4.2 A Systemic View on Tailings Dams Failure Processes
43
• Slope Instability;
• Earthquake and Mine Subsidence;
• Over-topping;
• Foundation;
• Seepage and Structural.
The Data
Data on tailings failure is scarce, sometimes tainted by biases and censorship, and
spread through various entities and databases of variable reliability (for example,
notice in Fig. 4.4 the very large number of “unknown” causes). In response to this
we adopted a “quick and dirty” engineering approach to estimates, preferring to
rapidly gain an understanding for the order of magnitude of the estimate rather than
waiting to get very precise “true” numbers. We saluted the actuarial effort published
in 2015 (Bowker and Chambers 2015) and were delighted to notice that our previous
estimates were in good agreement with the more precise numbers, although we
commented on some unfortunately “forced” linear regressions drafted by various
authors and to the tendency to use variable time intervals to jump to conclusions.
In this section we ensure coherence with our earlier, now proven correct, “quick
and dirty” engineering approach, but decided to also include uncertainties by using
two different sets of causal lists, namely those resulting from ICOLD 1994 and those
from a 1910–2009 compilation (Azam and Li 2010).
Failures Reported by ICOLD 1994
For the sake of this discussion it was necessary to re-interpret the data in the literature. Readily available records generally report number of failures versus cause
of failure (Figs. 4.4, 4.5) and are fraught by many “unknown causes” or statements
like “unusual weather” which allow plenty of room for conjecture. This discussion makes it necessary to attribute causality of the failures to the various phases
Unusual Weather
Management
FoundaƟon
Subsidence
Slope stability
Over-topping
Seepage
Structural Defects
0
5
10
15
20
25
30
35
40
45
50
Recorded Cause of Failure
Recorded Failures Number
Fig. 4.5 Dam failures versus cause of failure from various sources (Azam and Li 2010)
43
• Slope Instability;
• Earthquake and Mine Subsidence;
• Over-topping;
• Foundation;
• Seepage and Structural.
The Data
Data on tailings failure is scarce, sometimes tainted by biases and censorship, and
spread through various entities and databases of variable reliability (for example,
notice in Fig. 4.4 the very large number of “unknown” causes). In response to this
we adopted a “quick and dirty” engineering approach to estimates, preferring to
rapidly gain an understanding for the order of magnitude of the estimate rather than
waiting to get very precise “true” numbers. We saluted the actuarial effort published
in 2015 (Bowker and Chambers 2015) and were delighted to notice that our previous
estimates were in good agreement with the more precise numbers, although we
commented on some unfortunately “forced” linear regressions drafted by various
authors and to the tendency to use variable time intervals to jump to conclusions.
In this section we ensure coherence with our earlier, now proven correct, “quick
and dirty” engineering approach, but decided to also include uncertainties by using
two different sets of causal lists, namely those resulting from ICOLD 1994 and those
from a 1910–2009 compilation (Azam and Li 2010).
Failures Reported by ICOLD 1994
For the sake of this discussion it was necessary to re-interpret the data in the literature. Readily available records generally report number of failures versus cause
of failure (Figs. 4.4, 4.5) and are fraught by many “unknown causes” or statements
like “unusual weather” which allow plenty of room for conjecture. This discussion makes it necessary to attribute causality of the failures to the various phases
Unusual Weather
Management
FoundaƟon
Subsidence
Slope stability
Over-topping
Seepage
Structural Defects
0
5
10
15
20
25
30
35
40
45
50
Recorded Cause of Failure
Recorded Failures Number
Fig. 4.5 Dam failures versus cause of failure from various sources (Azam and Li 2010)