Hosted by Utah State University Libraries
A Computational Framework for Dam Safety Risk Assessment with Uncertainty Analysis
Abstract
dc:description.abstractEvent tree analysis is a commonly-used method in dam safety risk analysis. Event trees are used to obtain quantitative estimates of the probability of dam failure and its associated consequences. This can be done to evaluate the safety of an existing dam or to provide insights into the choice between risk reduction alternatives. In the past, the calculations have been performed using either generalized software developed primarily for business applications or purpose-built spreadsheets. However, these approaches lack generality, can require substantial effort, can be fragile, and are difficult to modify to represent risk reduction measures or to update using new information from investigations. These limitations can lead to using event tree structures that do not properly represent the failure modes and poor numerical precision in risk estimates. To overcome these limitations, the US Army Corps of Engineers (USACE) sponsored the development of a computational framework for efficient, flexible and generalized event tree analysis called DAMRAE (DAM safety Risk Analysis Engine). DAMRAE is now being used by the USACE and the Tennessee Valley Authority (TVA) while continued development takes place at Utah State University. The supporting research and development of DAMRAE was carried in two stages. First, a deterministic (or non-stochastic) version was formulated and developed and then, uncertainty analysis functionality was formulated and included. DAMRAE includes a graphical user interface (GUI) for developing and populating event tree inputs and functionalities for calculating and post-processing results. It provides estimates of the probabilities of various failure modes and their associated consequences for an existing dam. The post-processing step allows the user to combine results for various loading types (e.g. flood and earthquake) and to make comparisons against USACE tolerable risk guidelines. It can be applied to analyze structural and non-structural risk reduction measures, considered as alternatives or staged measures, including obtaining estimates of the risk reduction and the cost effectiveness of risk reduction. Various theoretical challenges were overcome in formulating the computational framework of DAMRAE and new computational concepts were introduced. These include: Connectivity and Pedigree matrices to quantify the user-drawn event tree structures; Common-Cause Adjustment for the non-mutually exclusive failure modes along with the new concepts of system response probability and consequences freezing; and separation of uncertainties in logic and event trees. Several new output presentation formats including the tolerable risk evaluation under uncertainty were introduced as risk analysis outputs. The developed computational framework was extensively verified using several risk analyses performed by USACE and by Dr. David S. Bowles and his group.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy (PhD)
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Civil and Environmental Engineering
- Grantor dc:publisher
- Hosted by Utah State University Libraries
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Srivastava, Anruag
- Contributors dc:contributor
-
- David S. Bowles
- Arthur J. Caplan
- James A. Bay