Where geology meets engineering in hydrothermally altered environments: considering veins in geotechnical engineering

Abstract

The mechanical properties of intact rock and fractures are key components that control rockmass behaviour in homogeneous rockmasses, which are the basis of many field characterization protocols, laboratory measurement, and numerical modelling approaches in geotechnical engineering. While effective in many cases of surface and underground excavations, these approaches are no longer adequate for particularly deep excavations in heterogeneous rockmasses, such as those with hydrothermal veins, stockwork, or other types of intrablock structures. Accounting for intrablock structures and other rockmass heterogeneity in rock engineering design has become critical for the safety and economic success of giant open pit mines, block caving operations, and other deep excavations. To do so, numerical modelling tools need to be integrated with improved characterization of heterogeneous rockmasses with all types of rockmass structures. In this work, the authors discuss rockmass characterization methodologies to address this challenge, including in field and laboratory settings, for application to numerical geotechnical design. Techniques to assess rock and vein mineralogies are discussed and examples illustrate the variety of mineralogies in porphyry and skarn ore deposits. Geotechnical field assessments of veined rockmasses in stress environments where shear-based failure and brittle spalling failure occur are presented. Geomechanical laboratory testing of veined rocks is discussed and a case study of unconfined compressive stress (UCS) tests on rocks from the Canadian Legacy skarn deposit is presented, where the interpretations of UCS test results are explained using mineralogical analyses. Geological interpretations of these heterogeneous rocks are an important part of understanding their geomechanical behaviours and extrapolating to predict rockmass behaviours at the excavation scale both empirically and numerically.