first steps ORE2 tailings workflow

First steps ORE2 tailings workflow: 2 – ARCHIVAL DOCUMENT SEARCH METHODOLOGY

The archival documents delivered by the client to Riskope should cover the TSFs and their dams including:

• geotechnical studies;
• geological studies;
• design reports;
• geotechnical analyses reports;
• inspections;
• correspondence;
• incidents descriptions;
• dam break analyses; FIA and GISTM consequences evaluations;
• emergency plans; and finally
• monitoring (equipment and measurements as applicable).

During the deployment we deliver an analysis of the archival gaps, i.e. a list of documents that are apparently missing or incomplete (Take away #1).

Semi-automated discovery platform

The semi-automated ORE2_Tailings™ discovery platform is a software accepting as input MS word or PDF files, images and emails. It is capable of Optical Character Recognition (OCR). After a preliminary read of a sample of the files by experienced analysts, a list of keywords is edited in order to ensure that the “language” used by the reports under scrutiny is covered. The list contains dam/slope assessment specific words such as: dam/slope, settlement, fill, buttress, deformation, decant, tower, erosion, piping, liquefaction and the software will search all the documents for those keywords, in singular and plural.

A check is then performed to see if the list of keywords has deficiencies:

• if deficiencies root into “local language” differences (e.g. Glory hole instead of decant, etc.), we add words to Riskope standard world list and re-run the scan;
• if real deficiencies exist, we contact the client to ask for possible additional “forgotten” reports that deal with the missing words;

If we receive new reports the cycle is repeated until we obtain a satisfactory result and finally, we deliver the archival gap analysis constituting Take away #1 of the study.

Different languages in the portfolio do not cause great difficulties provided the analysts are fluent in each language. Even the orthography (for instance the French accents on the letter “e” can be solved by using the “dummy character”).

We deliver the results of the semi-automated ORE2_Tailings™ discovery platform in “structured form”. Structured form means the results are already structured by document/author/themes/search strings.

We then use the structured results in Chapter 3, Portfolio description and Chapter 4, Knowledge base results. As a result, these feed the evaluation of the ORE2_Tailings™ KPIs/KRIs we use to evaluate the probability of failure p_f the consequences of potential failures.

First steps ORE2 tailings workflow: 3 – PORTFOLIO DESCRIPTION

Based on preliminary discussions and the analysis of archival information this chapter describes the TSF portfolio and each dam or homogeneous section of dam.

In this section we describe all the dams of the portfolio in general terms (see Table 2), i.e.:

• activity, status;
• year of commissioning (start of service);
• year of end of spigotting (if applicable);
• history of accidents/incidents and finally
• comments on work performed to date.

Knowledge base results (4)

This Chapter delivers the bulk of the rich-data information we gather using the Riskope archival document search methodology in terms of general knowledge.

It includes at first a commented (as applicable) qualitative judgement on the available information. This covers the headers ORE2_Tailings™ KPIs/KRIs, i.e:

• General;
• Storage;
• Ancillary water management;
• Tailings lines and traffic;
• Erosion controls;
• Liquefaction, piping and leaching;
• Investigations and geological model;
• Project and construction;
• As is status;
• Monitoring and inspections ad finally
• Consequences

Note that two “identically design and built” dams could have different values because, for instance, one has active erosion on the D/S face, and the other has no working inclinometers, etc. Reciprocally two different dams can have the same “note” due to very different reasons. That is why, we will discuss the quality of the dams further and deliver a potential failure causality analysis. The chapter closes with a gap analysis on archival documents (Take away #2).

Tagged with: archival document search, decision, GISTM, management, Optical Character Recognition, risk, Risk Assessment

Category: Consequences, Hazard, Mitigations, Optimum Risk Estimates, ORE2_Tailings, Risk analysis, Risk management, Tolerance/Acceptability