A surge in power use from artificial intelligence (AI) data centers is projected to substantially increase natural gas demand in the second half of the decade, according to a report by analysts of investment banking firm Tudor Pickering Holt & Co.
The report suggests that the spike in power consumption from AI-driven data centers could need an additional 8.5 billion cubic feet per day of natural gas to meet the heightened demand.
Rapid Rise of Generative AI
According to Reuters, concerns have been raised by US power and technology firms regarding the capacity of the country's electrical infrastructure to keep pace with the rapidly escalating energy requirements of technologies like Generative AI.
That has led some data center operators to bypass traditional utilities and instead forge direct agreements with power producers or develop their own energy sources.
The rise in overall energy demand has contributed to a growing backlog of nationwide power generation and energy storage projects seeking grid connections.
According to Lawrence Berkeley National Laboratory data, the queue for such projects expanded from 2,000 gigawatts in 2022 to 2,600 gigawatts in 2023.
READ NEXT : AI Companies Face Data Shortage After Using Up All Available Internet Resources for Model Training
Natural Gas Demand
The report predicts that natural gas prices could average $4 per million British thermal units in the second half of the decade. Many natural gas producers were forced to curtail their production after their prices reached a three-and-a-half-year low of $1.61 per mmbtu in February because of mild winter weather.
Currently, data centers consume approximately 11 gigawatts (GW) of power. However, according to the report's projections, this figure is anticipated to surge to 42 GW by 2030.
The report also estimates that, in its base case, around 2.7 billion cubic feet per day of additional natural gas will be required by 2030.
Join the Conversation