Uncertainty assessment in hydrodynamic modeling of floods generated by dam break

Arthur Da Fontoura Tschiedel,Rodrigo Cauduro Dias De Paiva

doi:10.1590/2318-0331.231820170074

Arthur Da Fontoura Tschiedel, Rodrigo Cauduro Dias De Paiva

Open Access

https://doi.org/10.1590/2318-0331.231820170074

Copy DOI

Journal: RBRH	Publication Date: Jul 30, 2018
Citations: 10	License type: cc-by

Affiliation: Universidade Federal do Rio Grande do Sul

Abstract

ABSTRACT The objective of the present work was to evaluate sources of uncertainty in dam break simulations for areas with contrasting characteristics: one in a narrow valley with high slopes, and another an open valley with low slopes. A sensitivity analysis of the hydrodynamic model Hec-Ras 5.03 was performed, varying the input data of the model (Manning coefficient, breach configurations, reservoir volume, type of valley topography and equation considered). These variations cause different changes in peak flow, peak time, maximum depth and a maximum speed for different sections of the downstream watercourse. It was concluded that there are uncertainties in determining the input data that impacts in many ways at the generated flood wave, considering both a hydraulic variable of interest, the distance of the section from the dam and the mean geomorphological characteristic of the downstream valley. Topography is not always the most important input data, which allows the possibility of use of low resolution topographies to estimate the peak time in some sections, depending on the valley. Finally, safety coefficients for dam break studies are suggested, aiming to represent uncertainties of the input data in the generated results.

Highlights

Dam construction in Brazil has grown year by year, increasing the risk associated with the implantation of these structures
Regardless of the arrangement and purpose of the dam, possible ruptures of these structures have a high potential for damage, involving economic damage, and many human losses at the downstream locations of these dams (BRASIL, 2010)
We propose preliminary safety coefficients to be applied in studies of this type when the input data used are not fully reliable

Summary

Introduction

Dam construction in Brazil has grown year by year, increasing the risk associated with the implantation of these structures. Several simulations were performed using different input data for factors such as Manning Coefficient, Reservoir Volume, Breach Configuration and Topography, as well as the type of simulation model, in order to evaluate the impact of the uncertainty of each of these factors in the model output results, such as Peak Flow, Peak Flow Time, Maximum Depth, and Maximum Velocity.

Objectives

Results

Conclusion