Hydropower is a core component of building a clean energy power system. As hydro-generator capacity and head gradually increase, plant structural vibration caused by unit operation has become increasingly prominent, directly threatening the long-term operational safety of hydropower stations. Most existing studies rely on vibration tests under short-term, specific operating conditions, lacking research on vibration characteristics and sources during long-term unit operation. To address this gap, a one-month continuous vibration test was conducted on a large-scale high-head hydropower station. By analyzing the time-domain and frequency-domain characteristics of vibration data, the plant vibration law was explored. An information entropy analysis model integrating long-term vibration data and low-frequency unit monitoring data was established to quantitatively analyze different vibration sources’ effects. The results show that (1) the contribution of hydraulic load to plant structure vibration decreases markedly from 76.7% at ∼130 MW to 16.4% at the rated 600 MW load, (2) vibration intensity decreases with increasing unit load, and vertical vibration is significantly greater than horizontal vibration; (3) Low-frequency tailwater pulsation and hydro-generator rotation are the main sources, dominating low-load vertical vibration and high-load horizontal vibration, respectively; and (4) the information entropy model is feasible for vibration source identification and quantitative analysis. The proposed information entropy model can be readily adopted by other hydropower station operators for real-time monitoring and identification of dominant vibration sources. This study provides guidance for hydropower station plants’ long-term operational safety and complex vibration source control.
Li et al. (Wed,) studied this question.