Page 29 - Reservoir Geomechanics

14 Reservoir geomechanics

conditions, if S hmin increases more rapidly than 0.6S v (as shown in Figure 1.4b), a more
compressional stress state is indicated and S Hmax may exceed S v , which would define
a strike-slip faulting regime. If the least principal stress is equal to the overburden, a
reverse faulting regime is indicated as both horizontal stresses would be greater than the
vertical stress (Figure 1.4c). As seen in Figure 1.4a–c, the differences between the three
principal stresses can be large and grow rapidly with depth when pore pressure is close
to hydrostatic. This will be especially important when we consider wellbore failure in
Chapter 10.Again, in all cases shown in Figure 1.4, the maximum differential stress
(S 1 −S 3 )is constrained by the frictional strength of the crust, as described in Chapter 4.
When there are severely overpressured formations at depth (Figures 1.4d–f) there are
consequently small differences among the three principal stresses. In normal and strike-
slip faulting domains S hmin , the least principal stress (S hmin = S 3 ) must increase as P p
increases because, with the exception of transients, the least principal stress can never
be less than the pore pressure. In strike-slip and reverse faulting regimes (S Hmax = S 1 ),
theupperboundvalueofS Hmax isseverelyreducedbyhighporepressure(seeChapter4).
Thus, when pore pressure approaches the vertical stress, both horizontal stresses must
also be close to the vertical stress, regardless of whether it is a normal, strike-slip or
reverse faulting environment.

Measuring in situ stress

Over the past ∼25 years, stress measurements have been made in many areas around the
world using a variety of techniques. The techniques that will be described in this book
have proven to be most reliable for measuring stress at depth and are most applicable for
addressing the types of geomechanical problems considered here. Stress measurement
techniques such as overcoring and strain relief measurements (Amadei and Stephansson
1997; Engelder 1993) are not discussed here because, in general, they are useful only
when one can make measurements close to a free surface. Such strain recovery tech-
niques require azimuthally oriented core samples from wells (which are difficult to
obtain) and analysis of the data requires numerous environmental corrections (such
as temperature and pore pressure) as well as detailed knowledge of a sample’s elastic
properties. If the rock is anisotropic (due, for example, to the existence of bedding)
interpreting strain recovery measurements can be quite difficult.
A general overview of the strategy that we will use for characterizing the stress field
is as follows:
Assuming that the overburden is a principal stress (which is usually the case), S v can

be determined from integration of density logs as discussed previously. In Chapter 8
we discuss how observations of drilling-induced tensile fractures are an effective way
to test whether the vertical stress is a principal stress.