Discover highlights from FY 2025 in this digital report. Covering July 2024 to June 2025, the report features actionable research, engagement with decision‑makers on emerging issues and hands‑on learning experiences that equip Duke students to lead.

The rise of artificial intelligence is accompanied by a rise in the need for huge data centers. Duke University researchers Merritt Cahoon and Ian Hitchcock talked with WAER's Deeper Shade of Green podcast about their recent report examining whether the build out of these data centers will be a boon or bust in resource uses and innovation.

Energy Secretary Chris Wright has directed the Federal Energy Regulatory Commission to standardize large load interconnections, but the advance notice of proposed rulemaking leaves open many critical questions. Former FERC Commissioner Allison Clements and Tyler Norris, lead author of a seminal Duke paper on load flexibility, talked with Latitude Media about some of those considerations.

Diesel generators are meant to be used rarely by data centers, but there is concern they could run more frequently if grid operators disconnect them during times of peak demand, reports Inside Climate News. An alternative future could include diversifying backup power options to include battery storage, renewables, natural gas or cleaner diesel products, noted a February report published by the Nicholas Institute.

Merritt Cahoon and Ian Hitchcock, researchers at the Deep Tech at Duke Initiative, joined the Policy 360 podcast to talk about a recent report they authored examining the sustainability challenges and opportunities for hyperscale data centers.

Enabling load flexibility is critical to achieving speed-to-power for data centers and other large energy customers while keeping the grid reliable and affordable for everyone else, write former FERC Commissioner Allison Clements and Roselle LLP partners Miles Farmer and Sam Walsh in an op-ed for Utility Dive. The three were co-authors of a policy brief that offers rules that could comprise a workable framework to achieve large load flexibility success.

As a Ph.D. student and research fellow, Tyler Norris’ work is shaping industry and policy conversations, with bipartisan policy efforts, academic impact and a new career opportunity, Duke Today reveals in a new profile. A study on rethinking load growth in U.S. power systems that Norris co-authored has made waves across the energy sector. His research has led to a spot on the 2025 TIME100 Climate list, a guest essay in The New York Times and a new position at Google focused on energy innovation—and more.

For 20 years, the Nicholas Institute has pursued the “quest for yes”— bringing people together across differences, grounding big ideas in evidence and co-creating durable solutions to environmental and energy challenges. On Oct. 23, the institute welcomed students, scholars, alumni and partners in the community to an anniversary celebration that blended reflection with forward momentum.

Last Thursday, Energy Secretary Chris Wright directed the Federal Energy Regulatory Commission to consider rulemaking to fast-track interconnection for large loads—as long as they agree to be curtailable or colocate with dispatchable generation. Former FERC Commissioner Allison Clements and Duke Ph.D. student Tyler Norris joined the Catalyst podcast to discuss what the proposal actually means for interconnection.

Before data centers come online, they need to have proactive plans to avoid drawing too much energy, said Duke University expert Tim Profeta, who co-authored a February analysis on how load flexibility could help manage rising U.S. energy demand. Regulators and utilities could require data centers to create those plans in exchange for jumping long queues to connect to the grid. “The biggest incentive is speed to interconnect to the grid,” Profeta told The Guardian.

Duke Energy plans to burn coal for two to four years longer than previously estimated, citing the state’s growing energy demand—driven by data centers and large manufacturers—among other factors, reports Inside Climate News. A February analysis by Duke University researchers posits that targeted, timed reductions in power demand could ease the strain on the grid without massive power-plant buildouts.

How can new data centers for artificial intelligence be sustainably and responsibly powered on the current U.S. electrical grid? A Duke University panel discussion, held at The Nest Main Stage during Climate Week NYC 2025, delivered insights into how to manage electrical loads flexibly and power the AI future sustainably.

At Climate Week New York City 2025, two themes resonated throughout Duke University’s presence: innovation and connections. Duke Today recapped some of the nearly dozen events across the week that Duke experts—including several from the Nicholas Institute—participated in. The events connected climate to finance, health, oceans, technology and more.

After a whirlwind of panels, side conversations, and informal discussions, Nicholas Institute Director Brian Murray shared the themes and insights that stuck with him from Climate Week NYC 2025. "If there was one unifying thread, it was this: the global energy transition is real, accelerating, and yet deeply uneven," Murray wrote.

During a symposium at Middlebury College, Nicholas Institute executive in residence Tim Profeta warned that U.S. data centers could double their share of national power consumption from 4% to 8% by 2030. Yet Profeta also suggested that AI’s high demand and hyperscaling might accelerate clean power adoption as tech giants promise to invest billions in geothermal, nuclear and storage technologies, reports The Middlebury Campus.