Electricity demand for AI data centers is soaring. The result won't be pretty.
Chart Notes: Industrial prices for 1980-2010 only for years divisible by 5. The rest are extrapolated. I estimated the consumer cost for 2025 from the BLS consumer electricity price index.
TRUTH LIVES on at https://sgtreport.tv/
Big Tech's A.I. Data Centers Are Driving Up Electricity Bills for Everyone
The New York Times reports Big Tech's A.I. Data Centers Are Driving Up Electricity Bills for Everyone That is a free link.
The tech industry's all-out artificial intelligence push is fueling soaring demand for electricity to run data centers that dot the landscape in Virginia, Ohio and other states. Large, rectangular buildings packed with servers consumed more than 4 percent of the nation's electricity in 2023, and government analysts estimate that will increase to as much as 12 percent in just three years. That's partly because computers training and running A.I. systems consume far more energy than machines that stream Netflix or TikTok.
Electricity is essential to their success. Andy Jassy, Amazon's chief executive, recently told investors that the company could have had higher sales if it had more data centers. "The single biggest constraint," he said, "is power."
Even as some corporate customers have been underwhelmed by A.I.'s usefulness so far, tech companies plan to invest hundreds of billions of dollars on it.
At the same time, the boom threatens to drive up power bills for residents and small businesses. Nationally, the average electricity rate for residents has risen more than 30 percent since 2020, after years of relatively modest increases. Much of that increase has been driven by utilities' catching up on deferred maintenance and hardening grids for extreme weather.
It is difficult to predict what that will mean for consumers' power bills. But recent reports expect data centers will require expensive upgrades to the electric grid, a cost that will be shared with residents and smaller businesses through higher rates unless state regulators and lawmakers force tech companies to cover those expenses.
A June analysis, from Carnegie Mellon University and North Carolina State University, found that electricity bills are on track to rise an average of 8 percent nationwide by 2030 and as much as 25 percent in places like Virginia because of data centers.
In some places, it is happening already. Starting in June, the electricity bill for a typical household in Ohio increased at least $15 a month because of data centers, according to data from a major local utility and an independent monitor of the electric grid that stretches across 13 states and the District of Columbia.
That has led to growing tensions.
The utilities pay for grid projects over decades, typically by raising prices for everyone connected to the grid. But suddenly, technology companies want to build so many data centers that utilities are being asked to spend a lot more money a lot faster. Lawmakers, regulators and consumer groups fear that households and smaller companies could be stuck footing these mounting bills.
Over the past two years, the 5-year load growth forecast has increased by almost a factor of five, from 23 GW to 128 GW, including Grid Strategies' estimate of recent update reports.
Sustaining growth in strategic economic sectors will require significant new investments in electric power sector infrastructure. Meeting the electricity requirements of new advanced manufacturing and data centers through new nearby generation could result in capital investments of billions of dollars per GW of new load. A more valuable and less costly approach to ensuring reliability is ensure the same reliability with increased transmission capacity to transfer power from one region to another. Capacity delivered through interregional transmission is likely to cost less than $300 million per GW of new load. With this added reliability, less new generation is required and it doesn't need to be located close to new facilities.
Is There a Load Growth Bubble?
A recent Wall Street Journal article asked, "Internet Hype in the '90s Stoked a PowerGeneration Bubble. Could It Happen Again With AI?" Overbuilding in the 1990s contributed to bankruptcy by major independent power generators, such as Calpine and NRG Energy. Yet the same article also notes that, "the drivers of electricity-demand growth are more tangible this time." Companies developing data centers are making both verbal and contract commitments to pay their "fair share" of generation and transmission expansion costs.
Wood Mackenzie sampled several utilities' announced data center demand, estimating 93 GW total through 2030. Based on bottom-up analysis of tech industry capacity, Wood Mackenzie forecasts a more realistic national five-year AI buildout to be just 23 GW. Data center developers may have such a large appetite for growth that their projects could use up any and all currently-unused grid capacity over the next five to ten years.
On the other hand, business revenues to cover the costs of the artificial intelligence investments have not yet been proven. This combination of exuberance and uncertainty raises the question of whether these projects could fail to sustain anticipated power demand. Such a failure could leave other customers with responsibility to cover costs for transmission and, in some regions, generation investments. There are economic and energy reliability risks, both from under- and over-investment.
Six Regions Drive Load Growth Through 2029
There is much more in the report that merits a closer look.
Electricity Percentage Costs for Steel, Copper, Aluminum
Alcoa says producers would need firm contractual energy-price commitments of about $30 per megawatt hour for at least 15 years to be globally competitive. (WSJ)
My lead chart shows an industrial average at or over $80.00 for the last four years.
Trump vs Reality
Canada is a global leader in aluminum production due to its abundance of renewable hydroelectric power.