Geology Reference
In-Depth Information
a)
b)
Well 1 Well 2 Well 3 Well 4 Well 5 Well 6
-300
50m
-200
model wells
scaler
-100
0
0 0 0 0 0
Apparent thickness (ms)
c)
d)
-0.08
40
model wells
calibration curve
-0.06
30
-0.04
20
model wells
unity
10
-0.02
0
0
0 0 0 0 0
0
10
20
30
40
Apparent thickness (ms)
Net pay thickness (ms)
Figure 10.15
Model well dataset illustration of Connolly's(
2007
) net pay prediction technique; (a) model dataset based on a real field
example, (b) scaler curve scaled to minimize error at wells, (c) tuning plot with calibration curve (i.e. inverse of scaler), (d) net pay prediction
using scaler compared with actual net pay in model wells (from Simm
2009
).
contrasting average bandlimited impedance with
sum of negative amplitudes. It is evident that whilst
SNA shows a tuning effect, the curve has a different
form and tuning occurs at around 130 ft in contrast to
50
40
30
-
60 ft with average bandlimited impedance.
20
10.3.4.2 Pseudo-wells and uncertainty
Neff (
1990a
,
1990b
,
1993
) considered the amplitude
scaling methodology of Brown et al.(
1984
,
1986
)to
be too simplified and determined that the problem of
analyzing the complex relationships between trough-
to-peak amplitudes, isochrons and reservoir param-
eters requires a more rigorous approach. Clearly the
nature of the relationship between amplitude and
thickness is specific to particular geological scenarios.
Thus it is useful to generate pseudo-wells that are
consistent with known geological relationships. Ini-
tially this was done by varying the amount of pay in
well-based synthetic models but this can be extended
using statistical rock physics and geological modeling
of layers and facies.
Figure 10.19
shows a case in
which thickening of the pay interval is associated with
model wells
unity
10
0
0
10
20
30
40
Actual net pay thickness (ms)
Figure 10.16
Model well dataset: net pay thickness prediction
using Connolly's(
2007
) technique with a linear scalar (after
Simm,
2009
).
horizon picking, it may be more appropriate to use an
attribute that is less sensitive to layer thicknesses. One
such attribute is sum of negative amplitudes (SNA).
Figure 10.18
illustrates a bandlimited wedge model
234
