Page 266 - Reservoir Geomechanics

247 Wellbore failure and stress determination in deviated wells

of anomalous stress where principal stresses deviate from their average orientation
◦
by ±10 .
As a brief historical note, it is worth pointing out that at the time when the integrated
stress maps presented in Chapter 1 were initially compiled by Zoback and Zoback
(1980; 1989) and Zoback (1992)itwas assumed that one principal stress was vertical.
The rationale for this assumption was that at relatively shallow depth in the crust,
the presence of a sub-horizontal free surface would require one principal stress to
be approximately vertical. At greater depth, it was argued that for nearly all intraplate
crustal earthquakes for which there are reliable focal mechanisms, either the P-, B-or T-
axis was sub-horizontal. While this is a relatively weak constraint on stress orientation,
the consistency of the apparent horizontal principal stresses in the compiled maps
indicated that the assumption appeared to generally correct. To date, drilling-induced
tensile fractures have been identified in scores of near-vertical wells around the world.
As such fractures are nearly always axial, it provides strong confirmation that the
assumption that principal stresses in situ are vertical and horizontal is generally valid.

Estimating S Hmax from breakouts and tensile fractures
in deviated wells

Several authors have addressed the subject of the relationship between the failure of
inclined holes and the tectonic stress field. Mastin (1988) demonstrated that breakouts
in inclined holes drilled at different azimuths were expected to form at various angles
around a well bore. Qian and Pedersen (1991) proposed a non-linear inversion method
to attempt to extract information about the in situ stress tensor from breakouts in an
inclined deep borehole in the Siljan impact structure in Sweden. Qian, Crossing et al.
(1994) later presented a correction of their results because of errors in the published
equations of Mastin (1988) (although the figures in Mastin’s paper are correct). Zajac
and Stock (1992) suggested that it is possible to constrain stress magnitudes from
breakout azimuths if there are observations from a number of inclined holes drilled
at various azimuths in a uniform stress field. This technique is conceptually similar
to a technique reported by Aadnoy (1990a,b) to estimate in situ stress from leak-off
test data in a number of inclined boreholes that assumes that peak pressures from
leak-off tests are hydrofrac fracture initiation pressures. For the reasons outlined in
Chapter 7, this is likely a questionable assumption. Using formal geophysical inversion
theory with observations of breakouts or tensile failures presumes that there will be
data available from multiple wells of varied orientation that are sampling a uniform
stress field. Whether such data are likely to be available is a questionable assumption.
There are several straightforward ways to use observations of tensile fractures and
breakouts in deviated wells for determination of the magnitude and orientation of
S Hmax at depth assuming that the vertical principal stress, S v , and minimum horizontal