Three gradients have been identified in regard to the average horizontal ground stress increase with depth. S Ha (0-900 m) = 9.86 MPa + 0.0371MPa/m S Ha (900-2200 m) = 33.41 MPa + 0.0111 MPa /m S Hae (extreme) = 12.36 MPa = 0.0586 MPa/m The extreme stress values are not caused by mining geometry and are of limited extent only. In some cases geological disturbances like faults and rock type boundaries suggest an explanation but in other cases the unusually high stresses cannot be explained readily. It is interesting to note that in the same locations the vertical stress components are very much higher than those derived from overburden weight and that the vertical stress gradient is similar to the gradient obtained from the extreme horizontal stress components.
Ground stress determinations have been carried out in mines in Ontario and Manitoba, Canada, over more than a decade (Herget, 1980). Most of the sites are located in the Superior and Southern Tectonic Province of the Canadian Shield, which consist of Archaen and Proterozoic rocks comprising volcanics, metamorphosed sediments and granites. The youngest orogenic deformation occurred during the Grenville Orogeny (955 million years ago).
Ground stress determinations in mines were carried out by overcoring methods, such as those using biaxial instruments, e.g., the USBM meter and CSIR doorstopper, and triaxial instruments, e.g., the triaxial strain cell developed by the CSIR (South Africa) and CSIRO (Australia). Fortunately the rock material in the Canadian Shield is generally very strong. This is beneficial because the magnitude of ground stresses which can be determined depends on the range of the elastic behaviour of the rock under load. This has permitted the determination of ground stresses up to a magnitude of 130 MPa at a depth of 2100 m. Discing of the drill core can make overcoring difficult at those stress magnitudes and the possibility exists that full strain recovery is not measured (Hast, 1979). In highly stressed ground it has been observed that the doorstopper method resulted in successful overcoring, whereas methods requiring the overcoring of an Ex annulus over 30-50 cm were unsuccessful. Overcoring methods have been very successful when carried out in fine-grained, isotropic solid rock and within 10 to 20 m of excavation boundaries. For measurements in drill holes beyond this depth, overcoring procedures become time consuming and very costly. In most of the locations where strain recovery was measured by overcoring, a redundancy of strain recovery data was available for each tensor determination, so that error determinations were possible with the method of least squares. Errors of + 10-15% for the stress components are common. Details of overcoring methods, quality testing to obtain a high degree of reliability, and determination of physical parameters are described elsewhere, (Leeman, 1969; Gray and Barron, 1969; Herget, 1973).
Many investigators have observed that the vertical stress component (SV) increases linearly with depth and that the increase is related to overburden weight.