Following extensive investments in wind and solar energy projects, major hyperscale companies such as Amazon, Google, Meta, and Microsoft are now increasingly relying on natural gas to fuel the data centers critical to their ambitious AI initiatives. However, a recent research report indicates that this pivot towards the fossil fuel might lead to future complications.
According to Noreva, an energy research firm, natural gas prices in certain U.S. regions could surge threefold in the coming years. This projection is based on the convergence of escalating demand from hyperscalers, a slowdown in supply growth, and an increase in liquefied natural gas exports. The report suggests that these hyperscale entities may be ill-prepared for potential future price volatility.
"I believe there's a prevailing complacency in the energy markets, a perception that gas prices are immune to significant increases," stated Peter Gardett, CEO of Noreva, in an interview with TechCrunch. He added, "Simple calculations reveal that we are heading towards a considerably tighter gas market compared to just a few years ago."
The availability of inexpensive natural gas has prompted hyperscalers to secure a significant portion of the market. In March, Meta announced plans for a substantial 7.5-gigawatt natural gas power plant in Louisiana to support its Hyperion data center. Shortly thereafter, Microsoft and Google each revealed intentions to construct their own gigawatt-scale gas power plants in Texas. Amazon is also participating, with plans for a 7.6-gigawatt gas power plant, also located in Texas.
These companies, traditionally averse to extensive capital expenditures, are now undertaking significant investments in physical infrastructure due to the data center construction boom. This also propels them deeper into the energy markets, an area with which they have considerably less experience.
Gardett noted that at least one investor he consulted expressed "surprise" regarding the extent of natural gas price risk hyperscalers are prepared to absorb. He commented, "They are engaging in activities uncharacteristic for a typical energy off-taker."
Noreva forecasts that natural gas prices could exceed $10 per million BTUs at specific hubs, which are critical delivery points for futures contracts. Current prices typically fluctuate between approximately $2 and $4.50 per million BTUs, with the benchmark Henry Hub in Louisiana trading at just below $3.
Given that fuel constitutes roughly half the cost of electricity from a large power plant, a twofold or threefold increase in natural gas prices would significantly elevate the operational expenses for "bring your own power" AI data centers. This could either lead to higher token costs for AI services or compel hyperscalers to integrate with existing grids, subsequently increasing overall electricity prices.
While natural gas prices currently appear stable in the near term, with futures contracts not predicting substantial shifts, Gardett acknowledged, "It’s not an unreasonable bet." However, he expressed skepticism about the accuracy of these projections.
Gardett explained that natural gas price stability has been maintained by years of consistent demand and a steady influx of new supplies, which has offset declining output from older wells. He anticipates that energy companies will continue to expand supplies, though at a slower pace than previously, compounded by the increasing cost of new well development.
"While those factors alone wouldn't fundamentally alter the economics, what is truly shifting the landscape is the increasing integration of the domestic gas market with the global market," he stated. "The second significant driver is the surging demand spurred by AI."
Hyperscalers have been drawn to Texas and Louisiana due to the lure of inexpensive natural gas. Especially in West Texas, where oil extraction has been the primary focus, natural gas has historically been a byproduct with limited market access. A scarcity of major pipelines to transport it out of the region led producers to offer discounted gas to local consumers. This dynamic, however, is now evolving.
"New pipelines have finally been constructed in that area, with a significant portion of the gas now directed towards export markets," Gardett observed.
With West Texas becoming increasingly integrated into national and international natural gas markets, local demand will exert influence on prices in other regions, and vice versa. Consequently, even minor price fluctuations in proximity to hyperscalers' large data centers could have amplified effects across broader markets.
"We will see situations where abundant gas supplies exist adjacent to areas with scarcity, leading to substantial price differentials," Gardett explained. He emphasized that these very differentials are what will push prices in certain regions above $10 per million BTUs for prolonged durations.
In such a scenario, even if hyperscalers are able to absorb increased costs, their substantial natural gas consumption could intensify public criticism regarding data centers. Already, 80% of consumers express concern about the impact of data centers on their utility bills, predominantly concerning electricity. This apprehension could extend to natural gas expenses.
In their urgent pursuit to power AI data centers, hyperscalers are rapidly becoming deeply embedded in the fossil fuel sector. This is an arena where they possess limited prior experience, yet it is poised to significantly affect their business operations in the near future.
"During upcoming Alphabet earnings calls, you may hear discussions about the correlation between natural gas pricing and Google's financial performance, which, while unusual, reflects our current reality," Gardett concluded.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.