Geology Reference
In-Depth Information
Amplitude
Phase
a)
AI (m/s.g/cc)
Pseudo-well
1900
5300
5260
5220
5180
5140
5100
5060
5020
4980
4940
4900
2000
0
2100
2200
0
Frequency
2300
TWT
Amplitude
AI (m/s.g/cc)
Pseudo-well
b)
1900
5300
5260
5220
5180
5140
5100
5060
5020
4980
4940
4900
'Average' Wavelet
2000
2100
2200
TWT
2300
Figure 9.17
Averaging wavelets with similar phase spectra.
Figure 9.18
Model-based inversion of a wedge model; (a) using
top horizon only as a constraint for model interpolation, (b) using
top and base of wedge as the constraint. Note how residual tuning
effects remain in the inverted result when a single boundary
constraint is used.
0.7
Figure 9.19
An acoustic impedance
histogram based on log data from six
wells.
Oil (<30%V
sh
, <30%S
w
)
Water (<30%V
sh
, 100%S
w
)
Mixed (30-70%V
sh
)
Shales (>70%V
sh
)
0.6
0.5
0.4
0.3
0.2
0.1
0
Acoustic impedance
(
Fig. 9.11
is a salutary lesson). The conclusion appears
to be that inversions which are based on a few widely
separated wells are likely to contain residual tuning
effects.
will result in a histogram of impedance values for
different lithologies and fluid fills, based on log data
sampled at a depth spacing of about 15 cm. An
example is shown in
Fig. 9.19
. It would be easy to
conclude in this case that there is so much overlap
between the various classes that inversion would not
be useful in mapping out the oil sands. In fact an
inversion clearly shows the oil sand distribution
(
Fig. 9.20
). The reason is that seismic data sees
9.2.4.5 Vertical scale
Very often, a preliminary study of well impedance
values is undertaken prior to deciding whether an
inversion is likely to give useful results. Often this
208
