Page 166 - Reservoir Geomechanics

149 Faults and fractures at depth

Although wellbore image data only provide information on the apparent aperture of
faults where they are intersected by a wellbore, empirical relations have been proposed
that attempt to relate fault length to aperture (e.g. Gudmundsson 2000) and enable one
to develop ideas about fracture networks from wellbore image data. However, it should
be pointed out that the aperture at the wellbore wall is not the actual width of the fault
plane. As discussed by Barton, Tessler et al.(1991), image logs are basically measuring
the smoothness of the wellbore wall, such that the logs are quite sensitive to any spalling
at the wellbore wall that occurs as the bit penetrates the fault plane. Thus, the apparent
aperture is always going to be larger than the actual aperture. Barton, Tessler et al.
(1991) point out that it is sometimes useful to qualitatively distinguish large from small
aperture faults during the analysis of image data.
One obvious limitation of imaging tools is that they will undersample fractures and
faultswhose planesarenearlyparalleltothewellboreaxis.Thisismosteasilyvisualized
for a vertical well – the probability of intersecting horizontal fractures in the formation
is one, but the probability of intersecting vertical fractures is essentially zero. Following
Hudson and Priest (1983), it is straightforward to estimate the correction to apply to
a fracture population of a given dip. If we have a set of fractures of constant dip, φ,
with an average separation D (measured normal to the fracture planes), the number of
fractures observed along a length of wellbore, L = D/cos φ, N obs (φ) must be corrected
in order to obtain the true number of fractures that occur in the formation over a similar
distance, N true (φ) via

N true (φ) = (cos φ) −1 N obs (φ) (5.6)

This is illustrated for densely fractured granitic rock by Barton and Zoback (1992),
who also discuss sampling problems associated with truncation, the inability to detect
extremely small fractures, and censoring, the fact that when an extremely large fault
zone is penetrated by a wellbore, it is often difficult to identify the fault plane because
the wellbore is so strongly perturbed that data quality is extremely poor.

Representation of fracture and fault data at depth

Figure 5.5 illustrates the commonly used parameters that describe the orientation of
faults and fractures in three dimensions. Fault strike refers to the azimuth of a horizontal
line in the plane of the fracture and is measured from north. Unless otherwise noted,
dip is measured from a horizontal plane and is positive to the right, when looking
in the strike direction. Dip direction is often used instead of strike to define fracture
orientation. Measured from north, dip direction is the azimuth of the projection of a
vector in the plane of the fault that is normal to the strike direction and points down
dip. Slip on the fault is also defined by a vector, which can be thought of as the