Abstract

This study presents the method for detection and isolation of component faults and
degradation modes in industrial gas turbine engine. Performance of gas turbine engines
gradually deteriorate over the service life due to degradation of the gas path components such
as compressor, combustor and turbines. These physical faults gradually evolve over a
prolonged period of operation and lead to degradation of the performance parameters, such
as efficiency and flow capacity of individual gas-path components.
Performance degradation, in turn, causes changes in the measurable engine parameters,
such as temperature, pressure, rotational speed, and fuel flow rate. Traditionally these
component degradation modes and faults in the engine have been detected by measuring the
changes in these observable parameters through appropriate usage of signal processing and
pattern recognition tools.
In this contribution model-based diagnostic approach has been applied, where
measurable parameters have been used to estimate so-called engine health parameters, i.e.
component efficiencies and flow capacities. Health parameter deviations from nominal
conditions are subsequently used to obtain health indices and the best signature match is then
used to identify likely component degradation modes and faults.
Performance diagnostic and fault isolation process is based on multidimensional complex
health vector space which contains generated health indices, i.e. component capacity and
efficiency indices for different gas turbine components. Simulated gas turbine degradation
modes have been diagnosed and isolated by comparing gas turbine health vector against bank
of fault signatures for different gas-path components.