The document summarizes seismic reservoir characterization services offered by ARCIS, including defining reservoir geometry, distributing physical properties within reservoirs, and correlating available details. The services include computing attributes like coherence, curvature, and reflectivity to analyze structures, stratigraphy, and thin layers. Other techniques include AVO analysis, spectral decomposition, and impedance inversion to understand lithology and fluid distribution.

1. ASSISTING OUR CLIENTS IN THE DEFINITION OF RESERVOIR GEOMETRY, THE DISTRIBUTION OF PHYSICAL
PROPERTY CHARACTERISTICS WITHIN A RESERVOIR, AND IN THE CORRELATION OF ALL AVAILABLE DETAILS,
ARCIS OFFERS A STATE-OF-THE-ART SUITE OF SEISMIC RESERVOIR CHARACTERIZATION SERVICES.
Coherence attributes
Curvature attributes
Multi-spectral estimates of curvatures
Reflector convergence
Reflector rotation
Fracture analysis using curvature attributes
Visualization of attributes
Thin-bed reflectivity
Correction for spurious phase via estimation of non-minimum phase wavelet
Rock physics analysis
AVO/LMR analysis
Spectral decomposition
Impedance inversion
A TGS COMPANY

2. A TGS COMPANY

3. Reservoir Analysis
Coherence attributes
Coherence computes a measure of similarity between adjacent traces. Identification and mapping of channel edges, reefs, faults and fracture systems in 3D
volumes becomes so much easier with the coherence volume. Structural and stratigraphic interpretation of seismic data get facilitated by using a coherence
volume. Arcis offers coherence computation services based on Energy Ratio algorithm (a modified eigen-decomposition of covariant matrices approach).
This algorithm has generated results better than those from the standard semblance as well as eigen-decomposition based algorithms.
Coherence using semblance
without dip-steering option
Coherence using
Energy Ratio
HIGHLOW HIGHLOW
Coherence using semblance
Coherence using Energy Ratio
and structure-oriented filtering
HIGHLOW
HIGHLOW
Data courtesy: Olympic Seismic Ltd., Calgary

4. www.arcis.com
Curvature attributes
Curvature attributes measure the degree of bending of seismic reflections along a surface or in a volume and help to improve interpretation and
structural understanding of 3D seismic data volumes. Different curvature attributes identify subtle faults, fractures, and other features better than other
attribute applications. Arcis offers volume computation of curvature, together with their spectral estimates.
Horizon slice through coherence volume Horizon slice through most-positive curvature volume
Horizon slice through most-negative curvature volume Horizon slice through merged volume comprising coherence, most-positive
curvature, and most-negative curvature volumes using transparency

5. Reservoir Analysis
Multi-spectral estimates of curvatures
Multi-spectral curvature estimates can yield both long and short wavelength curvature images, allowing an interpreter to enhance geologic features
having different scales. Short-wavelength curvature often delineates details within intense, highly localized fracture systems. Long-wavelength
curvature often enhances subtle flexures that are difficult to see in conventional seismic, but are often correlated to fracture zones that are below
seismic resolution, as well as to collapse features and diagenetic alterations that results in broader bowls.
Time slices 1160 ms
Coherence
Most-positive
(long-wave)
Most-positive
(short-wave)

6. www.arcis.com
Reflector convergence
Reflector convergence attribute is useful in the interpretation of angular unconformities. It is a measure of the change in reflector normal
about a more or less horizontal axis.
Time slices through a coherence volume at (left) t=2.700 s and (right) t=2.670 showing two channel system. The channel is clearly delineated (yellow
arrows) in the deeper slice (left) but not at the shallower slice (right).
The same two time slices at (left) t=2.700 s and (right) t=2.670 now co-rendered with reflector convergence. Note the change in convergence towards
the edges of the channel within the deeper slice shown in (left) and away from the center of the channel in the shallower section seen in (right).
A’
N
S
W E
A’
A’
A
N
N N
N
A’

7. Reservoir Analysis
Reflector rotation
Time slice at 1.190 s from the coherence volume and vertical slices through seismic amplitude co-rendered with vector rotation. Red indicates down to
the right across the fault, while blue indicates up the right across the fault.
Reflector rotation attribute quantifies the rotation of fault blocks across discontinuities such as wrench faults. It is a measure of the
change in reflector normal about a more or less vertical axis.

8. www.arcis.com
Fracture analysis using curvature attributes
Interpretation of lineaments corresponding to subtle faults and trends can be carried out on the most-positive or most-negative curvature
horizon slices as shown. The interpreted lineaments are then converted into a rose diagram which can be compared with a similar rose
diagram available from image logs.
NEG
POS
Most-positive curvature
Rose Diagram
90270
0
180
Interpretation on curvature displays

9. Reservoir Analysis
Visualization of attributes
A stratal fault skeleton from the most-positive curvature attribute being correlated
with seismic data volume in the increasing inline direction as indicated with the
white arrow
Strat-cube from the most-
positive curvature attribute
co-rendered with
coherence seen here in a
3D chair view
Strat-cube from the
most-negative curvature
attribute co-rendered with
coherence seen here in a
3D chair view

10. www.arcis.com
Thin-bed reflectivity
Arcis offers thin-bed reflectivity inversion using a spectral inversion technique that produces ultra-high resolution seismic data that
enhances mapping. Thin-bed reflectivity removes the deleterious effects of the seismic wavelet that cause degraded resolution.
Thin-bed reflectivity is performed without using well data, requires no a priori model, no interpreted horizons and no assumed
reflectivity spectra.
Relative acoustic impedance
derived from thin-bed
reflectivity shows features
that are of interest and that
otherwise are not seen on the
input data
Seismic
Input seismic Reflectivity convolved with bandpass wavelet
Reflectivity

11. Reservoir Analysis
Correction for spurious phase via estimation
of non-minimum phase wavelet
Deconvolution with minimum phase wavelet assumption usually leaves the data with spurious phase. The phase correction required to remove
the spurious phase is done by a simple parameterization of the underlying mixed phase wavelet, which involves estimation of an all-pass operator
coefficients via cumulant matching technique. The approach involves setting up a cost function with unknown shape and using a simulated annealing
algorithm for optimization.
Original data Phase corrected data
Original data. The inserted curve is the P-velocity log Phase corrected data.
The inserted curve is the P-velocity log

12. www.arcis.com
Rock physics analysis
For a better understanding of lithology and fluid differentiation, Arcis emphasizes the importance of rock physics analysis and
integrating that with the AVO/LMR processing flow. Arcis uses an informative and straight forward approach to exhibit the elastic and
petrophysical properties of reservoir rocks.
VP against VS
P-impedance against S-impedance
Lambda-Rho against Mu-Rho
VP against VP/VS
VP against Poisson’s ratio
Lambda-Rho against Lambda/Mu

13. Reservoir Analysis
AVO/LMR analysis
Arcis has the capability and innovativeness to accomplish AVO/LMR analysis including AVO friendly processing, QC and interpretation.
Overlay of angle information on offset gathers
Lambda-Rho
Mu-Rho
Input gathers Reconstructed gathers Difference
NEG
POS
LOW HIGH
3D Cross-plotting of
Lambda-Rho,
Mu-Rho and Fluid Stack.
Clusters associated with
gas anomalies (yellow)
separated out

14. www.arcis.com
Spectral decomposition
Spectral decomposition allows utilization of the discrete frequency components of the seismic bandwidth to interpret and understand the subtle
details of subsurface stratigraphy. Besides the traditional Fourier transform, Arcis offers the CWT (Continuous Wavelet Transform), Matching
Pursuit Decomposition (MPD) and Exponential Pursuit Decomposition (EPD) methods for transformation of data to frequency domain. The latter two
techniques offer more accurate analysis.
Horizon slice from a seismic volume
Horizon slice through the seismic volume and passing through the
Doig sandstone
Equivalent horizon slice from merged frequency volume
(with 20 Hz, 30 Hz, and 40 Hz) and using RGB with seismic
Waveform classification map generated by hierarchichal
unconstrained waveform classification. Notice the higher level of
facies detail that is seen on this display compared with the seismic
amplitude alone, within the sandstone boundary drawn in black.
Seismic facies classification
Seismic waveforms in the broad zone of interest are categorized and related to different depositional facies.
Arcis offers both unsupervised (statistical) as well as neural network (deterministic) methods for arriving at meaningful and convincing facies analysis.

15. Reservoir Analysis
Impedance inversion
Arcis offers several different techniques/methodologies to perform acoustic impedance inversion.
Post-stack impedance inversion
• Recursive inversion
• Model-based inversion
• Sparse-spike inversion
Pre-stack impedance inversion
• Simultaneous inversion
• Elastic impedance
• Extended elastic impedance
Segment of a section from acoustic impedance volume generated using model-based inversion
Segment of a section from acoustic impedance volume generated using simultaneous inversion

16. www.arcis.com
Synthetic seismogram tie for the (left) low-angle (~10°) near-stack. The
synthetic seismogram was generated by using the impedance log curve.
The correlation seems to be reasonably good.
Synthetic seismogram tie for the high-angle (~30°) far-stack. The synthetic
seismogram was generated by using the computed El(30°) log curve. The
correlation seems to be reasonably good. Notice the weakening of the
amplitudes at the location of the orange arrows.
Impedance inversion
Elastic impedance provides a convenient way of producing synthetic seismograms for variable angles of incidence, and combines
the benefits of working with inverted data with far-angle data where the fluid information resides.
Top: Segment of the acoustic impedance section
(with the overlaid AI log) shows anomalously low
values of impedance at the gas-producing zone (in
yellow highlighted zone).
Below: Equivalent segment from EI (30°) section
showing the anomaly as much more pronounced.
(Data courtesy: PetroNorte, Colombia)
Angle-dependent inversion

17. Reservoir Analysis
Extended elastic impedance inversion results for (top) bulk modulus, (middle) Young’s modulus, and (bottom) V-shale attributes. A good lithology and
porosity variation can be interpreted considering both bulk modulus and Young’s modulus. The middle part of the V-shale zone of interest is showing high
volume of shale as was expected.
Extended elastic impedance approach allows the projection of seismic amplitudes to angles of incidence outside of the recorded angles, which
helps express the rock properties in terms of impedance values. A relationship between the extended elastic impedance and reservoir properties is
investigated by correlation analysis (as a function of new angle, Chi) between EEI logs and available petrophysical logs such as V-shale, porosity and
saturation, as well as lithology logs, such as gamma ray. Once the optimum angle Chi is determined, AVO analysis is carried out to obtain intercept
and gradient attributes, and their linear combination allows the determination of the reservoir properties.
Extended elastic impedance inversion

18. www.arcis.com
P-impedance obtained from post-stack inversion. The inserted black curve is the P-impedance log
P-impedance estimated with probabilistic neural network. The inserted black curve is the P-impedance log
Impedance
(m/s² g/cc)
Multi-attribute analysis
Arcis offers neural network or cubic-b spline based multi-attribute analysis. Such an analysis is beneficial if the 3D volume has a reasonably good well
control and the wells are representative of the geology in a lateral sense. Application of Probabilistic Neural Network (PNN) to estimate the P-impedance
volume provides more detailed information compared to the conventional model-based P-impedance inversion. As seen in the examples below, the PNN
application for acoustic impedance shows better correlation with the impedance log than the conventional P-impedance inversion.

19. Reservoir Analysis
Arcis Seismic Solutions, a wholly owned subsidiary of TGS-NOPEC Geophysical
Company, offers seismic solutions to the energy industry including seismic data
processing, reservoir analysis, advanced imaging, multi-client surveys, geotechnical
services, project management, data marketing and access to an extensive data
library. Arcis offers one of the most current 2D and 3D seismic data libraries for
the Western Canadian Sedimentary Basin, including Northeast British Columbia. We
are committed to exceptional customer service, superior data quality, innovation, and
integrity; while maintaining a focus on health, safety, and environmental stewardship.
We think different. We think seismic.

20. 2100, 250 - 5TH
Street SW Calgary, Alberta, Canada T2P 0R4 403.781.1700 or 888.269.6840 info@arcis.com www.arcis.com
Printed in Canada
A TGS COMPANY