Quantification of Visual Inspection Results to be Integrated as a Diagnostic Tool for Hydrogenerators

Visual inspection of generators is a widely used and industry-proven diagnostic technique. It can detect or confirm many symptoms or anomalies related to the generator failure mechanisms and complements conventional diagnostic tools to provide a more comprehensive assessment of the condition of generators. However, a visual inspection tool poses two major challenges: quantifying the symptoms of observed degradation and managing the information. Over the past decade, Hydro-Quebec has developed a methodology for quantifying the observations made during visual inspection in the same way as any other diagnostic tools. The initial version of the visual inspection tool only included the stator components. Quantification algorithms from this first version were updated based on feedback from more than ten years of collecting stator visual inspection results, and rotor components were integrated into the diagnostic tool. The latest version of the visual inspection tool for stator and rotor has been implemented in a new in-house application called DIAAA. The aim of this paper is to present the methodology developed to objectively quantify the results of visual inspection of stator and rotor components. The structure is described for the automatic calculation of the condition index based on the classification of observable symptoms, subcomponents, and components. Results obtained from the visual inspection tool can now be judiciously combined with other tools to improve the diagnostic of generators and automatically identify stator and rotor failure mechanisms.