DETERMINATION OF THE WAKE FREQUENCY CONTENTS FOR A TWIN SHAFT GAS TURBINE VARIABLE NOZZLE GUIDED VANE

摘要

This paper focuses on improving maintenance techniques for GE twin shaft gas turbines equipped with variable Nozzle Guided Vanes (NGVs) by enhancing the ability to predict potentially harmful stimuli due to NGV misalignment. NGV, by varying the geometry, controls the energy split between the high-pressure turbine and power turbine, allowing higher operational flexibility and higher efficiency at partial load/speed. This flexibility is a requirement for GT mechanical drive applications. The alignment of the NGV according to specification is part of the GE procedure to guarantee reliable operation of the turbine. The orientation of the vanes is checked and guaranteed via the measurement of the S/T ratio (throat opening/pitch). The S/T distribution potentially affects the downstream buckets in terms of harmonic stimuli for natural frequencies that could result in bucket HCF failures. The scope of this work is to develop a Fourier series-based tool that, for any statistically possible distribution of S/T ratios, automatically computes the Fourier Coefficients of the corresponding wake pattern and scales each harmonic magnitude with respect to the maximum harmonic magnitude of the symmetric wake pattern. This tool has been specifically developed to drive field operations when some of the NGVs in the row must be replaced. It allows the detection of harmonic stimuli that could potentially increase downstream bucket resonance without making S/T measurements, reducing the forced outage cycle time and saving maintenance costs.