Over the past fifteen years, my colleagues and I have developed a suite of techniques for determination of the full stress tensor in arbitrarily-oriented wells and boreholes utilizing observations of non-catastrophic failures of the wellbore wall - compressive failures, drilling-induced tensile fractures and stress perturbations associated with slip on faults cutting through the wellbore. While these techniques have had extensive application in the petroleum industry, they have also been used in core holes drilled from excavations in mines to yield information about the state of stress within, and beyond, the area affected by the stress concentration surrounding the excavation. When possible, hydraulic fracturing is used to provide independent information about the magnitude of the least principal stress, but is not used to estimate the maximum horizontal principal stress. In this paper we review both the conceptual elements of this methodology and two challenging case studies. The first involves determination of the state of stress following the drilling of the first phase of the SAFOD project, a scientific borehole drilled through the San Andreas Fault in central California. These case studies document how observations of wellbore failure in deviated wells yield consistent stress orientations and magnitudes over appreciable depth ranges.
While a number of papers had been written about compressive and tensile failures in deviated wells, (Peska & Zoback, 1995) published the first systematic investigation of the tendency (and orientation) of wellbore failures in arbitrarily-deviated wellbores in normal, strike-slip and reverse faulting stress regimes. The principles outlined in that paper have proven to be quite useful for determination of the complete stress tensor in arbitrarily-oriented wells and boreholes, especially when combined with independent measurements of the magnitude of the least principal stress from hydraulic fracturing.