A recently completed ACARP (Australian Coal Association Research Program) project, “Quantification of fault uncertainty and risk management in underground longwall mining”, has developed a model aimed at quantifying these risks.
Professor Roussos Dimitrakopoulos, WH Bryan Mining Geology Research centre, University of Queensland, was the ACARP project leader. The findings of the ACARP project demonstrate that any risk assessment based on an exploration dataset "as is" can seriously underestimate risk and should be avoided, according to the final report.
“It should also be noted that traditional longwall design and planning based on exploration information only implicitly assumes and reflects such risk underestimation,” according to the report.
Dimitrakopoulos said a key element of the research project was the introduction of risk modelling that is by nature quantitative and in turn generates a different way of thinking and problem solving compared to traditional approaches. Furthermore, such an approach added flexibility because the uncertainty was explicitly quantified.
The project developed a new fault simulation algorithm based on power-law model for fault size distributions and length versus maximum throw relationships. Case studies from Moranbah North, Goonyella-Riverside and Newlands demonstrate the application of the methods developed in three different cases.
Benefits from initial application have delivered benefits including the use of fault probability maps to optimise longwall designs based on maximizing mineable coal reserves, and categorisation of coal reserves based on geological risk. Another benefit was the ability to combine quantified geological risk with different longwall layouts in order to assess and minimise fault impact on mine economics.
A key project deliverable is the development of a computerised model named GeoStructure, available to all ACARP members.
The project also identified other issues which require further development. These include the development of cost-effective methodologies for optimal data collection and data “worth” assessment, and the development of computerized methods that facilitate the integration of fault uncertainty into optimal longwall design.
Dimitrakopoulos is conducting a workshop at the Symposium on Geological Hazards, to be held in Newcastle on November 15-16. See symposium for further details. The workshop will demonstrate the financial benefits of quantifying risk and describe the approach to quantifying risk.
The workshop will also back-analyse North Goonyella and assess a case study at Goonyella Riverside.
