Technological improvements in geotechnical mapping, modelling and monitoring now allow a digital twin to be developed and maintained in parallel with mining operations. That is, the routine use of aerial photogrammetry, semi-automatic rock mass characterization, three-dimensional slope stability modelling and ground-based interferometric synthetic aperture radar monitoring allow near real-time comparison of modelled and realized ground conditions. This is critical for reconciling the effectiveness of geotechnical models for predicting future slope stability (or instability). This paper presents a semi-brittle slope failure case study from Pueblo Viejo gold mine which was managed using a digital twin approach. The near real-time comparison of the three-dimensional model with observed slope conditions allowed critical structures to be added to the geotechnical model as mining progressed. This led to the identification of a high risk area that would not have been identified if the digital twin was not implemented & maintained during mining.
Pueblo Viejo gold mine is located approximately 80 km north of Santo Domingo, in the Dominican Republic, on the island of Hispaniola in the Caribbean archipelago. The mine is located at an altitude of approximately 300 to 500 m above sea level. The Pueblo Viejo district has a tropical climate with the nearby cities of Cotuí and Bonao receiving 1843 mm and 1956 mm of average annual rainfall (Bar et al., 2021).
Gold mineralization in the Pueblo Viejo district, Dominican Republic, is spatially and temporally related to a series of Early Cretaceous volcanic domes (Nelson, 2000). Two principal deposits (Moore and Monte Negro) and a number of smaller deposits including Cumba, Mejita, Banco V, Arroyo Hondo I and II have contributed ore since surface mining commenced in 1975. The geological setting is structurally complex with several phases of thrust faulting and hydrothermal alteration. Due to the complex ground conditions and tropical setting, several instabilities and geotechnical challenges have occurred and were retrospectively reviewed for incorporation into future designs (Bar et al. 2022a; Bar et a. 2022b; Cobián et al. 2022).
