A strategy for advanced condition based maintenance of large generators

In this paper a general maintenance strategy for generators is discussed, which incorporates various modules to assess critical generator components on a regular basis within a dynamic time span. This new approach for maintenance, not only includes various on- and off-line diagnostic procedures, but also a "library of experience" represented by multiple design know-how, generator fleet experience, unit history, operational characteristics and OEM records. The capability for advanced diagnostic trending and assessment will be obtained by providing expert analysis of stator winding, rotor winding, and generator bearings through rapid remote communication with local stations and a remote server. The system comprises a so-called virtual working environment providing intelligent software agents to support human experts in diagnostic pattern visualization, analysis, trending and assessment.