Retrofitting Dams to Produce Hydropower in the U.S. Could Generate 15.2 TWh of Electricity, or 4 GW of Capacity, According to New Study by Oak Ridge National Lab

       The DOE’s Oak Ridge National Laboratory (ORNL) recently assessed the potential of powering non-powered dams in the U.S. to produce hydropower. Its assessment suggests that 4GW of capacity could be tapped, which could generate up to 15.2TWh of electricity annually. The Idaho National Laboratory (INL) was also involved in the study. ORNL notes that historically, only about 3% of U.S. dams have been powered by adding turbines and associated infrastructure. The 4GW in capacity was spread over 2600 non-powered dams (NPDs), with an average of 1.4 MW per facility.

     The engineers utilized an open-source software platform known as HydroGenerate to assess the NPDs. According to Interesting Engineering:

“The software models power production by pairing turbine performance curves with hydraulic head, which measures the potential energy created by water elevation behind a dam. It also calculates a site-specific design flow to indicate the exact turbine dimensions needed for peak operating efficiency.”

“To establish this metric, the system processes daily historical records from the US Geological Survey’s stream gage network alongside Dayflow, an Oak Ridge dataset that routes runoff across American river channels.”

     Previous assessments came up with higher estimates of between 12 and 30 GW but did not consider the practical constraints to powering NPDs. The figure below shows improvements made after the previous assessments. 

     The new assessment accounts for seasonal flow changes and existing structural limits.

“Many of these dams already serve other critical purposes, and their operational constraints — such as flood control or navigation — can significantly limit hydropower development,” added Carly Hansen, the project’s principal investigator and lead author of the report. “We’ve worked to reflect those limitations in a way that hadn’t been done before.”

     The flowchart below shows the methodology for estimating monthly hydraulic head.

     About 86% of the identified capacity was found to occur on federally owned property, with sites clustered heavily in the Great Lakes basin and along the upper Mississippi River. The Lower Ohio River Valley also had several sites

     They note that updates to the software are planned.

“Upcoming updates will introduce multi-decade precipitation and streamflow variations into the software, while expanding field evaluations to include non-powered dams throughout Alaska and Hawaii.”

     The researchers noted that hydropower development typically takes time, years to a decade or more, and precipitation and streamflow rates change.

“Hydropower development takes time, often decades, so we need to consider what future water availability might look like,” Hansen said. “In some regions, increasing precipitation could even create greater potential for hydropower in the future.”

     In the paper's conclusion below, they note the importance of feasibility in determinations of NPD powering potential.

 

References:

 

Powering 1.4 million US homes: Untapped dams could generate 15.2 TWh electricity. Aman Tripathi. Interesting Engineering. September 6, 2026. Powering 1.4 million US homes: Untapped dams could generate 15.2 TWh electricity

ORNL assessment uncovers hydropower opportunities in existing dams. Oak Ridge National Laboratory. September 1, 2026. ORNL assessment uncovers hydropower opportunities in existing dams

AN ASSESSMENT OF TECHNICAL HYDROPOWER POTENTIAL AT NONPOWERED DAMS IN THE UNITED STATES Carly Hansen, Juan Gallego-Calderon, Camilo Bastidas Pacheco, Cleve Davis, Scott DeNeale, Rohit Mendadhala, Jakob Meng, Glenn Russell, Sean Turner. Oak Ridge National Laboratory. Idaho National Laboratory. January 2026. 3014287