Recent failure at Pau Branco Mine, MG, Brazil
Jan 13th, 2022
The recent failure at Pau Branco Mine, MG, Brazil is a classic example of cascading events. At this Vallourec Mineração (VMN) iron ore mine the failure of an uphill structure/slope led to overtopping of a downstream retention structure. As a result a flood reached a highway and the environment.

Photograph provided by: Adolfo Tribst Corrêa
A bit of Pau Branco innovation history
Back in 2015 this mine was touted as a technological success when a new filter press module was installed. This module would maximize Pau Branco Mine sustainability. Indeed, instead of heightening the upstream Cachoeirinha Waste Dam, an in-house technical team proposed Filter Press. This was intended to “eliminate the stability risks associated to the traditional method”. Filter Press would withdraw water from tailings and allow for piling the drained waste eliminating the need for dam disposal.
The solution was considered as a panacea and here are a few comments made at the time:
- “…breaking the technical-finance feasibility paradigm of iron ore waste filtering compared to conventional dams use.”
- “The sustainability bias of the solution deserves even more prominence since this practice avoids the set up of new dams and heightening of the existing dams.”
- “This implementation directly benefits society, the environment, and the Company itself, as it ensures operational continuity and business feasibility based on a highly sustainable process”.
Technical papers also exist on the subject, for instance Gomes et al. (2016) paper: “The mining industry has seen several significant dam failures in recent years. Dam failures are associated with errors in design, implementation, operation, and monitoring (Azam, 2014). Dewatered stockpiling (dry stacking) is a safer alternative to tailings dams (Rico et al., 2008) for tailings disposal.”
What was missing?
Overall design of slopes apparently missed to identify natural hazards, long term view and residual risks. Technological innovation provoked generalized blindness. The mine even won the Brazilian Environmental Award in 2017.
However, any slope, anywhere and in any material has potential stability and drainage issues that one should carefully consider.
As a result, one cannot be overly optimistic as zero risk does not exist.
When reviewing the risks of such a project at another location a few years ago, we detected this hazardous “design euphoria”.
Of course, the client was not happy…
Business as usual vs. divergent events
Now, extreme rains generated the Pau Branco accident. But how extreme was this extreme? In Extreme rainfall and its impacts in the Brazilian Minas Gerais state in January 2020: Can we blame climate change? for example we read that despite its dreadful impacts on society, an event with the magnitude of the 2020 rains has a return period of ∼4 years. Incidentally, we were in Brazil for a risk assessment during the 2020 event… and seemed indeed “extreme” to us.
So, one could even argue that this type of extreme precipitation is the new “business as usual” . Neither an “Act of God”, nor the often cited “Black Swan”.
In our latest book we demonstrate how we can use risk assessment in the context of climate change and divergent events.
What was the design criteria for the slope uphill of the pond?
Lessons learned with respect to TCFD and TNFD
So, we have in an interesting scenario:
- communication to investors, regulators and the public is: innovative solution abates/eliminates risks.
- The reality is: solution may be missing some very important hazards at the site and residual risks are material.
Five years ago we were saying that NI43-101 should disclose operational risks.
Task Force on Climate-Related Financial Disclosures (TCFD) offer now and Taskforce on Nature-related Financial Disclosures (TNFD) will soon offer a framework of disclosure recommendations aimed at Investors, lenders, insurance underwriters etc. Risks should find their way into annual filings.
As a result, TCFD and TNFD industry has to base on “real” risk assessments and their quantitative results, so that the disclosures are somewhat representative of present and future conditions.
If we do not, we will see the mining industry lose credibility.
Tagged with: black swan, complex cascading events, decision, operational, risk, Risk Assessment, tailings
Category: Consequences, Hazard, Probabilities, Risk analysis, Risk management
Leave a Reply