To illustrate the scale, Hughes points to Buffalo, N.Y., whose peak electricity demand is roughly 500 MW. Emerging AI data centers are increasingly approaching or exceeding 1 GW of demand.

In response, NEMA, ASHRAE and Pacific Northwest National Laboratory on June 10 launched the AI Data Center Energy Performance Framework, describing it as a resource intended to guide planning, siting, design, commissioning, operations, retrofit strategies and grid-interactive design for AI facilities. 

The framework includes guidance on energy storage, microgrids, demand flexibility, liquid cooling and other technologies increasingly incorporated into large-scale projects.

The effort comes as NEMA's Grid Reliability Study, prepared by PA Consulting for the group, projects that data center electricity consumption will increase by roughly 300% over the next decade and account for 38% of net U.S. electricity demand growth through 2037.

Hughes says the framework emerges from concerns that electrical equipment is increasingly being deployed in applications for which existing codes and standards offer limited guidance.

"What we saw was electrical equipment being installed and used in ways that weren't necessarily designed for, and the building codes and standards didn't exist for that specific application," he adds.

Unlike formal standards, which can take years to develop and revise, the framework is designed as a living resource.

"Standards take time," Hughes says. "That does not match the pace of technology innovation and the speed of construction of data centers today."

For Kathryn Thompson, founder and CEO of Thompson Research Group, the scale of the buildout is often misunderstood.

"People ask if this is a bubble," Thompson says of data centers. "It's like asking if building the U.S. interstate system in the late '50s was a bubble. It was only just getting started."

Contractors increasingly describe the challenge in similar terms.

"Five years ago, the challenge was building a data center," said Abrar Sheriff, president of Turner Construction Co., in a statement to ENR. "Today, success depends on delivering an entire ecosystem of power, cooling, workforce, digital infrastructure, and supply-chain capacity at unprecedented scale."

Chris McFadden, senior vice president of global communications at Turner, points to a multibillion-dollar AI data center campus the company is building in the Southeast as an example of how projects have evolved beyond the construction of individual facilities.

"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 says.

Those investments can extend beyond the project itself. McFadden said large AI campuses often require upgrades to power generation, transmission and utility infrastructure that can have broader benefits for surrounding communities.

"It's really any of these multibillion-dollar projects [that] often result in a broader look at that development, beyond the build of a single data center," he adds.

The workforce demands can be equally significant. McFadden says Turner currently has roughly 7,500 workers on the AI data center campus compared with about 2,200 workers on a current NFL stadium project.

RELATED

Grid Access, Not Land, Emerges as Bottleneck for Data Center Construction

Power Moves to the Forefront

Access to electricity has emerged as the industry's dominant challenge, overtaking more traditional concerns such as site selection and building construction.

Data centers are projected to become the largest driver of U.S. electricity demand growth through 2050, accounting for 38% of net consumption growth through 2037, according to NEMA's Grid Reliability Study.

Graphic: Courtesy of NEMA

"This concept of speed to power, I think, is number one," Hughes says, referring to the industry's struggle to secure utility interconnections and electrical capacity. "It's getting through the utility interconnection queue."

Wood Mackenzie sees similar challenges.

"The planning processes of doing load and generation not together would probably be the top one," Allison Weus, the firm's global head of storage, says: "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, making it difficult to align new power supply with rapidly growing AI demand, she adds.

Developers are increasingly pursuing energy infrastructure alongside data centers themselves.

Weus says roughly 35% of planned data center projects, measured by power demand, include some form of associated generation, storage or flexibility resource intended to help accelerate deployment.

In some cases, batteries and storage systems are being used to reduce peak demand and secure faster utility approval. In others, developers are turning to gas-fired generation, microgrids or dedicated power systems.

"It is done because they are desperate to get online," Weus explains.

Developers would generally prefer traditional grid service because of its reliability and operational advantages, she adds, but prolonged interconnection timelines are forcing many projects to pursue alternatives.

Asked whether AI data centers increasingly should be viewed as infrastructure projects rather than real estate developments, Weus says many developers are now assuming responsibilities historically associated with power-sector companies.

"A lot of data center developers are turning into IPPs essentially now because that is the key challenge," she notes, referring to independent power producers.

RELATED

FERC Weighs Federal Oversight of AI Data Center Grid Connections

An Industrial-Scale Challenge

Yet, power is not the industry's only challenge: Hughes ranks workforce availability as the second-largest constraint facing AI infrastructure development.

The workforce challenge extends beyond construction labor. Contractors, manufacturers and analysts interviewed for this story all pointed to shortages among the workers needed to produce transformers, switchgear, energy storage systems and other electrical equipment.

"If you look throughout the whole supply chain for what it took to build that data center, it's really an industrial resurgence [demanded] in the United States," Hughes says. "We need more."

Thompson cited an example in Northern Virginia where an ambulatory care facility reportedly received only a single electrical bid as demand for electricians intensified amid data center growth.

Those labor pressures are helping accelerate interest in modularization and standardized design approaches.

Hughes says standardization can improve workforce productivity by reducing project-to-project variability. Thompson comes to a similar conclusion, saying, "The U.S. has talked for years about modularization of construction ... it's finally happening, because it has to."

Looking ahead, Hughes points to direct-current power distribution as one area where significant innovation is likely to occur.

"We've been talking about 800 volts DC to the rack, but we even had an internal meeting of our members a few weeks ago where they were planning for 1,200 volts or 1,500 volts DC power delivery," he says.

The higher-voltage architectures could improve efficiency by reducing repeated conversions between alternating current and direct current, Hughes adds, though they also raise new safety and engineering challenges.

Weus identifies another challenge: the operational complexity of managing generation and storage assets historically owned and operated by utilities.

"It's not so simple to just buy some gas and batteries, and then you're going to be good to be off-grid," she says.