Geology Reference
In-Depth Information
a)
b)
b
i
b
i-1
Higher
probability
of red
0.
5 m
ile
Lower
probability
of red
1
0
a
j-1
a
1
Channel probability
Attribute A
Figure 10.3
Estimating probabilities of channel sands in the fluvial Lance Formation of the Green River basin, onshore USA; (a) principle
behind the probability modelling (attributes A and B are P and S impedance respectively), (b) depth slices through 3D channel probability
cubes (left - using 39 wells to generate the conditional probability model and right - using four wells), after Michelena et al.(
2009
).
Probabilit
Well 10
Well 08
Well 07
1.0
0.9
300
0.8
350
0.7
0.6
400
0.5
450
0.4
500
0.3
0.2
550
0.1
600
0.0
Figure 10.4
Probability of sand occurrence based on 50 stochastic realisations (after Francis
2002
).
addressing the ambiguities that arise in AVO analysis.
For example in situations where wet sands have low
impedance, pre-stack simultaneous inversion may be
able to help in differentiating the relative contributions
of porosity and saturation as well as accounting for
thickness effects on amplitude.
A good example of the potential of pre-stack sim-
ultaneous inversion for reservoir property mapping
has been described from the Marlin Field in the Gulf
of Mexico by Russell et al.(
2006
). Both full stack
amplitudes and the fluid factor AVO attribute (effect-
ively a projection at
225
26.2
0
) show relatively poor
χ¼
