Retroanálise da ruptura por liquefação de barragens alteadas a montante

Abstract

The use of mining tailings disposal in a wet environment, such as in dams, has been widely used in the history of mining in Minas Gerais, and the presence of these structures generated the need for the creation of a national legislation proper to the knowledge of their safety. From Law No. 12334 (2010), the National Dam Safety Policy, until its update by Law No. 14066 (2020), there was a decade of application and proposals for improvements, especially after the disruptions in Mariana (2015) and Brumadinho (2019), where the Dam Safety Plans - PSB started to require more rigorous studies of dam failure, or hypothetical failure, primarily in the worst situations. In this context, this present work aimed to design a study of dam breakage in retro-analysis of these two cases, based on data taken from original management and audit documents of these dams. The hydrographs were created using the Hec-HMS® 4.7.1 software, for various methods of measuring the volume mobilized at the rupture, obtaining a maximum flow rate for Mariana, of 22,082.8 m³/s for 32 Mm³ of mobilized tailings. For Brumadinho, it was found a maximum flow of 27,178.9 m³/s, for 9.7 Mm³ dispersed. Flood spots were overcome by the Hec-RAS® 6.0 software, based on the digital elevation models SRTM (30 m resolution) and Alos Palsar (12.5 m resolution), obtaining a better measurement of hydrodynamic results and greater fidelity to the real situation by the 12.5 m model in both case studies, in the measurements of the wave arrival time, the velocity and the maximum flow depth reached, especially in the risk simulation. Finally, it was concluded that there were limitations with the use of Hec-HMS® due to the software's rupture mode only contemplating failures due to piping and overtopping, but satisfactory results were obtained using the “piping” mode. The use of Hec-RAS® with the new module for non-newtonian fluids has been shown to be very satisfactory, requiring few adjustments at the expense of the previous Newtonian module, which needs adjustments for calibration. The risk zones were well delimited when compared to the available technical data and engineering interpretation concepts, corroborating the use of these techniques with free and free material, for studies of dam failure