Development, validation and application of bonded block modelling approaches for support design in underground mining

Tunnel support systems are essential for ensuring worker safety and minimizing production delays caused by collapse incidents or the need for rehabilitation. The purchase and installation of support elements also represent a significant expense in the operational budget of any underground mine. Despite this importance, current design practices still rely on empirical methods developed decades ago, based on limited datasets that represent narrow ranges of rock mass and stress conditions, or on simplified kinematic models of rock–support interaction using wedge analysis and key-block theory. Experience shows these lack the robustness needed for today’s mining depths and safety standards. Advances in computing power and numerical modelling tools now enable the industry to move beyond these oversimplified methods and adopt new approaches that more accurately simulate failure mechanisms and interactions between the rock mass and support elements. A new methodology using three-dimensional Bonded Block Modelling (BBM) is presented and validated against case histories from operational mines. This thesis demonstrates the effectiveness of the BBM approach in simulating complex failure mechanisms—including pillar spalling and bulking, and the progressive failure of intact rock bridges along wedges formed from non-persistent joints—achieving closer agreement with field observations in deep mining environments. The approach was validated against operational case studies, including the Raglan Mine and Diavik Diamond Mine, with results showing a more accurate depiction of the mechanism leading to support system failure than could be achieved with conventional approaches. Unlike conventional empirical or kinematic models, the BBM approach enables explicit simulation of spalling, progressive rock bridge failure, and complex rock–support interactions, accurately reflecting stress-induced failure mechanisms.