Towards optimising ground support systems in underground mines

There is a large body of literature describing different methods for designing ground support systems for underground excavations. Ground support design methods are commonly grouped in three classes—namely, analytical, empirical and numerical modelling methods. A comprehensive review of practices in Australian and Canadian mines (Potvin & Hadjigeorgiou 2016) has shown that the design of ground support is largely the result of the evolution of and adjustments to an initial design. The initial design relies mainly on two methods: the Grimstad–Barton (Grimstad & Barton 1993) empirical chart and the RocScience software UnWedge, based on a limit equilibrium wedge analysis defined by continuous large geological structure. The optimisation process is largely reactive and involves modifying the ground support systems to cater for ground conditions that differ from those initially anticipated or replacing ground support elements to improve performance. Sometimes the decision is backed up by numerical modelling analyses in anticipation of stress changes as a result of mining. Nevertheless, a systematic engineering methodology is rarely employed to optimise ground support systems in underground mines. An optimum ground support system can be defined as the lowest cost system, including installation cost and productivity factors such as development mining rates, which can achieve a tolerable Probability of Failure (PoF) during the service life of an excavation. A further requirement is that such a system will minimise the need for rehabilitation. The above definition implies that no matter which design method is applied, a probabilistic assessment is required to ascertain whether the ground support design meets the target PoF. This paper outlines userfriendly mXrap-based tools developed within the scope of the Australian Centre for Geomechanics’s Ground Support System Optimisation research project.