Abgeschlossene Abschlussarbeiten
Eine Liste mit Themenvorschläge für Abschlussarbeiten zur Erlangung des Diplom- oder Mastergrades und des Bakkalaureats sowie für Projektarbeiten finden Sie auf der Seite des Instituts.
Folgende Abschlussarbeiten wurden an der Professur für Wasserbau betreut:
Probabilistic Safety Analysis of Dams - Methods and Applications
Art der Abschlussarbeit
Dissertation
Autoren
- Kassa, Negede Abate
Betreuer
- Prof. Dr.-Ing. habil. Hans-B. Horlacher
Weitere Betreuer
Univ.-Prof. Dr.-Ing. Jürgen Jensen (Universität Siegen)
Abstract
Successful dam design endeavor involve tasks of generating technical solutions that meet intended functional objectives and selecting the best among the alternative solutions. The selection criteria are mainly degree of design conformance with technical specifications and safety standards. The selection process involves evaluation on whether an optimal balance is set between safety and economy. This requires quantitative expression for lifetime dam performance and safety. Objective and numerical evaluation of safety and performance (structural and functional) of dams is a complex undertaking. Its domain involves much uncertainty (uncertainty in loads, hazards and strength parameters, boundary conditions, models and consequences). Nonetheless, there is no adequately demonstrated, satisfactorily comprehensive and precise method for explicit treatment, quantification and integration of all uncertainties in variables of dam design and safety analysis.
rnClassical design approaches employ deterministic design parameters and models, such as deterministic peak flood magnitude that is selected based on flood frequency analysis, static flood hydrograph, deterministic material characteristic, deterministic slope stability analysis and reservoir routing methods. Such an approach does not give the ability to design dams that live up to today’s societal safety-economy expectations. Moreover, using such methods, it is hard to keep uniform reliability across dams’ safety chain (for example, against overtopping and sliding). One or more weak links could exist in the safety chain that put the entire system only as safe as the weakest element in the chain.
rnTherefore, there is need for rigorous dam design and risk assessment methods. In this regard, probabilistic methods are arguably the main strategy to reasonably estimate dams’ performance and safety, for example hydrologic performance guaranteed in cases of extreme events or structural performance like slopes and foundation stability. Probabilistic uncertainty models and risk analysis provide the most complete characterization of dam performance and safety issues. Risk is a combined measure of the probability and severity of an adverse effect (functional and/or structural failure). Thus, risk analysis requires (1) determination of failure probabilities (Pf) and (2) probabilistic estimation of consequences.
rnThis dissertation identified various ways of analyzing and representing design parameters uncertainty pertinent to three dominant dam failure causes (sliding, overtopping and seepage), and it tests a suite of stochastic models capable of capturing these uncertainty to better facilitate the evaluation of Pf. It presents various probabilistic methods (first order second moment (FOSM), second order second moment (SOSM), Monte Carlo simulation method (MCSM) and a new to the purpose analytical method (ASDD)) for determination of Pf. Proves of the analytical methods basic principles are laid, generic implementation architecture is prepared, and its application is demonstrated. A multitude of tailor-made reliability equations, implementation architectures, and solution procedures that will enable the implementations of the above stochastic and analytical methods for the evaluation of Pf due to the three dominant dam failure causes are presented. The approaches are illustrated taking the case of Tendaho Dam, Ethiopia. The performance of the various methods and approaches are compared with each other and with that of the classical deterministic design approach. The presented probabilistic methods and technical approaches are proved to be applicable with varying degree of accuracy, suitability, reproducibility and theoretical foundation. They exposed facts that are obscured by the deterministic method. Eventually, several important conclusions and lessons are drawn from the analysis made.
rnClassical design approaches employ deterministic design parameters and models, such as deterministic peak flood magnitude that is selected based on flood frequency analysis, static flood hydrograph, deterministic material characteristic, deterministic slope stability analysis and reservoir routing methods. Such an approach does not give the ability to design dams that live up to today’s societal safety-economy expectations. Moreover, using such methods, it is hard to keep uniform reliability across dams’ safety chain (for example, against overtopping and sliding). One or more weak links could exist in the safety chain that put the entire system only as safe as the weakest element in the chain.
rnTherefore, there is need for rigorous dam design and risk assessment methods. In this regard, probabilistic methods are arguably the main strategy to reasonably estimate dams’ performance and safety, for example hydrologic performance guaranteed in cases of extreme events or structural performance like slopes and foundation stability. Probabilistic uncertainty models and risk analysis provide the most complete characterization of dam performance and safety issues. Risk is a combined measure of the probability and severity of an adverse effect (functional and/or structural failure). Thus, risk analysis requires (1) determination of failure probabilities (Pf) and (2) probabilistic estimation of consequences.
rnThis dissertation identified various ways of analyzing and representing design parameters uncertainty pertinent to three dominant dam failure causes (sliding, overtopping and seepage), and it tests a suite of stochastic models capable of capturing these uncertainty to better facilitate the evaluation of Pf. It presents various probabilistic methods (first order second moment (FOSM), second order second moment (SOSM), Monte Carlo simulation method (MCSM) and a new to the purpose analytical method (ASDD)) for determination of Pf. Proves of the analytical methods basic principles are laid, generic implementation architecture is prepared, and its application is demonstrated. A multitude of tailor-made reliability equations, implementation architectures, and solution procedures that will enable the implementations of the above stochastic and analytical methods for the evaluation of Pf due to the three dominant dam failure causes are presented. The approaches are illustrated taking the case of Tendaho Dam, Ethiopia. The performance of the various methods and approaches are compared with each other and with that of the classical deterministic design approach. The presented probabilistic methods and technical approaches are proved to be applicable with varying degree of accuracy, suitability, reproducibility and theoretical foundation. They exposed facts that are obscured by the deterministic method. Eventually, several important conclusions and lessons are drawn from the analysis made.
Zugeordnete Forschungsschwerpunkte
- Sicherheitskonzepte (Risikomanagement) für wasserbauliche Anlagen
Schlagwörter
Talsperren Versagen Risiko
Berichtsjahr
2010