Overtopping Risk Evaluation of Tabarak-Abad Dam based on Univariate and Bivariate Flood Frequency Analysis

Among the most important causes of dam failure, overtopping is the main causes. Analysis of dam failure due to overtopping is essential to calculate the failure risk considering the uncertainties. In this study, the application of risk analysis for earth dam failure due to overtopping based on univariate and bivariate flood frequency analysis has been investigated using log-normal distribution for the Tabarak-Abad earth-fill dam located in Khorasan Razavi province of Iran. Considering Qp – V combinations (flood peak discharge - flood volume) for the joint return periods of 50, 100 and 200 years, the results of the bivariate flood frequency analysis have been proposed in the form of six inflow hydrographs. The overtopping risk has been evaluated based on the univariate flood frequency analysis for all hydrographs resulted from bivariate frequency analysis with different return periods and six initial depth of water in the reservoir, considering quantile of flood peak discharge, initial depth of water in the reservoir and discharge coefficient of spillway as uncertain variables. Uncertainty analysis is conducted using Monte Carlo simulation method and Latin hypercube sampling technique. Finally, comparison of univariate and bivariate flood frequency analysis within different periods indicates that bivariate flood frequency analysis method resulted in greater estimated overtopping risk values in all return periods which is accompanied with higher degree of risk. Also, the V1-Q hydrograph, which has the highest volume of flood, is associated with a higher risk in all waterlevels in comparison to other hydrographs. Moreover, to evaluate overtopping risk based on univariate and bivariate frequency analyses, the increasing trend of risk values for rising water level in the reservoir is more tangible than that of increasing return periods. So that at the water level of 66 meters, overtopping risk for the dam is very high.