Analysis of caving and ground deformations in Malmberget using a coupled CAVESIM-FLAC3D model

Abstract

Large-scale sublevel cave mining (SLC) remains the mining method of choice for efficient underground mining of iron ore at the Luossavaara-Kiirunavaara AB (LKAB) mines in Kiruna and Malmberget, northern Sweden. However, SLC mining ultimately results in ground deformations above active mining areas, thus necessitating relocation of surface infrastructure and/or residential areas within the locations containing large and damaging deformations. Moreover, caving may also affect underground infrastructure and a prediction methodology for both surface effects and underground infrastructure is warranted. A coupled CAVESIM-FLAC3D model was developed for the LKAB Malmberget mine. Initially, a mine-scale model with the centrally located major orebodies was set up for detailed analysis of critical infrastructure. This was followed by an extended model in which all orebodies and production areas (approximately 20 of them) were included. Large-scale geological structures were included in the model, as well as the local geology. The coupled modelling approach enables simulating production and material flow, as well as the rock mass response outside the caved volumes. Production was simulated from the start of mining up until today and the model calibrated against observed cave cratering on the ground surface and inferred cave shapes from seismic monitoring, followed by model validation against measured ground surface deformations. The calibrated model was then run for future mining and a planned production increase, up until the year 2070. The results were evaluated with respect to: (i) surface cave cratering, (ii) ground surface deformations, (iii) strains on the ground surface and (iv) cave shapes and deformations around critical underground infrastructure.Alternative re-locations of underground infrastructure were compared and stability conditions quantified, as input for planning decisions made by the mine. For the ground surface, the model results provided a quantitative prediction of which areas will be affected, both spatially and temporally. For areas within the Malmberget township, these predictions are critical for planning and execution of the ongoing urban transformation. Moreover, the model provided predictions of ground surface affects within and near the industrial area, as input to future location of planned new surface infrastructure. The coupled modelling approach and the extensive calibration process was instrumental in developing reliable and accurate predictions for the continued mining at the Malmberget mine.