We use archival information to formulate the ORE2_Tailings™ quality note and causality analysis. This includes site visits, when possible, and photographic historical documentation. Information will evolve over time as new data becomes available. Thus, one can use ORE2_Tailings results to decide which aspects of the considered structure to study.
ORE2_Tailings results encompass various dam body probabilities. They correspond to the factors of safety (FoS) yielded by available engineering analyses and the causality analysis. Here, a link between ORE2_Tailings and conventional failure modes analyses is possible. If available, the effective stress, undrained strength and pseudostatic (with a return period) FoS are transformed into an annualized probability of failure using a proprietary algorithm. Thus, static drained conditions, liquefaction/residual strength and rapid drawdown modes of failure can be simulated. If the client has developed event trees, failure trees, bowties, etc., we can compare the results and discuss potential significant differences.
The ancillary water management (weirs, spillways, decants, etc.) also contribute to the annualized probability of failure by their design and possible deficiencies. This is similar to what an overtopping, scouring, erosion, etc. failure mode analysis would consider.
Water Management
Water management is a complex issue. ORE2_Tailings deals with it in a systematic manner by looking at the various internal and external elements of the water management subsystem, including external water courses. Each element is characterized by a design criteria defined in the project files and qualitative state of the system (e.g. excellent, fair, poor) based on known/perceived near-misses.
Their combination leads to the likelihood of a water management failure generated by external and/or internal system element failures.
The vulnerability of the structure to erosion, in the case of overtopping, leads to modifying the estimated likelihood of failure. This covers point (c) of the failure definition explained in Section 5.1 of the report.
Combining the Probabilities
At the end of the ORE2_Tailings evaluation, two families of results are available:
1. the probability of failure pfA of the dam body alone (without considering water management) which is a result of the:
- level of knowledge
- quality of the structure
- result of the engineering analyses (stability, seismic, liquefaction, etc.)
2. the probability of the water management system failure which we combine with pfA in order to evaluate the probability pfB of the system as is.
The probability pfA is a rather theoretical value. Indeed, it would correspond to an ideal state where all water management issues have been mitigated to a probability level lower than the dam itself. For example, if a dam body had a pfA=1.0E-05, one would need to mitigate the water management system to 1.0E-05 or lower for A to be valid. As water management generally has a pf=1.0E-03 or higher, it becomes self-evident that the system “as is” will generally obey to pfB>pfA.
6.1 Benchmarking of Probabilities of Failure (Takeaway #4)
The probabilities of failure (pfA, pfB) will be displayed alongside the worldwide benchmark for catastrophic failures for each dam/slope. Riskope has benchmarked dam hazards since 2013. Hazard benchmarking is based on the probability of failure regardless of the consequences. Indeed, this significantly differs from a full-fledged risk prioritization that will be developed in future posts.
We discussed this theme in detail in a blogpost called “Dam portfolio ORE2_Tailings support for decision-makers.” There, we clearly show that a rating based on just probabilities is not sufficient for rational decision-making.
Takeaway #4: The benchmarking exercise
6.2 First Rating of Risks (Takeaway #5)
We deliver this rating for the sake of completeness. However, it is generally not valid for risk-informed decision-making. That is because multiplying the probability of failure by the consequences leads to identical values if the probability is very large and the consequences small, or vice versa. That corresponds to neutral risk aversion, but we know this is the wrong approach for tailings.
This explains why common-practice−informed matrices that attribute similar or identical colors to the cells (pmax, Cmin and pmin, Cmax) are misleading. Simultaneously, one needs to apply a proper risk tolerance threshold.
Takeaway #5: The decreasing risks ranking.
7 Risk Tolerance Thresholds
In order to allow meaningful comparison, one needs to maintain a constant failure definition, consequence metric and risk tolerance throughout the entire mine/TSF/dam portfolio. We can assist the client in developing their own risk tolerance threshold. This is normally the object of a separate mandate.
8 Risk Prioritization
In this section we generally combine the risks from Chapter 6 with the tolerance example we gave in Chapter 7 to deliver a risk-informed prioritization.
Takeaway #6: Risks ranking in terms of risk tolerance level
Takeaway #7: Risk aggregation and interdependencies of the portfolio
