Stochastic evaluation of turbine-generator shaft fatigue due to system faults and faulty synchronization

Abstract

In spite of the fact that the nature of the interaction between transmission network disturbances and the turbine-generator shaft system is subject to a relatively large uncertainty, the analysis of turbine-generator shaft torques has been usually performed using a deterministic framework. This paper presents a more realistic probabilistic approach in studying this phenomenon. In this approach, an analytical technique is used to evaluate the maximum turbine-generator shaft torsional torques considering the uncertainties associated with system faults and faulty synchronization. Since the probabilistic nature of the basic initiating factors of events such as the error angle, the generator speed, the fault type, the fault location and line reclosing scheme are independent and mutually exclusive, the conditional probability theorem can be utilized to assess the turbine-generator shaft fatigue. A calculation method is shown for the expected life expenditure ratio and the expected shaft lifetime taking into consideration these factors and using the shaft strength characteristics.