RELIABILITY-BASED MODELING FOR MISSOURI RIVER DAM SYSTEM

Abstract

Recently, the dam failure types shift from traditional causes (such as system nature of incidents) to operational risk. These failures occur because of unforeseen combinations of usual conditions. Since the operational components have complex internal and external interactions, we take them into an integrated system. Moreover, the Monte-Carlo simulation method was applied to develop a reliability-based model to study the system performance. Our approach incorporates different sources of uncertainty, including failure probability of components such as turbines and spillway gate facilities. This model allowed us to evaluate the reliability and availability of the system. The system reliability analysis helps us understand the relationship between failure modes and safety decisions made. In further, the model also allows experimenting on operational strategies as well as maintenance guidelines. This thesis presents the framework we have developed and illustrated the results and analysis of our application in the Missouri River mainstem reservoir system. In addition, four scenarios which Corps engineers are going to consider, have been applied to explicit the impacts of modeling system with different maintenance strategies. Besides, we used the stochastic time-series inflow instead of our historical data to evaluate the system performance.