TAILINGS DAM FAILURES: A STATISTICAL DATA ANALYSIS

Abstract

Minerals processing produces two main products: a commercial product (concentrate, pellets) 
and tailings (waste from ore processing). The tailings are stored in a tailings storage facility (TSF), a 
special hydraulic structure for receiving and storing solid and liquid mineral processing waste. The 
focus  is  primarily  on  ensuring  the  reliability  and  safety  of  these  facilities  during  their  design, 
construction, and operation, as a failure of such structures can have catastrophic consequences. This 
paper computes the statistics of tailings dam failures using an up-to-date database on failures. The 
retrospective review covers the period from 1915 to 2024. During this time, about 373 cases of tailings 
dam failures were recorded. The historical trend regarding the number of failures has an average of 
3.42  failures  per  year.  The  majority  of  failures  occurred  between  1960  and  2000  (230  failures), 
accounting for approximately 62% of the total number of accidents. In the 2000s, a decrease in the 
number of incidents compared to previous decades was observed, while after 2010, there is a tendency 
for  an  increase  in  the  number  of  incidents  (5.9  failures  per  year).  The  accidents  in  tailings  dam 
breakages occur mostly in North America (41%). Statistical data analysis shows that the main causes 
of failures are overtopping (21%), slope instability (16%) and seismic instability (15%). The analyses 
of the distribution of the world’s tailings dam failures with regard to dam height show that 90% of 
the cases occurred in dams less than 50m in height and only 2% of incidents in dams higher than 100 
m. The correlations between stored and released volumes have been verified using a database. Based 
on the relationship found from the regression analysis, as the tailings volume increases by 1%, the 
release volume increases by 0.25%. Statistical data analysis of tailings dam failures allowed for the 
identification of the most vulnerable and least reliable elements of the structures of TSF, which will 
allow them to be used in the probability analysis of accidents at these structures.