Geoscience Reference
In-Depth Information
Fig. 6.27 Modelling of
thin beds: when it goes
wrong
2
1
Logs don't resolve
beds, Ø log
Block the incorrect
Ø log values
Ø rock
4
3
k
Ø
k
Ø
Apply wrong k/Ø
values at a larger
scale
Transform wrong Ø
values to k using a k/ Ø
relationship derived at a
different scale
6
5
forecast
actual
Pass to simulation
and RE makes
adjustments
Produce meaningless
forecast
6.5.4 Small-Scale Heterogeneity in
High Net-to-Gross 'Tanks'
partitioned by thin but extensive heterolithic
intervals with low permeability which would be
poorly resolved on logs (Fig. 6.29 ). The issue is
whether or not such heterolithic intervals would
have significant vertical permeability and how
laterally extensive the intervals would be, i.e.
do they constitute barriers or baffles?
Without very good log resolution a similar
logic is needed to that applied to thin beds -
effective permeability needs to be estimated from
small-scale modelling. This is a simpler exercise,
however, as the key issue is vertical permeability;
the horizontal permeability in the gross sand inter-
val will always be dominated by the high N/G
sands above and below the heterolithics.
For confined systems in which the reservoir is
detectable from seismic attributes, it is tempting
to work directly from seismic and treat the field
as a 'tank of sand', albeit an irregularly-shaped
one. What tends to be overlooked is the contribu-
tion of low-net thin beds within the generally
high N/G system; the inverse condition of the
'thin bed' scenarios described above.
An example of this is shown below from the
well-studied high N/G ratio reservoir analogues
of the Annot region in SE France (Pickering
and Hilton 1998 ). Massive sand intervals are
Search WWH ::




Custom Search