< Back to all clusters
[TECHNOLOGY] · Ireland, Netherlands, South Korea, Mexico, United States · 27 sources

started · updated

AI data center growth strains global power grids and water resources

The rapid expansion of artificial intelligence (AI) is placing unprecedented strain on global energy and water resources. According to the International Energy Agency (IEA), electricity consumption by data centers is projected to nearly double, rising from 485 TWh in 2025 to 950 TWh by 2030, accounting for approximately 3% of global demand.

The United Nations Economic Commission for Europe (UNECE) has warned that data center construction is outpacing the development of electrical grids. While new facilities can be operational within two to five years, upgrading high-voltage transmission lines often requires more than a decade due to regulatory and planning delays. This mismatch risks causing voltage oscillations and cascading failures, particularly in grids heavily reliant on intermittent renewable energy.

In response to these challenges, companies are seeking alternative solutions. Microsoft is planning to expand its computing capacity to 38 gigawatts by 2032, while others like Bloom Energy provide on-site power alternatives to bypass grid delays. Additionally, the environmental impact extends to water usage; Rystad Energy estimates that water demand for data center cooling could reach 644 billion liters by 2030. Tech firms like Nvidia are exploring closed-loop liquid cooling systems to mitigate this water footprint.

Entities

Gartner · International Energy Agency · Ireland · McKinsey · Microsoft · Netherlands · Nvidia · Risen Energy · Rystad Energy · Thomas Croll-Knight · United Kingdom · United Nations Economic Commission for Europe

Claims

What the coverage asserts, and how many sources carry each claim.

Sources

about 21 hours ago
about 9 hours ago