AI data centers now routinely demand more electricity than midsize cities. On June 10, an industry group launched a new energy performance framework as developers, utilities, and contractors face a buildout scale with no modern precedent in commercial real estate.
The National Electrical Manufacturers Association (NEMA), ASHRAE and Pacific Northwest National Laboratory released the AI Data Center Energy Performance Framework to guide planning, siting, design, commissioning, operations and grid-interactive design for facilities whose electrical loads increasingly exceed 1 GW. Patrick Hughes, senior vice president of technology at NEMA, said data centers "are fundamentally different than any other building type. You have to think of them more as cities than a building." NEMA's Grid Reliability Study projects data center electricity consumption will rise roughly 300% over the next decade and account for 38% of net U.S. demand growth through 2037.
Building the ecosystem, not just the structure
The framework arrives as contractors describe a fundamental shift in what it takes to deliver these projects. "Five years ago, the challenge was building a data center," Abrar Sheriff, president of Turner Construction Co., said in a statement. "Today, success depends on delivering an entire ecosystem of power, cooling, workforce, digital infrastructure, and supply-chain capacity at unprecedented scale."
That shift has real-world implications for real estate and construction firms that manage everything from site selection to skilled labor coordination. Resources focused on AI for Real Estate & Construction are emerging as the industry adapts, but the on-the-ground reality is that projects now require utility-scale power investments that once fell outside a developer's scope.
Chris McFadden, senior vice president of global communications at Turner, pointed to a multibillion-dollar AI campus in the Southeast where the company has roughly 7,500 workers on site - more than three times the workforce on a current NFL stadium project. "There's huge investments in workforce development. There's a huge investment in bringing people into the trades. There's a huge investment in power, and huge investment in providing all the utilities that are needed to make that data center run," McFadden said. Those upgrades often extend beyond the campus fence, improving generation, transmission, and utility infrastructure for surrounding communities.
Power access overtakes site selection as the top bottleneck
Access to electricity has eclipsed traditional concerns such as land costs and building design. "This concept of speed to power, I think, is number one," Hughes said, referring to the struggle to get through utility interconnection queues. "It's getting through the utility interconnection queue."
Allison Weus, global head of storage at Wood Mackenzie, identified the same problem: "The planning processes of doing load and generation not together would probably be the top one. The lingering issues of bringing new generation online. And then the cost of new generation." In many organized wholesale electricity markets, generation and load interconnection processes remain separate, slowing the alignment of new supply with AI demand.
As a result, data center developers are increasingly acting as independent power producers. Weus said roughly 35% of planned data center projects, measured by power demand, now include some form of associated generation, storage or flexibility resource. "It is done because they are desperate to get online," she said. Those alternatives include batteries to reduce peak demand, gas-fired generation, microgrids, and dedicated power systems - all of which add construction complexity and operational headaches traditionally handled by utilities.
Workforce and supply chain constraints reshape project delivery
Power may be the loudest challenge, but a shortage of skilled labor across the entire supply chain is a close second, Hughes said. The shortage extends from electricians and construction crews to the workers who manufacture transformers, switchgear, energy storage systems and other electrical equipment. Kathryn Thompson, founder and CEO of Thompson Research Group, described the buildout's scale in stark terms: "People ask if this is a bubble. It's like asking if building the U.S. interstate system in the late '50s was a bubble. It was only just getting started."
Those labor pressures are forcing contractors and developers to embrace modularization and standardized designs that reduce on-site variability and raise productivity. "The U.S. has talked for years about modularization of construction … it's finally happening, because it has to," Thompson said. Hughes sees similar momentum around high-voltage DC power distribution, with some NEMA members already planning for 1,200-volt or 1,500-volt architectures that could cut energy losses from repeated AC-DC conversions, though they introduce new safety and engineering questions.
Why this matters for real estate and construction professionals
AI data centers are no longer just a specialty asset class. They are infrastructure projects that demand the same scale of power planning, workforce mobilization, and supply chain coordination as a major public works program. Contractors and developers who treat these projects as larger versions of traditional commercial buildings will misjudge the lead times, interconnection rules, and utility partnerships required. The framework from NEMA, ASHRAE and PNNL is designed as a living document - not a static standard - precisely because the pace of change outstrips the years-long cycle of code development. For firms involved in site development, construction management, or skilled trade staffing, the data center buildout is reshaping regional power markets and labor availability in ways that will affect project costs and timelines across all commercial building types.
Your membership also unlocks:
