The SRK Vancouver Rock Mechanics team has been working at Cameco’s underground Cigar Lake Uranium Mine providing advanced multidisciplinary geotechnical modelling since the summer of 2015. SRK conducted geotechnical and structural geology reviews, which led to refining the existing structural and alteration model. This was critical for developing numerical models of the production drifts, which are located in highly variable geology – strong rock to weak clay alteration – and interactions from the overlying artificially induced ground freeze.
The project presents unique challenges, most notably the extreme variations of rock strength from moderate (50 MPa) to weak (less than 5 MPa), and a largescale artificial freeze from the surface that encapsulates the orebody. The ore is extracted above the development using an innovative method by jetboring 5m diameter stopes.
The large mass of frozen ground induces stresses and associated displacements that vary depending on the rock type, alteration, and water content due to volumetric expansion that occurs during the state change. These high and unevenly distributed stresses, combined with the highly variable geology and its contrasting physical properties, are adversely affecting the NATM production drives. This has resulted in interruptions to jet boring operations and significant expense to rehabilitate the damaged production drives.
SRK’s team is modelling the existing development and ground design to develop a more resilient design that does not interrupt the jet boring of the orebody. Using Itasca code FLAC3D, SRK has modelled the volumetric expansion brought on by the freeze and interactions with the tunnels below. The complexity increases with different ground support and liner material, limited clearance for the jet-boring machine, and a narrow pillar width.
The modelling process utilises FISH code to import the complex geology model developed in LeapFrog and the isotherms provided by Cameco and their frozen ground consultant. This approach has allowed us to move around the production development geometry and introduce the appropriate geological conditions.
The numerical models have been calibrated using monitoring data from the existing tunnels comprised of displacement and stress data as well as qualitative data from liner condition mapping. A significant challenge has been emulating the behavior of the sprayed concrete liner and HidCon yield elements with multiple excavation and rehabilitation phases.
Once the initial model was suitably calibrated, the code was modified to predict the behavior for two as-yet unmined production drives. These models allowed for optimising ground support and liner thickness for the long-term mine plan. The modelling results supported a reduction in ground support in the stronger rock units and in particular a significant reduction in ground support requirements as ground conditions are expected to improve. The high tunnel development cost for this project makes any reduction in ground support requirements and cycle time valuable with a resultant decrease in OPEX.
