Zero to 300 in thirty
Jun 8th, 2022
Zero to 300 in thirty is not the performance of a new sensational sport car. And we are talking about thirty years, not seconds!
Actually, zero to 300 in thirty years is the cumulative experience Riskope and its founders have dealing with tailings dams risk assessments all over the world.
You should not understand it as an “average” evaluation either, as the evolution has been trending towards exponential.
Let’s discover together what we are talking about.
Zero to 300 in thirty years
That experience started indeed thirty years ago, when the company was still a branch office of the original predecessor, founded in Switzerland. At that time Franco had already extensively published in the field of probabilistic analyses of natural and man-made slopes. He had completed his Ph.D. under the mentoring of Prof. M. E. Harr (Purdue University Geotechnical Engineering – Lyles School of Civil Engineering – Purdue University).
The company was working for:
- hydropower companies concerned by their infrastructures being subject to Alpine landslides,
- transportation corridors in difficult geotechnical conditions,
- communities and infrastructures exposed to rockfalls and finally
- environmental rehabilitations.
We gathered experience in the Middle East, the Philippines, Algeria, Gabon, Madagascar.
In 1992 a client in Vancouver called us in to evaluate rockfall potential and mitigation for a large development on the north shore. That was, as far as we know, the first rockfall mitigation project for residential houses in Canada, and likely in North America.
Shortly after that first job we started a cooperation with Iain Bruce and Wayne Savigny (the founders of BGC ). We brought to the team our knowledge of probabilistic approaches. We adapted methodologies we had developed in Europe for railroad, including Algerian and trans-Gabonese railroads, to north American clients (CN , CP , UP ). Then it was the turn of mining.
First experiences in mining
At that time we also completed a “first in the world” designing the environmental rehabilitation of an asbestos mine in Italy using probabilistic approaches and criteria. Actually, not long ago we were surprised by a request for information by a large US engineering office who told us that what we did at that mine in Italy was still unique.
We also collaborated more and more with Claudio Angelino on many quantitative risk assessments for landslides in the alps and related multi-hazard civil protection databases.
Then it was a frenzy of new applications we developed for our friends at BGC in the mining field:
- Placer Dome in Canada and US,
- Australia, US for Homestake, then
- Antamina, and finally
- Codelco, etc.
At that time we were using FMEAs and FTAs, ETAs, so we were following common practices. However, Riskope’s own clients were asking for more! They had questions that “standard approaches” could not answer.
That’s when Riskope’s R&D effort really kicked-in, supported already by over ten-years experience in quantitative risk assessments in a variety of industries.
Diverging from “cosi fan tutti”
From Central and Northern Canada to Latin America we started using symptom-based approaches to evaluate probabilities. These stemmed from studies we completed for road network owners (Italy) and the GICHD for unexploded ordnance and landmines (Cambodia, Laos PDR), as well as large insurers.
Right after the 2008 crash we introduced “convergent” ERM dashboard to a Major Canadian company, of course at the time we did not formalized it as “convergent” (our latest book delves into convergence ).
Then, in 2010 we formally introduced to a Canadian mining company the concept of risk tolerance. We already compared single dams risks against tolerance corporate and societal tolerance. Subsequently came another epiphany! Indeed, delivering “probabilities” as numbers was not enough. We needed to benchmark risks of a dam to a world-wide threshold. Thus, in 2013 we compared the world-wide performance of tailings dams to other industries and created benchmarks. In order to do that we used for the sake of that paper a methodology (Silva Lambe and Marr (2008) – Probability and Risk of Slope Failure | PDF | Risk | Geotechnical Engineering (scribd.com) ) that applies to slopes only, certainly not dams! Unfortunately, we have seen a number of instances where engineers are using these type of approach for “any dam”, and we feel it is time to formulate a clear warning.
Unless a dam can be “reduced” to a simple slope, meaning it has no active distribution line at the crest, it is not prone to liquefaction and is missing many other tailings dams usual features, using a method similar to SLM leads to potentially fatal underestimates of the probability of failure.
In our 2019-2020 tailings management book (https://link.springer.com/book/10.1007/978-3-030-19447-5 ) we showed the effect of misusing SLM and similar approaches. Indeed, a dam has to be considered as a system made of the dam itself and all it ancillary water management infrastructure. As a result, tailings distribution pipes, weirs, spillways, access roads, and their effect cannot be “lumped-up and oversimplified” by symptoms. In particular, ancillary water management facilities have design criteria and their status of maintenance defines their “as is” condition and related probabilities of failure.
Testing, calibrating, proving ORE2_Tailings
ORE2_Tailings™ was born thanks to the requests of our clients.
We also understood that the large data available for each tailings dam needed to be processed in order to be efficient. Indeed, simply reading all the archives was not sufficient. Hence we started our collaboration with a specialized legal solutions company, Epiq.
With each new job, each new tailings system we refined the methodology. We had the opportunity to test, calibrate and prove ORE2_Tailings™ results all over the world.
Needless to say we were delighted when ICMM published GISTM (Global Industry Standard on Tailings Management – Global Tailings Review) shortly after we published our book. As we have expressed in writing many times, GISTM has various points that could be enhanced and possibly even modified. However, it has created a solid based for discussion, world-wide.
And that base for discussion needs feeding with more than nice words. More than ever, our clients want to reinforce their position with numbers, while they remain very aware of the uncertainties.
We also recognized that climate change is also altering the future state of a structure and ancillary infrastructure, hence we started cooperating with Correntics to quantify these changes (Gloor, M., Halter, G., Oboni C., Oboni, F., Impact of climate change projections on tailings dams survivability, CIM Climate Change and Tailings Management, CIM2022, Vancouver Canada, May 2, 2022).
Indeed, those numbers bring value and comfort to GISTM compliance, the definition of ALARP mitigation level (Oboni C., Oboni, F., Tailings Dam Risk Mitigation Through Risk Informed Decision Making, PLANNING FOR CLOSURE 2022, Santiago Chile, May 11, 2022), future TCFD/TNFD disclosures (Oboni F., Oboni, C., TCFD and TNFD for mining: concepts, examples and caveats, CIM Mineral Economics and Finance, CIM2022, Vancouver Canada, May 4, 2022), etc.
We close this historic review with a few pointers that we consider paramount in tailings dams risk assessment and the selection of the approach.
- How does the approach analyzes the history of the considered system? Who does it? On which criteria? Avoid introducing biases in the knowledge base. Also, pay attention to attaining a homogeneous level of knowledge across your dams portfolio.
- When performing any type of risk assessment, pay attention to avoid oversimplification, abusive lumping-up and using “snake-oil” solutions.
- How are uncertainties included in the analysis? Are they explicit?
- Beware of “solutions” claiming to “simplify” and allowing “self assessment”.
- Can the approach deliver a defensible quantifiable ALARP definition?
- Ask how many hundreds of dams the approach evaluated before “yours”? Ask what are the failure criteria?
- Check about liquefaction, seismic loadings, external loadings and particular features, etc. inclusion. The list is long and you should not accept “trust me” replies. Finally,
tailings dams are complex systems that come in infinite flavors. Pigeonholing solutions based on oversimplified approaches are hazardous. As a result, third party review by analyst is paramount.
Tagged with: decision, management, risk, Risk Assessment, Risk Management
Category: Hazard, Mitigations, Optimum Risk Estimates, Probabilities, Probability Impact Graphs, Risk analysis, Risk management, Tolerance/Acceptability, Uncategorized