Advances in non-Newtonian dam break studies

Abstract

Tailings dam failures have and continue to cause large-scale devastation and environmental impacts. Historically these impacts have largely been predicted using Newtonian hydrodynamic modelling principles resulting in a general overestimation of potential consequences. However, since about 2014 the collective mining industry began developing complex tools to better predict the likely impacts of these failures. The industry has focused on leveraging the latest in computational flow dynamics modelling software and computational hardware to preform non-Newtonian tailings dam break assessments. However, as our tools become more sophisticated so does the requirement on input data. This paper discusses the past modelling approaches and the development of non-Newtonian tailings dam break models. The sensitivity of the flow behaviour is presented through four case studies, showing how this selection influences the outcomes and how previous approaches assuming Newtonian characteristics may present an overly conservative result. It is noted that additional knowledge and expertise will become available as non-Newtonian tailings dam break studies become the norm. In the interim, the uncertainty of these analysis needs to be analysed.