Open Collections

Deterministic and Statistical Analysis in the Definition of Triggered Action Response Plans in Tailings Dams

Tailings and Mine Waste Conference (2023 : Vancouver, B.C.); University of British Columbia. Norman B. Keevil Institute of Mining Engineering

One of the critical aspects of the safety management of tailings dams is the definition and implementation of systematic and integrated operational processes for the observation of their structural and geotechnical behaviour, do as to act preventively in risk assessment and to be able to implement mitigation measures when and if they prove necessary. In this work we present the tools and the methodologies used in the deterministic and statistical analysis of monitoring data from piezometers and water level indicators to define the criteria for the activation of Trigger Action Response Plans (TARPs) and to assess their level, which are essential for the safe operation of tailings dams. The TARPs define, at different levels, the actions to be taken in case of observed deviation from the normal safety operation conditions of the structures, and thus mitigate the associated risks and restore normal operating conditions. The readings of the different types of instruments installed in a tailings dam (piezometers and water level indicators) are compared with reference values defined in accordance with deterministic analyses and statistical analyses of the historical data. This helps in the evaluation of the global risk level of the structure at a specific point in time. The deterministic analyses, based on the stability analysis of the structure, allow us to determine the Control Levels of each sensor and to establish their level of risk. In the implemented methodology, the criterion for the activation of the TARPs was based on a weighted average of the level of risk of each sensor, taking into consideration its location and its relative importance in the global behaviour of the dam. The statistical analysis evaluates over time the historical readings of each sensor and identifies statistically abnormal behaviours, establishing zones of coherent and incoherent readings. In the methodology implemented, different classes of Tendency Alerts were defined.

Text

2023-12-08

Attribution-NonCommercialNoDerivatives 4.0 International

http://hdl.handle.net/2429/86823

Unreviewed

http://creativecommons.org/licenses/by-nc-nd/4.0/