All in the 4th estate have done an absolutely fantastic job unpacking the bits and pieces that are now in the public domain and trying to say the right things about them. Our work here at World Mine Tailings Failures is dedicated to compiling and vetting all the relevant information, having it presented and interpreted by experts as to cause of failure, early signs of failure, consequence of failure and what information was available pre failure that could have prevented the loss.
We have been working non stop translating, reading everything available to us which may or may not include whatever document Reuters is referring to. Reuters has not sent us that document. In my large sphere of global experts, none of them have seen the document Reuters is referring to.
What we can say so far is there is no question that the facility was in fact in a discoverable pre failure/extremely vulnerable state for sometime. What is not clear is whether that was appropriately measured and identified in what was required by law to satisfy required audit and stability reports or what was developed by Vale’s own engineers or consulting engineering staff.
What is clear in the first instance is that the government classified this dam as ” high hazard” in the event of failure using a pretty much universally used classification system intended to assess the potential “mining affected area” and the nature of adverse effects in the event of failure. A more refined severity measure developed by the Columbia Water Center puts the Brumadinho at #5 in terms of scale of impact in the event of failure. That fact(of brumadinhos hazard class and the degree of severity by Larrauri & Lall) has been sitting in my computer files for almost two years.
My impression from the bits and pieces we have is that Vale has been actively assessing the adequacy and capacity of its entire tailings portfolio for about 9 years. Mainly, it seems, they were looking for the lowest cost path to creating tailings capacity for a planned 50% output in production by 2023. The oft referenced document that may have been Reuters document and which is reported to estimate the cost of a failure at $1.5bn potential damage is likely Vale’s effort to prioritize financial risk to them in the event of failure and not a stability analysis of assessment of likelihood of failure. ( interestingly $1.5 b is what government officials demanded Vale put aside. Could be coincidence?)
We know that Vale says no depositions were made to Dam 1 for 4 years ( but aren’t clear when mineral production at Corrego d Faijao ceased).
That kind of assessment is not a risk assessment..it is not an assessment of risk of failure or present stability nor does it normally refer to specific failure conditions or problems known when the possible cost of a “worst case scenario” failure is calculated.
We are just beginning to post what we see at our web page specifically for the Corrego do Feijao Dam 1 failure. and we have formed no conclusion as to what Vale’s Board, its engineers, its consultants or the Government of Minas Gerais actually understood about the extreme vulnerability of this dam some time before its final collapse.
What would make that clear is the actual factor of safety measured for the dam over the last 10 years and an assessment of the quality of that analysis by experts. To my knowledge, the government of Minas Gerais has not made public any of what it has apparently accepted as evidence of satisfactory stability.
Whatever it may have said in Brazil’s or Minas Gerais’ law there is a universal standard that a high hazard dam should be designed and maintained to minimum factor of safety of 1.5. On the part of any engineer whether Vale’s or consultants there is a due diligence duty beyond law to know that and apply that. So far we don’t see that that happened. An engineer who holds a license and expects to keep it could never assert stability for a high hazard dam if they could not confirm a factor of safety of 1.5 or greater.
Brazils Law has since 1993 required that as a minimum standard. Both the TUV SUD 2017 and 2018 confidential report to Vale on Dam1 stability confirmed FS >1.5 in drained condition. The law then in effect in Brazil as revised 2017 required that the FS consider both drained and undrained conditions. TUV SUD in both reports found undrained FS <1.3 but asserted that the dam was not susceptible to failure by static liquefaction. These were both the public reports filed as required declarations of stability.
It is very likely , and seemingly apparent, that neither TUV SUD nor Vale’s inhouse geotechnical staff are up to date on what currently constitutes best knowledge on static liquefaction. Neither is Brazilian Law as of 2018.
I have copied all press who have been reporting on Brumadinho as a caution not to cite that Reuters claim as fact until people of authority and knowledge have vetted what that document actually was and what it means.
A 2010 thesis by Vale Engineer Washington Pirete and a later version of that in a published paper were both about using Dam1 as a case study to develop an estimate of static liquefaction trigger. Their finding of complete resilience to failure , per their trigger, is not in fact the same thing as an accepted stability analysis but it appears Vale’s technical staff and perhaps the Board itself may have believed this was a certification of stability. That is not clear.
We are posting at our web page what is authoritative and clear an describing what is not yet clear. I hope as you continue to over the Brumahindo you will check in there and see where things are.
February 15, 2019 Stonington Maine ( Revised March 15,2019)
Lindsay Newland Bowker, Executive Director
World Mine Tailings Failures
01 207 367 5145