Announced data center projects more than tripled in the first half of 2024, illustrating the immense rise in energy demand that is straining the national electricity grid amid the artificial intelligence race. 

A new report from data and analytics firm Wood Mackenzie highlighted the massive uptick in data center development. The report claimed the U.S. electric utility industry “has been caught flat-footed” by the unprecedented growth. 

“A large load like this has never really existed in history,” Chris Seiple, vice chairman of Wood Mackenzie’s power and renewables group, told the Financial Times. “It’s clear that there is a race going on across America to secure land and interconnection to energy to be able to build as much data center capacity as possible.”

Wood Mackenzie estimated that the number of new data centers announced in the first half of 2024 totaled a capacity of nearly 24 gigawatts, more than the entire amount announced in 2023, according to the Financial Times. Georgia, Texas, and Virginia are leading the charge of this growth, with Virginia still holding the title of data center hub of the United States.

These data centers, often focused on AI technology and developments, are filled with hundreds of thousands of computer servers that hold and process data used for sending emails, hosting Zoom meetings, liking social media posts, and sending queries in ChatGPT.

The facilities can vary in size, with small centers often only using between 1 and 5 megawatts of power. However, large data centers have been known to consume over 100 megawatts of power. The average size of data centers is also growing, Wood Mackenzie revealed. Since January 2023, average capacity has reportedly grown to around 8 megawatts of power. 

To put this in perspective, one megawatt is considered to be roughly the same amount of energy consumed by 400 to 900 homes in one year. 

At this rate, data centers are estimated to consume, on average, around 20-25 gigawatts of hourly energy demand, Wood Mackenzie said. This is roughly equivalent to 4-5% of all electricity consumption in the U.S. As tech companies such as Amazon, Meta, and NVIDIA all look to get ahead in the AI race, this is only expected to grow. 

“The adoption of AI software powered by large language models is expected to be one of the fastest adoption curves the market has ever seen, on a par with mobile phones. And the race is on to add more data-centre capacity and computational power fast,” the report said. “Tech company executives fret that if they don’t move fast enough, they will miss out on one of the greatest business opportunities of all time.”

Data centers aren’t the only thing responsible for straining the grid. In its Thursday report, Wood Mackenzie pointed to an imbalance of resources and demand, as coal continues to be phased out and renewables prioritized by the Biden administration have yet to overtake as dominating power sources.  

The firm explained that some grid operators, such as PJM and SERC, are expecting to see energy demand grow by 10% in the coming years. However, those regions are also expecting to see coal capacity decline by an amount equal to 10% of demand. 

“In reality, one likely result of increased demand growth will be the deferral of planned retirements,” the firm predicted. 

CLICK HERE TO READ MORE FROM THE WASHINGTON EXAMINER

As a result, Wood Mackenzie predicted that electricity prices would be under pressure in the U.S. as utilities and tech giants sought to secure power from alternative sources like nuclear. 

Thursday’s report comes the same week that Google and Amazon announced they are entering agreements to purchase nuclear energy from several small modular reactors. Other companies, such as Microsoft, have begun the process of restarting retired nuclear plants, such as Three Mile Island, to support their AI data center developments.