Abstract
Dam break assessment (DBA) is a widely used method to estimate the potential consequences of an eventual failure of a tailings dam on the downstream affected area, including inundation extent, characteristics, and flood arrival times.
Performing it requires defining a series of input parameters, among which release volume and peak flows are as crucial as indeterminate. As opposed to water reservoir failures in which total impoundment is discharged, tailings properties and breach characteristics impact the flow properties and the amount of tailings mobilized, thus generating epistemic uncertainties on the estimation of consequences.
Despite this, the current general approach for tailings DBAs still considers the effects of a failure caused by a single deterministic release volume. This leads to any subsequent emergency preparedness and response plan (EPRP) and risk-managed decisions to be subject to results of an analysis that has large uncertainty associated to it.
In this presentation, the authors proposed a probabilistic approach to DBAs that allows mitigating uncertainties by incorporating a level of variability for a series of key parameters affecting the failure characteristics, and a Montecarlo simulation of scenarios. In the opinion of the authors, this approach provides tools for better developing EPRPs, and may potentially enhance risk-based decision making.
Furthermore, an understanding of how incremental consequences are associated to a decreasing probability of occurrence may help defining limits for an adequate level of risk tolerance based on an ‘as low as reasonably practical’ (ALARP) perspective.
This approach was applied in mock case study to showcase the scenarios that are typically ignored with the single-deterministic approach. This example is intended to illustrate that the proposed procedure can be used to assess the potential incremental consequences associated to increasingly improbable scenarios and defining limits for ALARP.
