Addressing the variability of elongate support performance

Elongates are commonly used as stope support in South African gold and platinum mines. To evaluate the effect of the inherent strength, stiffness and yield variability associated with elongates, a statistical method is presented which addresses and quantifies the support performance at various levels of certainty. The analysis, which is based on a normal distribution, makes use of forcedeformation curves determined from multiple laboratory compression tests on the same support type. The applicability of the method is illustrated making use of actual performance capabilities of various types of elongates. In order to ensure a high probability that support units installed underground exceed the standard support performance curve established by means of laboratory tests, the statistical mean of a suite of tests based on the same support type is downgraded. The i?½correction factori?½ is a function of the mean and standard deviation of the test data, as well as the sample size of units deemed to control the local rock mass stability. It is shown that downgraded performance curves based on a high probability (90 to 95 per cent) of support units exceeding this performance level are suitable to ensure that the design of support systems will invariably meet the design criteria. In the highly discontinuous hangingwall rock mass typically associated with intermediate- and deep-level mining operations, the interaction between adjacent support units is comparatively limited (at the support spacings commonly used). Therefore, the appropriate sample size (n) of the support units controlling the local rock mass stability should be n = 1. In shallow mining conditions, sample sizes of up to n = 5 may be warranted, depending on the spacing of joints and other discontinuities in the hangingwall.