Geology Reference
In-Depth Information
b)
d)
1 00
-100
200.0
8 0
-50
a)
6 0
0
Lith
Phi
AI Best
Syn
Well
4 0
50
100
150
-200.0
2 0
ΔΥ (m)
0
× Well location
+ Selected trace location
Amplitude
c)
e)
0.75
0
0.5
-10
0.25
0
-20
-0.25
-0.5
-0.75
0
-100
-175
Synthetic shift (ms)
Frequency (Hz)
Figure 4.14
Example well tie generated from the well matching technique; (a) synthetic trace (
'
Syn
'
) compared to the best match location
trace (
'
Best
'
); red line defines the time segment for the analysis, (b) extracted wavelet, (c) cross-correlation over 500 ms time segment, (d)
PEP
map showing well location and best match location, (e) wavelet spectra.
Based on the well tie in Fig. 4.14a phase shift of
+70° and +16 ms time shift was applied to the seis-
mic. A subsequent well tie now shows the extracted
wavelet being generally symmetrical ( Fig. 4.15 ) and
the top and base of the low impedance sand are
identifiable as a trough and peak respectively. The
asymmetry in the wavelet may be due to the incorp-
oration of noise into the wavelet estimation and it is
questionable as to whether this should be included
for example in a zero phase operator design. The
wavelet has been edited to give the zero phase equiva-
lent of the amplitude spectrum and the resulting tie is
shown in Fig. 4.16 . There is clearly no degradation in
the tie and the modelled AVO gather is unambiguous
in the responses expected from the principal litho-
logical boundaries.
Figures 4.17 and 4.18 show how the adaptive
technique might work with this dataset. The same
time
assumption that this might indicate that the data has
negative standard polarity an initial tie can be made
with a 180° wavelet (with constant phase) derived
from the estimated amplitude spectrum ( Fig. 4.14e ).
The cross-correlation is 0.65 and at this point the
interpreter might be happy enough with the result.
However, the cross-correlation plot shows significant
asymmetry ( Fig. 4.17b ), indicating that the wavelet is
unlikely to be zero phase. Rotating the phase to
50°
and time shifting the synthetic relative to the seismic
gives the tie in Fig. 4.18 , with an improved cross-
correlation of 0.71. As would be expected, it is a
similar tie to that shown in Fig. 4.14 based on the
well matching technique.
In most situations similar results are obtained
using the well matching technique and the adaptive
technique. It might be argued, however, that when
good time
depth data are available the well matching
technique offers greater flexibility in evaluating the
data as well as describing the improvements in a tie
following log editing ( Chapter 8 ) or seismic condi-
tioning ( Chapter 6 ).
-
depth relationship (based on checkshots) was
used as in the previous figures. It is evident that the
top of the main sand appears to be represented by a
peak on the original dataset ( Fig. 4.17a ). On the
-
48
 
 
Search WWH ::




Custom Search