Análise de incerteza dos parâmetros de brecha aplicada à modelagem probabilística do hidrograma de ruptura de barragens de terra

• Ano de defesa: 2021
• Autor(a) principal: Alexandre Ângelo Carmo Luiz da Silva
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ENG - DEPARTAMENTO DE ENGENHARIA HIDRÁULICA; Programa de Pós-Graduação em Saneamento, Meio Ambiente e Recursos Hídricos; UFMG
• Palavras-chave em Português: Estudo de ruptura hipotética; Brecha de ruptura; Analise de incertezas; Análise probabilística; Hidrograma de ruptura probabilístico; Engenharia sanitária; Recursos hídricos - Desenvolvimento; Barragens e açudes - Ruptura; Incerteza; Barragens de terra
• Link de acesso: http://hdl.handle.net/1843/55743

Dam failure is a complex event not fully understood. The mathematical modeling of the phenomenon requires a myriad of simplifications. It is reasonable to consider that uncertainty is part of dam break studies. These uncertainties are propagated to dam breach results, used in the definition of flood risk management. It has an important socioeconomic and environmental impact due to the large number of dams constructed around the world. Uncertainty analysis in dam breach is considered a good practice. This study seeks to quantify the uncertainty of earthfill dam breach parameters applied in breach hydrograph estimates. A long database, with almost 4,000 cases and compiled by BERNARD-GARCIA & MAHDI (2020) was verified to quantify the breach parameters uncertainty. Probability distributions were fitted to breach parameters samples to quantify uncertainties. The samples were analyzed by groups. The results demonstrated that some variables affect parameters behavior. This was the case of failure mode on breach slope and dam shape at bottom width. The probability distributions identified allowed the development of a study case to quantify uncertainties in peak dam breach outflow hydrograph. So, these models can be applied in future researches about this subject and in a construction of probabilistic inundation maps. Different probability distributions were identified. These were applied in a probabilistic dam breach modeling of the case study developed in ZENZ & GOLDGRUBER (2013). The Monte Carlo Method was used and the breach parameters were considered as random variables. The results allowed to quantify the uncertainty in peak of the breach hydrograph. The results of probabilistic analysis were used to quantify uncertainties in deterministic dam breach modelling being observed that this models are, in majority, conservatives.