Geoscience Reference
In-Depth Information
Open hole
logs
Surface
curvature
Layer
thickness
Core
plugs
FMS logs
Remove shale
and anhydrite
Resisitivity
logs
Fracture
density
logs
Matrix
porosity
Check
vs.
tot. por
Fracture
porosity
K/phi
transform
Core
observations
Check
vs. core
Fracture
aperture
Fracture
density
model
Matrix
permeability
Check
vs. core
Check
vs. tests
geometric
average
arithmetic
average
Fracture
permeability
Fracture
spacing
Matrix
porosity
Matrix
kh
Matrix
kv
Clip
aperture,
recalculate
porosity
Directional
fracture
permeability
Fracture
orientation
CALIBRATE
TO WELL
TESTS
Fracture
porosity
Eclipse
kx frac
“Sigma”
“DPNUM”
Ky frac
Fig. 6.62
Workflow for implicit static description of fractures, feeding into standard dual-permeability simulations
combined with standard workflows for modelling
the matrix (the right hand side of Fig.
6.62
).
The same logic can be applied to the dynamic
model, in which the effective flow properties of a
'fracture REV' can be established using small-scale
models, and applied directly in a standard single
porosity, single permeability simulator. Numerous
assumptions must be made along the way, and it is
important to check whether the underlying fracture
concept has been compromised in the process. This
is particularly the case for ensuring an appropriate
level of network connectivity on the cell-to-cell
scale: if the concept is for field-wide fracture conti-
nuity in a particular direction, is that appropriately
represented in the dynamic model?
conductive than fractures which are in compres-
sion. In reservoir modelling, conditioning static
fracture models to dynamic data (well tests and
production data) often acts as a proxy to stress-
modelling, since the dynamic data indicates
which fractures are actually flowing. However,
this may not be very predictive, and so it may be
preferable to try and forward-model or 'forecast'
which fractures are most likely to be conductive.
Predicting the effects of the stress field on
fracture flow properties is a significant challenge,
so that a more realistic modelling objective is to
allow fracture flow properties to be 'stress sensi-
tive' - that is to try to capture the relationship
between fracture conductivity and orientation.
Bond et al. (
2013
) successfully demonstrated
this approach by modelling fracture anisotropy
as controlled by the stress field, for a CO
2
storage
modelling study. They used dilation tendency T
d
,
as the controlling factor:
6.7.2.5 Handling the Effects of Stress
A prescient challenge for fracture modelling, for
both the explicit and implicit approaches, is cap-
turing the effects of stress. The present-day stress
system will act on the inherited sets of fractures
to determine their fluid flow properties. Fractures
which are favourably aligned to the present-day
stress field will
T
d
¼ ˃
1
˃
n
ð
Þ ˃
3
˃
n
ð
Þ
where
˃
n
is the normal stress on the fracture plane.
tend to be more open and
