Natural caving systems as potential analogues to mining induced caving

Abstract

A rising demand for metals has forced the mining industry to target deeper orebodies. With increasing depth, lower grade orebodies become uneconomical because of higher extraction and production costs. Hence, the industry is transitioning to more cost-effective underground mass mining methods such as block caving (BC). In BC, the ore is undercut by blasting to create a void and then drawing off the broken rock to induce failure and subsidence of the overlying rock mass (cave back). This initiates an upward progression of the cave either due to failure along natural discontinuities (gravity caving) or due to stresses induced in the cave back (stress caving). Detailed knowledge of the geology and structure is required at all scales to manage the mine caving processes. This involves predicting the caveability and fragmentation of the rock mass and the effects of subsidence-induced deformation mechanisms on the surrounding rocks, and at surface. BC is a ‘blind’ mining method and direct observations of the mined zone are limited. To enhance the understanding of the processes operating in gravity induced cave systems, solution collapse breccia pipes (SCBPs) and sinkholes were studied on the Colorado Plateau, USA, as potential analogues to mining induced caving. SCBPs form under the influence of gravity, when rock masses progressively collapse into a developing void produced by water flow and the dissolution of rocks such as limestone. These processes result in an upward stoping process through the overlying rocks, which eventually forms a vertical, pipe-shaped column of broken rock. After lithification, this structure becomes an SCBP. Recent erosion that formed the Grand Canyon system has exposed many SCBPs in section on canyon walls.This study employed close-range photogrammetry and photo-interpretative mapping to study the features of well exposed but mostly inaccessible SCBPs and sinkholes. Georeferenced 3D models allowed accurate data to be collected for structural analysis, inferences to be made about rock mass behaviour, and the identification of processes that operated during subsidence. This was the first approach to study SCBPs by applying close-range photogrammetry. The results enhance the understanding of SCBP geological evolution over time.Four zones were identified in typical SCBPs, with each exhibiting different deformation and collapse mechanisms and/or flow processes. From the centre to the periphery, typical SCBPs are comprised of, (1) the breccia body, a clast to matrix-supported breccia with angular to sub-rounded clasts, (2) the pipe margin, a zone of intense fracturing and shearing, (3) the deformation zone, a region of faulted and displaced rocks surrounding the breccia body, and (4) the undeformed wall-rocks. The mechanics of SCBP collapse are most strongly influenced by lithology and both pre-existing and stress-induced discontinuities. Placing studied SCBPs in a regional context showed that their location and shape is controlled by basement faults and joint patterns.The analysis of clasts in SCBP breccia bodies shows that complex flow features developed when they were mobile, including differential flow, ‘pipes within pipes’ and shearing along the pipe margin contacts prior to lithification. Clast size and shape was determined by competent wall rocks, which formed larger, angular blocks. Less competent rocks produced smaller, rounded fragments. Joint densities in wall-rocks, and bed thickness also influenced clast size and shape. Large blocks interlocked and enhanced the downward percolation of fines through preferred pathways. The spatial relationship of larger clasts derived from the same host unit was maintained during subsidence. Some clasts were rotated along their long axes, particularly at the pipe margin. In the deformation zone, the effects of subsidence varied from large- to small-scale faulting along with displacement and joint development, particularly evident in competent layers, to subtle folding of less competent rocks. Zones of higher porosity within the breccia body were identified and are mostly concentrated around the margins. These zones enhanced water movement associated with alteration and introduction of minerals. Some SCBPs intersected the surface, producing topographic depressions that were many times larger than the diameter of the subsiding column.SCBPs and a typical BC mine were compared in terms of equivalent zones recognised in both systems, each exhibiting characteristic geological influences on operating processes. The results show that natural systems are analogous to mining induced caving in several ways. The nature and variability of host lithologies, pre-existing and induced discontinuities (including bedding) significantly impact the caveability and fragmentation of a rock mass, and behaviour of the broken material in a subsiding column. SCBPs are most closely analogous to the standard ‘single drawpoint’ Discrete Element Method (DEM) flow modelling simulations that have guided the theoretical understanding of cave mining flow processes (e.g. Hancock, 2013). This study is the first real-world validation of the theoretical assumptions inherent in DEM models. The knowledge gained can be used to inform BC failure mechanism simulations and highlights the importance of a detailed understanding of the litho-structural architecture of the rock mass at all scales.