The variability of in situ rock stress is discussed in relation to different scales: the tectonic/regional scale, site scale, excavation scale, borehole/measurement scale, and microscopic scale. The factors causing the stress variability are rock inhomogeneity, rock anisotropy, discontinuities and free faces. A computer simulation of inhomogeneity-induced stress variation is included. The variation of stress with depth is made clearer through the use of the first stress invariant. Rock discing in boreholes at the JinPing II site is illustrated. Computer modeling of the rock mass is recommended as an enhancement to stress measurements.
The magnitude and orientation of the in situ rock stress components are required for the design of underground rock engineering projects. In recent years, there has been a move to use numerical modeling/simulation methods in conjunction with rock mass classification as the main support for rock engineering design. All these numerical methods require information concerning the in situ stress at the project site as boundary condition information. Thus, there has also been an increased requirement to establish the local site stress conditions during the site investigation process. This then raises questions about the stress measurement techniques and indeed which values to assume for the in situ stress field parameters for the numerical modeling. In this paper, we discuss the reasons for such in situ stress variability in relation to the tectonic/ regional scale, site scale, excavation scale, borehole/ measurement scale, and microscopic scale and and we review the factors causing the variability: rock inhomogeneity, rock anisotropy, discontinuities and free faces. This will lead to a recommended approach to the subject through an understanding of the host geology and associated numerical modeling, providing guidance on the degree of stress variation that would be expected at a particular site and hence howto approach a stress estimation campaign.