Pennsylvania’s data center boom raises big questions about power, water and AI
Pennsylvania is experiencing a significant boom in data center development driven by AI, raising concerns about electricity and water demands. Governor Josh Shapiro has issued an executive order to implement new regulations, requiring developers to cover infrastructure costs and ensure environmental protections. Experts from Bucknell University highlight the state's appeal to developers and the escalating resource consumption challenges posed by AI-powered data centers.
Pennsylvania is experiencing a rapid expansion in data center development, driven by the increasing demands of artificial intelligence. This boom has raised significant concerns among communities, such as Archbald and Allenwood, regarding the substantial electricity and water consumption, development costs, and local environmental impacts.
In response to this accelerated growth, Governor Josh Shapiro issued an executive order on August 18, establishing new "guardrails" for data center development. These regulations mandate that developers bear the costs for new electricity generation and grid upgrades, rather than shifting them to other ratepayers. The order also emphasizes environmental protections, transparency, and community involvement in the development process.
Experts from Bucknell University provided insights into the phenomenon. Brian King, Xin Faculty Fellow in Computer Science and Engineering, noted Pennsylvania's attractiveness to developers due to its available land, robust energy resources, and strategic location near major East Coast population centers. King highlighted that data centers' electricity consumption, particularly for AI, is projected to surge from 4.4% of all U.S. electricity in 2023 to 12% by 2028, placing substantial stress on the grid. He suggested that adopting smaller, specialized AI models could significantly reduce power usage.
Benjamin Hayes, Program Director of Watershed Sciences & Engineering, addressed the profound environmental implications of water usage, explaining that even seemingly small withdrawals can severely impact local aquatic environments, especially when combined with the water needed for power generation. Lily Li, Jane W. Griffith Faculty Fellow and Assistant Professor of Mechanical Engineering, explored the potential for more distributed computing loads to alleviate infrastructure strain. The consensus among the experts and state officials is that balancing the transformative power of AI with its considerable environmental and infrastructural costs represents a complex and urgent challenge for Pennsylvania.