A new report, “Rethinking Load Growth: Assessing the Potential for Integration of Large Flexible Loads in U.S. Power Systems,” published by Duke University’s Nicholas Institute for Energy, Environment & Sustainability,” notes that rapid U.S. load growth (driven by unprecedented electricity demand from data centers, industrial manufacturing, and electrification of transportation and heating), is colliding with barriers to timely resource expansion.

“Protracted interconnection queues, supply chain constraints, and extended permitting processes, among other obstacles, are limiting the development of new power generation and transmission infrastructure,” said the report. “Against this backdrop, there is increasing urgency to identify strategies that accommodate rising demand without compromising reliability, affordability, or progress on decarbonization.”

The report noted that aggregated U.S. winter peak load is forecasted to grow by 21.5 percent over the next decade, rising from approximately 694 GW in 2024 to 843 GW by 2034, according to the “2024 Long-Term Reliability Assessment” of the North American Electric Reliability Corporation (NERC). This represents a ten-year compound annual growth rate (CAGR) of two percent higher than any period since the 1980s, according to NERC. Meanwhile, the Federal Energy Regulatory Commission’s (FERC) latest five-year outlook forecasts 128 GW in peak load growth as early as 2029, which is a CAGR of three percent.

As noted, the primary catalyst for these updated forecasts is the surge in electricity demand from large commercial customers. While significant uncertainty remains, particularly following the release of DeepSeek (a Chinese AI company product), data centers are expected to account for the single largest growth segment, adding as much as 65 GW through 2029 and up to 44 percent of US electricity load growth through 2028. Artificial intelligence (AI) workloads are projected to represent 50 percent to 70 percent of data center demand by 2030, up from less than three percent at the start of this decade, with generative AI driving 40 percent to 60 percent of this growth.

“Analysts have drawn parallels to the 1950s through the 1970s, when the United States achieved comparable electric power sector growth rates,” said the report. “Yet these comparisons arguably understate the nature of today’s challenge in the face of stricter permitting obstacles, higher population density, less land availability, skilled labor shortages, persistent supply chain bottlenecks, and demand for decarbonization and greater power reliability.”

While historical growth rates offer a useful benchmark, the sheer volume of required new generation, transmission, and distribution capacity forecasted for the United States within a condensed timeframe appears unprecedented.

The immensity of the challenge underscores the importance of deploying every available tool, especially those that can more swiftly, affordably, and sustainably integrate large loads. The time-sensitivity for solutions is amplified by the market pressure for many of these loads to interconnect as quickly as possible.

In recent months, the U.S. Secretary of Energy Advisory Board (SEAB) and the Electrical Power Research Institute (EPRI) have highlighted a key solution: LOAD FLEXIBILITY.

The premise is that the unique profile of AI data centers can facilitate more flexible operations, supported by ongoing advancements in distributed energy resources (DERs). Flexibility, in this context, refers to the ability of end-use customers to temporarily reduce their electricity consumption from the grid during periods of system stress by using on-site generators, shifting workload to other facilities, or reducing operations.

“When system planners can reliably anticipate the availability of this load flexibility, the immediate pressure to expand generation capacity and transmission infrastructure can potentially be alleviated, mitigating or deferring costly expenditures,” said the report. “By facilitating near-term load growth without prematurely committing to large-scale capacity expansion, this approach offers a hedge against mounting uncertainty in the U.S. data center market in light of the release of DeepSeek and related developments.