Reservoir Simulation

Chapter 1: Introduction
9 What is simulation
9 Why reservoir simulations
9 History and Evolution
9 Main Application of simulators
9 Classification of Reservoir simulations
9 Black oil simulations
9 Compositional simulations
Chapter 2: Basic Concepts
9 Introduction
9 Fluid flow in porous media
9 Fluid properties overview
9 Reservoir Rock properties overview
 Chapter 3: Differential Equations
Introduction
Single phase flow
Multiphase-Multidimensional flow
Black oil model
Formulation
Discretization
Linearization
Linear Algebraic Equation
Method of Solution
Method of Grid Construction
Boundary Conditions
Truncation
Stability
Multiphase formulation & Solution
IMPES Method
SS Method
Method of Solution
Compositional model
 Chapter 4: Practical Aspects of Reservoir
Simulation
9 Study objectives
9 Data analysis
9 Model construction
9 History matching
9 Reservoir performance predictions
Chapter 5: Commercial Simulators
9 Introduction
9 CMG
9 ECLIPSE
References
Basic Applied Reservoir Simulation,
by T. Ertekin, J. H. Abou-Kassem and G. R. King, SPE
Textbook Vol. 7, ISBN 1-55563-089-8, 2001.
Petroleum Reservoir Simulation,
by K. Aziz and A. Settari, Elsevier Applied Science
Publishers, ISBN 0-85334-787-5, 1979.
Reservoir Simulation,
by C.C. Mattax and R.L. Dalton, SPE Monograph Vol. 13,
ISBN 1-55563-028-6, 1989.
Fundamentals of Numerical Reservoir
Simulation,
By D. W. Peaceman, Elsevier Scientific Company, 1977
Introduction
What is Simulation?

Reservoir Simulation is a blend of

Engineering
Physics
Chemistry
Mathematics
Numerical Analysis and Computer Programming
Experience and Practice
Why Use Simulators?
No other solutions available (complex physics)
Cheaper or more reliable than other methods
It is always possible to simulate the reservoir, but
you can produce it only one time!
Increase profitability through improved reservoir
management
Assess economic and technical risks
Enhance credibility with third parties
Respond to safety, environmental and regulatory
concerns
Optimize well locations, type and spacing
Why Reservoir Simulation?
Ultimate recovery and performance can be studied
under different modes of operation
When should secondary recovery be initiated to
maximize recovery?
What type of EOR is appropriate?
What is the effect of location and spacing of wells
What is the effect of producing rate on recovery
What type of properties have the most effects on
recovery
What is a Reservoir Simulator?

Computer Program that solves equations for

heat and mass flow in porous media, subject
to appropriate initial and boundary
conditions
Uses discretization in time (timesteps) and
space (gridblocks) to represent the physical
domain (oil/gas reservoir)
Brief History and Evolution

Traditional Reservoir Engineering (1930 -1960)

o Representation of reservoir by single block (Tank
models)
o One dimensional, analytical solutions for linear two
phase and radial single-phase flow
Early Reservoir Simulation (1960 - 1970)
o First generation of digital computers
o Simulation in Research Labs, high costs
o Limited by speed and storage
o Poor reliability and confidence in technology
Brief History and Evolution

Modern Reservoir Simulation (1970 - 1985)

o Decreasing hardware costs
o Increasing confidence in technology
o 3D Models, large numbers of grid cells
o Availability of supercomputers
o Applications available to Reservoir Engineers in
operating companies
o Multi-component fluid descriptions
Brief History and Evolution
1985 - today
o Graphical User Interfaces
o Personal Computers
o Massive Parallelization
o Multi-Purpose Simulation Models
o Internet applications
o Reservoir Simulation has become a tool for
Reservoir Management
o Integration, .
Main Uses of Simulators

Forecasting Reservoir Performance

Improving Reservoir Description

(History Match)

Development of simple models and

correlations
Forecasting Reservoir Performance

Full Field Studies for depletion planning

and field development (number and
location of wells, injection/production
rates, injection fluids)
Sensitivity studies (EOR process
selection)
Assessment of uncertainty in forecasting
reservoir performance
Reservoir Management
Improving Reservoir Description

History Matching
Identification of flow units, barriers and
aquifer influx
Near well properties
Dual porosity behavior (fractures)
Unstable displacements, heterogeneity
Development of Simple Models

Coning Studies

Steam injection

Inflow performance relationships

(IPRs)
Modern Reservoir Simulation System

Suite of Software Applications for:

Pre-Processing
Data Loading and Importing
Gridding
Processing
Numerical Calculations
Post-Processing
Visualization of calculated results
Overall Workflow

Pre- Post-
1 2 4
Processing Processing Processing

1: Importing Geological Data

2: Parameterized Grid, Neighborhood
Connections
3: Rock-, PVT-, and Well data
4: Simulated Results for Visualization
Options and Features of a Modern
Pre-Processor

Automatic Grid generation

Cartesian, Radial and Corner Point grids
Local Grid Refinement, Aquifer Modeling
Flexible Well Modeling (arbitrary directed wells)
Fault Modeling (vertical, sloping)
Data Importing from various sources and formats
Graphical User interfaces and Visualization
Data preparation modules (Rock, PVT and Well
data)
Validation of Geological Model
Options and Features of a Modern
Processor
Black-Oil Models (single phase to three phase models)
Extended Black Oil (Salt, Polymer, API)
Compositional Model (n components, EOS model)
Thermal Model
Solution Techniques
IMPES, Fully implicit, adaptive implicit method
Dual Porosity Dual Permeability Models
Rock compaction and hysteresis models
Analytical and numerical aquifer treatment
Flexible and multilevel well management controls
Options and Features of a Modern
Post-Processor

Visualization of calculated results on grid

distribution plots for any object

Visualization of results vs. time

well plots, well group plots, reservoir plots
Animation of results

Supporting all major graphic formats

Exporting maps
Reservoir Simulator Classification

Type of reservoir and reservoir fluids

Black oil
Compositional
Recovery processes
Conventional recovery (black oil)
Chemical flood
Thermal recovery
Miscible displacement
Coordinate system used in the model
Number of dimensions in space
Number of phases
 Classification of Reservoir Simulators
Hydrocarbon Recovery Methods

Primary Recovery Secondary Recovery Tertiary Recovery

Chemical
Polymer Flooding
Solution Gas Drive Waterflooding
Surfactant Flooding
Reverse Wetting Agent
Gas Cap Expansion Pressure Maintenance Carbonated Waterflood

Gravity Drainage Thermal

Steam Injection
Rock Expansion Hot Water Injection
In-situ Combustion
Water drive Well bore Heating
Miscible
Vaporizing Gas Drive
Enriched Gas Drive
Alcohol Flooding
CO2 Flooding
Miscible Hydrocarbon
Flooding
Black oil simulators

Are used in situations where recovery

processes are insentive to compositional
changes in the reservoir fluids.
Mass transfer is assumed to be strictly
pressure dependent
Bo, Bg,and Rs govern PVT behavior
Compositional simulators

Recovery processes are sensitive to

compositional changes. These situations
include primary depletion of volatile oil and
gas condensate reservoirs.
Multiple contact miscible processes are
generally modeled with compositional
simulators.
Cubic equation of state governs the PVT
behavior.
Classification according to geometry and
dimension
3-D in rectangular coordinates are used for
full field study
2-D in rectangular coordinates can be used
for areal applications
2-D models in cylindrical coordinates can be
used for single well coning applications.
1-D for laboratory core floods.
Artifacts of Reservoir Simulation

Numerical errors
Non uniqueness of solution

Grid orientation effects

Averaging problems

Discretisation
Numerical Errors

Calculations are made with a finite

number of significant digits. Round off
errors can accumulate during run time

Material Balance errors will appear for example: ,

initial Oil, actual Oil, produced Oil, injction = 0

Numerical solution is inexact partial differential

equations are linearized causing truncation errors
Numerical Dispersion- Rapid
Saturation Changes
Numerical Dispersion can cause severe
distortions in simulations of processes in
which relatively rapid saturation changes
occur.
In case of water flood, at least some water
will move forward one gridblock each time
step.(see next slide).
A four-gridblock Waterflood Model

a) Hypothetical reservoir
and its water saturation
distribution

b) Model and simulated

water saturation distribution
Numerical Dispersion
Discontinuities at displacement
fronts will be smeared.

Increase number of blocks to

reduce smearing
Non Uniqueness of Solution
History Matching is an Inverse Modeling
approach.
I.e. the unknowns are the input parameter. We
attempt to find the best set of input data to
reproduce past performance.
Many different sets of input data may reproduce
the same performance. Even non-physical
values!
Inverse problems have no unique solution!
Dependent on good engineers judgment.
Grid Orientation Effect
Orientation of the grid may have considerable
influence on the results.

It is important when the displacing phase is much more

mobile than the displaced phase.
Averaging Problems
What is the permeability for the flux
between those 2 blocks, if
k1 = 100 md, k2 = 200 md
k1 = 100 md, k2 = 0 md
k1 = 100 md, k2 = 100 md

K1 K2
Example: Averaging
Case1 Case2 Case3
K1 K2 k1=100 k1=100 k1=100
k2=200 k2=0 k2=100

Arithmetic Average: 150 50 100

Geometric Average: 141 0 100

Harmonic Average: 133 0 100

Discretization
Approximation of continuous domain by discrete
volumes (gridblocks)

Cross Section:
Layer Top

Layer Bottom
Input Data for Black Oil Simulators

I) Cell Data
Absolute permeability
Porosity
Thickness
Elevation
Block or grid dimensions
Rock compressibility
Initial Reservoir pressure
Initial phase saturations
Input Data for Black Oil Simulators

II) Pressure dependent data

Oil, water and gas formation volume factors
Oil, water and gas viscosities
Solution gas/oil ratio
Oil, water and gas densities
Input Data for Black Oil Simulators

III) Saturation dependent data

Relative permeability for each phase
Oil/water capillary pressure
Gas/oil capillary pressure
Input Data for Black Oil Simulators

IV) Well data

Arial location and producing intervals
Production rate
Bottom hole pressure versus time
Input Data for Compositional Oil
Simulators
Same as Black OIL Simulator data except
PVT Properties
Composition
Equation of State (EOS)
Planning a Simulation Study

Problem Definition
Data Review
Selecting The study approach
Model design
Programming Support
History Matching
Predicting performance & analyzing results
Reporting
Definition of the problem

Define the reservoir performance

Operation problems
Adequate background data should be assimilated
Preliminary analysis of the reservoir fluid mechanics
should be conducted
Identify the principle depletion mechanisms
Identify factors dominate reservoir performance (ex coning,
gravity, heterogeneity)
Determine level of complexity
Identify required data
Data Review

Review of the available data will almost always

reveal gaps and inconsistencies
Decision on the quantity of the data
Decision on the quality of the data
Collection of data

Note: Many objectives can be met even with poor

data by evaluation of the sensitivity of reservoir
performance to reservoir description or other
parameters over a range of values believed to
encompass the actual values.
Selecting the Study Approach

Factors that influence the approach to a study

Availability of simulators that can adequately solve

the reservoir mechanics problems
Programming changes that must be made to the
simulator to model wells & facilities
Type & number of simulator runs needed to meet
study objectives
Calendar time, manpower, computing, and financial
resources available for the study
Model Design

Design of a simulation model will be influenced by:

Type of process to be modeled

Difficulty of the fluid mechanics problem
Objectives of the study
Quality of the reservoir description data
Time and budget constraints
Level of credibility needed to ensure acceptance of
the study results
Number of gridblocks & the detail included in the
treatment of individual wells
Programming Support

Well management program must calculate

and account for constraints imposed by
surface facilities, wellbore and flowline
hydraulics, artificial lift, and production
targets.
Editing package to summrize simulation
results such as: production & injection
summaries, Well status report, saturation &
pressure contour maps, recovery efficiency
by area
History Matching

After a reservoir model has been constructed, it

must be tested to determine whether it can
duplicate fluid behavior

Historical production data

Pressure data
Past performance of individual well

Model input parameters must be adjusted until

match between the model and the field is obtained
History Matching

Describe real reservoir as accurately as

possible with the available data
Typically one adjust

1. Reservoir permeability to match field pressure gradients

2. Permeability and areal extent of shales or other low permeability
zones to match vertical fluid movement
3. Relative permeability/saturation relationships to match dynamics
saturation distribution and pressure gradients
4. Aquifer size, porosity, thickness, and permeability to match the
amount and distribution of natural water influx
Predicting Performance & Analyzing
Results
Oil production rate
WOR & GOR Performance
Well &well workover requirements
Reservoir pressure performance
Position of fluid fronts
Recovery efficiency
Estimation of ultimate recovery
General information( timing of gas lift, water
handling , and low pressure separation)
Reporting

The final step in a reservoir simulation study

is to assemble results and conclusions in a
clear and concise report
The report should include:

1. The objective of the study

2. Description of the model
3. Data used
4. Results and conclusions
Activities in a Typical Reservoir
Simulation Study