Page 36 - Reservoir Geomechanics

20 Reservoir geomechanics

(Zoback and Zoback 1980, 1989), demonstrate that large regions of the North Amer-
ican continent (most of the region east of the Rocky Mountains) are characterized by
relatively uniform horizontal stress orientations. Furthermore, where different types of
stress orientation data are available, see, for example, the eastern U.S., the correlation
between the different types of stress indicators is quite good. The distribution of data is
quite uneven throughout North America as the absence of data from wells, earthquakes
or young geologic phenomenon in the much of the intraplate region leave large regions
where the state of stress is unknown. In contrast, well-constrained earthquake focal
plane mechanisms are ubiquitous in southern California such that the data are so dense
that individual data points cannot be identified at the scale of this map.
Two straightforward observations about crustal stress can be made by comparison
of different types of stress indicators. First, no major changes in the orientation of
the crustal stress field occur between the upper 2–5 km, where essentially all of the
wellbore breakout and stress measurement data come from, and 5–20 km where the
majority of crustal earthquakes occur. Second, a consistent picture of the regional
stress field is observed despite the fact that the measurements are made in different
rock types and geologic provinces. Finally, the criterion used to define reliable stress
indicators discussed in subsequent chapters appears to be approximately correct. Data
badly contaminated by non-tectonic sources of stress or other sources of noise appear
to have been effectively eliminated from the compilations. The state of stress in the
crust at very shallow depth (i.e. within ∼100 m of the surface) is not discussed here
for two reasons. First, this topic is outside the scope of this book (see, for example,
Amadei and Stephansson 1997). Second, in situ stress measurements at shallow depth
cannot be used in tectonic stress compilations because tectonic stresses are very small at
shallow depth (because of the low frictional strength and tensile strength of near-surface
rock) and a number of non-tectonic processes, including thermal effects, strongly affect
in situ stresses near the earth’s surface (Engelder and Sbar 1984). In general, only in situ
stress measurements made at depths greater than ∼100 m seem to be independent of
rock type, are spatially uniform and consistent with earthquake focal plane mechanism
data coming from much greater depths. This means that techniques applied in wells and
boreholes,and earthquakedatacanbeusedtogether(withsufficient care) tocharacterize
the crustal stress field.
It is important to point out that the relative uniformity of stress orientations and
relative magnitudes observed in Figure 1.5 is also seen at a variety of smaller scales.
Forexample, the stress field in central California near the San Andreas fault (an actively
deforming fold and thrust belt in a transpressional plate tectonic setting) is generally
quite uniform (Figure 1.6, after Castillo and Zoback 1994). With the exception of
the southernmost San Joaquin valley (which is discussed below), an overall NE–SW
maximum horizontal stress direction is implied by both wellbore breakouts (inward
pointing arrows) and earthquake focal mechanisms (lines with open circles) correlate
extremely well. Both sets of data are consistently perpendicular to the trend of currently