Reservoir Characterization & Modelling

Đặc trưng hóa và Mô hình hóa Vỉa dầu khí

GE3105

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 1

 Overview and Introduction

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 2

Content
The objective of this course is to teach the basic science, technology and related
assumptions involved in building reservoir models. It will also introduce the problem of
uncertainty in reservoir modelling. The course aims to develop a critical-thinking
approach (rather than a prescriptive, “recipe” approach) wherein students are aware
of the “toolkit” available to them and can appropriately bring this to bear on reservoir
characterization studies. It will prepare students to understand and interpret
techniques that are most frequently used in commercial software (but will not teach
software usage itself). The core of the course is around constructing 3D static models.
Data sources, quality, relevance and choice of modelling techniques will be covered.
This is followed by classical gridding, mapping and contouring. Kriging is introduced as
a data-driven (variograms) form of classical mapping (estimation) and as a means of
data integration. Simulation techniques are introduced as a means of modelling
heterogeneity. Scaling (grids and properties) for the purpose of simulation is the final
topic. The integration and application of all the major ideas will be illustrated by case
studies, software demonstration to Petrel.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 3

Objectives
• Learn the main terminology, concepts, tools, and techniques used for generating 3D
static reservoir models.
• Understand some of key issues in reservoir characterisation & modelling, particularly
uncertainty & heterogeneity
• Practice using these tools – computer exercises
• Develop a critical-thinking and problem-solving approach to modelling, rather than a
prescriptive or “recipe” approach
• Develop & practice skills in data analysis and using Excel (key tools for all engineers)
• Illustrate how the main tools & techniques are used for real problems through case
studies
• If time, demonstrate some commercial software that is commonly used in the O&G
industry (Petrel)
• As an effective member of a team, research a paper/topic related to the course material
and present a critical review of it
• Read and understand a range of papers and articles related to reservoir characterization

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 4

Course learning outcomes
1. Understand and explain the main terminology, concepts, tools,
and techniques used for generating 3D static reservoir models
2. Understand key issues in reservoir characterization and
modelling, particularly uncertainty and heterogeneity
3. Practice using common used tools (Excel and commercial
softwares) for characterization exercises
4. Practice using commercial software to characterize and build a
static model through case studies
5. Analyze results and calibrate 3D static reservoir model. Oil and
Gas Reserve Estimation Applications and Uncertainty
6. Work as an effective team member to present a critical review
of a static model in the literature

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 5

A bit about me
•Education:[Link]., Hydrogeology & Engineeringgeology, Ho Chi Minh City University of Technology,
Vietnam, 2009.
 MSc.D., Petroleum Geology_Application, Ho Chi Minh City University of Technology,Vietnam, 2012.
•Lecturer at The University of Adelaide, 2013– Present
• Academic Awards
 Outstanding graduates 2009
 Fellowship of spe 2005
 Typical young teachers
 Chaired many scientific topics
 Published many scientific articles in journals and conferences
• Work Experience
Description: teaching and participating research in the field: Reservoir Engineering, Reserve Estimation,
Modeling and Simulation, Applied Hydrogeology for O & G exploration, O & G formation evaluation,
Geotechnics, Geostatistics, Applied Mathematics for Petroleum Science.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 6

Request
list of board staff, address and contact phone
subgroups: 4-5 students per group, group chief, information
email
Contact:
Thai Ba Ngoc – Geology Petroleum Department
Phone: 0972 42 77 55
[Link]: tbngoc@[Link]

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 7

Course Structure
•Reservoir Characterization and Modeling: Overview and
Introduction;
•Data Sources, Quality and Analysis;
•Reservoir Characterization;
•Modeling: Structural Model, Stratigraphic Model, Facies Model,
Property Model;
•Volumetrics and Uncertainty;
•Simulation and Upscaling;
•Excel and Petrel Training;
•Integrated Reservoir Characterization and Modeling Case Study;

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 8

Assessment
•Overall Purpose
–to encourage you to revise, learn, read more broadly
–to test your understanding and learning - not just your ability to
recall (that is assumed)
•Assessment Summary
–The course will be assessed with a weighting of
•10% in-class tests and assignments,
•30% practical project,
•60% final exam.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 9

Course Information: Approximate Schedule
1. Reservoir Characterization and Modeling: 1st Week
Overview and Introduction;
2. Data Sources, Quality and Analysis; 2nd Week
3. Reservoir Characterization; 3rd - 4th Week
4. Modeling: Structural Model, Stratigraphic Model, 5th - 8th Week
Facies Model, Property Model;
5. Volumetrics and Uncertainty; 9th - 10th Week
6. Simulation and Upscaling; 11th - 12th Week
7. Excel and Petrel Training; 13th - 14th Week
8. Integrated Reservoir Characterization and Modeling 15th Week
Case Study;

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 10

Useful Text
1. Michael J. Pyrcz, Clayton V. Deutsch, 2014, Geostatistical Reservoir
Modeling 2nd, Oxford University Press.
2. Steve Cannon, 2018, Reservoir Modelling: A Practical Guide 1st Edition,
Wiley-Blackwell
3. Sanjay Srinivasan, Juliana Y. Leung, 2022, Petroleum Reservoir Modeling
and Simulation. Geology, Geostatistics, and Performance Reduction,
McGraw-Hill Education
4. Stephen Tyson, Steve Begg and Jochen Kassan 2006, Reservoir
Characterisation for Petroleum Exploration and Development, Adelaide.
5. Y. Z. Ma, 2019, Quantitative Geosciences: Data Analytics, Geostatistics,
Reservoir Characterization and Modeling, Springer International
Publishing.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 11

Course Structure

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 12

 Overview of Reservoir
Characterization & Modelling

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 13

We need a model (computer model) of the
reservoir! (eg is a porosity model)

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 14

Why: E.G. Using 3D Model for Well-Planning

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 15

Field Development Plan

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 16

Input to Reservoir Simulation (Dynamic) Model

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 17

What: Reservoir Characterization
“Reservoir characterization is a process for quantitatively assigning
reservoir properties, recognizing geologic information and
uncertainty in spatial variability.” (Lake and Carroll, 1986)
“……a quantitative description of reservoir properties to a numerical
simulator with enough detail so that the predicted fluid-flow
characteristics match the historical performance and predict future
performance of the reservoir.” (Fowler et. Al., 1999)
“Reservoir Characterization encompasses all techniques and methods
that improve our understanding of the geological and petrophysical
controls of fluid flow.” (Yarus, 2003)

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 18

What: Reservoir Characterization “Core”
•Two Main Elements
–Generating a reservoir description (spatial model of reservoir structure and flow &
storage properties)
–Transforming the description into a model that is fit for the business purpose.
•Key Flow Properties to Model
–Porosity, Permeability (X, Y, Z), Saturations, (Net:Gross)
•Typical Data
–Seismic (Surface data: horizons, faults. Volume data: attributes)
–Core (description and poro-perm measurements)
–Well Logs
–Analogues (outcrops & modern systems)
–Conceptual Geological Models
–Well-tests, Production Data

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 19

What: Integration of multiple data types to build a static model

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 20

What & Why: Reservoir Characterization in Context of Business
Decision Making

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 21

Why & When: with respect to Oil-field Life-Cycle Decisions

Level of res. char. activity and/or impact

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 22

Why: Motivation for Static Reservoir Models
•There is a need for reliable estimates of the original volume of
hydrocarbon in the reservoir. These insitu volumes are important to
–allocate equity among multiple owners,
–compare the relative economic merits of alternative reservoirs, and
–determine the appropriate size of production facilities.
–determine the economic viability of producing a given reservoir,
–Identify drilling targets and plan well paths
•Reconcile an abundance of soft data (say, from a 3-D seismic survey or
historical production data) with a limited amount of hard well data.
•Well locations must be selected to be economically optimal and robust
with respect to uncertainty in the reservoir description. What type of wells
(horizontal, vertical, multilateral, ...? How many wells?

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 23

Why: Motivation for Static Reservoir Models

• Assess the potential for bypassed oil and the value of infill wells by comparing
current versus original in place volumes
• Starting point for developing a dynamic reservoir model (reservoir simulator)
that predicts reservoir performance with different production scenarios.
– The use of flow simulation was initially hampered by the limited resolution
of flow models (primarily a computer hardware consideration) and
simplistic geological input.
• Modern decision analysis tools require an assessment of the uncertainty in future
production. One of the greatest sources of uncertainty is uncertainty in the
geological inputs.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 24

Why: Economic Importance …

• Oil recovery from most reservoirs as a result of standard primary and secondary
recovery ranges from 5% to 80% of the OOIP (average 35%)
(National Petroleum Council, 1984).

• Petroleum engineers first recognized reservoir heterogeneity as a significant

factor influencing secondary and tertiary oil recovery efficiencies during the
1950’s.

• “The use of oversimplified geological models based on data from a limited

number of widely spaced wells is probably one of the most important reasons for
the failure in predicting field performance.”
(Damsleth et. al., 1992)

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 25

Main Steps in a Typical Reservoir Characterization Study
1) Have clear understanding of the business purpose of the model (decisions it will
support)
– allows you make reasonable simplifying assumptions
2) Determine fundamental geological controls on flow properties
– classify poro-perm to maximize difference between classes and minimize variation within
classes
– best done with core data, otherwise log, seismic & analog
3) Model spatial distribution of those geological controls
– often hierarchical, may be stochastic and/or deterministic
4) Use 2) to convert 3) to a spatial distribution of poro-perm
– may need to do some implicit re-scaling or use limit rules
5) Up-scale to get poro-perm at model scale
– if using for simulator grid-blocks, may need to fit heterogeneities explicitly to grid (e.g. faults,
shales)

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 26

High Level Modelling Procedure:
[Link] are the business questions / decisions
Example: Reservoir Management questions & decisions
• What is the drive mechanism?
– Will the reservoir have a natural water drive? If so, what’s the aquifer geometry, continuity,
and strength?
• Are there areas of by-passed HC?
• How should the wells be perforated and completed?
• Will water or gas injection be needed and when?
• Will recoveries be better by water or gas displacement?
• Should we plan for EOR processes – if so, when?
• What production should we predict for next year?
• What will oil prices do in the future and how will they effect my reservoir?

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 27

High Level Modelling Procedure:
2. Determine Geological Properties that Control Flow
- minimize variance within categories and
- maximize difference between categories
Why: because it allows us to use some simple,
robust ways of assigning effective values to
each category

How much thought do we put into doing this?

minimize

maximize
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 28
High Level Modelling Procedure:
[Link] Geological Properties that Control Flow
10000

1000

100

permeability (mD) 10

Lagoon Prodelta
Tidal Flats Distributary
1 channel
Tide affected channel
Proximal delta front
0.1

0.01
0 5 10 15 20 25 30 35
porosity (%)
2025 29
Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT
High Level Modelling Procedure:
3a. Framework: Combine Horizons and Faults
Faults

Structural Model

Horizons

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 30

High Level Modelling Procedure:
3a. Framework: Example

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 31

High Level Modelling Procedure:
3b. Grid (shown here in 2D vertical cross-section)

100m elevation
Surface 4
Surface 5
Surface 3

Surface 2

Surface 1

0m elevation

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 32

High Level Modelling Procedure:
3c. Model Geological Properties that Control Flow
2D Horizontal Cross section

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 33

High Level Modelling Procedure:
4. Convert Geological to Petrophysical Properties
2D Horizontal Cross section

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 34

High Level Modelling Procedure:
5. Upscaling to simulation grid (x-section view)

From: VoluMetrix

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 35

Key Problems/Issues: Data

• Data Reliability, Resolution, Scales

– Hard data: actual measurements
– Soft data: interpreted, or calibrated data & “knowledge”
– In reality the distinction is blurred

• Scarcity of Hard Data

– Direct measurement of key flow properties on one trillionth of the reservoir
• Resolution & Uncertainty of Soft Data
• Need for Integration of Different Data Types
• Variable Scales of Data Measurements
– ‘volumes’ on which the measurements are made

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 36

Key Problems/Issues: Multiple data sources & scales, multiple
modelling scales
We measure at one scale, but model at another

Measure Model Measure

106
1 1010
Core Geological
Well Test

104
1013
Log
108 Reservoir

Reservoir
Simulation
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 37
Scales: Common Terminology

Micro Macro Mega Giga

Thin Section Core Simulation Reservoir

SEM Porosity Tracer studies Outcrop studies
Mineralogy Permeability Wireline logging Seismic surveys
Pore throats Rel-perm curves Pressure transients Reservoir geometry
Cementation effects Capillary pressures Recovery efficiency Sequence stratigraphy
Pore-size distribution Residual saturations Reserves distribution Depositional modelling

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 38

Be thorough …

• Resist the temptation to:

– Assume that you have all the data for a particular reservoir
– Assume that reported results of log analysis, core analysis are accurate
– Assume that previous reservoir models are accurate
 Layering
 Controls on porosity distribution
 Permeability predictors
 Distribution of reservoir properties

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 39

Key Problems/Issues: Heterogeneity

• In real reservoirs the properties are heterogeneous (spatially

variable at a scale smaller than the data separation)
– One sample is not enough to describe the reservoir properties
– often the scale of heterogeneity is much smaller than the separation
between our sample points
• Reservoir heterogeneity impacts estimates of
– Oil in place
– Recovery Factor
– Well productivity
• ….. which in turn impact economics and hence investment decisions
• Due to low sampling rates, heterogeneity is also the prime
source of sub-surface uncertainty

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 40

Heterogeneity and Unrecovered Mobile Oil …
Estimate of remaining unrecovered mobile oil by depositional environment. Note that depositional systems with increasingly
complicated facies architecture have larger relative amounts of mobile oil remaining.

100
Drive Strength

Strong
80
Unrec. Mobile Oil % Moderate

Weak
60

40

20

From Tyler, 1988 and Tyler and Finley, 1991

0

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 41

Reservoir Heterogeneity
Horizontal Heterogeneity

Vertical Heterogeneity

2025 42
Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT
Reservoir Heterogeneity – after Tyler & Finlay (1991)

Horizontal Heterogeneity

Vertical Heterogeneity

2025 43
Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT
Reservoir Heterogeneity and Data Density
Assessing Uncertainty
Geologists are
idiots!!

?
It’s perfect! The Reservoir
Engineer

Now it’s perfect!!

The Geologist

The Same
Geologist
2025 Middle Jurassic deltaic series, Yorkshire coast. Well spacing of 300 m (after Homewood, et al., 1992) 44
Reservoir Heterogeneity and Data Density

Porosity Map

55 Wells and
7 Wells
Seismic

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 45

Heterogeneity and measurement scale (volume on
which measurement is made)
• Due to heterogeneity, property values (and their variability)
are a function of the measurement, or calculation, scale
1
2nd region
of homogeneity

1st region of
homogeneity
(absolute or
0 statistical)
physical size of sample or model

The key to successful and efficient characterization is to

exploit natural scales where the reservoir can be
considered homogeneous, or statistically homogeneous
2025 46
Key Problems/Issues: Uncertainty

• Data related
• Reservoir Heterogeneity (Spatial)
• Computational/Model Limitations

=> Uncertainty
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 47
Dealing with uncertainty: requirement for Stochastic
(Probabilistic) Modelling
Multiple Reservoir Data Data
Interpretations Heterogeneity + Scarcity “Fuzziness”

Uncertainty

Need
Probabilistic Approach to
Modelling
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 48
Uncertainty, Accuracy & Precision

Unknown Real World Unknown Real World

Inaccurate & Imprecise Accurate & Imprecise

Unknown Real World Unknown Real World

Which is better? At what cost?
Feasible?
Is it worth it?

Inaccurate & Precise Accurate & Precise

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 49
Key Reservoir Characterization Concepts

• Petrophysical properties:
- need to construct high resolution 3D models of lithofacies, porosity, and permeability
(and saturation)

- Hard truth measurements are lithofacies assignments, φ, and k inferred from

cores. All other data types including well logs and seismic are called soft data and
must be calibrated to hard data.
• Analog data:
- rarely enough data to provide reliable statistics and horizontal measures of continuity.
Need to borrow data from analog outcrops, densely drilled fields, and geological
expertise to infer spatial statistics.
• Modeling scale:
- not possible nor optimal to model reservoir at resolution of hard data. Core data must
be scaled to intermediate geological modelling scale, which may be scaled to even
coarser resolution for flow simulation.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 50

Key Reservoir Characterization Concepts
• Uniqueness and smoothing:
– conventional mapping devised to create unique and smooth maps to reveal large scale
geological trends; for fluid flow problems, smooth maps would yield optimistic results,
extreme high and low values have large effect on flow response.
• Uncertainty:
– exists because of our ignorance/lack of knowledge. It is not an inherent feature of
the reservoir (it is deterministic)

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 51

Dealing with uncertainty: requirement for Stochastic
(Probabilistic) Modelling
Multiple Reservoir Data Data
Interpretations Heterogeneity + Scarcity “Fuzziness”

Uncertainty
Models Integrates
Spatial Multiple Data
Continuity Types
Need
Probabilistic also
Approach to Geostatistics
Modelling
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 52
Geostatistics: a collection of tools that enable us to deal
with previous problems

• Data Analysis
• Heterogeneity Modeling
• Gridding & Up-scaling
• Data Integration (Fusion)
• Uncertainty Assessment

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 53

What is Geostatistics?

• Geostatistics is not the general application of statistical

methods to geology
• It is
– The statistics of observations located in space or time
– The application of statistical tools to determine the
spatial distribution of geologic variables
– A branch of applied statistics offering a set of tools which quantify and
model spatial variability

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 54

Why Geostatistics?

Geostatistics provides quantitative methods for:

1. Evaluating the spatial continuity and variability of
geological and petro-physical properties (heterogeneity)
2. Interpolating properties between wells and
extrapolating away from wells
3. Integrating different property measurements,
– with different spatial resolution and spatial sampling,
– and different degrees of reliability
4. Quantifying the uncertainty in the reservoir model due to sparse data
sampling

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 55

Why Geostatistics?:
Modelling Heterogeneity and Spatial Continuity
The two images below both obey the data points quite different in
terms of their heterogeneity (continuity)

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 56

Why Geostatistics ?:
Data Integration
Well Logs and Core Seismic Conceptual Geology
surface locations surfaces / stratigraphy / fluids stratigraphic interpretations
lithofacies/geological data attributes for porosity/facies observation from modern
porosity/permeability 4-D seismic monitoring analog fields / outcrops

Interpretation Interpretation Interpretation

Geostatistical
Interpretation Modeling Interpretation

Engineering Data Forward Sediment Modeling

DST/RFT data stacking patterns
pressure transient / tracer geometric data for facies
historical Q, P data spatial information for φ/k

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 57

Why Geostatistics? :
Data Integration
Seismic Velocity Porosity Map

Porosity Map constrained by Seismic

2025 58
Why Geostatistics ?:
Assessing uncertainty

Distribution of Rock
Multiple
& Fluid stochastic
Properties models of
reservoir

Numerical
model of
Recovery Process process

Distribution
of possible
Field Response responses

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 59

 One Generic Workflow …

Reservoir Characterization

Data
Audit,
Analysis
High
Resolution
Model
Uncert.
Assess

Uncert.
Flow Assess
Simulation Upscaling

2025 60
 Expanded Workflow …

QC
Data
Analysis

Conceptual Geological Model Data Structural

Model Stratigraphic
Spatial Audit, Model
Analysis
Analysis
Facies
HRM Model

Petrophysical
Probab Model
Assess

Flow Probab
POD
Simulation Assess Multiple
Realizations
Upscaling
History Ranking
Matching Selection
Simulation
Validation Grid

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 61

 Project Planning …

Start Early…

Data Mining

HRGM
Fine-scale Model

Probability Assessment

Upscaling

FlowSimulation
Flow Simulation

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 62

 On Models ……

“…. when you can measure what you are speaking

about and express it in numbers, you know
something about it. But when you cannot express it
in numbers, your knowledge is of a meagre and
unsatisfactory kind. It may be the beginning of
knowledge, but you have scarcely in your thoughts
advanced to the state of science, whatever the matter
may be.”
Lord Kelvin
Taken from Davis, 1986

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 63

 On Models ……

What could be cuter

Than to feed a computer
With wrong information
But naïve expectation
To obtain with precision
A Napoleonic decision

Major Alexander P. de Seversky

Taken from Davis, 1986

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 64

 On models: don’t confuse models with reality

Reality Model
(what we will get) (predictions for decisions)

Single true
Distribution of Rock
Multiple
distribution & Fluid stochastic
Properties models

Actual process Numerical

implemented Recovery Process model of
process

Single true Distribution

response Field Response of possible
responses
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 65
 On Models .…

“All models are wrong, some models

are useful”
Box
• There is no substitute for:
• Critical independent evaluation on the part of the
geoscientists and engineers to assure the success
of a reservoir modeling project.
Most failures occur because a basic
assumption was found to be wrong.
2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 66
Summary: Reservoir Characterization

• Based on the principles that:

– Sequence stratigraphy controls the 3D stratigraphic environment
– Depositional facies control petrophysics
– Heterogeneity needs to be modeled
– Uncertainty needs to be incorporated
– Up-scaling should be appropriate for model purpose & account for
heterogeneity

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 67

Typical (non-numeric) exam question
In reservoir characterization studies, give 2 reasons as to why it is beneficial to categorize
petro-physical sample data into sub-groups. What is the basic principle that should be applied
(or objective to be achieved) when doing this categorization. (6 marks)

Two reasons it is beneficial to categorize petro-physical sample data into sub- groups are:

1)because it splits a the population into groups or classes that control flow and different groups
might have different statistics or spatial continuity – 2 marks
2) it helps with up-scaling because we might be able to apply some simple rules
–and if each group is homogeneous enough then it doesn’t really matter what up-scaling rule is
used – 2 marks
– or other valid reason

The basic principle is to minimize the variance within groups and maximize the difference
between the groups, ie such that each group is as homogeneous as possible and as different from
other groups as possible – 2 marks

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 68

EXERCISES

1. What do you expect a reservoir characterisation project to tell you about your reservoir?

2. Explain the difference between reservoir characterisation and reservoir description.

3. Explain why you would want to construct a model of a petroleum reservoir.

2025 Thai Ba Ngoc – Faculty of Geology & Petroleum Engineering - HCMUT 69