This is an example of limitations, disclaimers and risks. Below is the summary of the case.
Like all other consultants, we emphasize that our reports address a specific project or situation, thus no one should generalize them to other operations, even if they appear similar.
In particular, the verbiage aims at protecting the client from the temptation to “run away” with a piece of intellectual property (IP). That is because if poorly understood and then poorly generalized, that piece of IP may lead the client and the public to unwanted exposures.
Thus, it is with deep consternation that we heard of a client explaining to an international audience a generalization to an entire portfolio of mines from a specific report. Out of professional courtesy and care, we immediately warned the client of numerous omissions we immediately spotted. We urged them to be cautious while congratulating them for “their idea”.
Of course, we were at first angered because they did not reference the source of the IP. However, we then decided it was better they did not mention Riskope. Indeed, we do not want to be associated with such a flawed approach.
Risk is a Job for Specialists
The numerous omissions we spotted show a fact many have recognized as we discuss at the end of this text.
Even well-respected and highly experienced engineers cannot improvise and develop quantitative risk assessments. Indeed, there is a large volume of knowledge that has to be acquired before anyone can claim to be able to perform a sensible risk assessment on a tailings facility.
It is interesting to see that, for the time being, the new Global Industry Standard on Tailings Management (GISTM) explains the role and qualifications of key personnel. However, and very unfortunately, it fails to explain who should perform a risk assessment and what a risk assessment should deliver. We will discuss this in more detail soon, although we know that Professor Norbert Morgenstern is reportedly working on filling this gap.
For the time being, we will note that the flawed approach developed by our client does indeed deliver results that could be accepted by the new code while being extremely misleading.
Example of Risk Assessment Flaws We Have Seen Over the Years
1. At Riskope, we do not think it is ethical to evaluate probabilities below credibility (10-6) using a semi-empirical method like Silva Lambe Marr (2008). Thus, we impose a cap, even if the engineers declare very high Limit Equilibrium Method (LEM)−derived Factors of Safety (FoS).
2. We also note that, for example, a number of Class 5+ nuclear accidents have occurred for reasons not included in very detailed analyses. Examples of these analyses are the Walsh report and French nuclear safety studies in the mid-70s. Therefore, the factual rate of Class 5+ nuclear accidents is several orders of magnitude higher than the fault tree/event tree−derived theoretical values (Oboni & Oboni 2013).
3. Incidentally, we also have concerns in accepting engineers evaluating FoS>2 for any geo-structure, because of LEM intrinsic limitations.
4. Silva Lamb Marr developed their approach for slopes, not necessarily for dams. Thus, it does not include liquefaction considerations, nor potential hazards from tailings distribution lines, etc. As a result, its application to dams is very limited. We have used the methodology in some specific studies for clients, including examples in our book. However, we added a discussion of the effect of liquefaction on the overall probability.
5. We cannot imagine characterizing a dam by a single FoS. A dam can indeed succumb to many hazards.
A Couple Philosophical Thoughts
At Riskope, we adhere to the ideas in Ralph Peck’s 1980 lecture “Where has all the judgment gone?” (Canadian Geotechnical Journal, 17(4), pp. 584−590). Peck’s ideas are related to failure modes. They were confirmed by the most recent post-mortem analyses of the Brazilian and Canadian accidents. They also cover oversights versus theoretical analyses. Please note that Peck was talking about hydro-dams; today, well-designed hydro-dams have likely gone one order of probability of failure (PoF) magnitude down (10-5) with respect to the 10-4 Peck quotes as “historic” in 1980. Tailings dams are a different story, of course.
Regarding about the oversight and excessive reliance on analyses, at Riskope we would not consider a loading case as “completely not credible,” particularly in the case of a known defect in a geological formation the engineers may want to discard as not credible. What about possible pressurizing a pre-sheared foundation zone, for example? There are good examples of this in dam failure history.
Closing Remarks on Limitations, Disclaimers and Risks
Of course, this is just the tip of the iceberg.
As Terry Eldrige said in his keynote lecture at Tailings and Mine Waste 2019, risk is a job for specialists, and cannot be improvised. See in particular:
- 4:38−5:36 Success requires avoiding many separate causes of failure; failure mechanism and cause may not be the same; failure caused by complex factors.
- 25:22−26:10 Shortcomings of engineers performing risk assessments.
We would be delighted to discuss these topics with you. Please contact us.
