Data Centers

AI deployment is driving rapid growth in both the number of data centers and their capacity, and that growth is in turn driving an unprecedented surge in electricity demand.¹ Global electricity demand from data centers is projected to more than double by 2030, to around 945 terawatt-hours, roughly equivalent to Japan’s entire current electricity consumption.² In the United States, the largest and fastest-growing data center market, meeting this demand will require generating 10% more electricity nationally by 2030 than would otherwise be needed.³

Energy supply, not capital or demand, has become the binding constraint on data center growth. In the United States, more than 2,200 GW of generation and storage capacity remains stuck in interconnection queues, nearly twice the capacity of the country’s current power plant fleet, with the typical project now taking 4.5 years to reach commercial operation.⁴ The legacy alternative, on-site gas turbines, offers little relief: lead times for new combined-cycle gas power plants have stretched to roughly five years in 2025, up from three and a half years in 2023, with wait times for large turbines reaching as long as seven years in some cases.⁵ Turbine manufacturers are facing multi-year backlogs, with delivery slots increasingly constrained by surging demand.⁶ Gas turbines also produce direct CO₂ emissions, while their electrical efficiency is generally lower than that achievable by high-temperature fuel-cell systems.

For data center developers, the result is a structural mismatch: electricity demand is accelerating faster than either grid infrastructure or turbine manufacturing capacity can expand. 84% of developers now rank power availability among their top three site selection considerations, having recently overtaken fiber connectivity as the leading factor. It has become the top concern in data center site selection and project planning, ahead of land, capital and connectivity.⁷