Prognostic Enhancements to Diagnostic Systems (PEDS) Applied to Shipboard Power Generation Systems

Numerous advancements have been made in gas turbine health monitoring technologies focused on detection, classification, and prediction of developing machinery faults and performance degradation. Existing monitoring systems such as ICAS (Integrated Condition Assessment System), which is the Navy’s program of record and is deployed on many US Navy ships, employ alarm thresholds and event detection using rule-based algorithms. Adding the capability to predict the future condition (prognostics) of a machine would add significant benefit to the Navy practice. The current paper describes a framework and development process that allows more “plug ‘n play” integration of new diagnostic and prognostic technologies using evolving Open System Architecture (OSA) standards. Although many modules were developed in the PEDS framework, specific gas turbine modules that focus on compressor and nozzle degradation algorithms are discussed. The modules use statistical prediction algorithms and were developed using seeded fault data generated by the Navy engineering station. The modules are fully automated, interact with the existing monitoring system, process real-time data, and utilize advanced forecasting techniques. Such an advanced prognostic capability can enable a higher level of condition-based maintenance for optimally managing total Life Cycle Costs (LCC) and readiness of assets.