Data Center Power Demand Drives US Grid Cost Debates
US data center operators are using peak shaving and special contracts to manage electricity costs, sparking regulatory debates over whether residential customers are subsidizing AI infrastructure.
The rapid expansion of AI data centers is creating a volatile economic impact on the US power grid, leading to a conflict between industry operators and residential consumers. While some data centers may lower bills for others by covering fixed maintenance costs, the PJM Interconnection reported that data center growth has already increased system costs by $29.4 billion. In July 2024, PJM saw capacity auction prices jump from $28.92 to $269.92 per megawatt per day due to rising demand and plant retirements.
To manage these costs, operators are employing peak shaving, a practice of sharply cutting grid usage during high-demand hours. The Data Center Coalition argues this reduces the need for new infrastructure and lowers costs for all customers. However, critics and residential users contend this shifts the financial burden of grid maintenance onto households. In Virginia, regulators testified that residential customers paid the largest share for a transmission project serving a data center.
Regulators and utilities are responding with new mandates to ensure equitable cost distribution. Dominion Energy established a rate class requiring new data centers to sign 14-year contracts with minimum payments. In Texas, Google challenged transmission charges that could reach $1.4 billion over 20 years. Conversely, the Michigan Public Service Commission projected that a special contract with DTE Energy could save other customers $300 million annually by spreading shared system costs.