What LNG and AI mean for drilling in 2030

FORD BRETT, CONTRIBUTING EDITOR

Prior columns discussed the drilling advances needed to make hot dry rock geothermal a major part of the world’s energy picture. The elevator pitch: IF (note the big ‘if’) we could construct hot dry rock geothermal wells as efficiently as we build shale wells, the world would go crazy for geothermal drilling.  Drilling advances would need to happen, but it appears to be physically possible—even if it still feels a bit fantastic. 

Refer to my last columns if you’re interested in that story. I try to make each of these helpful as stand-alone missives, but there is a red thread running through them all, and prior articles might be useful for new readers.  

Fig. 1. U.S. crude oil production, MMbpd.

This column will start a series that will explore a less fantastic situation: What will the growth in U.S. LNG export capacity and electricity demand for AI mean for drilling?   

This column will kind of set up the situation that the U.S. natural gas business will soon face, and future columns will explore impact outside the U.S., and what that could mean for drilling people, process and technology. 

Unconventional oil makes U.S. natural gas a “trash” hauling business.  

The effect of unconventional oil production on world prices is pretty well known.  Figure 1 shows how unconventional oil production has grown in the U.S.  That growth accounts for over half of the world’s increased production since 2015.  This, of course, has had the effect of keeping the oil price lower than it would have otherwise been.   

Fig. 2. European and U.S. natural gas prices since Covid.

A bit less well-known is the effect that increased oil production has had on the price of natural gas. Each unconventional oil barrel comes with associated natural gas (~4 Mcf/bbl in the Permian, and ~2 Mcf/bbl in the Bakken; at an average of 2Mcf/bbl, unconventional oil accounts for ~30Bcfd or ~30% of U.S. gas production). All this associated gas, along with the imperative to limit flaring and limited capacity to move treated gas to market, has caused the price of gas at the wellhead in unconventional resource plays to remain close to zero for years.  Sometimes, it has actually gone negative. For example, the Permian’s Waha hub spot natural gas price was negative for over a month in 2024 and has been negative for 57 trading days in 2025.  Something with a negative price is the definition of “trash.”        

The Permian’s “trash problem” has been a boon for U.S. gas consumers, and we should give thanks to the folks exporting LNG to gas consumers in Europe and Asia.  Figure 2 shows European and U.S. natural gas prices since Covid.  All the “trash” gas from shale plays has kept the price of U.S. consumers’ natural gas quite low, except for the bump caused by the situation in Ukraine, which made the price of natural gas in Europe volatile. U.S. (and Qatari and other) drillers were able to ramp up production quickly, quickly and  replace much of the gas Europe had received from Russia. 

Fig. 3. North American operational and sanctioned LNG export capacity.

This low gas price has kept drilling for unassociated gas lower than it would otherwise be.  Since most Russian gas went offline, Europe’s gas price has been about five times higher than that in the U.S.   

How LNG and AI will turn “trash” into “gold.”  

The ~5x higher European gas price (as a surrogate world price) has caused U.S. LNG exporters to get very busy.  Sanctioned projects in the works will almost double U.S. LNG export capacity by 2030, Fig. 3.  The bottom line for drilling is there will need to be an additional 17.5 Bcfd to feed additional LNG demand.  

LNG exports are only part of the demand. U.S. and global data center demand for electric power is set to quadruple over the next decade, Fig. 4. Per this Bloomberg projection, the U.S. will need an additional 300 TWhr by 2030. This power will need to be dispatchable—which means 24/7. Wind and solar, no matter how cheap, won’t fit the bill without expensive, nonexistent storage. Nuclear or coal could solve the problem, but coal has 2.25 x the CO₂ emissions of combined cycle gas, and nuclear takes years and years to get going.  It seems to me that most of this 300 TWhr will be provided by combined cycle gas. That would mean another 6.5 Bcfd by 2030 and yet another 5 Bcfd again by 2035.   

Fig. 4. Global data center power demand outlook by market.

LNG and AI data centers will demand ~25 Bcfd more of U.S. production by 2030.  Figure 5 shows what drillers needed to get done in 2024, just to keep production flat at about 107 Bcfd, it took 123 rigs.   By 2030, to meet AI and LNG export demand, the curve will need to grow faster than it has ever grown and extend off the frame up to 140 Bcfd or so. This all begs the question: is the gas even there to be had? 

The molecules are there.  Are the rigs?  Are the crews?  

All this demand means that the world would like the U.S. to produce more natural gas. The first question is whether there is even enough gas in the U.S. to meet that demand.  It seems the short answer is “yes.” Figure 6 shows a Texas Bureau of Economic Geology ‘TORA’ study, showing the quantities of unconventional gas and oil resource by region.  The EIA estimate of 86 years is based on "technically recoverable resources," which includes both proved reserves and unproved resources that are estimated to be recoverable with current technology.  The EIA estimates proved reserves at about 46 years.  The molecules are there.  

Fig. 5. U.S. Lower 48 production by well vintage (2010-2024).

(Note: The Tight Oil Resource Assessment (TORA) program, the premier research group on U.S. unconventional plays, provides basin-wide and granular analysis of resource potential.  

They have good stuff at Tight Oil Resource Assessment | Bureau of Economic Geology — perhaps you should check them out). 

The next question: can it be produced and still make money?  That depends on what it costs to get it—drilling and completion—and what it costs to move it to market.   

Short side note on what it costs to move gas to market. Kindra Snow-McGregor, gas processing expert, summarizes LNG economics as roughly $2.25 to 3.50/MMbtu to liquify; depending on vintage of facility; $1 to $3 to transport by LNG carrier, depending on distance; and about $0.15 to 0.30 to regasify. That means it costs $3.40 to $6.80/MMbtu to get gas from the fence of the LNG facility to the overseas distribution pipeline.  So, for example, if the price of gas is $10/MMbtu at the point of sale, the producer could get anywhere from $6.60 to $3.20/MMbtu for their gas.  

Fig. 6. U.S. unconventional resources and production.

We’re going to have to leave what drilling and our colleagues in production operations, completions, geology, and land and finance, etc., must do to ensure that this is possible. But, it seems to me that the likelihood of yet another unconventional gas boom is less fantastic than hot dry rock geothermal.   

What does this mean for drilling?  

Demand for more LNG exports and 24/7 electrical power for data centers means that U.S. natural gas production could need to grow by ~25% by 2030.  That’s a FASTER growth rate than ever before, with more absolute additions than in any previous five-year time frame.  With unconventional oil production likely plateauing, there won’t be free “trash” gas to haul off anymore. The next few editions of Drilling Advances will explore what this gas demand will mean for drilling activity in gas provinces and what it could mean for drilling People-Process-Technology advances we may need to make it all happen. 

Until next time, I hope to continue our conversation on how we can all help drilling advance.  If you have any questions, ideas, comments or corrections, please email me at ford.brett@petroskills.com and I promise I’ll respond.