Influence of draft tube water injection system on cavitation behaviour in a full-scale Francis turbine with visual access

29th IAHR Symposium on Hydraulic Machinery and Systems

Vilberg, I.K.^1^2 ; Kjeldsen, M.^2 ; Escaler, X.^3 ; Ekanger, J.V.^2 ; Nielsen, T.K.^1

Norwegian University of Science and Technology, Trondheim, Norway^1

Flow Design Bureau AS, Stavanger, Norway^2

Universitat Politècnica de Catalunya, Barcelona, Spain^3

关键词: Acoustic emission sensors;    Blade passing frequency;    Cavitation intensity;    Demodulation techniques;    Long term observations;    Modulation frequencies;    Pressure fluctuation;    Water injection systems;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/240/2/022015/pdf
DOI  :  10.1088/1755-1315/240/2/022015

来源: IOP

【 摘 要 】

This paper describes the results of a comprehensive and ongoing project to control unsteady fluid-dynamic behaviour and associated effects at Svorka hydro power plant (25 MW, 260 m, 600 rpm), operated by Statkraft in Norway. This Francis unit had an original air admission system installed on the draft tube wall and turbine shaft. However, these systems did not perform satisfactorily to prevent high intensity pressure fluctuations due to vortex rope at part load. To mitigate them, a complementary draft tube water injection system, designed by Flow Design Bureau AS (FDB), was installed in 2010. Moreover, the runner suffered from cavitation pitting on the blades and a measuring campaign with several accelerometers and an acoustic emission sensor concluded that the unit was susceptible to high load cavitation erosion. Given the complexity of the turbine flow, it was decided to install four transparent acrylic glass windows on the draft tube, allowing for visual access to the runner blades and outlet flow. To evaluate the influence of the draft tube injection system on the cavitation behaviour, a series of measurements of cavitation intensity were carried out. High frequency data from the sensors were processed with demodulation techniques at various expected modulation frequencies, particularly the blade passing frequency and the frequency of the draft tube vortex rope. Furthermore, the pressure pulsations were measured and quantified with pressure sensors on the draft tube wall. The results indicate that draft tube water injection has little or no effect on the measured cavitation intensity in the runner. However, other aspects of the machine, including the seasonal and daily variation of the submergence level, must be taken into consideration. In conclusion, it appears that additional long-term observations are needed to clarify the global dynamic behaviour along the entire operating range.