Regression examination technique: Correlation between nozzle velocity and vibration level of hydro-turbine in SHP run-off-river scheme

Abstract

Regression Examination is one of the most popular statistical analysis method used to determine the correlation between examined variables. It can also be used to predict the value of a dependent variable based on the value of at least one independent variable and this will allow explaining the impact of changes in an independent variable on the dependent variable. With the measured data of flow rate, output power and efficiency of turbine system, the nozzle velocity can be calculated using an appropriate mathematical equation. Along with vibration level data at every measurement axes, they are tabulated and imported into MATLAB program for the correlation determination. Three different set of data are used in this analysis which is based on the variation of water level at different month (low, typical and high) due to rainfall distribution within one year period. The objective of this study was to examine two questions: first is the effect of water level to the measured flow rate and generated output as well the calculated nozzle velocity. Second, with these variation of nozzle velocity, how significance their correlation with the vibration level at multiple axes of the turbine and generator? From the results, it has been discovered that there are significant correlation between nozzle velocity and vibration level, particularly during low level of water (lower rainfall distribution).