Seismic Data Interpretation

Part 2
 Interpretation of Seismic Data
Identifying and making maps of hydrocarbon bearing
subsurface horizons (i.e. reservoirs)
Geological interpretation of seismic images
• 2D and 3D interpretation
• Structural, Stratigraphic
• Quantitative seismic modeling:
• Reflection Coefficient, Velocity, Density
• Convolutional Model, Synthetic Seismogram
• Seismic resolution

Determining reservoir properties (rock & fluids)

Geophysical and petrophysical interpretation of seismic images
• Amplitude interpretation, DHI, AVO
• Seismic attributes
• Inversion
 Interpretation of Seismic Data
Identifying and making maps of hydrocarbon bearing
subsurface horizons (i.e. reservoirs)
Geological interpretation of seismic images
• 2D and 3D interpretation
• Structural, Stratigraphic
• Quantitative seismic modeling:
• Reflection Coefficient, Velocity, Density
• Convolutional Model, Synthetic Seismogram
• Seismic resolution

Determining reservoir properties (rock & fluids)

Geophysical and petrophysical interpretation of seismic images
• Amplitude interpretation, DHI, AVO
• Seismic attributes
• Inversion
 Interpretation of Seismic Data
Determining reservoir properties (rock &
fluids)  Geophysical and petrophysical
interpretation of seismic volume.

Seismic Amplitude interpretation – DHIs

(Direct Hydrocarbon Indicators)

• Identify rules for recognizing an amplitude

(Hydrocarbon Indicators)
• Validate amplitude composition
• Relate amplitude to lithology and pore-fluid
content
• Synthetic seismogram generation
• Petrophysical modeling
 Hydrocarbon Indicators
Amplitude changes on stacked sections
Amplitude brightening (bright spot)
Dimming (dim out)
Multiple reflections associated with shallow gas
Amplitude shadow underneath hydrocarbon zones
Velocity changes
Lowering of velocity in hydrocarbon zones
Time sag underneath gas accumulations
Stacking velocity variations at edge of reservoirs
Wavelet changes
Polarity reversal (phase change at reservoir edges)
Phase shifts
Frequency changes
Lowering of frequency immediately beneath reservoir due to attenuation
Flat spot (horizontal reflection) produced by fluid-interface reflection
Gas chimney effects
Deterioration of data quality
Time sag
Distortion of reflectors
Changes in amplitude with offset
Gas and high GOR reservoir may exhibit larger reflection amplitudes as the distance between
source and receiver increases (i.e. class II AVO).
after Hilterman, 2001
Geological Constraints for Hydrocarbon
Indicators
Structural conformity
HCI limits must honor trapping mechanisms such as
rollover, up-dip pinch outs, trapping faults etc.

Down-dip limits
The down-dip limits of HCI’s are expected to exhibit a flat
event if a pore fluid interface is anticipated.
The HCI limits must conform to the structural down-dip
contours

Amplitude interpretation must make sense geologically.

Apply less “principle of least-squares” and more “principle
of least astonishment.”

after Hilterman, 2001

Amplitude interpretation
DHI’s ?

Nahm J.W. and M.P. Duhon (2003)

Bright spot
Structural conformity
Down dip limits
 DHI’s ?
P-wave C-wave

Nahm J.W. and M.P. Duhon (2003)

P-wave (left) Converted-wave (right)

Gas “chimney” or “cloud” NONE
Multiple shallow gas reflections
 DHI’s ?
P-wave C-wave

Nahm J.W. and M.P. Duhon (2003)

P-wave C-wave
Bright spot NONE
Flat spot
Amplitude dimming below gas reservoir
 DHI’s ?

Bacon M., Simm R. and T. Redshaw (2009)

Bright spot
Flat spot
Tuned gas sand
DHI’s ?

Bacon M., Simm R. and T. Redshaw (2009)

Validation of Hydrocarbon Indicators

from Brown, 1999

Amplitude Variation with Offset – AVO
Amplitude Variation with Angle- AVA

from Hilterman, 2001

AVO AVA– history and background
AVO AVA– history and background
 Propagation: how much
• interfaces
A1 Z 2  Z 1 
R 
A 0 Z 2  Z1 
A2 2 Z1
T 
A0  Z 2  Z1 

Zi  Vi  i

Z 2  Z1 2
ER 
Z 2  Z1 2
4 Z1 Z 2
ET 
Z 2  Z1 
 Propagation: how much
• interfaces
For the continuity of
stress and strain, at
the interfaces
between media with
different acoustic
impedance,
converted waves
are generated.
 Propagation: how much
• interfaces
Seismic energy partition at interfaces

(-A0 + A1)cosq1 – B1sind1 = -A2cosq2 – B2sind2 Bacon M., Simm R. and T. Redshaw (2009)
(A0 + A1)sinq1 + B1cosd1 = A2sinq2 – B2cosd2
(A0 + A1)Z1cos2d1 – B1w1sin2d1 = A2Z2cos2d2 + B2w2sin2d2
(-A0 + A1)(b1/a1)w1sin2q1 +B1w1cos2d1 = -A2(b2/a2)w2sin2q2 + B2w2cos2d2
A0= displacement of incident P-wave A1= displacement of reflected P-wave
A2 = displacement of transmitted P-wave B1 = displacement of reflected S-wave
B2 = displacement of transmitted S-wave
a P-wave velocity b  S-wave velocity
Z = acoustic impedance (p-wave) W = acoustic impedance (s-wave)
 AVO AVA– history and background
During several decades the presence of “bright
spot” - high-intensity seismic reflections – have been
qualitatively interpreted as indicators of hydrocarbon
accumulations, particularly gas. In the 80s, bright
spot exploration was considered has significantly
increasing the success ratio for wildcat gas wells.
1984 – W.J. Ostrander on the basis of previous work
by Zoeppritz, Muskat e Meres (1940), Koefoed
(1955), describes the amplitude variations with
incidence angle / offset as a function of parameters
that can be correlated with the presence of gas.
 Amplitude Variation with Offset – AVO
Simplification of Zoeppritz equations

from Hilterman, 2001

Koefoed, 1962; Ostrander, 1984; Aki and Richards, 1980; Shuey, 1985;
Rutherford and Williams, 1989; Castagna, 1993
 Poisson’s ratio
Poisson’s ratio: measure of incompressibility. The negative ratio of the transverse
strain to longitudinal strain.
s values range from 0 to 0.5.
Incompressible fluids have s  0.5.
Water has high s; a sponge has very low s.
s is sensitive to pore fluid.
At a particular depth, shales have greater s than sands, especially gas saturated
sands

from Hilterman, 2001 Bacon M., Simm R. and T. Redshaw (2009)

AVO AVA– principles
 AVO AVA– principles
Using the simplified Zoeppritz equations given by Koefoed
(1962), one can show that four independent variables exist at a
single reflecting/refracting interface between two isotropic
media:
1. P-wave velocity ratio between the two bounding media;
2. density ratio between the two bounding media;
3. Poisson’s ratio in the upper medium;
4. Poisson’s ratio in the lower medium.
These four quantities govern plane-wave reflection and
transmission at a seismic interface.
AVO AVA– principles
If Poisson’s ratio is the
same for both media, one
would conclude that angle
of incidence has only minor
effects on P-wave reflection
coefficients over
propagation angles
commonly used in reflection
seismology.
 AVO AVA– principles
But, if we impose a Poisson’s ratio variation…
AVO AVA– principles

The reason for these

variations is related to
the fact that bulk
modulus depends on
fluid properties, while
shear modulus
depends only on the
solid matrix properties.
 AVO AVA– principles

Reflection coefficients as a function of angle of incidence are

calculated for both the top and base of the gas sand.
Sand top reflection coefficients vs. angle of incidence - 0.16  -0.28
Sand bottom reflection coefficients vs. angle of incidence 0.16  0.26
 Amplitude response and AVO classifications

from Hilterman, 2001 Bacon M., Simm R. and T. Redshaw (2009)

Class I – Dim out: Decreasing amplitude with increasing angle of incidence. Sands have a
higher impedance than the encasing rock.
Class II – Phase reversal: Sands have nearly the same impedance as the encasing rock.
Zero-offset reflection is near zero and therefore large changes in reflectivity from
near to far offset can occur. A polarity change with increasing offset will occur if NI is
positive.
Class III – Bright spot: Sands have a lower impedance than the encasing medium. Will
usually have a large decrease in acoustic impedance compared to a water-filled
sand so a "bright" reflection amplitude anomaly will be observed.
Class IV – Large negative amplitude decreases slightly with offset.
from Rose and Associates, SAAM, 2003;Hilterman, 2001; Bacon M., Simm R. and T. Redshaw (2009)
AVO Interpretation – Partial Angle Stacks
Class 1 AVO – Dim out - GOM 16000’ deep turbidite

from Hilterman, 2001

from Hilterman, 2001
from Hilterman, 2001
 AVO interpretation
AVO interpretation can
be also performed
through inversion
procedures that are
based on the synthetic
seismogram computation
starting from a model of
the physical parameters.
 AVO Analysis: Intercept-Gradient Crossplots
To simplfy AVO analysis, the Shuey equation (1985) is used that
expresses the reflection coefficient in terms of the normal incidence
reflectivity R0 (referred as intercept) and the AVO gradient G
R(q) = R0 + G sin2q
Crossplot: R0 and G are calculated on every CMP gather in a seismic
survey by measuring the amplitude and calculating the fit to a plot of R(q)
vs. sin2q. The resulting R0 and G pairs are charted on the crossplot and
may give information about lithology and fluid fill.

Brine and oil sands show

separate trend lines.

Porosity (increasing to
the left) causes brine and
oil points to align along
the trend lines.

Bacon M., Simm R. and T. Redshaw (2009)

AVO – cross-plots
 Attributes
A seismic attribute is a quantitative measure of a
seismic characteristic of interest.
Analysis of attributes has been integral to reflection
seismic interpretation since the 1930s when geophysicists
started to pick travel-times to coherent reflections on
seismic field records.
There are now more than 50 distinct seismic attributes
calculated from seismic data and applied to the
interpretation of geologic structure, stratigraphy, and
rock/pore fluid properties.
 Seismic interpretations
Conventional logging programs provide sparsely sampled one-
dimensional (or “vertical”) measurements  many of the above
properties are not measured at all in a well but need to be
estimated.
A good seismic attribute either is directly sensitive to the desired
geologic feature or reservoir property of interest or allows us to
define the structural or depositional environment and thereby to
infer some feature or properties of interest.
Interpretation is based on the identification of features by
comparing them to a mental database of examples (mix of art and
science  conceptual models  pattern recognition).
One of the goals of seismic attributes is to capture interpreter
expertise by quantifying the amplitude and morphological features
seen in the seismic data through a suite of deterministic
calculations performed on a computer.
 Seismic interpretations
Many modern techniques make simultaneous use of
multiple attributes. Appropriate multiple attributes
should be properly selected:
•Independent of one another;
•associated with physical properties and features of
interest to our play or reservoir
•each attribute should capture only one type of
physical property or feature, which can then be
combined intelligently through geostatistics or other
multi-attribute analysis tools.
 Seismic attributes classification
Taner et al. (1994)  two general categories:
• geometrical attributes  enhance the visibility of the geometrical
characteristics of seismic data - dip, azimuth, and continuity
• physical attributes  relate to lithology - amplitude, phase, and frequency.
• the classification may be further divided into poststack and prestack
attributes.
Brown (1996b, 2004)  four main categories:
• time  information on structure
• amplitude  information on stratigraphy and reservoir
• frequency
• attenuation
• poststack and prestack.

Barnes (1997)  complex-trace attributes  amplitude and phase are the

fundamental attributes from which all others are derived:
• 1D, 2D, or 3D
• time or depth
• instantaneous or local
 Attributes classifications
Liner et al. (2004):
• general attributes: measures of geometric, kinematic,
dynamic, or statistical features derived from seismic data
 reflector amplitude, reflector time, reflector dip and
azimuth, complex amplitude and frequency, generalized
Hilbert attributes, illumination, edge detection/coherence,
AVO, and spectral decomposition. These general
attributes are based on either the physical or
morphological character of the data tied to lithology or
geology and are therefore generally applicable from basin
to basin around the world.
 Attributes classifications
Liner et al. (2004):
• specific attributes have a less well-defined basis in physics
or geology. While a given specific attribute may be well
correlated to a geologic feature or to reservoir productivity
within a given basin, these correlations do not in general carry
over to a different basin. There are literally hundreds of specific
attributes.
Chopra and Marfurt (2005)
• “composite” attributes (or meta attributes)  sums,
products, or other combinations of more fundamental general
attributes.
• used to display more than one attribute at a time (cross-
plot)
• combined using geostatistics, neural nets, or other
classification technology
 Seismic attributes evolution
The evolution of seismic attributes is closely linked to advances in
computer technology.
• 1960s: advent of digital recording  improved measurements of
seismic amplitude and correlation between hydrocarbon pore
fluids and strong amplitudes (“bright spots”);
• 1970s: introduction of color printers  color displays of reflection
strength, frequency, phase, and interval velocity to be overlain
routinely on black-and-white seismic records.
• 1980-90s: interpretation workstations  ability to interact quickly
with data to change scales and colors and to easily integrate
seismic traces with other information such as well logs.
•Today: very powerful computer workstations  integrating large
volumes of diverse data and calculating numerous seismic
attributes are a routine tool used by seismic interpreters seeking
geologic and reservoir engineering information from seismic data.
 Complex-trace analysis

In 1975, T. Taner and F. Koehler studied seismic wave propagation and

developed an energy-based procedure to compute the envelope of a
seismic trace. N. Neidell, introduced the Hilbert transform approach for
the same task.
The seismic trace amplitude is treated as the real part of the (complex)
analytical signal while the imaginary part of the signal is computed by
taking its Hilbert transform.
The envelope is computed by taking the square root of the sum of the
squares of the real and imaginary components, whereas the phase is
computed by taking the double argument inverse tangent of the
imaginary and real components.
Finally, the frequency is computed as the rate of change of the phase.
 Complex seismic trace

A complex seismic trace consisting of a real part x(t), which is the actual
seismic trace, and an imaginary part y(t), which is a mathematical function
calculated from the real part by a Hilbert transform. When the real and
imaginary parts are added in a vector sense, the result is a helical spiral
centered on the seismic time axis (t). This helical trace is the complex
seismic trace.
 Seismic attributes
Instantaneous seismic attributes – amplitude a(t), phase (t) and
frequency (t) that can be calculated from a complex seismic trace
using the listed equations
 Complex-trace analysis

By 1975, three principal attributes—envelope, phase and

frequency—were established and used in stratigraphy
studies:
1) Instantaneous envelope (reflection strength) is
sensitive to changes in acoustic impedance and thus to
lithology, porosity, hydrocarbons, and thin-bed tuning.
2) Instantaneous phase is useful for tracking reflector
continuity and, therefore, for detecting unconformities, faults
and lateral changes in stratigraphy.
3) Instantaneous frequency is useful in identifying
abnormal attenuation and thin-bed tuning.
Seismic attributes – Instantaneous Phase

• Good indicator of lateral continuity

• Has no amplitude information, hence all events are represented
• Detailed visualization of stratigraphic elements.

The instantaneous phase makes strong events clearer and is effective at

highlighting discontinuities of reflectors, faults, pinch-outs, angularities and
bed interfaces. Seismic sequence boundaries, sedimentary layer patterns
and
51 regions of onlap/offlap patterns often exhibit extra clarity.
 Seismic coherence (continuity)
By computing and mapping a normalized crosscorrelation
between adjacent traces in the same survey, the variability
of source-wavelet amplitude and phase could be eliminated,
and waveform continuity could be quantified (Finn’s (1986)
M.S. thesis).  Faults were easily seen and could be
tracked on the 2D section.  Seismic coherence.
The development and application of coherence to 3D
seismic data, in the form of “coherence cube” technology,
represented a noticeable improvement. Bahorich and
Farmer (1995) state that their coherence methodology was
the “first published method of revealing fault surfaces within
a 3D volume for which no fault reflections had been
recorded.”
 Seismic coherency
Their volume of coherence coefficients computed from the seismic
amplitudes on adjacent traces using a crosscorrelation technique,
clearly portrayed faults and other stratigraphic anomalies on time and
horizontal slices. The coherence images distinctly revealed buried
deltas, river channels, reefs, and dewatering features. The remarkable
detail with which stratigraphic features show up on coherence
displays, with no interpretation bias and some previously unidentifiable
event with close scrutiny, appealed to the interpreters. They had a new
view of their data.
According to SEG’s citation recognizing this contribution, “this
significantly changed the way geophysicists interpret 3D seismic data
and the way oil industry management views geophysicists’
contributions to the industry.”
Continuity (or Coherency cube)

x y “HIGH”
y

x y “LOW”
continuity ≈ 1 y

continuity ≈ 0
 Seismic coherency

An example from offshore East Coast of Canada, where northwest-southeast

faults and fractures, apparently difficult to interpret, show up clearly on
coherence time slices. Overlaying coherence on a seismic time slice provides
the interpreter with the capability to more easily name and link master and
antithetic faults.
 Continuity (or Coherency cube)
Field B
Field A

Coherency cube

56
 Dip steered continuity
continuity should be high, but is
masked by strong dip

dip-steered
input continuity
seismic
continuity

dip-steered continuity takes care of structural dip

 Curvature
• Curvature attributes are related to the presence/absence of changes in the
geometric dip of reflectors.

after Roberts, 2001

• They can map deformations related features: folds, flexures, faults.

• The combined use of continuity (i.e. coherency) and curvature can provide a
more comprehensive picture
• Always check seismic expression of curvature attribute features: what do
they correspond to in the input volume?
 Curvature
The structural geology relationship between curvature and fractures is
well established (Lisle, 1994) though the exact relationship between
open fractures, paleostructure, and present-day stress is not yet
clearly understood. Roberts (2001), Hart et al. (2002), Sigismondi and
Soldo (2003), Masaferre used seismic measures of reflector curvature
to map subtle features and predict fractures.
Curvature (a 3D property of a quadratic surface that quantifies the
degree to which the surface deviates from being planar) attribute
analysis of surfaces helps to remove the effects of regional dip and
emphasizes small scale features that might be associated with primary
depositional features or small-scale faults.
 Curvature

Figure shows minimum curvature draped over a near-basement reflection in

part of the San Juan Basin. A prominent north-south-trending incised valley is
apparent, as are some faults that strike approximately northwest-southeast.
 Curvature

Figure a) shows a time-structure map of the top of a Tertiary incised channel-levee

complex. Figure b) and c) show the dip component of curvature overlain on a 3D
representation of the horizon with shaded relief to enhance features. Note the
improvement compared to the time-structure map.
 Curvature
Curvature
attributes are
volumetric, i.e:

– do not require
interpreted
horizons
– available as 3D
volumes.
Coherency vs. Curvature
Coherency vs. Curvature
 Spectral decomposition

Spectral decomposition started with Greg Partyka

who generated some enticing images of poorly
resolved reef plays in Canada using short-window
Fourier transforms. He applied his technique to
3D data where, like many other attributes
displayed in 3D map and horizon slices, it made a
significant interpretational impact (Partyka et al.,
1999; Peyton et al., 1998). Spectral
decomposition was born and continues actively
today using several spectral estimator.
 Spectral decomposition
The concept behind spectral decomposition is that the seismic reflection from a thin bed has a
characteristic expression in the frequency domain that is indicative of its thickness in time

Note the different seismic

response at 40 Hz as
compared to 20 Hz;
much more detail can be
ascertained with the 40
Hz wavelet, however, the
20 Hz wavelet still
manifests information
about temporal bed
thickness and the
stratigraphic nature of the
deposit.

After all datasets have been produced, the reservoir interval of interest can then be
scrutinized in greater detail. This is carried out by capturing the seismic response at each
frequency subset (15.3 Hz, 29.6 Hz, 44.4 Hz, etc.) - essentially, a "screen-capture" of the
seismic image for each of these intervals can be input into an animated sequence from
lower frequencies to higher frequencies, thus revealing spatial changes in stratigraphic
thickness otherwise impossible to ascertain from the full frequency dataset. Spectral
decomposition reveals details that no single frequency attribute can match.
 Spectral decomposition
10 Hz Peak Frequency
30 Hz Peak Frequency
55 Hz Peak Frequency

Final Model
Lobe systems stacking pattern

67
Spectral Decomposition – A method to visualize it…

Simultaneous display of 3
attributes will therefore
show more of the data

40 Hz  Blue
50 Hz  Green
60 Hz  Red
 Optical stack
A “flat spot” is an horizontal seismic event
that is not following the other strata
dipping

“Flat-spots” can be enphasized by special

seismic computations (optical-stack) that
highligth horizontal continuous signal
 Optical stack

optical stacking

input seismic optically stacked

Gas chimneys (Congo offshore )

Black=Through
Red=Peak
 Gas chimneys (Congo offshore )

Gas chimneys ?

Gas chimneys

Top geobody 1
 Texture attributes
More recently the idea of studying seismic textures has been revived.
While the term was earlier applied to seismic sections to pick out
zones of common signal character (Love and Simaan, 1984), studies
are now underway to use statistical measures to classify textures
using gray-level co-occurrence matrices (Vinther et al., 1995; Vinther,
1997; Whitehead et al., 1999; West et al., 2002; Gao, 2003, 2004).
Some of the statistical measures used are energy (denoting textural
homogeneity), entropy (measuring predictability from one texel or
voxel to another), contrast (emphasizing the difference in amplitude of
neighboring voxels), and homogeneity (highlighting the overall
smoothness of the amplitude).
Energy, contrast, and entropy have been found to be the most effective
in characterizing seismic data.
 Texture

A comparison of the amplitude and energy horizon slice at the same

stratigraphic level. Notice that the channel/levee deposits can be
recognized, mapped, and detected more effectively from the energy
volume than from the amplitude volume.
 3D attributes

Rijks and Jauffred (1991)

introduced two concepts that are
now commonplace in the
interpretation workplace:
dip/azimuth maps and amplitude
extractions.
These images not only showed the
value of 3D seismic data, they also
established standard workflows
that are still accepted as best
practices today.
 Enhanced visualisation

Display of the strat cubes (subvolumes bounded by two not necessarily parallel
horizons) generated from the seismic and the coherence volumes. The
coherence strat cube indicates the north-south channel very clearly, the east-west
fault on the right side, as well as the downthrown side of the north-south fault on
the left.
Data Visualization:
Tool requirement – 3D Volume Sculpting

77
Data Visualization: 3D Volume Sculpting

78
Data Visualization:
Tool requirement – Opacity/Transparency Rendering

79
G&G Data Visualization:
Tool requirement – Multi-Volume rendering (i.e. Seismic + Continuity)

80
Geobodies

Structure visible
on timeslice

Canale interpretato in
modalità automatica
Multi-Attribute Co-Rendering (i.e. Amplitude + Coherency)

Opacity
Opaque

0
Neg 0 Po
Amplitude s

1 Opaque

Opacity
Transparent

0
Low High
Coherence
 Ant Tracking : Co-rendering Ant Tracking / Smoothed
seismic (crossline direction)

83
84
85
86
 Crossplotting of attributes
Crossplotting of attributes was introduced to visually display the
relationship between two or three variables (White, 1991).
Verm and Hilterman (1994) used crossplots in AVO analysis, which have
been used since as AVO anomaly indicators. AVO attributes are
interpreted in cross-plot diagrams that allows lithology and rock
properties to be retrieved.
When appropriate pairs of attributes are crossplotted, common lithologies
and fluid types often cluster together, providing a straightforward
interpretation. The offtrend aggregations can then be more elaborately
evaluated as potential hydrocarbon indicators.
Extension of crossplots to three dimensions is beneficial, as data clusters
hanging in 3D space are more readily diagnostic, resulting in more
accurate and reliable interpretation.
3D crossplotting Use of modern crossplotting
software of three attributes that
help identify a gas anomaly: λ-ρ
on the x-axis, μ-ρ on the y-axis,
and fluid stack on the z-axis.a -
we indicate a gas anomaly on a
time slice through the λ-ρ volume
by a blue patch. We then draw a
red polygon on the time slice
(outline) to select live data points
to be displayed in the crossplot.
The red cluster of points in Figure
b corresponds to the red polygon
and five time slices (two above
and two below the one shown).
As the crossplot is rotated toward
the left on the vertical axis, the
fluid stack shows the expected
negative values for the gas sand
(Figure c). The yellow and
magenta clusters in Figure b and
c are the corresponding
contributions from the yellow and
red polygons in Figure 17a.
Automated pattern recognition on attributes
The attribute proliferation of the 1980s resulted in an
explosion in the attribute alternatives available to
geophysicists. Besides being overwhelming, the sheer
volume of data defied attempts to gauge the information
contained within those data using conventional analytical
tools and made their meaningful and timely interpretation a
challenge.
For this reason, one school of geophysicists examined
automated pattern recognition techniques (de Figueiredo,
1982) wherein a computer is trained to determine the
patterns of interest and sift through the available bulk of
data seeking those patterns. A second school of
geophysicists began combining attributes sensitive to
relevant geological features through multi-attribute
analysis.
 Neural Network application for
multi-attribute analysis
One attempt at automated pattern recognition took the form
of neural networks (Russell et al., 1997), wherein a set of
input patterns is related to the output by a transformation
that is encoded in the network weights.
An example of how multivariate statistical analysis can be
used in determining whether the derived property volumes
are related to gas saturation and lithology is presented
(Chopra and Pruden, 2003).
For the case study from southern Alberta, it was found that
the gamma-ray logs in the area were diagnostic of sands,
and there was a fairly even sampling of well data across the
field. A nonlinear multiattribute-determinant analysis was
employed between the derived multiple seismic-attribute
volumes and the measured gamma-ray values at wells.
NN interpretation
In Figure a and b, we
show the λ-ρ and μ-ρ
sections with the anomaly
enclosed in yellow
polygons. The crossplots
for these two attributes
are also shown (Figure c).
The yellow dots on the
crossplots represent the
values within the
polygons in Figure a and
b.
Multi-attribute
The magenta polygon in c
indicates where we would
expect to find gas sands in λ-ρ
and μ-ρ space in Figure a and b,
respectively.
 Multiattribute analysis
By training a neural network with a statistically representative
population of the targeted log responses (gamma ray, sonic, and bulk
density) and the multiple seismic-attribute volumes available at each
well, a nonlinear multiattribute transform was computed to produce
gamma-ray and bulk-density inversions across the 3D seismic
volume.
Cubic B-spline curves (mathematical representation of the
approximating curves in the form of polynomials) have also been used
for determination of mathematical relationships between pairs of
variables for well logs; those relationships were then used to invert
attribute volumes into useful inversion volumes such as gamma ray
and porosity (Chopra et al.,2004).
 Inversion
The principle objective of seismic inversion
is to transform seismic reflection data into a
quantitative rock property, descriptive of the
reservoir. In its most simple form, acoustic
impedance logs are computed at each
CMP. Compared to seismic amplitudes,
inversions show higher resolution and
support more accurate interpretations.
 Seismic impedance inversion

Inverted seismic section from the Swan Hills Devonian reef bank for
prediction of carbonate porosity.
 Seismic inversion revisited
The original recursive or trace-integration seismic inversion technique
for acoustic impedance also evolved during the late 1980s and 1990s,
with developments in model-based inversion, sparse-spike inversion,
stratigraphic inversion, and geostatistical inversion providing accurate
results (Chopra and Kuhn, 2001). The earlier techniques used a local
optimization method that produced good results when provided with an
accurate starting model. Local optimization techniques were followed
by global optimization methods that gave reasonable results even with
sparse well control. Connolly (1999) introduced elastic impedance,
which computes conventional acoustic impedance for nonnormal angle
of incidence. This was further enhanced by Whitcombe (2002) to reflect
different elastic parameters such as Lame’s parameter λ, bulk modulus
κ, and shear modulus μ.
 Seismic inversion

From: Berge et al., 2002, TLE

 Seismic inversion

From: Berge et al., 2002, TLE

 Seismic inversion

From: Berge et al., 2002, TLE

Inversion