PREDICTION OF COMPRESSOR FOULING RATE UNDER TIME VARYING OPERATING CONDITIONS

Performance of the compressors deteriorates due to detrimental effects of fouling on the aerodynamic flow characteristic. The compressors need periodic clean up services to re-gain the designed performance. Apart from the operating time, the ambient and the operating conditions affect the fouling phenomenon making accurate scheduling for predictive maintenance very difficult. In this work, the symptoms of compressor fouling are captured through the evolution of the compressor map in terms of loss of isentropic efficiency and mass flow decrease. Compressor mass flow and the rate of humidity. condensation at the inlet of the compressor are identified as the effective factors on the fouling rate. Humidity condensation has a competing effect on fouling rate; increment of the condensed humidity up to a certain level accelerates the fouling rate, while additional mist has an inverse effect. The complex effect of the condensed humidity along with the air mass flow is extracted through training an adaptive neuro-fuzzy inference system. The resulting model reveals how the efficiency and the mass flow of the compressor map vary as a result of fouling development, given the mass flow and the humidity condensation history. The methodology is verified using data from a similar compressor commissioned at a different period.