Experience coming from the Kola super deep and Cajon Pass wells clearly indicates the wellbore breakout analysis and the differential velocity analysis appear most promising and feasible methods of in situ stress estimation at great depth, provided an adequate data set of core compressive/tensile strength, compressional/shear wave velocity, and anisotropy with respect to both strength and elastic properties is available. Reliability of the horizontal principal stress orientation and magnitude determination using breakout analysis is strongly affected by the symmetry and degree of rock strength anisotropy as well as orientation of this symmetry with respect to the borehole axis. Utilization of differential velocity analysis appears especially justified within elasticly compliant lithologies where accuracy of 25–30% in the vertical stress magnitude estimates may be attained. Using this method, we managed to recognize deep-seated brittle fault zones with a considerable in situ vertical stress relief as compared to gravitational loading.
Deep scientific drilling programs in the USSR, West Germany, and USA provide an excellent opportunity for in situ stress estimations at great depth shedding light on key issues of tectonophysics and geodynamics. Failure to account for these factors inevitably results in a large degree of uncertainty in the interpretation of such stress estimation techniques as hydraulic fracturing, breakout analysis, as well as purely petrophysical methods such as anelastic strain recovery measurements. In this paper we consider some effects of these factors in wellbore breakout (WBA) and differential velocity analyses (DVA) using case histories from the two most comprehensively studied deep holes - the Kola superdeep and Cajon Pass wells.
Stress-induced wellbore breakouts first discussed by Bell and Gough (1979, 1982) are characterized by formation mechanism similar to the one invoked by mining engineers facing spalling and rockburst problems in both horizontal and vertical mines (e. g., Turchaninov et al., 1978). With the advent of the four- and eight-arm caliper tools and, especially, the acoustic borehole televiewer the breakout analysis have developed into a promising means of in situ stress studies (Zoback et al., 1985; Plumb and Hickman, 1985). Much of the controversy around WBA (e. g., Byerlee and Lockner, 1988; Blenkinsop and Sibson, 1988) disappears when breakouts are defined as symmetric zones of spalling due to the shear failure of the rock matrix in response to the stress concentration around the hole. Our goal in this section is to show that even at this strict definition breakouts may not necessarily reflect the direction of the maximum in situ principal stress in a simple way suggested by the circumferential compressive stress distribution around the hole. The Kola superdeep well drilled uncased to a depth of about 12 km with continuous coring in crystalline rock of a wide range of lithologies presents a perfect example to consider. Figure 1 is a schematic data reduction first presented by Kazansky et al. (1981) and later by a group of authors in the monography "The Superdeep Well of the Kola Peninsula" (1987).