AI & digital twins |
Right: Real-time data from a hydropower generation unit at Alder Dam on the Nisqually River in western Washington state, US was used to build a digital twin that accurately represents real life
Dam operators can customise their digital twin to
reflect the uniqueness of each facility; they just need to upload their facility’s data and the Digital Twins for Hydropower Systems dashboard handles the analysis. “Each dam requires a unique maintenance
strategy to improve efficiency, and the new digital twins platform can provide those solutions,” said Chitra Sivaraman, PNNL principal investigator of the project. “The platform is both extensible and scalable - capable of adapting to new facilities, data and models.”
Digital twin models also tackle another challenge.
“The average age of the nation’s dams is around 60 years, meaning multiple generations of employees have worked on each turbine. And knowledge is inevitably lost as seasoned employees retire and new employees join the team,” says Scott Warnick, electrical and automation systems engineer at PNNL and the technical lead on the digital twins project. Hydropower digital twins can record and simulate all changes made to the dam over subsequent years – passing down knowledge and helping future generations make decisions. “The digital twins solution enables hydropower
operators to simulate different scenarios, such as low water flow or varying water levels, and predict future performance or maintenance needs,” Warnick adds. To build a digital twin that accurately represents real life, the team used real-time data from a hydropower generation unit at Alder Dam on the Nisqually River in western Washington state, operated by Tacoma Public Utilities (TPU).
Leading the modelling development is Hong
Wang, principal investigator of the project at DOE’s Oak Ridge National Laboratory. He and the team collected data, such as the river’s pressure as it enters the hydropower facility, how fast the turbines spin and how much power the dam generates over time. In the original version of the digital twin, dam
operators could only observe how normal or expected conditions affected the dam’s mechanical parts. Now in version 2.0, operators have more control. They can
16 | January 2025 |
www.waterpowermagazine.com
adjust water levels, flow rates and turbine speed that might change based on weather, droughts or energy demand. With the ability to simulate both normal conditions and higher or lower water flow, the dam operators can home in on potential problems before they arise. The updated digital twins model addresses another emerging need. As the US moves towards a sustainable grid and embraces more wind and solar energy, hydropower systems must be adaptable and responsive to support a stable grid. The platform is capable of extending the lifespan of US dams while at the same time integrating additional sources of renewable energy to the grid. With a digital twin solution, operators can simulate and review real-world power demand fluctuations. If the model shows that the conditions are optimal to run the turbines, operators can feel confident about proceeding, which maximises revenue. “The digital twins dashboard paves the way for the digitalisation of hydropower systems -providing a critical tool for operators to simulate and optimise grid operation for increased penetration of renewables, such as solar and wind,” Wang said. With continual collaboration with TPU, the team gains insight into ways to improve the model. While helping TPU operate more efficiently, updates to the dashboard mean it can represent a wider diversity of turbines. “The PNNL and ORNL teams have the mathematical and practical skills needed to solve complex digital twin problems,” said Greg Kenyon, automation engineering manager at TPU. The team is also working with Chelan County Public Utility in north-central Washington state to collect and analyse years of operation data records from the Rocky Reach Dam to develop a digital twin. Just like for Alder Dam, Chelan County’s hydropower digital twins will provide operators with the ability to review performance monitoring and analysis, perform predictive maintenance and optimise energy production – all at zero cost.
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45
