Sie sind auf Seite 1von 4

The AIGI crossplot as an aid to AVO analysis and calibration

David N. Whitcombe*, BP, Aberdeen, UK,


John G. Fletcher, BP, Sunbury-on-Thames, UK.
mathematical transform, it can be visualized as a rotation
angle on the AIGI crossplot.

Summary
The AIGI crossplot has been found to be an effective way
of visualizing AVO behavior and tying rock properties to
seismic data. Gradient Impedance, GI, is displayed on the
y-axis, while Acoustic Impedance, AI, is displayed on the
x-axis. The crossplot is therefore the impedance form of the
traditional intercept vs. gradient reflectivity crossplot.
While AI is a function of the p-wave velocity, , and
density, , GI is additionally a function of s-wave velocity,
.

Theory
Whitcombe et. al. (2000) introduced Extended Elastic
Impedance, EEI, as:
EEI() = p q r

..(1),

where:
Extended Elastic Impedance, EEI, was defined to range
between AI and GI controlled by the variable angle
, which is related to , the angle of incidence at the target
horizon. The angle can be regarded as a rotation angle on
the crossplot. EEI can be considered as a projection in
AIGI space.
The AIGI crossplot allows a seismic AVO projection to be
easily determined to maximize a particular property, such
as the separation between hydrocarbon and brine sands, or
between sands and shales.

p = (cos + sin),
q = -8Ksin,
r = (cos-4Ksin)

and K is a constant, usually set to the average value of


(/)2 over the log interval of interest.
and are related (Simm et. al., 2000) by the relationship:
tan = sin2

Introduction
Traditionally, AVO work is done in the reflectivity domain,
and the intercept vs. gradient (or A vs. B) crossplot is an
established tool for the identification and analysis of
hydrocarbon anomalies, particularly those associated with
gas sands (Castagna and Swan, 1997).
Connolly (1999) introduced Elastic Impedance to enable
the calibration of inverted far offset data sets. EI is a
function of , , , and . EI tends to AI as tends to zero.
EI allows the benefits of inversion (Buxton Latimer et. al,
2000) to be enjoyed while using data that is better tuned to
hydrocarbon effects. Whitcombe (2001) refined the
definition of EI to remove the dependence of its
dimensionality on the angle , by using normalizing
constants. Whitcombe et. al. (2000) introduced Extended
Elastic Impedance, EEI, as the generalization of Elastic
Impedance, which allows inversions to be carried out on
data that is tuned either for lithology or fluids.
Additionally, the concept of Gradient Impedance, GI, was
introduced. A change in GI results in the generation of the
gradient reflection coefficient, B. EEI is a function of , ,
, and . EEI tends to AI as tends to zero, and tends to GI
as tends to 90 degrees. While was introduced as a

..(2),

.(3).

Whitcombe (2001) refined the definition of EI to remove


the dependence of its dimensionality on the angle , by
using the normalizing constants 0, 0, 0, which are set to
average values. This normalization can also be applied to
EEI:
EEI() = oo [(/o)p ( /o)q (/o)r]

....(4),

Gradient impedance is now the value of EEI at = 90


degrees. At this angle the exponents are:
p=1
q = -8K
r = -4K

.(5).

Example
We now take a reservoir sand and fluid substitute between
the gas, oil and brine states. Table 1 shows the , ,
values for the three fluid states for a N Sea reservoir.

SEG Int'l Exposition and Annual Meeting * San Antonio, Texas * September 9-14, 2001
Downloaded 11 Mar 2010 to 210.212.83.82. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

The AIGI crossplot

(g/cm3)
2.15

Oil sand

2772

1470

2.23

Brine sand

3057

1456

2.31

Table 1: , , values from a N Sea reservoir.


Taking the averages of these values we obtain o=2843
m/s, o =1475 m/s, o =2.23 g/cm3 and K=0.272. The AI
and GI values may now be calculated:

Gas sand

(m/s*g/cm3)
5807

GI
(m/s*g/cm3)
6053

Oil sand

6182

6230

Brine sand

7062

6751

Table 2: AI and GI values calculated from the values

R = A + B tan

.(7)

Note that we have established the fluid projection, without


reference to any shales or shale baseline. Figure 2 shows
the same AIGI crossplot but now contains a point for the
reservoir shales.

8000
7500
7000

Brine

6500

id
on
Flu jec ti
pr o

Oil

Gas
6000

5500
5000
5000

5500

6000

6500

7000

7500

8000 8500

Acoustic Impedance (m/s*g/cm3 )


Figure 1: The AIGI crossplot generated using rock
properties from a N Sea field. The general relationships
shown here are common to all reservoirs, although the
exact angle , and the separation of the points will depend
on the details of a particular reservoir.
8500
8000

Shale

7500
7000

Brine

6500

Oil

Gas
6000

5500
5000
5000

5500

6000

id
on
Flu jec ti
o
pr

id
-flu
th o o n
Or jec ti
pr o

Figure 1 shows the AIGI crossplot for these points. The


details of the orientation differ for different reservoirs
depending on the rock properties, but the general form of
this relationship is common to all reservoirs: As we move
from gas sand to oil sand, both the AI and GI values
increase. As we move from the oil state to the brine state,
both impedances further increase. These points do not fall
exactly on a straight line; the gas to oil gradient is usually
slightly less than the oil to brine gradient. However, to a
first approximation these points can be considered to fall on
a straight line, in this case orientated at 29 degrees to the
horizontal. We call this the fluid projection. We maximize
the fluid effect by projecting data onto this line. This is
equivalent to viewing the data from a direction at 90
degrees from the projection line. It is also equivalent to
calculating the EEI of the data (equation 4) at a projection
of =29 degrees. Because of the relationship between
and (equation 3), this is equivalent to inverting a
seismic data set calculated using:

8500

Gradient Impedance (m/s*g/cm3 )

(m/s)
1499

Gradient Impedance (m/s*g/cm3 )

Gas sand

(m/s)
2701

6500

7000

7500

8000 8500

Acoustic Impedance (m/s*g/cm3 )


Figure 2: The AIGI crossplot additionally showing the
position of a reservoir shale and the ortho-fluid projection,
designed to minimise the separation between the three fluid
phases. In this case the ortho-fluid projection provides a
good shale-sand, or lithological projection.

SEG Int'l Exposition and Annual Meeting * San Antonio, Texas * September 9-14, 2001
Downloaded 11 Mar 2010 to 210.212.83.82. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

The AIGI crossplot

8000
7500

Shale

7000

Brine

6500

Oil

Gas
6000

5500
5000
5000

5500

6000

6500

7000

Acoustic Impedance

Region C

Region B

Region D

id tion
flu jec
p ro
45 o

Region A

Acoustic Impedance

ui d
-fl
th o o n
Or jec ti
o
pr

Gradient Impedance (m/s*g/cm3 )

8500

Region D is the worst possible place for shales to plot as


the hydrocarbon sands will have impedances that lie
between the shales and the brine sands. Region C is better,
as the hydrocarbon sands will have the low impedances,
and the shales the high impedances, with brine sands as
intermediate values.

Gradient Impedance

Additionally we display the ortho-fluid projection. This is


the projection at right angles to the fluid projection. We see
that imaging along the ortho-fluid projection minimizes the
fluid effect: the three different fluid phases will appear to
have the same impedance when projected onto this line.
The sands, however, will appear to have a different
impedance to the shales. This projection is therefore a good
sand-shale discriminator, or lithological projection. By
generating the fluid and ortho-fluid projections, we are
therefore able to separate lithology and fluid effects.

7500

8000 8500

(m/s*g/cm3 )

Figure 3: The AIGI crossplot showing the case where the


reservoir shales fall on the fluid projection. In this situation
there is poor separation of lithology and fluids using fluid
and ortho-fluid projections.

Consider now a further case, where the reservoir shales plot


along the fluid projection (Figure 3). An ortho-fluid
projection is now no longer is able to distinguish lithology.
We have lost the ability to independently separate the
effects of lithology and fluid. The extra dimension that
AVO provides has no longer bought us an extra degree of
freedom over what we could achieve with just AI data.
These concepts enable us to define four regions in AIGI
space (Figure 4). If the shales plot in regions A and B
relative to the reservoir rocks we have the potential to
separate out lithology and fluid effects using fluid and
ortho-fluid projections. If the shales plot in regions C or D,
we have poor potential for lithology and fluid separation
using fluid and ortho-fluid projections.

Figure 4: Illustrating how AIGI space can be divided into


four regions, A, B, C and D. The ability to separate
lithology and fluids using fluid and ortho-fluid projections
is best when the shales are located in Region A relative to
the sands, and worst in Region D.
Region A is the best possible place for shales to plot
relative to the sands, as any porosity increase in the sands
will drive the sand points further away from the shales,
thereby increasing the impedance separation. In Region B,
a porosity improvement has the undesirable effect of
driving the sand points towards the shales.
Thus the order A, B, C and D represents a best to worst
ranking order for the being able to separate lithology and
fluid effects. Note also that the impedance separation
between the sands and the shales will generally improve the
ability to separate lithology and fluid effects using fluid and
ortho-fluid projections.

Further example
In this example we take log data from a N Sea well. The
AIGI crossplot is shown in Figure 5. In this diagram we
highlight the oil sands in red, the brine sands in blue and
the shales in green.
The shale data fall in Region B and it should be possible to
generate two seismic projections that separate lithology and

SEG Int'l Exposition and Annual Meeting * San Antonio, Texas * September 9-14, 2001
Downloaded 11 Mar 2010 to 210.212.83.82. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

The AIGI crossplot

fluid effects. The caveat is that within Region B, significant


porosity variations may corrupt the lithological calibration,
with high porosity sands looking like shales.
In Figure 5 we also show a fluid projection line through
these data, calculated from average properties for the
reservoir sands. This direction maximizes the fluid effect.
The ortho-fluid projection will provide a lithological
projection, independent of fluid type.

Gamma Ray

Pay
Sand

Shale:Sand:Oil:Water

into one area of the AIGI crossplot, rather than being split
into two quadrants, as is the case for reflectivity data.
This approach is subject to the same approximations that
underlie EI and EEI theory. The approximations provide
high accuracy when the reflection coefficients are small.

References
Buxton Latimer, R., Davidson, R., and van Riel, R., 2000,
An interpreters guide to understanding and working with
seismic-derived acoustic impedance data: The Leading
Edge, 19, No 3, 242-256.
Castagna, J., and Swan, H., 1997, Principles of AVO
crossplotting: The Leading Edge, 16, No 4, 337-342.
Connolly, P., 1999, Elastic Impedance: The Leading Edge,
18, No. 4, 438-452.

Gradient Impe dance(m/s*g/cm3 )

Simm, R., White, R., and Uden, R., 2000, The anatomy of
AVO crossplots: The Leading Edge, 19, No 2, 150-155.

Key
Shale

Whitcombe, D.N., Connolly, P.A., Reagan, R.L. and


Redshaw, T.L., 2000, Extended elastic impedance for fluid
and lithology prediction: 70th Ann. Internat. Mtg., Soc.
Expl. Geophys., Expanded Abstracts, xxx-xxx.
Whitcombe, D.N., 2001, Normalizing Elastic Impedance:
Geophysics, In press.

Oil Sand
Brine Sand

Acknowledgements

Acoustic Impedance (m/s*g/cm3)

We thank Roger Reagan, Pat Connolly and Terry Redshaw


for constructive comments on this work. We would like to
thank the Board of BP for permission to publish this paper.

Figure 5: Log sections and AIGI crossplot derived from a well in


the N Sea. The fluid projection (shown) will maximise the
separation of oil and brine sands. The ortho-fluid projection
provides lithological (sand-shale) discrimination.

Conclusions
The AIGI crossplot is a very convenient way of viewing
rock properties and leads to a quick understanding of the
way to image the data for a desired outcome, such as
maximizing the hydrocarbon or lithological effects.
Crossplot interpretation of seismic anomalies should be
easier in the impedance domain than the reflectivity
domain, as the anomalous hydrocarbons will be focused

SEG Int'l Exposition and Annual Meeting * San Antonio, Texas * September 9-14, 2001
Downloaded 11 Mar 2010 to 210.212.83.82. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

Das könnte Ihnen auch gefallen