Geological Modeling:

Deterministic and Stochastic Models

Irina Overeem
Community Surface Dynamics Modeling System
University of Colorado at Boulder

September 2008
1
 Course outline 1

•  Lectures by Irina Overeem:

•  Introduction and overview

•  Deterministic and geometric models
•  Sedimentary process models I
•  Sedimentary process models II
•  Uncertainty in modeling

•  Lecture by Overeem & Teyukhina :

•  Synthetic migrated data

2
 Geological Modeling: different tracks
Reservoir Data
Seismic, borehole and wirelogs

Data-driven modeling Process modeling

Deterministic Sedimentary
Stochastic Model Process Model
Model

Static
Reservoir Model

Upscaling

Flow Model
 Deterministic and Stochastic Models

•  Deterministic model - A mathematical model which contains no

random components; consequently, each component and input is
determined exactly.

•  Stochastic model - A mathematical model that includes some

sort of random forcing.

•  In many cases, stochastic models are used to simulate

deterministic systems that include smaller- scale phenomena that
cannot be accurately observed or modeled. A good stochastic
model manages to represent the average effect of unresolved
phenomena on larger-scale phenomena in terms of a random
forcing.

4
 Deterministic geometric models

•  Two classes:
•  Faults (planes)
•  Sediment bodies (volumes)

•  Geometric models conditioned to seismic

•  QC from geological knowledge

5
 Direct mapping of faults and
sedimentary units from seismic data

•  Good quality 3D seismic data allows recognition of subtle faults and sedimentary
structures directly.
•  Even more so, if (post-migration) specific seismic volume attributes are
calculated.

•  Geophysics Group at DUT worked on methodology to extract 3-D geometrical

signal characteristics directly from the data.

6
L08 Block, Southern North Sea

Cenozoic succession in the

Southern North Sea consists
of shallow marine, delta and
fluvial deposits.

Target for gas exploration?

Seismic volume attribute analysis

of the Cenozoic succession in the
L08 block, Southern North Sea.
Steeghs, Overeem, Tigrek, 2000.
Global and Planetary Change, 27,
245–262.
August 5, 2009 7
Cross-line through 3D seismic amplitude data,
with horizon interpretations (Data courtesy Steeghs et al, 2000)
 The numerous faults have been interpreted as synsedimentary
deformation, resulting from the load of the overlying sediments.
Pressure release contributed to fault initiation and subsequent fluid
escape caused the polygonal fault pattern.

Combined volume dip/azimuth display at T = 1188 ms.

Volume dip is represented by shades of grey. Shades of blue indicate the
azimuth (the direction of dip with respect to the cross-line direction).
 Fault modelling
Fault surfaces

•  from retrodeformation (geometries of

restored depositional surfaces)

10

Example from PETREL COURSE NOTES

More fault
modelling
in Petrel

•  Check plausibility of implied stress

and strain fields
11

Example from PETREL COURSE NOTES

 Fan

Fan Feeder
channel

Delta Foresets

Combined volume dip / reflection strength slice at T=724 ms

 Delta front slump channels

Delta Foresets

Combined volume dip / reflection strength slice at T= 600 ms

 Gas-filled meandering channel

Combined volume dip / reflection strength slice at T= 92 ms

Deterministic sedimentary model
from seismic attributes
Object-based Stochastic Models

•  Point process: spatial distribution of points (object centroids) in

space according to some probability law

•  Marked point process: a point process attached to (marked with)

random processes defining type, shape, and size of objects

•  Marked point processes are used to supply inter-well object

distributions in sedimentary environments with clearly defined
objects:
•  sand bodies encased in mud
•  shales encased in sand

16
 Ingredients of marked point process

•  Spatial distribution (degree of

Width
of
sandbodies NAM

clustering, trends)
•  Object properties (size, shape,
orientation)
M ulti-storey sandbodies
Single storey sandbodies

100.000 100.000

10.000 10.000
•  Object-based stochastic
geological model conditioned to
1.000 1.000
Width (m)

Width (m)

100 100

10 10 wells, based on outcrop

1
1,0 10,0
Thickness (m)
100,0
1
1,0 10,0
Thickness (m)
100,0 1.000,0 analogues
(Source:
Shell
database
for
width/thickness
ratios)

17
An example: fluvial channel-fill sands

•  Geometries have become more sophisticated, but conceptual

basis has not changed: attempt to capture geological knowledge
of spatial lithology distribution by probability laws

18
•  Examples of shape characterisation:
•  Channel dimensions (L, W) and
orientation
•  Overbank deposits
•  Crevasse channels
•  Levees
 Exploring uncertainty of object
properties (channel width)

•  W = 100 m •  W = 800 m
•  W = 800 m •  R = 800 m

How can one quantify the differences between different realizations?

•  Major step forward:
object-based model of
channel belt generated by
random avulsion at fixed
point

•  Series of realisations
conditioned to wells
(equiprobable)
Stochastic Model constrained by
multiple analogue data

•  Extract as much information as possible from logs and

cores (Tilje Fm. Haltenbanken area, offshore Norway).

•  Use outcrop or modern analogue data sets for facies

comparison and definition of geometries

•  Only then ‘Stochastic modeling’ will begin

22
Lithofacies types from core
Example: Holocene Holland Tidal Basin

Tidal Channel Tidal Flat Interchannel

August 5, 2009 23
 #
Diamond Modern Ganges
SELECTED tidal
WINDOW
harbour
FOR STUDY
delta, India
Kulpi
#

Haldia # Raidighi

distance 50 km
#

Kakdwip
#
iver
li R
Hug

Matla River
MuriGang a

River
Saptamukhi River

an
Thakur

Channel width
 Tidal channels
Interchannel Branching
heterolithics main tidal
channels

Tidal flats Fractal Conceptual model of tidal basin

pattern of
tidal creeks (aerial photos, detailed maps)

Growth of fractal channels is governed by a branching rule

Quantify the analogue data into relevant
properties for reservoir model

•  Channel width vs distance to shoreline

Tidal channel width vs distance to shoreline

2000
1800
1600
1400
Width [m]

1200
1000
800
600
400
200
0
10 15 20 25 30 35
distance to shoreline [km]
The resulting stochastical model……
Some final remarks on stochastic/
deterministic models

•  Stochastic Modeling should be data-driven modeling

•  Both outcrop and modern systems play an important
role in aiding this kind of modeling.

•  Deterministic models are driven by seismic data.

•  The better the seismic data acquisition techniques
become, the more accurate the resulting model.

28
References

•  Steeghs, P., Overeem, I., Tigrek, S., 2000. Seismic

Volume Attribute Analysis of the Cenozoic Succession
in the L08 Block (Southern North Sea). Global and
Planetary Change 27, 245-262.

•  C.R. Geel, M.E. Donselaar. 2007. Reservoir modelling

of heterolithic tidal deposits: sensitivity analysis of an
object-based stochastic model, Netherlands Journal of
Geosciences, 86,4.

29