Geology Reference
In-Depth Information
to the top of the water filled sand (
Fig. 5.50b
). So there
is the possibility that the initial interpretation has
deviated from the top of the water filled sand outside
closure and onto the contact reflection over the
culmination. Several observations would appear to
justify this interpretation. The peak is fairly flat over
the central portion of the culmination consistent with
a contact. In addition, the isochron between the
picked peak and the prominent peak at the top of
the section shows a thickening over the culmination.
This is unusual as there is no evidence of structural
inversion in this part of the basin. Ghosting-in the top
reservoir based on isochronous stratigraphic layering
(
Fig. 5.50c
) would make sense of the flat portion of
the picked peak as the oil water contact. Further
evidence for the interpretation of a contact is also
given by the presence of the apparent reflector ter-
minations below the peak (
Fig. 5.50c
). These features
are characteristic of modelled contact effects in thick
sands in this area. The feature was drilled and found
to be oil bearing.
An example of a more detailed application of a
rock model has been described by Avseth et al.(
2003
)
in which the statistics of the rock model have been
used to predict the probability of seismic responses
representing particular
a)
Effective angle
of stack
b)
Oil sand on
water sand
Shale on water
sand
0.1
0.05
0
Rc
-0.05
-0.1
Shale on oil sand
-0.15
lithofacies
scenarios. The
workflow applied is as follows.
Angle of incidence (
θ
)
(1) Determine lithofacies and elastic parameter
statistics at the depth of interest.
(2) Establish layer configurations (boundary types
and vertical facies associations) based on an
understanding of the geology
c)
Isochron of overlying
shale unit remains fairly
constant
in this case the
target is in a deep sea submarine fan setting.
(3) For each boundary type use Monte Carlo
simulation to generate intercept vs gradient scatter
plots (a statistical analysis of these results provides
the probability of a given value of intercept and
gradient representing a particular boundary type).
(4) Generate intercept and gradient from the seismic
trace data and crossplot.
(5) Perform AVO calibration (i.e. scale the intercept
and gradient to the model values). As the gradient
is not a directly scalable quantity this step in
essence is performed by scaling reflectivities and
recalculating the gradient (see
Section 6.3.3
). The
scalars are determined for the
-
Oil water
contact
Top sand
Reflector terminations
near contact
Figure 5.50
Migrated stack seismic interpretation aided by an
AVO plot constructed from data in a nearby well; (a) original
interpretation of top reservoir, (b) AVO plot showing responses for
water sand, oil sand and oil water contact, (c) revised
interpretation taking into account the AVO model and other
seismic observations.
'
background
'
responses.
(6) Blind test at wells (determining acceptable mis-
classification error).
91
