JOGMEC-TRC

Reservoir Engineering Course

Lecture Notes

Geological Modeling (Part 1)

28th Feb, 2022

Lecturer: Machiko Tamaki

Japan Oil Engineering Co., Ltd.

1
Outline of the Lecture
Geological Modeling
 Part 1
 Introduction
 1. Data collection and assessment
 2. Framework modeling
 3. Property modeling

 Part 2
 3. Property modeling
 3-1. Facies modeling
 3-2. Petrophysical property modeling

2
Outline of the Lecture
Geological Modeling
 Part 1
 Introduction
 1. Data collection and assessment
 2. Framework modeling
 3. Property modeling

 Part 2
 3. Property modeling
 3-1. Facies modeling
 3-2. Petrophysical property modeling

3
Introduction: What is the reservoir modeling?
“Reservoir modeling involves the construction of a computer model of a petroleum reservoir,
for the purposes of improving estimation of reserves and making decisions regarding the
development of the field, predicting future production, placing additional wells, and evaluating
alternative reservoir management scenarios.”
From Wikipedia (https://en.wikipedia.org/wiki/Reservoir_modeling)

 Reservoir model should be built to answer a

specific aspect of the subsurface that impacts
on hydrocarbon distribution or fluid flow

https://www.software.slb.com/

4
 Reservoir evaluation workflow
Potential evaluation
 Typical reservoir modeling and simulation workflow
Make an economic or engineering decision
Reservoir model construction

Data analysis Static model Dynamic model Flow simulation

Integration Up-scale Input

Log/Core/Seismic 3D Geological model Add dynamic properties

History matching, Prediction study
Sensitivity/Uncertainty analysis

Multi-iterative work

Volumetrics & uncertainty analysis

5
 Static model (geological model)
 Static model generally means a geological model defined by the initial reservoir properties of
geological aspects
 3D geological modeling
Integration of all geological, geophysical, petrophysical and interpreted/conceptual information about a reservoir into a
single 3D numerical description of the reservoir

Eltom et al (2020)

Kharitontseva et al (2021)
Geophysical interpretation
Stratigraphic
interpretation
Petrophysical
Structural Integration work interpretation
interpretation

Melehan et al (2021)

6
Key issue for geological modeling
 Reservoir heterogeneity
• One of the most important factors for
geological modeling because it affects the Well data (2 wells)
fluid flow

 Critical issue for geological modeling

• How incorporate the reservoir
heterogeneity by using limited data Case A

Predicted oil recovery

Case B
Recovery

Case A Case B

Production/injected period

7
Reservoir heterogeneity classified
by depositional environment
 Permeability variability for each
depositional environment

 Carbonate reservoir has a large

heterogeneity because diagenesis will
usually destroy any primary deposition
that controls on reservoir quality.

Redrawn from Corbett and Jensen (1992)

8
Multiple scales of geological heterogeneity
 Various scale of heterogeneity are incorporated into geological model

Hierarchical modeling
Fault structure
large scale

larger scale

Framework modeling
(structural/stratigraphical) Zone & Layer

middle scale

Facies modeling Facies

smaller scale
small scale

Petrophysical Porosity
property modeling

Redrawn from Weber (1986)

9
Geological modeling workflow
1. Data collection 2. Framework modeling
3. Property modeling
& Management
3-1. Facies modeling
3D grid cells

Facies

3.2 Petrophysical property modeling

4. Volumetric assessment
& Uncertainty analysis
Porosity Permeability Water saturation

HCPV

10
Commercial modeling software
 Geological models in this class were constructed from Petrel

https://www.software.slb.com/

11
Outline of the Lecture
Geological Modeling
 Part 1
 Introduction
 1. Data collection and assessment
 2. Framework modeling
 3. Property modeling

 Part 2
 3. Property modeling
 3-1. Facies modeling
 3-2. Petrophysical property modeling

12
1. Data collection and management

13
Data collection and management
 Preparing data for loading and checking for the inconsistencies
 Decision should be made when setting up the project in the modeling software

Project settings Input data

• What is the objective of the model? • What type of input data is to be used?
• What units are to be used in the model? • What format will the data be provided in?
• What is the coordinate system? • How and where are the data stored?
• What is the area of interest
• How many wells are there?
Geological settings
• Does it include an aquifer?
• What is the geological environment?
• Are there any spatial or vertical trends recognized in the data?
• Are there any analogues that might be used guide the modeling?

14
Input data for geological modeling
 Seismic data
 Horizons
 Fault sticks and polygons
 Seismic volume attributes
 Velocity model for depth conversion

 Well data
 Wellbore path
 Well log data (e.g., facies and petrophysical properties)
 Well markers (e.g., stratigraphic zone boundaries)
 Core data (e.g., SCAL data)

Kharitontseva et al (2021)

15
Outline of the Lecture
Geological Modeling
 Part 1
 Introduction
 1. Data collection and assessment
 2. Framework modeling
 3. Property modeling

 Part 2
 3. Property modeling
 3-1. Facies modeling
 3-2. Petrophysical property modeling

16
2. Framework modeling

17
Framework modeling
 The objective of framework modeling is to build 3D geocellular grids

Large scale
 Framework modeling includes two components:
structural modeling and stratigraphic modeling
 Hierarchical approach from large to middle scale of heterogeneity

Heterogeneity
Fault modeling Pillar Gridding Make Horizon Make Zones Layering

Middle scale
Large scale Heterogeneity Middle scale

Redrawn from Weber (1986)

18
Fault modeling and pillar gridding
 Fault modeling is to make fault sticks from seismic data and define the segments bounded by fault system
 Pillar gridding is the process of defining the 3D mesh within the segments along the fault pillar
 Output of the pillar gridding is skeleton framework that consists of three skeleton grids (top, mid and base)

Fault modeling Pillar gridding

19
Zoning and layering
 Make horizon to capture the large-scale geological structures based on seismic interpretation
 Make zone and layering is the process of defining the vertical resolution of the model
 Make zones process defines the major changes of stratigraphic intervals between horizons
 Layering process defines the internal vertical resolution of the 3D grids

Make horizon Make zone Layering

Proportional Fractions Follow base Follow top

20
Framework modeling workflow
 Schematic summary Pillar gridding Make horizon
Fault modeling

Make zone Layering

Geo-cellular grids
21
Outline of the Lecture
Geological Modeling
 Part 1
 Introduction
 1. Data collection and assessment
 2. Framework modeling
 3. Property modeling

 Part 2
 3. Property modeling
 3-1. Facies modeling
 3-2. Petrophysical property modeling

22
3. Property modeling
Main topics

23
Property modeling
 Property modeling is to assign the reservoir properties in each grid by interpolation of input data
 Geostatistics is used to integrate the various data-set and describe the reservoir heterogeneity

Input data 3D distributed model

interpolation
log facies

phi
core
NTG

seismic perm
geostatistics
Sw

24
Geostatistics ~a tool box of various interpolation algorithms~
 Definition of geostatistics
Statistical technique that accounts for spatial relationships of
variables in estimating values at unsampled locations. For example, 5 obtained well data (porosity)

 Basic steps in geostatistical modeling

Step 1: Data analysis using input data to model spatial relationship
Step 2: Interpolate the reservoir property in 3D grids

 Two approaches in geostatistics

Deterministic Stochastic
e.g., Kriging estimation e.g., Conditional simulation

25
Kriging estimation
 Kriging is an interpolation technique based on a weighted average of the existing data
 Variogram is used to calculate those weights. Important factor to control geological continuity

unknown n(h) known Variogram (γ)

Z * (u0 )  0   i Z (ui ) Correlation function to quantify the spatial relationship (geological continuity)
i 1 and described as a function of distance between two points

γ
？ Three parameters:

Dissimilarity
Range: Distance until the variogram
curve saturates
Sill: Height at the range (=1)
Nugget: Height at the 0 distance

Distance (h)

26
Stochastic simulation  Comparison of porosity distributions

Kriging estimation
 Stochastic simulation aims at more realistic
global representations of reservoir heterogeneity Well A Well B
than estimation.
 Basic concept: use variogram and random error
 multiple realizations

Realization # Krigged value Random error Stochastic simulation

Z(cl) (uo )  Z * (uo )  r (l) (uo ) Well A Well B

Krigged Surface
Kriging estimation Conditional Simulation
deterministic stochastic
smoother heterogeneous
locally accurate globally accurate

27
Workflow for property modeling
 Target properties in the static geomodel 3D dynamic model
Facies model and petrophysical properties (porosity, permeability, initial saturation)

3D facies model
Input data
Facies
interpolation

geostatistics

3D petrophysical property model

Porosity Permeability Water saturation

Volumetrics & Uncertainty analysis

28
Summary
 Key aspects for geological modeling
 Describing reservoir heterogeneity is important for modeling because it affects on
fluid flow.
 Multiple scale of heterogeneity are incorporated into the model by hierarchical
approach that is framework modeling, facies modeling, and petrophysical property
modeling.

 Geostatistical approach for property modeling

 Statistical interpolation technique that accounts for the spatial relationships of
variables.
 Deterministic method for smoother reproduce and stochastic methods for more
heterogeneous using variogram that is a spatial correlation between two points.

29
Thank you!

Do you have any questions?...

30