Hyperscalers May Reevaluate Natural Gas Usage If New Forecasts Prove Accurate

After years of investment in renewable energy initiatives, leading tech firms like Amazon, Google, Meta, and Microsoft are increasingly relying on natural gas to fuel the data centers that underpin their ambitious AI projects. However, a recent research report raises concerns that this growing dependency on fossil fuels may result in adverse effects for these companies.

Noreva, an energy research agency, indicates that natural gas prices in certain U.S. regions might potentially triple in the coming years. This surge would be driven by escalating demand from hyperscalers, stagnant supply growth, and rising liquefied natural gas exports, which could expose hyperscalers to sudden price increases.

“I think many in the energy markets have been misled into believing that gas prices will not increase,” said Peter Gardett, CEO of Noreva, to TechCrunch. “A simple calculation reveals that the gas market is much tighter than it was a few years ago.”

Significant Investments by Hyperscalers

The availability of cheap gas has encouraged hyperscalers to capture large portions of the market. In March, Meta unveiled plans for a major 7.5-gigawatt natural gas plant in Louisiana to support its Hyperion data center. Shortly after, Microsoft and Google announced their intentions to develop gigawatt-scale gas facilities in Texas. Amazon also revealed plans for a 7.6-gigawatt gas plant, also in Texas.

These companies, typically cautious regarding significant capital investments, now find themselves compelled to invest substantially in physical infrastructure while navigating complex energy markets.

Gardett noted that at least one investor was taken aback by the level of natural gas price risk that hyperscalers are willing to accept. “They are engaging in behaviors that are atypical for an off-taker,” he remarked.

Noreva forecasts that natural gas prices could exceed $10 per million BTUs at specific delivery points, contrasting with current prices ranging from $2 to $4.50 per million BTUs, with the commonly quoted Henry Hub in Louisiana priced just under $3.

Given that fuel costs account for nearly half of the electricity expenses tied to a large power plant, a doubling or tripling of natural gas prices could significantly raise operational costs for “bring your own power” AI data centers. This situation might result in elevated token prices or cause hyperscalers to connect to the grid, further increasing electricity expenses.

In the immediate future, natural gas prices seem stable, as futures contracts do not indicate significant fluctuations. “It’s a reasonable bet,” Gardett remarked, although he harbors lingering doubts about its accuracy.

Surging Demand

Natural gas prices have remained stable primarily due to years of steady demand and a gradual increase in supply, counterbalancing declining production from aging wells, according to Gardett. He believes energy companies will manage to augment supplies, albeit not at previous rates. Moreover, drilling new wells is becoming costlier.

“This alone wouldn’t dramatically change the economics. The crucial element is that we’re at last connecting the domestic gas market with the global one,” he elaborated. “Additionally, there’s rising demand driven by AI.”

Hyperscalers have been drawn to Texas and Louisiana due to low natural gas prices. In West Texas, for instance, many wells focus on oil extraction, with natural gas as a byproduct that has limited market access. Previously, a lack of significant pipelines hindered the transportation of this gas, forcing producers to sell it at lower prices. However, improvements are finally occurring.

“New pipelines have recently been built, with a considerable portion dedicated to export markets,” Gardett clarified.

As West Texas becomes more integrated with both national and international natural gas markets, local demand will begin to impact prices elsewhere, and vice versa. Even minor price shifts near hyperscaler data centers could have widespread repercussions in other regions.

“There will be areas where gas is plentiful compared to places where it is scarce, leading to significant price disparities,” Gardett pointed out. Such differences could push prices in some regions above $10 per million BTUs for extended timeframes.

In this landscape, even if hyperscalers can manage rising prices, their consumption of natural gas may lead to increasing public opposition to data centers. Currently, 80% of consumers express concern about data centers’ effects on their utility bills, especially regarding electricity costs. This worry may extend to natural gas expenses as well.

By hastily powering their AI data centers, hyperscalers are becoming deeply intertwined with the fossil fuel industry—a sector in which they possess relatively limited expertise, yet which could significantly affect their operations.

“In forthcoming earnings calls for Alphabet, you may hear discussions about the relationship between natural gas prices and Google’s performance, which is unusual, but that’s the reality we are currently facing,” Gardett concluded.

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.