A Bayesian approach for a damage growth model using sporadically measured and heterogeneous on-site data from a steam turbine

Abstract

Accurate prediction of the remaining useful life (RUL) of plant turbines is a major scientific challenge for effective operation and maintenance in the power plant industry. This paper proposes an RUL prediction methodology that incorporates a damage index into the damage growth model. A Bayesian inference technique is used to consider uncertainties while estimating the probability distribution of a damage index from on-site hardness measurements. A Bayesian approach is proposed for the damage growth model for use with aged steam turbines. The predictive distribution of the damage index is estimated using its mean and standard deviation. As a case study, real steam turbines from power plants are examined to demonstrate the effectiveness of the proposed Bayesian approach. The results from the proposed damage growth model can be used to predict the RULs of the steam turbines of power plants regardless of load types (peak-load or base-load) of the power plant.