Risk Assessment Method for Steam Turbines Based on the Reliability Analysis of Field Data

摘要

By arranging the field maintenance database of steam turbines with operation history and by conducting probabilistic safety analyses, a quantitative risk assessment method was established. The method includes three steps: (1) event tree analyses of parts failures which require any maintenance action, (2) assignment of unreliability functions of operation time or start cycles for each item in those event trees, and (3) calculation of total risk function by summing up recovery costs multiplied by the unreliability functions obeying the event trees and comparison of the risk cost with preventive maintenance cost. Those three steps were applied to actual turbine components and the effectiveness of proposed method was demonstrated with a risk simulation analysis.