Using stochastic crosshole seismic velocity tomography and Bayesian simulation to estimate Ni grades: Case study from Voisey's Bay, Canada

Abstract

Modeling of grades is a key step and the major source of error in appraisal stage of mining projects. We used a geostatistical approach to explicitly integrate seismic travel time data, as well as acoustic and core logging data into the estimation of nickel grades in the Voisey's Bay deposit. Firstly, the crosshole seismic travel times are inverted using a stochastic tomographic algorithm. This algorithm allows for the inclusion of acoustic log data and seismic covariance into the inverse problem, leading to high-resolution velocity tomographic images of the orebody. Secondly, grade realizations are generated using a Bayesian sequential Gaussian simulation algorithm, which integrates the ore grades measured on the core logs and the previously inverted tomographic data. The application of the presented method to the Voisey's Bay deposit yields an improved knowledge of the geology setting and generates grade models with realistic spatial variability compared to conventional methods.