Geology Reference
In-Depth Information
a)
b)
Figure 10.29
Pressure and saturation
effects on an initially oil filled sandstone
reservoir; (a) AI vs GI crossplot showing
changes in AI and GI relative to an initial
oil filled sand (green point) (modified
after Ricketts et al.,
2008
); (b) AI vs V
p
/V
s
crossplot showing change as the ratio of
monitor to baseline values (modified after
Andersen et al.,
2006
. Copyright 2006, SPE,
reproduced with permission of SPE,
further reproduction prohibited without
permission). Note that
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.7 0.8 0.9 1.0 1.1 1.2 1.3
AI change
+
+
Δ
S
w
+
Δ
P
+
Δ
S
w
+
Δ
P
+
Δ
P
+
Δ
S
w
χ
Δ
GI
-
Δ
P
P ¼
pore pressure,
+
Δ
S
g
+
Δ
S
g
-
Δ
P
S
w
¼
water saturation and
S
g
¼
gas
+
Δ
S
g
-
Δ
P
saturation.
-
-
+
Δ
AI
high blue amplitudes in
Fig. 10.28
) the original oil
water contact and the new oil water contact are deter-
mined by the lower and upper zero crossing picks
respectively. The zone of no difference to the right in
the figure is interpreted as a zone of remaining oil,
representing a potential infill target. Clearly the inter-
pretation of a
water injection) will give a response that falls between
the fluid and pressure vectors. The opposite will occur
if the pressure drops at a producer and oil is replaced
by gas. In principle the 4D effects of fluid change and
of pressure change can be separated by using appro-
priate chi angle projections (
Chapter 5
). Relative
changes in saturation will be optimized at relatively
small projection angles whereas pressure effects will
require large projection angles (e.g. Nunes et al.,
2009
). Inevitably this means that the pressure signals
are likely to be more prone to noise (
Chapter 5
). Using
absolute impedance inversions, changes in saturation
and pressure can be deduced by the relative changes in
acoustic impedance and Poisson
zone depends to a large
extent on the understanding of the reservoir based on
the baseline interpretation. Relatively low N:G zones
which are poorly connected to the main reservoir units
might give a similar signature.
The whole range of reflectivity and inversion tech-
niques described so far in this topic can be brought to
bear on time-lapse interpretation. For example, Con-
nolly
'
no difference
'
s ratio (
Fig. 10.29b
).
Clearly, the fact that time-lapse signatures are often
tuned responses gives rise to potential ambiguities in
interpretation. For example, in a layered reservoir
where the sweep is uneven it may not be possible to
determine in which layer the time-lapse effects have
been generated. Stochastic inversion methods discussed
in
Chapter 9
could be used to address this problem (e.g.
Gawith and Gutteridge,
2001
;Veireet al.,
2007
).
Rather than estimate reservoir changes directly
from AVO signatures or inverted impedances, some
authors approach the problem by using nonlinear
multi-attribute methods in which correlations
between production data (i.e. changes in fluid satur-
ation and effective pressure) and a variety of seismic
attributes are established (e.g. Sønneland et al.,
1997
;
MacBeth et al.,
2004
, Ribeiro and Macbeth,
2006
).
Figure 10.30
shows an example of saturation and
pressure mapping based on such an approach. Ribeiro
and Macbeth (
2006
) describe time-lapse inversion
directly for the fluid modulus and pore pressure.
Uncertainty estimation can be incorporated into the
'
s(
2007
) net pay technique might be used to
estimate the change in net pay from the types of
responses shown in
Fig. 10.28
. Saturation and pres-
sure changes are often approached from the perspec-
tive of AVO, either in the reflectivity or inverted
impedance domain (e.g. Tura and Lumley,
1999
;
Landrø,
2001
). The AVO component can help in
addressing potential interpretation ambiguities, for
example determining the difference between a rise in
impedance due to pressure drawdown and an increase
in water saturation due to oil production.
Figure 10.29
shows a schematic diagram illustrat-
ing the relative elastic effects of pressure and satur-
ation on a consolidated sandstone reservoir. If an oil
sand positioned at the origin of the
'
GI plot
(
Fig. 10.29a
) is replaced by water, the point will move
to the upper right (thin blue vector), whereas replacing
oil by gas moves the point along the thin red vector.
Pressure changes are likely to be orthogonal to the
fluid change vectors. The combined effect of increas-
ing water saturation and pore pressure (i.e. during
Δ
AI/
Δ
243
