AI Data Centers: Energy and Water Impact FAQ
What is the precise impact of AI data centers on natural, energy, and economic resources? Based on the ACES Symposium discussions, the following data points provide clarity on the energy and water metrics shaping the future of data centers in Pennsylvania and across the nation.
National electricity consumption by data centers is projected to grow from roughly 4% to 12% by 2030. In terms of raw power, U.S. data center electricity consumption is expected to rise from 183 Terawatt-hours (TWh) in 2024 to 426 TWh by 2030. This rapid growth is primarily driven by the energy-intensive nature of generative AI.
The surge in demand has led to historic price increases in regional markets like PJM. The July 2024 capacity auction cleared at roughly $270/MW-day, an 800% increase from the previous year’s price of $30/MW-day. By the July 2025 auction, prices reached the market cap of nearly $330/MW-day, reflecting the strain on the grid.
Power densities for advanced AI racks are scaling exponentially. In 2023, the average density was between 25kW and 40kW. By 2025, that figure is expected to exceed 200kW, with densities of 1MW per rack projected for the near future. This shift requires radically different cooling and power delivery infrastructure compared to traditional data centers.
Data centers are already one of the top 10 water-consuming industries in the United States. Current growth rates suggest that by 2030, AI data centers could drain between 731 and 1,125 million cubic meters of water. This is equivalent to the annual household water usage of 6 to 10 million Americans.
Approximately 30% of U.S. households experience some form of energy insecurity due to the rising costs of natural gas and electricity. Currently, 1 in 6 electricity customers struggle to pay their utility bills on time, highlighting the social importance of managing the "electricity shed" to ensure large-scale data center growth doesn't disproportionately affect residential affordability.