Page 234 - Reservoir Geomechanics

215 Determination of S 3 from mini-fracs

higher than the values shown (if the measurements were not carefully made), it might
be the case that S 3 would appear to be equal to S v such that a reverse-faulting regime
would be indicated. If one were concerned with the propagation of hydraulic fractures
(to stimulate production in low-permeability reservoirs, for example), this point is quite
important. As noted above, if S 3 ≡ S hmin ,vertical hydrofracs would be initiated at the
wellbore wall. However, if S 3 ≡ S v ,vertical fractures would be expected to form at the
wellbore wall (when the increase in wellbore pressure causes σ θθ to go into tension).
However, the hydraulic fracture will rotate into a horizontal plane (perpendicular to S v )
as the fracture propagates away from the wellbore (Baumg¨artner and Zoback 1989).
After fracture propagation away from the wellbore, the FPP or ISIP can be used to
determine S 3 .In cases where S 3 ∼ S v it is particularly important to carefully integrate
density logs to determine S v and to determine if S 3 corresponds to S v or S hmin with
confidence. In fact, in the Visund field, considerable effort was taken to estimate rock
density at extremely shallow depth to derive the curve shown. Had this not been the case,
it would have been extremely difficult to determine whether or not the least principal
stress was less than, or equal to, the vertical stress.
Another way to measure the least principal stress is to conduct step-rate tests. In such
tests injection into the well is performed at a number of fixed flow rates as illustrated
in Figure 7.5.Itis easy to detect the pressure at which a hydrofrac opens; injection can
take place at increasingly higher flow rates with only minimal increases in wellbore
pressure. Prior to the hydrofrac opening, there is a strong increase in pressure with flow
rate as expected for a system dominated by diffusion into the formation and/or a closed
hydraulic fracture.
A number of methods for the analysis of shut-in pressure data for determination
of the least principal stress have been proposed over the years. A discussion of var-
ious techniques was reviewed by Zoback and Haimson (1982), Baumg¨artner and
Zoback (1989), Rummel and Hansen (1989), Hayashi and Haimson (1991) and Guo,
Morgenstern et al. (1993). An alternative way to measure the least principal stress is to
pressurize the wellbore in steps and measure the rate of pressure decrease after pressur-
ization is stopped. The logic behind such tests is that once a fracture has formed, the rate
of pressure decrease with time will be faster. If a sufficient number of closely spaced
pressure steps is used, the magnitude of the least principal stress can be determined
with corresponding accuracy. Similarly, there are other techniques referred to as pump
in/flow back methods (see Raaen and Brudy 2001) that yield reliable results. Hayashi
and Haimson (1991) discuss the interpretation of shut-in data from a theoretical per-
spective and present a technique they argue is most optimal for determination of the
least principal stress.
Unfortunately, many LOT’s are conducted using extremely poor field procedures.
When trying to analyze such tests, two questions must be kept in mind to know whether
the test can be used to obtain a measure of the least principal stress. First, is there an
indication that the LOP was reached? If so, the LOP can be considered an approximate