
For nearly two decades, electric utilities planned around flat demand and incremental upgrades. That era is over. Executives from utilities, grid reliability organizations, engineering firms, and commercial real estate said this week that hyperscale AI data centers are forcing the electric industry to rethink how it studies, prices, and connects large customers, according to a report from Data Center Knowledge covering a Cleveland energy forum hosted by Thompson Hine.
“The data centers are ubiquitous. They are coming,” said Brian Thiry, director of strategic engagement and reliability at ReliabilityFirst, a NERC regional entity. He noted that PJM Interconnection load forecasts have risen sharply, and even a fraction of projected growth would exceed anything the grid has absorbed in decades.
Beyond Generation: Deliverability Is the Bottleneck
Panelists stressed that finding new generation capacity is no longer the core challenge. “It’s not just about the generation and the load,” Thiry said. “It’s about the deliverability and the infrastructure.” That means new transmission lines, substations, transformers, and switching equipment, much of which takes years to permit, manufacture, and install.
Rachel Lindesmith of FirstEnergy urged developers to engage utilities as early as possible in project planning. FirstEnergy has introduced a two-stage load study process to give developers early cost and schedule estimates while filtering out speculative projects before they consume engineering resources. She noted that some transformers and breakers now take four to five years to obtain. FirstEnergy is also partnering with American Electric Power on more than 300 miles of new 765-kV transmission across Ohio.
Power Now Drives Site Selection
Commercial real estate broker Terry Coyne said industrial site-selection conversations have flipped: developers once led with questions about highways and workforce, but now lead with megawatts. Manufacturers, robotics firms, and defense projects are increasingly requesting 20 MW or more, competing directly with hyperscale AI campuses for the same substations and transmission capacity. “The power is tapped out,” Coyne said of Ohio’s data center land market.
Reliability Extends Beyond the Fence Line
Behind-the-meter generation and small modular reactors were discussed as partial solutions, but panelists said neither eliminates the need for grid integration. “Bringing your own generation doesn’t necessarily solve all of our reliability questions,” Thiry said, noting that private power plants still require modeling and operating studies to manage power flows during outages or disturbances.
ReliabilityFirst is working with NERC on proposed standards treating large AI campuses as computational load entities, effectively recognizing them as operating elements of the bulk power system. The discussion followed a recent Federal Energy Regulatory Commission order requiring regional transmission organizations to justify or reform large-load interconnection rules, though panelists cautioned that policy changes alone cannot shorten equipment lead times or speed up permitting and manufacturing constraints.