Next

Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
OCN-3205
Stability Analyses
Reservoir equations (Computer Programming for Oceanographers)
Reservoir performance

Lectures- 31 & 32

Developers References
Back Next

Topic Overview
Topic overview
Introduction

Differential equations
Gridding An introduction to standard numerical solution
Difference approximation
Discretization error
techniques for Sea Basin flow equations.
Stability Analyses
Reservoir equations

Reservoir performance

Developers References
Back

Introduction
Topic overview
Introduction

Differential equations
Gridding
Gridding
Difference approximation
Stability analyses
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

Differential Numerical
equations Difference
Model Approximation
for mass flow
Basin Performance

Difference Approximation Discretization Error

For more information click on the subject you want to learn more about.

Developers References
 Discretization Techniques
Topic overview
Introduction

Differential equations
Gridding General partial differential equations for reservoir
Difference approximation
Discretization error
fluid flow must be discretized before they can be
Stability Analyses treated computationally.
Reservoir equations

Reservoir performance

The most common techniques are:

- finite differences
- finite elements

We will in in this module learn about the finite

difference technique.

Developers References
 Finite Differences
Topic overview
Introduction

Differential equations
Gridding
Finite difference approximations are used in most
Difference approximation commercial reservoir simulation software to solve
Discretization error
Stability Analyses
fluid flow equations numerically.
Reservoir equations

Reservoir performance
Main steps in a discretization procedure:

- replace differential operators by algebraic

ciexpressions
- compute approximate solution at given points and
iiispecified times

Developers References
Next

Differential Equations for Mass Flow

Topic overview
Introduction
Mass conservation equations for Black Oil models:
Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations
𝜕𝜁 𝜕̱
𝑢 𝜕̱
𝑣
+ + =0 Equation of Continuity
𝜕𝑡 𝜕𝑥 𝜕𝑦
Reservoir performance

Zonal Component of Momentum Eqn.

𝜕̱
𝑢 𝜕 𝜕 𝜕𝜁 1 𝜕𝑝𝑎 𝐹𝑠 𝑐𝑓̱
𝑢 1
+ 𝑢) +
(𝑢̱ (𝑣̱
𝑢) − 𝑓̱
𝑣 = −𝑔(𝜁 + ℎ) − (𝜁 + ℎ) + − (𝑢2 + 𝑣 2 )2
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜌 𝜕𝑥 𝜌 (𝜁 + ℎ)

Meridional Component of Momentum Eqn.

𝜕̱
𝑣 𝜕 𝜕 𝜕𝜁 1 𝜕𝑝𝑎 𝐺𝑠 𝑐𝑓 ̱
𝑣 1
+ 𝑣) +
(𝑢̱ 𝑣) + 𝑓̱
(𝑣̱ 𝑢 = −𝑔(𝜁 + ℎ) − (𝜁 + ℎ) + − (𝑢2 + 𝑣 2 )2
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜌 𝜕𝑦 𝜌 (𝜁 + ℎ)

Discretization Techniques
Developers References
Back Next

Difference Approximations
Topic overview
Introduction
Taylor series can be used to derive a difference
Differential equations
Gridding formula for single and double derivates.
Difference approximation
Discretization error
Stability Analyses Taylor series of f(x+x) and f(x-x) are given by:
Reservoir equations

1 1
f ( x)x 2 +  + f ( k ) ( x)x k
Reservoir performance
f ( x + x) = f ( x) + f ( x)x +
2! k!

1 1
f ( x − x) = f ( x) − f ( x)x + f ( x)x 2 +  + f ( k ) ( x)x k
2! k!

With these expansion we can deduce:

- first order approximation of f ’
- second order approximation of f ’
- second order approximation of f ’’

Developers References
 Next

First Order Approximation of f’

Topic overview
Introduction
From the expansion of f(x+Δx) we get an expression for f’(x):
Differential equations
Gridding
f ( x + x) − f ( x)
Difference approximation
f ( x) = +R
Discretization error
x
Stability Analyses
Reservoir equations

Reservoir performance From the expansion of f(x-Δx) we get an expression for f’(x):

f ( x) − f ( x − x)
f ( x) = +R
x

This difference formula is used

for discretizing time derivative
in the mass equations

x x
Click on the box to see how the approximation changes when the step size is halved.

Developers References
 Difference Formula
Topic overview
Introduction

Differential equations

A first order approximation of ut at the point

Gridding
Difference approximation
Discretization error
n+1 is given by:
Stability Analyses
Reservoir equations

u n +1 − u n
(ut )
Reservoir performance n +1
=
t

The time axis is divided into points at

distance Δt:
n n +1

t t + t

Developers References
Back

First Order Approximation of f’

Topic overview
Introduction

Differential equations
Gridding
Difference approximation
From the serie f(x+Δx):
Discretization error
Stability Analyses
f ( x + x) − f ( x)
Reservoir equations f ( x) = +R
Reservoir performance
x

From the serie f(x-Δx):

f ( x) − f ( x − x)
f ( x) = +R
x
x x
2 2

The step size reduction produces more accurate

approximations.

Developers References
 Back Next

Second Order Approximation of f’

Topic overview
Introduction
Adding expansion of f(x+Δx) and f(x-Δx) results
Differential equations
Gridding in the approximations:
Difference approximation
Discretization error f ( x + x) − f ( x − x)
Stability Analyses f ( x) = +R
Reservoir equations 2x
Reservoir performance

X X
Click on the box to see how the approximation changes when the time step is halved.

Developers References
 Back

Second Order Approximation of f’

Topic overview
Introduction
The sum of f’(x) of the series f(x+Δx) and f(x-Δx):
Differential equations
Gridding
Difference approximation
Discretization error
f ( x + x) − f ( x − x)
Stability Analyses f ( x) = +R
Reservoir equations 2x
Reservoir performance
Step size reduction
will produce
more accurate
approximations.

x x
2 2

Developers References
 Back

Second Order Approximation of f’’

Topic overview
Introduction

Differential equations The sum of the Taylor series f(x+Δx) and

f(x-Δx) is used to deduced a second
Gridding
Difference approximation
Discretization error
order approximation of f’’:
Stability Analyses
Reservoir equations

f ( x − x) − 2 f ( x) + f ( x + x)
Reservoir performance
f ( x) = +  (x 2
)
x 2

This approximation is frequently used and

the numerator is written:

2 f = f ( x − x) − 2 f ( x) + f ( x + x)

Developers References
Back

Difference Approximation
Topic overview
Introduction
Uxx can be approximated at each point i by the
Differential equations
Gridding formula:
Difference approximation

(u xx )i = ui −1 − 2ui2 + ui +1
Discretization error
Stability Analyses
Reservoir equations x
Reservoir performance

i −1 i i +1

x − x x x + x

Developers References
Back Next

Discretization Error
Topic overview
Introduction
The order of a difference approximation can by
Differential equations
Gridding analysed using Taylor expansions.
Difference approximation
Discretization error
Stability Analyses The discretization error approaches zero faster for
Reservoir equations

Reservoir performance
a high order approximation then for a low order
approximation.

Developers References
 Staggerring Grid Points
Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

Developers References
 Next

Coastal Configuration
Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

(Not active yet)

Developers References
 Back Next

Cyclone Tracks
Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

(Not active yet)

Developers References
 An Imposed Grid
Topic overview
Introduction
Main criteria for grid selection:
Differential equations
Gridding
Difference approximation
- The ability to identify saturations and pressures
Discretization error ii at specific locations (existing and planned well i
Stability Analyses
iiiilocations).
Reservoir equations

Reservoir performance - The ability to produce a solution with the i

iiiirequired accuracy (numerical dispersion and
iiiigrid orientation effects).
- The ability to represent geometry, geology and
iiiphysical properties of the reservoir (external
iiiboundaries, faults, permeability distribution
iiiincluding vertical layering).

- Keep the number of grid blocks small in order to

iiimeet requirements of limited money and time
iiiavailable for the study.

Developers References
 Back

Initial Fluid Distribution

Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

(Not active yet)

Developers References
Back Next

Stability Analyses
Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

Stable Unstable

(Not active yet)

Developers References
Back Next

Stable
Topic overview
 t 
Introduction
Animation of the stable solution  2 = 0,4 
Differential equations  x 
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

Developers References
Back Next

Unstable
Topic overview
Introduction
 t 
Animation of the unstable solution  2 = 0,6 
Differential equations
Gridding
 x 
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

Developers References
Back

Animation of 3D-Storm Surge

Topic overview
Introduction

Differential equations
Gridding
Difference approximation
Discretization error
Stability Analyses
Reservoir equations

Reservoir performance

Sound not active yet

Developers References
Topic overview Developers
Introduction

Differential equations
Gridding
Difference approximation OCN-3205
Discretization error
Stability Analyses
Reservoir equations
(Computer Programming for Oceanographers)
Reservoir performance

Lectures- 31 & 32

Sujit Kumar Debsarma

Visiting Professor, DU & BSMRMU
Former Scientist, SMRC
Former Principal Meteorologist & System Manager, BMD
Email: sujit.debsarma@gmail.com,
sujit.ocn@bsmrmu.edu.bd

Developers References