INTRODUCTION
Mining for gold in South Africa takes place at depths down to 3,5 km below surface. The deposits mined are sedimentary in origin and the auriferous conglomerate reefs occur in strong quartzite host rocks interbedded at approximately 0,5 m intervals with thin shale layers. Dykes and faults are common. The reefs are mined from narrow, 1 m wide, shallow dipping stopes. The stopes are very extensive, in some areas mining has extended for several kilometres on strike and dip. It has been well substantiated that for the most part the rock behaves as a continuous elastic mass. Thus, for these mining configurations at depth, large stress concentrations exist ahead of the stope with the consequence that the rock inevitably fractures. Figure 1 shows the fractured rock at a stope face.
All stoping takes place within this environment of fractured rock, therefore it is important if mining methods are to be improved that the generation of fractures and deformation of the failed rock be fully understood. To address this problem numerous observations of fracturing both in and around stopes have been made and seismic activity has been monitored by means of a seismic network which had an accuracy of location of better than 10 m and which was capable of detecting events of Richter magnitude -1,5. This paper reviews the findings of these studies. It deals with the classification and distribution of fractures around stopes and the influence of certain of the factors which affect the development and propagation of fractures. Some of the findings are not entirely in agreement with other workers in this field.
Kersten (1969) was the first person to classify mining-induced fractures which form in deep-level gold mines. He recognised two basic types but made allowances for three classes,:- namely
Class I. Fractures which reveal no relative movement parallel to the fracture surface and which were thought to have formed as a result of a tensile stress.
Class II. Fractures which represent intermediate types and can, for example, refer to a class I fracture which has subsequently been subjected to later movements.
Class III. Fractures which reveal distinct signs of movement, for example, striations or powdered rock material on the fracture surface.
Kersten was very careful not to imply that his Class III fractures were shear fractures. This view differed from that of Pretorius (1958) who identified fractures which were inclined to the vertical with a distinct component of displacement. The fracture planes comprised zones of broken and comminuted rock material and were called "burst fractures", with the implication that they formed in association with rockbursts.
McGarr (1971a, 1971b) divided mining induced fractures into two basic types, that is, "types 1 and 2". If fractures in the hangingwall of a stope are considered then his type 1 fractures dip in the direction of face advance and type 2 fractures in the opposite direction.
McGarr regards both types as shear fractures although type 1 fractures reveal less evidence of shear and comminution than type 2.
