The results of the large-scale power grid auction in the United States are released. The surge in energy consumption of AI data centers has pushed up the electricity budgets of 13 states
On July 14, the power capacity auction results released by PJM, the largest power grid operator in the United States, showed that due to the surge in power demand from artificial intelligence (AI) data centers, tens of millions of households and businesses in 13 states in the eastern United States and Washington, DC are expected to face a total of $6.3 billion (approximately 45.6 billion yuan) in additional electricity bills in the next three years. This data highlights the deep-seated contradiction between the expansion of AI computing infrastructure and the insufficient carrying capacity of traditional power grids in the United States. The imbalance between supply and demand intensifies: the siphon effect of data center energy consumption appears PJM Power Grid covers a vast area from Virginia Beach to Chicago, serving a population of 67 million, including the world's largest data center cluster - the Northern Virginia Data Center Cluster. In this annual power capacity auction, power generation companies have significantly increased their bids to transmit power to the grid during peak power consumption periods. Due to the disorderly expansion of high-energy-consuming terminals such as data centers, the growth rate of electricity demand in the region far exceeds the growth of power generation capacity, directly pushing up the wholesale price of electricity, and these costs will eventually be passed on to end consumers. PJM President and CEO David Mills pointed out in a statement that the results of this auction clearly reflect that the growth of electricity demand has significantly exceeded the growth of supply. Mills emphasized that the power grid is working with government and industry representatives in many aspects to accelerate the introduction of new developments. Machinery and reasonable management and control of new loads, with a view to reestablishing the balance between power supply and demand as soon as possible. However, data disclosed by Monitoring Analytics, a third-party independent market regulator, shows that since 2024, the increase in electricity prices due to data center construction has cumulatively increased electricity costs by approximately US$29 billion for utility users in the PJM jurisdiction. Local conflicts intensify: New York State halts new construction, Pennsylvania seeks legal action High electricity costs and severe grid load are causing strong dissatisfaction among local governments, political circles and social utility advocacy organizations in the United States. Because PJM is directly under the jurisdiction of the U.S. Federal Energy Regulatory Commission (FERC), local state governments and regulatory agencies lack direct intervention methods in its electricity pricing mechanism, resulting in frequent local legal proceedings. Patrick Cicero, counsel for the Pennsylvania Public Utilities Legal Project, bluntly stated that the growth trend in electricity demand cannot be reversed in the short term, and the normalization of high electricity prices is a foregone conclusion. Previously, in view of the high energy costs that seriously harmed the interests of consumers, Pennsylvania Governor Josh Shapiro sued PJM in December 2024, and finally reached a settlement agreement to set a price cap, and then reluctantly locked in some financial relief for consumers in the state. Facing the huge pressure caused by computing power facilities on energy and the environment, local governments have begun to take administrative intervention. The New York State government officially announced on Tuesday that it will implement a one-year ban on new large-scale data centers across the state. This is also the first state-level data center construction moratorium in the United States, aiming to allow a buffer period to evaluate the substantial impact of such high-energy-consuming projects on the local ecological environment and energy networks.