AI Data Center Expansion: Texas Oil Field Landowners Rush to Seize New Opportunities
Taylor Wilson
Three landholding companies control roughly 1.4 million acres of West Texas desert and are pivoting from oil-and-gas leasing to AI data centers — the region's land, water, power, and near-zero community opposition make it an underappreciated front in the compute-infrastructure race.
Who is betting, and on what exactly?
Texas Pacific Land (TPL), LandBridge (LB), and EagleRock (EROK) together hold about 1.4 million acres — more than seven times the area of New York City — and are still expanding.
The core model: lease land to data-center operators while bundling water, sand, and limestone as add-on services. This means → they are not just selling plots; they are selling a turnkey site package.
In plain terms = these landlords used to feed off the oil patch; now they want a second meal from AI compute.
Why this particular stretch of desert?
EagleRock CEO Greg Pipkin Jr. notes that data-center projects nationwide struggle with water, power, and transmission — the Permian Basin has all three in abundance.
The area has roughly 500,000 residents; if it were a state, it would be the least populous in the U.S. — large-scale projects face almost no community pushback.
Oil-and-gas producers have long faced a natural-gas pipeline bottleneck and are eager to sell surplus gas to tech companies for power generation; the high-salinity wastewater from drilling can be treated and reused to cool data centers. This means → the oilfield's "byproducts" and "chronic headaches" become the data center's raw inputs.
Why are investors willing to pay premium valuations?
TPL is up roughly 14% this year at a forward P/E of about 37×; LandBridge is up about 56% at roughly 39×; EagleRock completed its IPO earlier this year, raising $320 million.
In plain terms = unlike drilling, which demands massive upfront capex, land leasing is asset-light — low spend, high margins, and investors naturally love that.
Bryan Loocke, energy partner at Vinson & Elkins, puts it bluntly: "There is too much money chasing this one big fish." This reflects how the AI-infrastructure narrative now pulls capital far harder than traditional oil and gas.
Which real-money deals have already landed?
Chevron and Microsoft signed a twenty-year power-purchase agreement to build a gas-fired plant in the Permian Basin powering a Microsoft data center; TPL sold the land to Chevron.
AI startup Poolside and cloud-infrastructure firm CoreWeave announced plans to jointly build a data-center complex on a large West Texas ranch.
TPL invested $50 million in Bolt, a data-and-energy infrastructure company co-founded by former Google CEO Eric Schmidt, to advance data-center development in West Texas. This means → the buyer lineup now spans Big Tech to AI startups — this is past the "just a story" stage.
Where is the risk — the gap between concept and scale?
About $386 million of TPL's revenue last year still came from legacy oil-and-gas services (sand, limestone, water sales, and brine-disposal royalties); data-center revenue is not yet a main driver.
P/E multiples of 37–39× already price in aggressive growth — any shortfall in scaled deployment would invite a meaningful pullback.
In plain terms = the narrative is compelling, but the projects on the ground are still isolated dots; whether they connect into a full map is the real test for these premium-valued stocks.
Content is for reference only, not financial advice.