Skip to main content

3C2: Hydraulic design challenges for replacement runners at Barkley Dam Power Plant

July 13, 2022
Room 110
Equipment and Technology
This article describes the challenging, yet successful hydraulic design performed by ANDRITZ for the Kaplan runner replacement included in Barkley Dam Power Plant’s rehabilitation. Barkley Dam Power Plant is located on the Cumberland River in the State of Kentucky, U.S.A. ANDRITZ received in 2020 an order from the U.S. Army Corps of Engineers to supply four new Kaplan runners, rewind of three generators, general turbine inspection and refurbishment. The turbines have a very wide operating range in head, ranging from a minimum net head of 18 ft (5.5 m) up to a maximum net head of 57 ft (17.4 m). In terms of power, the units are to operate from 12000 to 63605 HP (8.9 to 47.4 MW). This new output is almost 10% higher than the original one, thus creating interesting hydraulic and mechanical challenges. As a matter of fact, the runner being the sole main new item in the water passage, it has to run in the existing ‘’barrel type’’ discharge ring, while ensuring an acceptable cavitation behavior. Since the existing mechanical governor is kept, the blade torques and blade operating range (limited to only 10º) had to be carefully optimized during the design phase to match the existing system’s capacity and servomotor stroke. While numerical flow simulations are thoroughly used and calibrated for designing turbines, the development of a Kaplan turbine remains challenging, especially for a particular rehabilitation project like Barkley Dam. The need of a model test is of high importance to verify all performance aspects of the turbine. Fully homologous model tests were performed in 2020 in the ANDRITZ HYDRO laboratory in Linz/Austria and were made in accordance with IEC 60193 “International Code for Model Acceptance Tests of Hydraulic Turbines”. Tests regarding efficiency, cavitation, runaway speed, pressure pulsation, axial thrust, runner blade and guide vane torques were performed during the homologous model test.

300 x 250 Bottom