AI firms are constructing large natural gas plants for data center energy. Risks?
Image Credits:wenbin (opens in a new window) / Getty Images
The AI Bubble: A FOMO Phenomenon in the Tech Industry
Who doesn’t feel the thrill of FOMO (Fear of Missing Out)? Across various eras of technology, from the dot-com boom to Web 2.0, virtual reality, and blockchain, the tech industry has always been prone to trends that spark urgency and fear of exclusion. Today, the AI bubble stands out as the most significant manifestation of this phenomenon, prompting a surge in investments aimed at securing power for data centers.
The Natural Gas Rush: An Offspring of the AI Bubble
Recent months have seen tech giants scrambling to secure natural gas supplies, marking an alarming trend. Microsoft has partnered with Chevron and Engine No. 1 to construct a natural gas power plant in West Texas, which aims for a capacity of 5 gigawatts. In North Texas, Google is collaborating with Crusoe for a 933 MW facility. Meanwhile, Meta is expanding its Hyperion data center in Louisiana by adding seven more natural gas plants, pushing its total capacity to 7.46 GW—enough to power South Dakota entirely.
These investments are primarily concentrated in the southern United States, which is rich in natural gas deposits. According to the U.S. Geological Survey, one region alone could theoretically provide energy to the entire country for ten months. With every data center operator eager to stake a claim, it’s a race against time and resources.
The Economic Implications of Natural Gas Supply and Demand
However, this growing demand for natural gas is creating complications. A critical shortage of turbines necessary for power plants is anticipated, with prices expected to soar by 195% compared to the pre-2019 costs, according to Wood Mackenzie. Turbines alone can account for 20-30% of a power plant’s overall expenses. Adding to the challenge, new orders for turbines won’t be fulfilled until 2028, and deliveries are taking up to six years.
This situation indicates that tech companies are banking on the idea that the AI trend will remain strong, requiring ever-increasing amounts of energy. They believe that natural gas will be a vital component for success in this AI-driven future.
The Risks of FOMO in Energy Investments
While natural gas is currently abundant in the U.S., the production growth rate has decelerated significantly in the primary shale gas regions that contribute to three-quarters of national output. This raises the question: how insulated are tech companies from price fluctuations? Without clear contractual terms disclosed, it’s challenging to gauge their vulnerability to market volatility.
Even if they have fixed-price agreements, companies may still run into complications. Natural gas sources account for approximately 40% of the electricity generated in the U.S., as reported by the Energy Information Administration. This close relationship means that any rise in natural gas prices is likely to impact electricity costs, directly affecting tech companies’ operations.
Behind-the-Meter Controversies
In a temporary bid to evade scrutiny, tech firms may shift their natural gas power plants behind the meter, opting to connect them directly to their data centers instead of the grid. While this could obscure their resource usage from public view, it merely reallocates energy demands from one grid to another—the natural gas grid.
If companies expand their energy ambitions without adequate resource management, this may lead to higher power prices for everyone. Ordinary households, as well as industries heavily reliant on natural gas that lack renewable alternatives, could express discontent. After all, powering a data center with wind or solar energy may be manageable, but transitioning a petrochemical plant to renewables is not as straightforward.
Weather’s Unpredictable Impact
The instability of natural resources is further compounded by external factors, such as the weather. A particularly harsh winter could spike household energy demands, as seen in Texas in 2021 when wellheads froze, significantly curtailing gas supplies. In those moments, suppliers will face a choice: either sustain the operations of AI data centers or ensure heating for homes.
The Consequences of Resource Mismanagement
By aggressively pursuing natural gas and moving their operations behind the meter, tech companies can claim they’re “bringing their own power” without straining the electrical grid. However, the reality may be more complicated. This new paradigm has unveiled how constrained the digital infrastructure truly is, raising questions about the wisdom of betting on a finite resource.
The rush to capitalize on the potential of AI may ultimately lead tech companies to regret their decisions influenced by FOMO. As natural gas supplies become scarcer and prices fluctuate, the ambition to harness this resource could turn into a liability.
Conclusion: A Cautious Path Forward
The AI bubble may represent the latest chapter in the ongoing saga of tech industry trends spurred by FOMO. While the rush for natural gas power reflects immediate market demands, the long-term sustainability and ethical implications of these strategies warrant scrutiny. Companies must balance their thirst for power against the realities of dwindling resources and potential public backlash.
As they navigate this complex landscape, the tech industry must consider the sustainability of their energy choices. The future of AI may depend not only on technological advancements but also on how responsibly these companies manage their energy needs in a world with finite resources.
Thanks for reading. Please let us know your thoughts and ideas in the comment section down below.
Source link
#companies #building #huge #natural #gas #plants #power #data #centers #wrong
