A Comprehensive Global Database of Mine Tailings Flows

Tailings are finely ground, often toxic mine waste rock produced from mineral ore processing that are conventionally discharged as slurries into dammed storage facilities. Given that tailings disposal is necessary insofar as mining operations continue, tailings dams are perpetually evolving structures that are raised in stages over the span of several years to decades, with some impoundments being among the most voluminous anthropogenic structures in the world. The failures of some tailings impoundments have received considerable public attention in recent years due to the catastrophic human and environmental consequences caused by mass movements, as evidenced by the 2019 Feijão disaster in Brazil (272 deaths) and numerous precedents over the past century. Developing an inventory of these mass movements - or tailings flows - is an essential step to understanding the (i) magnitude-frequency statistics, (ii) internal and external preconditioning and trigger variables, (iii) breach-outflow mechanisms and (iv) downstream inundation geometry and geomorphic processes. Moreover, constructing a global database helps establish a consistent framework for reporting various properties of tailings dam failures and ensuing mass flows. Upon reviewing over 350 historical mine waste dam failures in pre-existing secondary datasets, we have developed a comprehensive database of 63 tailings flows from 1928-2020 (while following a strict case selection criteria) with the support of compiled satellite imagery, digital elevation models (DEMs) and source literature. Using a novel runout zonation method (developed by Ghahramani et al., 2020), the satellite images and DEMs were analyzed on geographic information systems (GIS) platforms to independently estimate runout distances, inundation areas and travel path angles. Contingent upon data availability or quality, we also summarized the background information, impoundment conditions and hydro-geotechnical indices to provide site-specific context to case histories. It is anticipated that the compiled data and review will broaden the scholarly understanding of breach-runout behaviour and enhance hazard-risk assessment strategies including empirical and numerical modelling. However, the compiled data must be treated with prudence. Tailings impoundments are extremely variable depending on the locality, and site-specific conditions (internal and external) are strong controls on post-breach behaviour. Even though the presented dataset is comprehensive, large-scale statistical approaches may not satisfactorily reflect the "true" site-specific field behaviour. As such, it is recommended that this dataset be used purely as a basis for screening-level assessments, case analog comparisons and academic research. For site-specific prediction studies undertaken by practitioners, targeted field observations, laboratory investigations and numerical models are essential.