Johns Hopkins study warns US grid-equipment shortages driven by data centres could start next year

Published in Cell Reports Sustainability, the study projects a roughly 15% shortfall in transformers and 30% in other grid equipment, with data centres responsible for up to two-thirds of the gap.
Why it matters
Power, rather than chips, is increasingly the bottleneck for AI data centres, and transformers are among the slowest parts to source. Fortune reports that researchers Yury Dvorkin and Xinyi Yao estimate US electricity use could rise 21% by 2030 and that lead times for transformers have stretched from about six months to more than two years. Trade disruption would widen the shortfall to 25% for transformers and 36% for other equipment, partly because only one US mill makes grain-oriented electrical steel. Watch for utilities delaying connections or pushing up costs for new data-centre projects.
Key facts
- Published in Cell Reports Sustainability
- Authors: Yury Dvorkin and Xinyi Yao, Johns Hopkins
- Shortfall ~15% transformers, ~30% other equipment
- 25% and 36% with trade disruption
- US electricity use could rise 21% by 2030
- Data centres up to two-thirds of the shortfall