Application of Pseudo-Poincaré Maps to Assess Gas Turbine System Health

Abstract

The transition between two operational modes in a dynamic system often occurs under strict measures such that repeatability is more a target than a metric for performance evaluations. The concept of Poincaré maps may be applied to health monitoring systems. In this paper, a diagnosis strategy based on Poincaré maps will be presented to evaluate dynamic systems’ behavior with application to power generating gas turbines based on the repeatability of the transition trajectories. Using this approach, extensive operating data for three 85 MW simple cycle power generating gas turbines were evaluated to assess and analyze their performance. The proposed strategy was capable of isolating failed starts; automatic unloads and load rejection events as major events during the test period. Results and conclusions drawn in this study regarding the health condition of the rotating equipment demonstrate the opportunity for system monitoring.