As a brief review, the disaster occurred in 2015 at a mine (The Samarco iron ore joint venture) owned by BHP Billiton and Vale. The collapse of a tailings dam “killed 19 people…(and) severely polluted the Rio Dose river all the way to its outlet at the Atlantic Ocean”.
I wrote an blog article on this in 2016, shortly after the dam failure, including what was known then of the cost imposition on the owners of the mine. See https://www.derekviner.com/samarcos-mine-woes-operational-risk-and-level-of-responsibility/ At that time the State had imposed damages on the owners of $5.2b, required an immediate start on rehabilitation of 500km of waterway and the companies had paid $260m in immediate compensation to victims.
The BBC news app on 27/1/2024, see https://www.bbc.co.uk/news/business-68102511 carried an item “Mining giant told to pay $9.7bn over dam disaster” for “non-material harm, such as emotional distress”. At the time of reporting, the owners of the dam had already paid out $7bn, up from a payment of $260m immediately after the disaster.
The BBC also reported (23/7/2018) that BHP face a claim by USA and by Australian shareholders, “for failing to disclose the risk of the dam’s failure to the stock market.” This is in addition to a civil lawsuit settlement of $47.5bn.
(Another tailings dam owned by Vale collapsed in 2019 in the same State, this time resulting in 270 deaths. A BBC report dated 4/2/2021 said that communities affected by this failure will receive a $7bn payout and Vale senior staff “are facing murder charges”.)
This sketch of the state of affairs gives real meaning to the significance of the Likely Worst Consequence in assessing the adequacy of control measures before a disaster occurs and the standards of care one might expect from the courts if it did occur..
My 2016 blog article (link above) included a discussion on how a formal Occurrence analysis would see the case. To understand formal analysis, see Chapters 3, 4 and 9.
There is a second aspect to this disaster that I’d like to explore in this update.
Following the 2015 disaster, I was speaking to a senior mining engineer who told me that he and his colleagues around the world felt ‘let down’ by ‘safety’ people on whose advice they had been relying. He had never heard of the idea that ‘general safety’ should be separated from ‘operational safety’. It appears I am a lone voice in drawing attention to this. I have written two articles about this in on my web site: see the commentary for Chapter 5 of my 2015 text.
It is easy for anyone with experience in industry to imagine a typical ‘safety’ process that may have been involved. However, imagination is not fact and none of what follows is intended to criticise those engineers and managers who were involved in the actual cases. Let’s go with my imagination, in order to illustrate something of a likely safety method and how it might have played out in this case.
Imaginary image 1: The site has a risk register, within which the risk itself is likely to be described in a rather unclear manner; subjective words on a risk matrix are used to characterise the “inherent” and “residual” risks.
Imaginary image 2: Subjective assessments of these risks are made that reflect the fact that the engineers who have designed the dam feel their design is a good one.
Imaginary image 3: The engineers involved have not received formal education in the science of Risk, which includes how to define a Risk and what Mechanism and Outcome analysis is.
Imaginary image 4: Neither the Risk Management department nor the Safety Department have a clear idea of how to define Risk, what inherent and residual mean in this (if anything) and of the literature pointing out the problems associated with use of the risk matrix.
Of course, we don’t miss what we are not aware of and it is easy to imagine the busy-ness surrounding the project meetings obscuring a clear understanding of the risk: the risk and safety people feel that the engineers have made good use of the risk register and the engineers feel that the risk register has helped them in a meaningful way to reflect on the risks of the mine, including the tailings dam.
In the absence of an awareness of formal damage process analysis it is easy to understand how the matter of a preferred Outcome pathway might have been neglected, as it appears to have been.