Failure-Based Maintenance Planning Using Bayesian Networks: A Case Study Hydraulic Turbine

Abstract

The assessment of existing infrastructures in the energy sector is of great economic importance for the world. The extension of the power generation life of hydroelectric power plants depends on decisions regarding the maintenance and renewal of the equipment. For this purpose, a Bayesian network (BN) has been applied to evaluate the failures in the hydraulic turbine to calculate the failure of the turbine. Forty-six nodes have been identified that will affect the operation of the system. Preventive measures have been established for failures with the highest posterior probability. By creating four different cases, failure probabilities and the change of the main fault have been calculated. How much savings could be made in each case is determined. This proposed framework will be guided in determining the maintenance strategies for hydroelectric power plant operators.