more heating demand. When it’s hot in the summer, you get aircon demand. You have a lot of seasonality, low demand in the shoulders. But essentially, both data centers and LNG are going to be competing for baseload demand, and then everyone else is that marginal demand on top of that.

Shayle Kann: And are we seeing that affect prices on a regional level?

Eric McGuire: So I mean, we’ve definitely seen it in terms of Henry Hub on LNG. So if we think about the structure of the market, the majority of US natural gas production used to be in the Gulf. And what would happen is everything would flow into this tiny little location in Southeast Louisiana called Henry Hub, and then it would flow out from there. And so the way the natural gas market trades is, when we think about natural gas prices, we actually use Henry Hub as the kind of national balancing point for the US.

And historically, what happened was Henry Hub was one price, and then all the regional prices traded as a premium to Henry Hub. So everything else was a positive spread. If you’re looking at spreads now, Henry Hub, compared to a lot of these locations, Henry Hub is the premium price, and a lot of these locations are trading at a discount to Henry Hub because the production’s moved out of the Gulf. It’s now in West Texas, it’s in Colorado, it’s in the Marcellus region in the Northeast, and that gas is actually flowing back down to Henry Hub.

So with the growth of LNG, we’ve definitely seen that. We’ve now seen that Henry Hub is a premium location. Now, there are other locations, you get extreme market locations like the Northeast, New England, California, those are still premiums to Henry Hub, but all the other regions are now—we’re seeing a shift to being discounts to Henry Hub.

Shayle Kann: Interesting. Let’s talk about building new pipelines. I mean, again, the reason why all this behind-the-meter stuff is happening at data centers is in part because it’s hard to interconnect to the electricity system increasingly, but also because it’s really, really hard to build new transmission lines. Historically, it’s been a lot easier to build new natural gas pipelines. Has that changed? Is it—and where is it easy, where is it not easy?

Eric McGuire: I mean, it depends on where you look. And so in our analysis recently, which I just mentioned, you know, we looked at brownfield pipelines, and then greenfield kind of developments, and in the brownfield areas where you’re building on kind of capacity that’s already there, it’s 8 months to get it up and running. In some areas, particularly if you look at Mountain Valley Pipeline in the Northeast, that took a very long time because you’re dealing with environmental regulations, you’re crossing a lot more states when you’re out there. When you’re dealing just with Texas, it’s a lot easier.

And so I would say, in general in the South, we see that it is easier to develop new pipelines than it is in the Northeast. But then you also have the incentive of, right now, Permian in West Texas is one of the fastest growing producing regions in the entire United States, and we’re actually seeing a bit of a slowdown in Marcellus. And so when you think about areas that you want to build in, Permian becomes really attractive because of abundant supply, because of how quickly some of these new pipelines have come online even this year, and then you also remember that at the beginning of this year, Permian gas prices were actually negative on a fixed price.

So what happened was oil prices got so high, and a lot of the production, it’s oil production out there, that people are making enough on their oil production that they were actually paying people to take away their natural gas production because the natural gas was constrained. Now that’s flipped around, now these export pipelines have come online and it’s getting out to the Gulf, but between pricing and ability to build pipeline interconnects, Texas is attractive.

Shayle Kann: Right, and so back to the data center point, we’ve got a gigantic pipeline of data centers in West Texas which are in various stages of development. It sounds like, from a natural gas supply perspective for those data centers, that’s probably the right place to put it, actually. It might be the easiest place to get new capacity for some period of time.

Eric McGuire: Yeah, and then your question about building interconnects and then building new pipelines, it kind of comes back to that question of, getting new pipeline built is not necessarily easy depending on how long the pipeline is, how many states you have to cross, how many different populated regions you have to go through. Now, if you go out to a region that’s not very highly—that’s not highly populated, that you don’t have to build a long pipeline to connect to your supply, suddenly things get a whole lot easier. All of your regulatory approvals, your environmental approvals, there’s just a lot less hoops you have to jump through if you’re out in some of these producing regions.

Shayle Kann: Back to the bigger picture of, is the rise of behind-the-meter, or actually, honestly, the rise of natural gas generation in general as a result of load growth, not all of it is going to be behind the meter, like, is it going to meaningfully affect natural gas supply? Is natural gas pipeline capacity going to meaningfully be a constraint? I mean, you know, if you said we’re doing 120 Bcf and you think maybe there’s 5 Bcf coming from behind the meter, that sounds meaningful, but not market-shaping. Is that about right?

Eric McGuire: I mean, it’s meaningful, but you have to look at it in context also of the growing LNG. We still have a lot of—we don’t have a lot of new LNG export capacity coming on next year, but it’s going to continue to grow over the next decade. We’re going to see this data center demand growth ramp up, and so it is going to be meaningful. It already has been meaningful, and it will be meaningful.

And we’ve already seen pipeline constraints. I mentioned that Permian was constrained earlier, and so gas was trading at a negative fixed price, right? Producers are paying people to take their gas.

Right now in the Northeast, we’re seeing Northeast gas prices trade under a dollar, and we’re seeing what looks like production shut-ins because there is not enough capacity to get that gas out in the short term as demand has dropped. Right now, people are shutting in gas because it’s oversupplied, so you already are seeing supply constraints all over the US.

Shayle Kann: Yeah, it’s interesting, that’s so different from the electricity kind of—I mean, I guess it’s not, right? Like we see curtailment in electricity when you have oversupply and insufficient demand or insufficient interconnection capacity, I guess. But is the expectation that you’ll see, like, because of the rise of behind the meter, in combination with a number of other factors, data centers are moving closer to where the energy is. So would you expect then to see a big boom in data center development in the Permian, for example? Like if you’re closer to the supply, does that make the pipeline constraint easier?

Eric McGuire: It does. I mean, 100% it makes it easier, and again, it puts you closer to the supply, it helps you avoid a lot of the constraints, it removes the demand-on-demand competition. I mean, not that you would, but if you put a data center on the Gulf, then you’re competing on those long-haul pipelines for the same capacity that LNG facilities are. If you’re sitting there right in the middle of the Permian, you’re not doing that.

And so I think that you will see data centers—and we’ve already kind of, that’s what we’re doing in the risk-assessing of these new proposed data centers. We have almost—I think over 900 gigawatts of proposed data centers, but when we risk-assess them, the number that we think are going to come to market are significantly less than that, but the ones that we think are the highly likely ones are the ones that are in the basins.

Shayle Kann: Yeah, that’s interesting. What is your outlook on pricing for natural gas given all of this? Clearly we have a rising demand environment through a combination of LNG and this load growth, this electricity load growth. You tell me what will happen is happening with production, but is the expectation we’ll see rising prices?

Eric McGuire: I mean, what we’ve seen so far, so we’ve seen, again, the natural gas market’s doubled in the last decade, right? We went from like 65 Bcf a day of production to right now we’re sitting—we hit a high I think of about 112 recently, 113 Bcf a day of production. So it’s increased significantly. But what we’ve seen in that time as demand has increased is that supply has rose to meet it.

And so, what’s been really interesting, the shift in the market is that it’s grown exponentially, and so supply has kept up, and we used to as a market be really scared of running out of gas. We’re actually not really scared of running out of gas anymore. And so, for traders out there, they generally look at their March-April spread, and the March-April spread represents the fear that you’re going to run out of gas at the end of winter. And so that used to be—it’s called the widowmaker—and it used to blow out and you get really wide spreads. The March-April spread I think is 10 cents right now for this coming for March-April ’27. So the market’s sending the signal they’re not scared of running out of gas.

But you get a lot of fear of running out of gas in January and February. And so the new spread is that January-to-March or the February-to-March spread, because what’s happened is the market has grown, storage capacity hasn’t grown.

And so I think the big thing that we’re going to see is, especially if a lot of this demand is located in producing regions, that you’ll see production rise to meet it, because with breakevens where they are, there’s plenty of incentive to continue to drill and to meet the demand that’s growing there. But I think you’re going to continue to see an issue in terms of deliverability in those peak months. And so who’s going to get the gas when it’s really cold? What happens when production drops off? And you have data centers, and you have LNG, and then you have everyone else calling on gas for heating, what happens in those events?

So I think, just me speaking personally, we probably see some support for pricing going forward. Right now the international arb between European gas prices and US prices is really wide, and so as you get new LNG export capacity coming online, you would expect that spread to narrow, and so we do expect a lot of new LNG to come online. But historically, what we’ve seen is as gas prices go up, producers step in and happily take that premium to produce more gas.

Shayle Kann: That’s interesting on storage capacity. So like, okay, again assume that you’re going to get a bunch of new demand in producing regions, that implies there will need to be some new pipeline capacity to serve those loads, but it’s not big long pipelines in terms of miles. However, you would need a lot of new storage capacity in those regions as well. Is it hard to build new storage?

Eric McGuire: It is something we have not seen a lot of. I mean, there’s multiple kinds of storage you have, but a lot of times it’s depleted reservoir storage, so it’s empty fields that are being used then and pumped back full of gas and then pulled back out. And so what you need, you’d need those spreads to justify the investment in it. And so right now, relatively low prices, I think you do have—you’re getting those wider storage spreads, so I think we’re seeing new projects come online and be proposed, but then the question is in that high winter, if the data center wants reliability and the local cities, citygates, utilities need that gas in max heating season, you’re going to have to pull on something.

And so right now, what the market would tell you based on pricing is that there will be peak days in the winter that we don’t have enough supply to meet peak demand already. We’re not scared of running out of gas by the end of winter, but there are going to be peak days where gas prices go crazy.

So let’s look at this past winter. When Winter Storm Fern hit, Henry Hub had a high print, I think Henry Hub had a high print close to $50 an MMBtu, 5-0, and it averaged that day $30 an MMBtu. And Permian I think got to a high of $17 or $18 an MMBtu before then trading negative in the coming months.

And so you can see there’s going to be a whole lot of volatility on the day in those peak winter days when the peak demand hits. That doesn’t mean that we’re going to have—and so that’s kind of I think what we’re going to see more going forward. As this demand grows, I’m not necessarily super bullish outright price—I think it’ll increase some, but I don’t think we’re going to go back to the $10 natural gas days when I started trading years ago—but I think that what you will see is you’ll see a lot more volatility in the market. So I think the story’s volatility, it’s not high prices, it’s how do you manage the volatility on the day.

Shayle Kann: Do LNG buyers who are going to produce LNG, do they ramp up and down their purchases based on price? Like on those days, do you see demand pull back? Because I know data centers won’t, right? They’re price-insensitive here.

Eric McGuire: So on the LNG side, I think we’ve only seen LNG shut-ins, you know, so what you’re saying, are LNG shippers going to turn down their supply in order on those high-price days? Well, first off, let’s talk about an LNG contract. These are long-term contracts, right? They’re doing 10-year, 15-year SPAs, these aren’t day-to-day contracts, and they’re going to meet demand elsewhere. A lot of these contracts also, the LNG market, it’s not quite as commoditized as what we see in natural gas and oil, even though LNG is essentially just transport for natural gas. These contracts are bilateral contracts, and so if I’m going to sit here and strand a contract, one, I’m stranding back gas that has to be delivered someplace else in another country 15 to 30 days from now, and then I need some way—I need to pick up the phone, call that person, it’s not like I can do with oil and gas on the screen and just click a button and unwind it. So it’s a bit harder to unwind those contracts.

I think we’ve really only seen LNG shut-ins once, and that was back in COVID, and essentially what happened was the entire world was oversupplied. We saw the Asian market, JKM, go flat to TTF, and then we saw TTF go flat to Henry Hub. But LNG didn’t start getting shut in when your transport economics were closed, and so, it costs a certain amount to ship it, and so when TTF got within a certain dollar range of Henry Hub, we didn’t see any shut-ins. It wasn’t until TTF went flat. And once that European price went flat to Henry Hub, then we saw LNG shut-ins.

I think it’s the only time when we’ve seen LNG shippers really economically turn off gas, but it was very fundamentally driven. We do see LNG exports decrease when we see these large cold snaps. What’s really interesting, if you look at the profile of the LNG exports and you look at the profile of the freeze-offs that are happening, they’re right on top of each other, they have a very high correlation. So my view is that what we’re seeing in the short-term events, one, I think they happen so quickly it’s difficult to take advantage of economically as a shipper, but what I think is happening is we’re actually seeing security-of-supply losses where that production disappears for a few days, and that impacts what’s being exported over those same days.

Shayle Kann: All right, so I guess to wrap it up, stepping back for a second, let me tell you my takeaway here and you can tell me if this seems right to you.

In the electricity world right now, the rise of data centers is like the biggest thing that’s ever happened. We haven’t seen this kind of load growth over this period of time since the 1950s, everybody’s scrambling and trying to figure out what to do with it, it’s a huge deal.

In the natural gas world, it seems meaningful, 4% or 5%-ish for behind the meter alone over the next 5 years, right, 5 out of 120 roughly. That seems meaningful, but actually two things seem different. One is, you have an even bigger factor, which is LNG, which is a bigger demand driver today and has been a bigger demand driver, and it sounds like you think will continue to be a bigger demand driver.

And two, it seems the natural gas structural market is better able to ramp. You can probably build new pipelines easier, you can ramp up supply faster, production faster. And so my read is that what you’re telling me is that it’s a meaningful factor in the context of the natural gas market, but it’s not quite as earth-shattering as it is in the electricity sector. Does that resonate?

Eric McGuire: I don’t know, I mean, I think maybe I’ve done data centers a disservice in the ramp-up, because pure numbers, LNG’s bigger, and so in that context, LNG’s bigger. But what we’re talking about is this marginal demand unit, and anyone who’s traded before knows that the market prices off the marginal demand, it doesn’t price off the baseload, right?

And so I think that this is going to be a big issue, and so I mentioned we’re already seeing constraints throughout the US market. In fact, I don’t think the US market really trades as one market anymore, it trades as lots of regional markets in the gas market, so more similarly to how power trades. I think we’re going to see more and more of a shift that way, and so I think we’re going to see more disconnects in regional pricing like Permian was trading negative, like right now you’re seeing shut-ins in the Northeast.

Now, I think that the data center demand where it hits is going to be revolutionary for those regions, right? I mean, you suddenly drop a huge amount of demand in the middle of the Permian with data centers, then all the LNG facilities are going to be saying, “Well, okay, now where do we get our supply?” Same with the Northeast, Northeast you drop a whole bunch of data centers right there in kind of Western Pennsylvania, and then you’re going to have an outlet for supply there and you’re going to have demand competition, and a lot of that gas is flowing to Boston or New York in the winter, right? So what happens there?

And so I think it will have a huge impact on the markets because what we’re talking about is the marginal demand molecule, and especially have a huge impact in the winter when suddenly all this baseload demand is competing with the grandmas and the grandpas and the kids and everyone else in all these populated cities who need the gas to heat their homes. And so I think it’ll be a big question of then, how does that balance happen in those peak demand periods?

Shayle Kann: All right, Eric, super informative, really appreciate your time. Thanks for joining.

Eric McGuire: Thanks, Shayle, appreciate it.

Shayle Kann: Eric McGuire is the Director of Natural Gas and LNG Trading Analytics at WoodMac.

This show is a production of Latitude Media. You can head over to latitudemedia.com for links to today’s topics.

This episode was produced by Max Savage Levenson. Mixing and theme song by Sean Marquand. Anne Bailey edits the video version of the show. Stephen Lacey is our Executive Editor.

All of our episodes are on YouTube. Subscribe to Latitude Media for episodes of this show and Open Circuit. You can also find the audio version of the show anywhere you get your podcasts.