Author(s): Abate. K. Negede; Hans-B. Horlacher
Linked Author(s): Hans-B. Horlacher
Keywords: No Keywords
Abstract: Overtopping is cause for 30-35% of registered dam failures world wide. Consequently, there are several published attempts towards establishing methods for estimating dam overtopping probabilities (Pf OT). These attempts use Monte Carlo Simulation Method (MCSM) mostly with peak flood inflow (Qp) and in few cases initial water level (ho) and wave parameters too assumed to be random variable (RV). Often influences of other flood hydrograph shape (FH) parameters than Qp are not accounted. Here FH is modeled using an existing gamma function. Effects of taking all or part of FH parameters as RV is evaluated. Besides, in Pf OT, evaluation wave parameters and ho (as influenced by continuous steady river flow, abstractions, evaporation, environmental release and sedimentation) are considered as additional RVs. Applicability of three conventional probabilistic methods (MCSM, FOSM, SOSM) and an analytical solution for finding derived distributions (ASDD) are tested using field data and a reliability function. The probabilistic methods and the proposed intermediate approaches proved to be applicable with varying precessions and ease. All of them predicted the reality in a much better way than the classical deterministic approach. Reference and scrutiny is made to earlier research attempts in this area.
Year: 2009