Hyperscalers Reevaluate Natural Gas Consumption Based on Updated Forecasts
As significant investments continue to flow into renewable energy, major tech corporations like Amazon, Google, Meta, and Microsoft are increasingly relying on natural gas as a key energy source for their data centers, which are essential for AI operations. However, a new study has raised concerns about the risks associated with this growing reliance on fossil fuels by these tech giants.
Noreva, a research firm focused on energy, warns that natural gas prices across various regions in the U.S. could potentially triple in the near future. This expected increase is largely driven by heightened demand from hyperscalers, stagnant supply growth, and an uptick in liquefied natural gas (LNG) exports, which may expose hyperscalers to unpredictable price shifts.
“Many in the energy sector have underestimated future gas prices,” stated Peter Gardett, CEO of Noreva, in an interview with TechCrunch. “Basic calculations show that the gas market is significantly tighter than it was just a few years ago.”
Major Investments by Hyperscalers
Current low natural gas prices are prompting hyperscalers to broaden their market reach. In March, Meta announced its intention to construct a 7.5-gigawatt natural gas facility in Louisiana to support its Hyperion data center. Soon after, Microsoft and Google revealed their plans to develop gigawatt-scale gas facilities in Texas. Additionally, Amazon also announced plans for a 7.6-gigawatt gas facility in Texas.
Known for their historically cautious spending, these companies now find themselves needing to invest significantly in physical infrastructure amid intricate energy market dynamics.
According to Gardett, at least one investor expressed surprise at the level of risk hyperscalers are willing to take regarding natural gas pricing. “Their behavior is quite unusual for an off-taker,” he remarked.
Noreva forecasts that natural gas prices may exceed $10 per million BTUs at specific delivery points, representing a stark contrast to current rates between $2 and $4.50 per million BTUs, with the renowned Henry Hub in Louisiana currently priced just below $3.
Given that fuel expenses account for nearly half of the electricity costs for large power plants, a doubling or tripling of natural gas prices could considerably raise operational expenses for AI data centers that generate their own power. This scenario could drive up token prices or necessitate a greater dependence on the power grid, ultimately increasing their electricity costs.
In the short term, natural gas prices appear stable, with futures contracts showing minimal volatility. “It’s a reasonable assumption,” Gardett noted, although he expressed concerns about long-term stability.
Rising Demand
The current stability in natural gas prices is primarily due to a prolonged period of steady demand matched with limited supply growth, offsetting declines from aging wells, according to Gardett’s analysis. He argues that while energy companies can boost supply, it is unlikely to return to historical levels, and drilling new wells has become increasingly costly.
“This alone wouldn’t drastically shift the economic landscape. The vital factor is that we are now linking the domestic gas market to the international market,” he stressed. “Additionally, there is a surge in demand driven by AI.”
Hyperscalers are concentrating their efforts on Texas and Louisiana, where natural gas prices are more advantageous. Many gas wells in West Texas primarily focus on oil extraction, with natural gas being an associated byproduct that suffers from limited market demand. Historically, insufficient pipeline infrastructure has restricted the transportation of this gas, compelling producers to sell at lower prices. However, substantial infrastructure improvements are currently in progress.
“Recent pipeline developments are largely oriented toward export markets,” clarified Gardett.
As West Texas strengthens its ties to both regional and global natural gas markets, local demand will increasingly shape prices in different areas and vice versa. Minor price fluctuations near hyperscaler data centers could substantially influence costs in neighboring regions.
“Some areas will observe gas surpluses, while others will face shortages, leading to considerable price variances,” Gardett reiterated. Such fluctuations could result in prices in certain regions soaring above $10 per million BTUs for prolonged periods.
In this evolving landscape, even if hyperscalers adapt to higher costs, their increased natural gas consumption may raise public concern over data centers. Currently, 80% of consumers worry about the effects of data centers on their utility bills, particularly in relation to electricity costs. This concern is likely to extend to natural gas pricing as well.
By swiftly electrifying their AI data centers, hyperscalers are becoming more entwined with the fossil fuel industry—an area where they have relatively limited expertise, yet one that could significantly impact their operational dynamics.
“In upcoming earnings reports from Alphabet, we might see discussions about the connection between natural gas prices and Google’s performance, which would be unusual yet indicative of the circumstances we face,” Gardett concluded.
Please be aware that clicking on links within our articles may generate a small commission for us, without affecting our editorial independence.
