Construction of Structurally and Stratigraphically Consistent Structural Models using the Volume-Based Modelling Technology (IPTC-18216-MS)

IPTC-18216-MS
Construction of Structurally and
Stratigraphically Consistent Structural Models
Using the Volume-Based Modelling Technology:
Applications to an Australian Dataset
Laurent Souche1, Gulnara Iskenova1, Francois Lepage1,2 and David
Desmarest1
1 2

Problems Solution Application
© 2014 Schlumberger. All rights reserved.
An asterisk is used throughout this presentation to denote a mark of Schlumberger.
Other company, product, and service names are the properties of their respective
owners.
IPTC logo by the International Petroleum Technology Conference (http://iptcnet.org)
The following slides were presented at the International Petroleum Technology
Conference in Kuala Lumpur, in December 2014.
They have been edited to include “Presenter notes”
The corresponding paper is available from the OnePetro online library:
https://www.onepetro.org/conference-paper/IPTC-18216-MS

OUTLINE
IPTC-18216-MS• Construction of Structural Models Using the Volume-Based Modelling Technology • Laurent Souche
1 Problem(s)
Interpretation
Model
“This presentation addresses the general problem of
building a geological model from interpretation data.”

OUTLINE
1 Problem(s)
Interpretation
Model
“Three topics will be covered in more details:
• The interpolation of the model in
presence of sparse/incomplete data ;

OUTLINE
1 Problem(s)
Interpretation
Model
• The incorporation of dense data, and of
complex geological constraints ;

OUTLINE
1 Problem(s)
Interpretation
Model
• The incorporation of dense data, and of
complex geological constraints ;
• The improvements that an initial model
can bring to geological interpretations.”

OUTLINE
1 Problem(s)
Interpretation
Model
2 Solution
Relative
geological age
“The technical solution we are
proposing is based on the
interpolation of a relative
geological age attribute in the
volume of interest. This is the
volume-based modeling (VBM)
technique.”

OUTLINE
1 Problem(s)
Interpretation
Model
2 Solution
Relative
geological age
3 Application
x
“A case study detailing the
construction of a 3D model from
a dataset located offshore
Australia will be presented to
demonstrate this technique.”

OUTLINE
1 Problem(s)
Interpretation
Model
2 Solution
Relative
geological age
3 Application
x

FULLY UTILIZED ROCK PHYSICS INFORMATION
Seismic signal
Problems Solution ApplicationProblems
“Many algorithms, workflow and software have been developed for integrating rock
physics information coming from seismic data into facies and petrophysical property
models. This is an area in which the seismic data is utilized to its full extent.”

Seismic signal

• Acoustic impedance
• AVO
• Seismic inversion
• Quantitative
interpretation

Facies & petrophysical
property modeling

UNDER-UTILIZED STRUCTURAL INFORMATION
Seismic signal
“Conversely, only limited structural data is commonly extracted from seismic and
integrated into structural models, in the form of interpreted horizon and fault surfaces.
However, much more information could be exploited: any small “shape” seen in the
seismic signal should indeed be used to guide and refine modelled geometries.”

Seismic signal

Seismic signalTypical
interpretation

interpretation
Structural
Model
“One consequence is that structural models are too often inaccurate or over-simplified.”

interpretation
Structural
Model
Unexploited
structural
information Above and below
the reservoir
In-between
interpreted horizons
Can help “guiding”
horizon geometry

interpretation
Structural
Model
Unexploited
structural
information Above and below
the reservoir
In-between
interpreted horizons
Can help “guiding”
horizon geometry
How to integrate this
information into a structural
framework?
Without creating 100s of
horizon surfaces…
Patches extracted
automatically from
seismic using the
Extrema technology
(Borgos et al. 2003)

Well logs
“To some extent, the same observation applies to well data...”

Well logs

Well logs
Well top
interpretation
& correlation

Well logs
Well top
interpretation
& correlation
Structural model

Well logs
Well top
interpretation
& correlation
Structural modelRefined
correlation
Bring additional
information on the
structure/thicknesses
framework
How to
facilitate
interpretation? How to
integrate into
the structural
framework?

Problems
EXISTING SOLUTIONS - STATE OF THE ART
Existing techniques (“Global interpretation”)
• Mostly automated
 Little user input
• Usually relying on one type of data (e.g. seismic)
 Lack of integration
[Lomask 2007]
[Pauget 2009]
[Hoyes 2011]
…
Solution Application
“Existing “global interpretation” or “automated correlation” techniques are often limited
to one single type of data (e.g. seismic) and do not fully take advantage of the additional
information (well correlation, thickness maps, etc.) used to constrain a structural model.
They would benefit from better integration with the structural model construction.”

Problems
[Lomask 2007]
[Pauget 2009]
[Hoyes 2011]
…
Interpretation
Static model
Dynamic model

Problems
[Lomask 2007]
[Pauget 2009]
[Hoyes 2011]
…
Interpretation
Static model
Dynamic modelHistory
matching

Problems
[Lomask 2007]
[Pauget 2009]
[Hoyes 2011]
…
Interpretation
Static model
Dynamic model
Mostly
Manual
History
matching

Problems
Our
solution
[Lomask 2007]
[Pauget 2009]
[Hoyes 2011]
…
Interpretation
Static model
matching
“The current reservoir modeling workflow tends to follow a waterfall approach: while
history matching techniques allowing to modify the static model as a function of dynamic
results are well known, little has been done to facilitate the refinement of interpretation
data from static models.”

Problems
Our
solution
[Lomask 2007]
[Pauget 2009]
[Hoyes 2011]
…
Interpretation
Static model
matching

PRINCIPLES OF VOLUME-BASED MODELING
Solution ApplicationProblems
Independent
surfaces
“Surface-based modeling techniques consist in modeling horizons and faults surfaces
consecutively, then gluing these surfaces together to form a watertight model.”

Independent
surfaces
Sealed model
glue
“Surface-based modeling techniques consist in modeling horizons and faults surfaces
consecutively, then gluing these surfaces together to form a watertight model.”

Independent
surfaces
Sealed model
glue
Initial volume

Independent
surfaces
Sealed model
glue
Relative geological
age (RGA)
Initial volume

Independent
surfaces
Sealed model
glue
Relative geological
age (RGA)
Low
High
Initial volume

Independent
surfaces
Sealed model
glue
split
Relative geological
age (RGA)
Low
High
Initial volume
“The volume-based modeling technique we are proposing consists in interpolating first a
relative geological age attribute in 3D, and then using iso-surfaces of this attribute to
partition the volume into several geological layers.”

Independent
surfaces
Sealed model
glue
split
Relative geological
age (RGA)
Low
High
Initial volume

High
Low
Relative geological
age

High
Low
Relative geological
age
Structural modeling = property modeling exercise
Isovalues = horizon surfaces
Gradient:
• Direction = dip
• Magnitude = thickness
Controls:
“Volume-based modeling turns structural modeling into a property modeling exercise:
interpolating the relative geological age (RGA) based on interpretation data. Not only does
this improve the robustness of the modeling and the geological consistency of the results,
but it also makes it possible to better control the geometry of the model through advanced
constraints on the value and the gradient of the RGA.”

High
Low
Relative geological
age
Gradient:
• Direction = dip
Controls:
Challenge: discontinuities
Unconformities Faults
“In implementing this approach, one major challenge is to properly account for structural
discontinuities such as faults and unconformities. This is solved by interpolating the RGA on a
discontinuous support, which is represented by an unstructured mesh.”

High
Low
Relative geological
age
Gradient:
• Direction = dip
Controls:
Solution: unstructured meshChallenge: discontinuities
“In implementing this approach, one major challenge is to properly account for structural
discontinuities such as faults and unconformities. This is solved by interpolating the RGA on a
discontinuous support, which is represented by an unstructured mesh.”

High
Low
Relative geological
age
Gradient:
• Direction = dip
Controls:
Solution: unstructured meshChallenge: discontinuities

“Here is a graphical summary of the volume-based modeling technique…”

Interpretation
Problems
VOLUME BASED MODELING – BASIC CONSTRAINTS
ApplicationSolution
Missing
interpretation
Imperfect
contacts

Interpretation
Problems
Basic constraints:
• Control value
• Smooth gradient
Relative geological age
ApplicationSolution
Missing
interpretation
Imperfect
contacts

Interpretation
Problems
Basic constraints:
• Control value
• Smooth gradient
ApplicationSolution
Missing
interpretation
Imperfect
contacts
1 horizon = 1 age
Minimize variations of dips and thicknesses

Interpretation
Problems
ApplicationSolution
Watertight geological model
Missing
interpretation
Imperfect
contacts
“Global”
technique

Interpretation
Problems
ApplicationSolution
Watertight geological model
Missing
interpretation
Imperfect
contacts
“Global”
technique
“Using volume-based modeling, all the horizons that conform to each other (i.e. that
belong to the same sequence) are interpolated simultaneously, yielding a watertight
model. Thanks to the regularization constraint which maximizes the smoothness of the
gradient of the RGA, layer thickness variations are minimized.”

Problems
VOLUME BASED MODELING – EXAMPLES
ApplicationSolution
(*)
(*)
(*)
(*) Physical (sandbox) models courtesy Fault Dynamics Research Group, Royal Holloway
University of London.

Problems
VOLUME BASED MODELING – ADVANCED CONSTRAINTS (1)
ApplicationSolutionSolution

Problems
ApplicationSolution
“Basic”
constraints only
Inaccurate
interpolation
Solution

Problems
ApplicationSolution
With “Fault
displacement”
constraint
Same “unknown”
relative geological age
Solution

Problems
ApplicationSolution
With “Fault
displacement”
constraint
Same “unknown”
Solution
Constraint on the
RGA value

Problems
ApplicationSolution
With “Fault
displacement”
constraint
Same “unknown”
Solution
Constraint on the
RGA value
“One advanced numerical constraint, similar in concept to the one described in Calcagno
et al. (2007) enables to constrain the value of the RGA at several arbitrary points so that
these points correspond to the same value of the RGA, without specifying this value. This
constraint makes it possible to interpret and control fault displacement (or apparent
displacement) on seismic sections.”

Problems
• “Dip direction”
constraint
• “Fold axis”
constraint
“Ideal”
surfaces
Modeled
surfaces

Problems
constraint
• “Fold axis”
constraint
Constraints on
the RGA gradient
“Ideal”
surfaces
Modeled
surfaces

Problems
constraint
• “Fold axis”
constraint
Constraints on
the RGA gradient
“Ideal”
surfaces
Modeled
surfaces
“Other advanced constraints allow to control the orientation of the RGA gradient, and
therefore the dip of the interpolated layers or the orientation of structural axis. Thank to
such constraints, models can be created from very sparse data – such as dip measurements
along wellbores. This slide presents a small synthetic example in which a cylindrical fold is
reconstructed from two wells.”

Problems
VOLUME BASED MODELING – ADVANCED OUTPUTS
3D model
Correlation for
faulted models or
deviated wells
RGA
Volume-based model

Problems
VOLUME BASED MODELING – ADVANCED OUTPUTS
3D model
Correlation for
faulted models or
deviated wells
RGA
Volume-based model
“Beyond the integration of advanced structural data to build structural models, volume-
based modeling yields attributes which can be used to refine interpretation data. In this
example, the RGA attribute is extracted along well paths and used to guide and refine well
correlation.”

Problems
CASE STUDY – AUSTRALIAN MODEL (1)
Dataset
• Area located offshore Western Australia
• 3D seismic block
• 8 wells with interpreted tops
Challenges:
• Complex stratigraphy (baselaps,
discontinuity and erosion)
• Complex structure: X, Y and λ faults
• Model from sea-bottom to underburden
Seismic data courtesy Geosciences Australia

Problems

Problems
“2D seismic reconstruction – which exploits a 2D version of volume-based modeling –
helped to interpret the seismic, and to validate the final interpretation.”

Problems
• 165 faults interpreted from seismic attributes (Ant tracking)
• Refining horizon interpretation iteratively by comparing
reconstructed horizons and seismic reflectors
• Add 4 horizons based on sparse data:
– From available well tops
– Without using isochore maps
Seismic data courtesy Geosciences Australia
1
2
3

Problems
“The model was created iteratively: first, a very sparse interpretation was created, and an
initial coarse model was built from this interpretation. The initial model was then used to
refine further the seismic interpretation. In parallel, patches extracted using the Extrema
technology were extracted, filtered, and integrated into the refined model using a
numerical constraint similar to the one described previously for controlling fault
displacement. This iterative workflow allowed to optimize both the model construction
time and the accuracy of the results.”

Problems

Problems
ApplicationSolutionSolution Application

Problems
CONCLUSION
SolutionSolution ApplicationApplication
• New “Global” modeling technique
• Based on interpolation of Relative Geological Age
• Allows controlling dip, thickness and “correlation”
• Can integrate large amounts of data and heterogeneous data types
• Ties with “Global interpretation” (e.g. Extrema technology)
• Can be used to enrich interpretation

Acknowledgements / Thank You / Questions
Acknowledgements
• Schlumberger Information Solutions and Rocksoft
• Geosciences Australia (case-study seismic dataset)
• Fault Dynamics Research Group, Royal Holloway University of London
(physical sandbox models)
• Montpellier Technology Center and Structural Geology Team

For the technically minded, the two next slides give a bit more details about
the interpolation of the Relative Geological Attribute.
They have been extracted from the following presentation:
Volume Based Modeling - Automated Construction of Complex
Structural Models, L. Souche, F. Lepage and G. Iskenova, EAGE 2013

CONSTRAINTS OF VOLUME-BASED MODELING
• Control points constraint:
 Sets the value of the implicit function (IF)
 At a given point in space  

Control
Point
0
1
23
N
Common
face
• Smooth gradient constraint:
 Constrains the gradient of the IF
 Ensures smooth variations
• Discontinuities in tetrahedral mesh:
 Break continuity of the interpolation
 Used to model e.g. faults

CONSTRAINTS OF VOLUME-BASED MODELING
• Control points constraint:
 Sets the value of the implicit function (IF)
 At a given point in space  

Control
Point
0
1
23
N
Common
face
• Smooth gradient constraint:
 Constrains the gradient of the IF
 Ensures smooth variations
• Discontinuities in tetrahedral mesh:
 Break continuity of the interpolation
 Used to model e.g. faults











0
gradient
tpoincontrol
F 

& 










0
gradient
tpoincontrol
F 

Numberofconstraints
Number of nodes
 Linear system of equations
 Unknowns 𝜑(𝛼𝑖)= IF values at the nodes of the tetrahedral mesh
 Solved using incomplete QR factorization

















equationsgradientconst
equationsgradient
equationstinpocontrol
C
C
C
A
.

Construction of Structurally and Stratigraphically Consistent Structural Models using the Volume-Based Modelling Technology (IPTC-18216-MS)

More Related Content

What's hot (20)

Viewers also liked (19)

Recently uploaded (20)

Construction of Structurally and Stratigraphically Consistent Structural Models using the Volume-Based Modelling Technology (IPTC-18216-MS)