100% found this document useful (1 vote)
393 views17 pages

Solution Examination Paper For TPG4160 Reservoir Simulation: Department of Petroleum Engineering and Applied Geophysics

Here are the coefficient matrices for the given systems: a) One-dimensional (x), one phase flow: | a b 0 0 0 | | c d e 0 0 | | 0 f g h 0 | | 0 0 i j k | | 0 0 0 l m | The non-zero diagonals are labeled: a - main diagonal b, c - subdiagonal 1 d - main diagonal e, f - superdiagonal 1 g - main diagonal h, i - subdiagonal 1 j - main diagonal

Uploaded by

Patrick
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
393 views17 pages

Solution Examination Paper For TPG4160 Reservoir Simulation: Department of Petroleum Engineering and Applied Geophysics

Here are the coefficient matrices for the given systems: a) One-dimensional (x), one phase flow: | a b 0 0 0 | | c d e 0 0 | | 0 f g h 0 | | 0 0 i j k | | 0 0 0 l m | The non-zero diagonals are labeled: a - main diagonal b, c - subdiagonal 1 d - main diagonal e, f - superdiagonal 1 g - main diagonal h, i - subdiagonal 1 j - main diagonal

Uploaded by

Patrick
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

Department of Petroleum Engineering and Applied Geophysics

SOLUTION

Examination paper for TPG4160 Reservoir


Simulation

Academic contact during examination: Jon Kleppe


Phone: 91897300/73594925

Examination date: June 4, 2013


Examination time (from-to): 0900-1300
Permitted examination support material: D/No printed or hand-written support
material is allowed. A specific basic calculator is allowed.

Other information:

Language: English
Number of pages: 6
Number of pages enclosed: 0

Checked by:

____________________________
Date Signature
Final Exam page 2 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 1 (3+3+3 points)

This question relates to the Gullfaks H1 Segment project work.

a) Which geological factors are causing the good communication in the Lower Brent Group
of the H1 Segment of the Gullfaks Field
b) Describe briefly how chemical injection is accounted for in the Eclipse simulations that
you did
c) What are the main uncertainties in the simulation results?

Solution

The student should show that he/she really participated in the group work
Final Exam page 3 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 2 (4,5+2+4+2+3+5+2 points)

The simple, one-dimensional, linear, horizontal, one-phase diffusivity equation may be


written as:
∂ 2 P φµc ∂P
=( ) .
∂x 2 k ∂t
a) List the steps involved in deriving the diffusivity equation
b) Sketch the one-dimensional, horizontal porous system that the equation applies to, in both
continuous and discrete form.
c) Using Taylor series expansions, derive the finite difference approximations needed for the
discretization of the equation (for constant grid block size).
d) What are the error terms associated with these approximations?
e) Write the difference equation on explicit form, and outline a procedure for pressure solution.
f) Write the difference equation on implicit form, and outline a procedure for pressure solution.
g) Why is the explicit form seldom used?

Solution

a) Only a list in text or using equations is required


∂ ∂
-continuity equation − ( ρu) = (φρ )
∂x ∂t
k ∂P
-Darcy´s equation u = −
µ ∂x
1 ∂V 1 ∂ρ dρ
-fluid compressibility c f = −( )( )T = ( )( )T ⇒ ρc f = ( )
V ∂P ρ ∂P dP
1 ∂φ
-rock compressibility c r = ( )( )T ⇒ dφ = φc r
φ ∂P dP
-assume constant permeability and viscosity
∂ ⎛ k ∂P ⎞ ∂
-substitution: ⎜ρ ⎟ = (φρ)
∂x ⎝ µ ∂ x ⎠ ∂t
∂ dφ ∂P dρ ∂P ∂P ∂P
-right side: (φρ ) = ρ +φ = ( ρφc r + φρc f ) = ρφc
∂t dP ∂t dP ∂t ∂t ∂t
∂ ⎛ k ∂P ⎞ k ∂ ⎛ ∂P ⎞
-left side: ⎜ρ ⎟= ⎜ρ ⎟
∂x ⎝ µ ∂ x ⎠ µ ∂x ⎝ ∂x ⎠
-then:

∂ ⎛ ∂P ⎞ ∂ ⎛ ∂P ⎞ ⎛ ∂P ⎞ dρ ∂P ⎛ ∂ 2 P ⎞ ⎛ ∂P ⎞ 2 dρ ⎛ ∂ 2 P ⎞ ⎛ ∂P ⎞ 2
⎜ ρ ⎟ = ρ +
⎜ ⎟ ⎜ ⎟ = ρ⎜ 2⎟ ⎜ ⎟ + = ρ⎜ 2 ⎟ + ⎜ ⎟ ρc f
∂x ⎝ ∂ x ⎠ ∂x ⎝ ∂x ⎠ ⎝ ∂x ⎠ dP ∂x ⎝ ∂x ⎠ ⎝ ∂x ⎠ dP ⎝ ∂ x ⎠ ⎝ ∂x ⎠

-assume that
∂ ⎛ ∂P ⎞ ⎛ ∂ 2P ⎞
-so that ⎜ ρ ⎟ ≈ ρ⎜ 2⎟
∂x ⎝ ∂ x ⎠ ⎝ ∂x ⎠
∂ 2 P φµc ∂P
-thus, the final equation becomes: =
∂x 2 k ∂t
Final Exam page 4 of 17
TPG4160 Reservoir Simulation, June 4, 2013

b) -continuous system

-discrete system
1 i-1 i i+1 N

Δx

c) Space derivative
At constant time, t+Δt (alternatively, t may be used), the pressure function may be
expanded in forward and backward directions:

Δx (Δx) 2 (Δx) 3
P(x + Δx,t + Δt) = P(x,t + Δt) + ′
P (x,t + Δt) + ′′
P (x,t + Δt) + P ′′′(x,t + Δt) + .....
1! 2! 3!
(−Δx) (−Δx) 2 (−Δx) 3
P(x − Δx,t + Δt) = P(x,t + Δt) + P ′(x,t + Δt) + P ′′(x,t + Δt) + P ′′′(x,t + Δt) + .....
By adding these two expressions, 1! 2!
and solving for the second derivative, we get 3! the
following approximation:

P(x + Δx,t + Δt) − 2P(x,t + Δt) + P(x + Δx,t + Δt) (Δx) 2


P ′′(x,t + Δt) = 2
+ P ′′′′(x,t + Δt) + .....
or, by employing the grid index system, (Δx)and using superscript to indicate 12 time level:
∂ 2P P t +Δt − 2Pit +Δt + Pi−1
t +Δt
( 2 ) ti +Δt = i+1 + O(Δx 2 ) .
∂x (Δx) 2
Time derivative
At constant position, x, the pressure function may be expanded in backward direction
in regard to time:
Δt (Δt) 2 (Δt) 3
P(x,t) = P(x,t + Δt) + P (x,t + Δt) +′ ′′
P (x,t + Δt) + P ′′′(x,t + Δt) + .....
1! 2! 3!
By solving for the first derivative, we get the following approximation:
P(x,t + Δt) − P(x,t) (Δt)
P ′(x,t + Δt) = + P ′′(x,t + Δt) + .....
Δt 2
or, employing the index system:
∂P P t +Δt − Pit
( ) ti +Δt = i + O(Δt) .
∂t Δt
d) Error terms are derived above:
∂ 2 P t +Δt
( 2 ) i : O(Δx 2 )
∂x
∂P t +Δt
( ) i : O(Δt)
∂t
e) Explicit
Using the approximations above at time level t we get the explicit difference equation
(shown only in general form; end blocks will be different):
t
Pi+1 − 2Pit + Pi−1t
φµc Pit +Δt − Pit
≈ ( ) , i = 1,...,N
Δx 2 k Δt
Final Exam page 5 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Since the equation contains only one unknown, it may be solved explicitly:
Δt k
Pit +Δt = Pit + ( 2 )( t
)(Pi+1 − 2Pit + Pi−1
t
), i = 1,...,N
Δx φµc
f) Implicit
Using the approximations above at time level t+Δt we get the explicit difference
equation:
t +Δt
Pi+1 − 2Pit +Δt + Pi−1t +Δt
φµc Pit +Δt − Pit
= ( ) , i = 1,...,N
Δx 2 k Δt
which is a set of N equations with N unknowns, which may be solved simultaneously,
using a number of solution methods, for instance Gaussian elimination:
t +Δt
ai Pi−1 + bi Pit +Δt + c i Pi+1
t +Δt
= di , i = 1,...N
g) The explicit formulation is seldom used because it becomes unstable for large time steps.
It has the following stability requirement:
1 φµc
Δt ≤ ( )Δx 2 ,
2 k
Final Exam page 6 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 3 (2+3+3+4 points)

Sketch the coefficient matrix for the following systems, indicating non-zero diagonals with
approximate lines. Label the diagonals. What is the bandwidth?

a) One-dimensional (x), one phase flow, with the pressure equation:

ai Pi−1 + bi Pi + c i Pi+1 = di , i = 1,N


applicable to the following grid system:

1 i-1 i i+1 N

b) Two-dimensional (x,y), one phase flow, with the pressure equation:

ei, j Pi, j −1 + ai, j Pi−1, j + bi, j Pi, j + c i, j Pi+1, j + f i, j Pi, j +1 = di, j i = 1,N x , j = 1,N y

Applicable to the following grid system:


i

1 2 3 4 5 6
4 4

j 7 8 9 10 11 12
4

13 14 15 16 17 18
4

19 20 21 22 23 24
4

25 26 27 28 29 30
4

31 32 33 34 35 36
4

37 38 39 40 41 42
4

43 44 45 46 47 48
4

c) As question b) above, but now the numbering of the grid starts in the j-direction.

d) Three-dimensional (x,y,z), one phase flow, with the pressure equation:

gi, j,k Pi, j,k −1 + ei, j,k Pi, j −1,k + ai, j,k Pi−1, j,k + bi, j,k Pi, j,k

+ c i, j,k Pi+1, j,k + f i, j,k Pi, j +1,k + hi, j,k Pi, j,k +1 = di, j,k i = 1,N x , j = 1,N y ,k = 1,N z

applicable to the following grid system (grid blocks numbered in the sequence of x,y,z)
Final Exam page 7 of 17
TPG4160 Reservoir Simulation, June 4, 2013

i,j,k-1

k i-1,j,k i,j,k i+1,j,k

i,j,k+1

Nz Ny
j
1 i Nx 1

Solution

a) Bandwidth=3
b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b c
a b

b) Bandwidth=2Nx+1 =13

b c f
a b c f
a b c f
a b c f
a b c f
a b c f
a b f
e b c f
e a b c f
e a b c f
e a b c f
e a b c f
e a b c f
e a b f
e b c f
e a b c f
e a b c f
e a b c f
e a b c f
e a b c f
e a b f
e b c f
e a b c f
e a b c f
e a b c f
e a b c f
e a b c f
e a b f
e b c f
e a b c
e a b c
e a b c
e a b c
e a b c
e a b
Final Exam page 8 of 17
TPG4160 Reservoir Simulation, June 4, 2013

c) Bandwidth= 2Ny+1=17
b f c
e b f c
e b f c
e b f c
e b f c
e b f c
e b f c
e b c
a b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b c
a b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b c
a b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b c
a b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b f c
a e b c
a b f
a e b f
a e b f
a e b f
a e b f
a e b f
a e b f
a e b

d) Bandwidth=2NxNy+1 or 2NxNz+1 or 2NyNz+1 depending on direction of numbering

a b c
e
g
Final Exam page 9 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 4 (5+5+5 points)

For two-phase flow of oil and gas in a horizontal, one dimensional, linear porous medium, the
flow equations may be written as:

∂ ⎛ kk ro ∂Po ⎞ ∂ ⎛ φSo ⎞
⎜ ⎟ − qo′ = ⎜ ⎟
∂x ⎝ µo Bo ∂x ⎠ ∂t ⎝ Bo ⎠
and
∂ ⎛ kk rg ∂Pg kk ∂P ⎞ ∂ ⎛ φS φS ⎞
⎜⎜ + Rso ro o ⎟⎟ − q′g − Rsoq′o = ⎜⎜ g + Rso o ⎟⎟ ,
∂x ⎝ µg Bg ∂x µo Bo ∂x ⎠ ∂t ⎝ Bg Bo ⎠
where
Pcog = Pg − Po

S o + S g = 1.

a) Write the two flow equations on discretized forms in terms of transmissibilities, storage
coefficients and pressure differences (no derivations).
b) List the assumptions for IMPES solution, and outline briefly how we solve for pressures
and saturations
c) Outline briefly how we can solve for pressures and saturations by Newtonian iteration (ie.
fully implicit solution).

Solution

a)
Txoi+1 2 ( Poi+1 − Poi ) + Txoi−1 2 ( Poi−1 − Poi ) − q′oi
= Cpooi ( Poi − Poit ) + Csgoi ( Sg i − Sg it ), i = 1,N

Txg i+1 2 [( Poi+1 − Poi ) + ( Pcog i+1 − Pcog i )] + Txg i−1 2 [( Poi−1 − Poi ) + ( Pcog i−1 − Pcog i )] − q′gi
+( RsoTxo) i+1 2 ( Poi+1 − Poi ) + ( RsoTxo) i−1 2 ( Poi−1 − Poi ) − ( Rsoqo′ ) i
= Cpog i ( Poi − Poit ) + Csgg i ( Sg i − Sg it ), i = 1,N
b) IMPES solution
Assumptions:
Txo t ,Txg t
Cpoo t ,Cpog t
.
Csgo t ,Csgg t
Pcog t ,Rso t

Having made these approximations, the discretized flow equations become:

2 ( Poi+1 − Poi ) + Txoi−1/ 2 ( Poi−1 − Poi ) − q′oi


t t
Txoi+1/
= Cpooit ( Poi − Poit ) + Csgoit ( Sg i − Sg it ), i = 1,N
Final Exam page 10 of 17
TPG4160 Reservoir Simulation, June 4, 2013

t
Txg i+1/ [
2 ( Poi+1 − Poi ) + ( Pcog i+1 − Pcog i )
t
]
+Txg i−1/
t
[
2 ( Poi−1 − Poi ) + ( Pcog i−1 − Pcog i ) − q′gi
t
]
+( RsoTxo) i+1/ 2 ( Poi+1 − Poi ) + ( RsoTxo) i−1/ 2 ( Poi−1 − Poi ) − Rso t qo′
t t
( ) i

=C t
pog i (P oi −P t
oi ) + C (S t
sgg i gi − S ),t
gi i = 1,N

IMPES pressure solution

The pressure equation for the saturated oil-gas becomes:

{T t
xoi+1/ 2 [
+ α i Txg i+1/
t
2 + ( RsoTxo ) i+1/ 2
t
]}(P oi+1 − Poi ) +

{T t
xoi−1/ 2 + α [T i
t
xg i−1/ 2 + ( RsoTxo ) i−1/ 2
t
]}(P oi−1 − Poi )

+α iTxg i+1/ 2 ( Pcog i+1 − Pcog i ) + α iTxg i−1/ 2 ( Pcog i−1 − Pcog i )
t t t t

(
−q′oi − α i q′g + Rso q′oi =
t
) i

(C t
pooi + α iC t
pog i )(P oi − Poit ), i = 1,N

where

α i = −Csgoit /Csgg it .

The pressure equation may now be rewritten as:


ai Poi−1 + bi Poi + c i Poi+1 = di , i = 1,N
and solved for pressures using a number of solution methods, ie. Gaussian elimination.

IMPES saturation solution


Having obtained the oil pressures above, we need to solve for gas saturations using either the
oil equation or the gas equation. Using the oil equation yields:
Sg i = Sg it +
1
t
Csgoi
t
Txoi+1/ [
2 ( Poi+1 − Poi ) + Txoi−1/ 2 ( Poi−1 − Poi ) − q′oi
t
− Cpooit ( Poi − Poit ) , i = 1,N ]
c)
Solution by Newtonian iteration
Let us express the oil equation as Foi and the gas equation as Fg i . Each equation will depend on
pressures and saturations in blocks i-1, i and i+1., as indicated below.

Foi (Poi−1,Poi ,Poi+1,Sg i−1,Sg i ,Sg i+1 ) = 0

Fg i (Poi−1,Poi ,Poi+1,Sg i−1,Sg i ,Sg i+1 ) = 0 , i = 1,N

By first-order Taylor series expansions, we obtain the following expressions, where iteration
level is given by k:
Final Exam page 11 of 17
TPG4160 Reservoir Simulation, June 4, 2013

∂Foi ∂Foi ∂Foi


Foik +1 = Foik + k +1
(Poi−1 − Poi−1
k
)+ (Poik +1 − Poik ) + (Po k +1 − Poi+1
k
)
∂Poi−1 ∂Poi ∂Poi+1 i+1
∂Foi k +1 ∂Foi k +1 ∂Foi k +1
+ (Sg i−1 − Sg i−1
k
)+ (Sg i − Sg ik ) + (Sg i+1 − Sg i+1
k
)
∂Sg i−1 ∂Sg i ∂Sg i+1

∂Fg i ∂Fg i ∂Fg i


Fg ik +1 = Fg ik + k +1
(Poi−1 − Poi−1
k
)+ (Poik +1 − Poik ) + (Po k +1 − Poi+1
k
)
∂Poi−1 ∂Poi ∂Poi+1 i+1
∂Fg i k +1 ∂Fg i k +1 ∂Fg i k +1
+ (Sg i−1 − Sg i−1
k
)+ (Sg i − Sg ik ) + (Sg i+1 − Sg i+1
k
)
∂Sg i−1 ∂Sg i ∂Sg i+1
i = 1, N
Thus, for a one-dimensional system we have 2N equations and 2N unknowns, and we can
easily solve for estimates of oil pressures and gas and water saturations. By applying
Newtonian iteration until we converge on a solution within some tolerance, we may obtain a
solution to the equations. Our linear equations for iteration step k+1 would then take the
form:
k +1
apooi Poi−1 + bpooi Poik +1 + cpooi Poi+1
k +1 k +1
+ asgoi Sg i−1 + bsgoi Sg ik +1 + csgoi Sg i+1
k +1
= doi

k +1
apog i Poi−1 + bpog i Poik +1 + cpog i Poi+1
k +1 k +1
+ asgg i Sg i−1 + bsgg i Sg ik +1 + csgg i Sg i+1
k +1
= dg i
i = 1, N
 k +1 ˆ  k +1  k +1 
or, on a compact form: aˆ i X i−1 + bi X i + cˆ i X i+1 = di , i = 1,N

where

apooi asgoi bpooi bsgoi  Po k +1  do


aˆ i = bˆi = X ik +1 = ik +1 di = i
apog i asgg i bpog i bsgg i Sg i dg i

The equations are solved for pressures and saturations iteratively, updating coefficients after
each iteration.
Final Exam page 12 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 5 (27x0,5 points)

Explain briefly the following terms as applied to reservoir simulation (short sentence and/or a
formula for each):
a) Control volume
b) Mass balance
c) Taylor series
d) Numerical dispersion
e) Explicit
f) Implicit
g) Stability
h) Upstream weighting
i) Variable bubble point
j) Harmonic average
k) Transmissibility
l) Storage coefficient
m) Coefficient matrix
n) IMPES
o) Fully implicit
p) Cross section
q) Coning
r) PI
s) Stone´s relative permeability models
t) Discretization
u) History matching
v) Prediction
w) Black Oil
x) Compositional
y) Dual porosity
z) Dual permeability

Solution

a) Control volume small volume used in derivation of continuity equation


b) Mass balance principle applied to control volume in derivation of continuity
equation
c) Taylor series expansion formula used for derivation of difference approximations
h h2 h3
(or formula: f (x + h) = f (x) + f ′(x) + f ′′(x) + f ′′′(x) + .... )
1! 2! 3!
d) Numerical dispersion error term associated with finite difference approximations
derived by use of Taylor series
e) Explicit as applied to discretization of diffusivity equation: time level used in
Taylor series approximation is t
f) Implicit as applied to discretization of diffusivity equation: time level used in
Taylor series approximation is t+Δt
Final Exam page 13 of 17
TPG4160 Reservoir Simulation, June 4, 2013

g) Stability as applied to implicit and explicit discretization of diffusivity equation:


φµc
explicit form is conditional stable for Δt ≤ 1 2
, while implicit form is
2 k (Δx)
unconditionally stable
h) Upstream weighting descriptive term for the choice of mobility terms in
transmissibilities
i) Variable bubble point term that indicates that the discretization og undersaturated
flow equation includes the possibility for bubble point to change, such as for the
case of gas injection in undersaturated oil
j) Harmonic average averaging method used for permeabilities when flow is in series
k) Transmissibility flow coefficient in discrete equations that when muliplied with
pressure difference between grid blocks yields flow rate.
l) Storage coefficient flow coefficient in discrete equations that when muliplied with
pressure change or saturation change in a time step yields mass change in grid
block
m) Coefficient matrix the matrix of coefficient in the set of linear equations
n) IMPES an approximate solution method for two or three phase equations where all
coefficients and capillary pressures are computed at time level of previous time
step when generating the coefficient matrix
o) Fully implicit an solution method for two or three phase equations where all
coefficients and capillary pressures are computed at the current time level
generating the coefficient matrix. Thus, iterations are required on the solution.
p) Cross section an x-z section of a reservoir
q) Coning the tendency of gas and water to form a cone shaped flow channel into the
well due to pressure drawdown in the close neigborhood.
r) PI the productivity index of a well
s) Stone´s relative permeability models methods for generating 3-phase relative
permeabilities for oil based on 2-phase data
t) Discretization converting of a contineous PDE to discrete form
u) History matching in simulation the adjustment of reservoir parameters so that the
computed results match observed data.
v) Prediction computing future performance of reservoir, normally following a history
matching.
w) Black Oil simplified hydrocarbon description model which includes two phases (oil,
gas) and only two components (oil, gas), with mass transfer between the
components through the solution gas-oil ratio parameter.
x) Compositional detailed hydrocarbon description model which includes two phases
but N components (methane, ethane, propane, ...).
y) Dual porosity denotes a reservoir with two porosity systems, normally a fractured
reservoir
z) Dual permeability denotes a reservoir with two permeabilities (block-to-block
contact) in addition to two porosities, normally a fractured reservoir
Final Exam page 14 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 6 (3+3+3 points)

The discretized form of the left hand side of the oil equation may be written in terms of
transmissibilities and pressure differences, as

T xoi +1 / 2 ( Poi +1 − Poi ) + Txoi−1 / 2 ( Poi−1 − Poi )

Using the following transmissibility as example,

2ki −1 / 2 λ oi−1 / 2
T xoi −1 / 2 =
Δ xi (Δ xi + Δ xi −1)

a) What type of averaging method is normally applied to absolute permeability between grid
blocks? Why? Write the expression for average permeability between grid blocks (i-1) and
(i).
b) Write an expression for the selection of the conventional upstream mobility term for use in
the transmissibility term of the oil equation above for flow between the grid blocks (i-1) and
(i).
c) Make a sketch of a typical Buckley-Leverett saturation profile resulting from the
displacement of oil by water (ie. analytical solution). Then, show how the corresponding
profile, if calculated in a numerical simulation model, typically is influenced by the choice
of mobilities between the grid blocks (sketch typical curves for saturation profiles computed
with upstream or average mobility terms, respectively).

Solution

a) Harmonic average is used, based on a derivation of average permeability of series


flow, assuming steady flow and using Darcy´s equation

⎧ λo if Poi−1 ≥ Poi
b) λoi−1/ 2 = ⎨ i−1
⎩ λoi if Poi−1 < Poi
c)

Qw
Sw
1-Swir
exact
average
upstream
S wir

x
Final Exam page 15 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 7 (3+1+1+1+1+3 points)

For a one-dimensional, horizontal, 3-phase oil, water, gas system, the general flow equations
are (including well terms):

∂ ⎛ kk ro ∂Po ⎞ ∂ ⎛ φSo ⎞
⎜ ⎟ − qo′ = ⎜ ⎟,
∂x ⎝ µo Bo ∂x ⎠ ∂t ⎝ Bo ⎠
∂ ⎛ kk rg ∂Pg kk ∂P ⎞ ∂ ⎛ φS φS ⎞
⎜⎜ + Rso ro o ⎟⎟ − q′g − Rsoq′o = ⎜⎜ g + Rso o ⎟⎟
∂x ⎝ µg Bg ∂x µo Bo ∂x ⎠ ∂t ⎝ Bg Bo ⎠
∂ ⎛ kk rw ∂Pw ⎞ ∂ ⎛ φSw ⎞
⎜ ⎟ − q′w = ⎜ ⎟
∂x ⎝ µw Bw ∂x ⎠ ∂t ⎝ Bw ⎠

a) Explain briefly the physical meaning of each term in all three equations.
b) What are the criteria for saturated flow? What are the functional dependencies of
Rso and Bo ?
c) What are the primary unknowns when solving the saturated equations?
d) What are the criteria for undersaturated flow? What are the functional dependencies of
Rso and Bo ?
e) What are the primary unknowns when solving the undersaturated equations?
f) Rewrite the equations above for undersaturated flow conditions.

Solution

∂ ⎛ kk ro ∂Po ⎞ ∂ ⎛ φS o ⎞
a) ⎜ ⎟− q′o = ⎜ ⎟
∂x ⎝ µo Bo ∂x ⎠ ∂t ⎝ Bo ⎠
transport of oil well potential accumulation of oil
∂ ⎛ kk rg ∂Pg kk ∂P ⎞ ∂ ⎛ φSg φSo ⎞
⎜⎜ + Rso ro o ⎟⎟ − q′g − Rsoq′o = ⎜ + Rso ⎟
∂x ⎝ µg Bg ∂x µo Bo ∂x ⎠ ∂t ⎜⎝ Bg Bo ⎟⎠
transport of transport of gas well oil well pot. accumulation. accumulation
free gas sol. gas potential (solution gas) of free gas of solution gas
∂ ⎛ kk rw ∂Pw ⎞ ∂ ⎛ φSw ⎞
⎜ ⎟− q′w = ⎜ ⎟
∂x ⎝ µw Bw ∂x ⎠ ∂t ⎝ Bw ⎠
transport of water well potential accumulation of water

b) Saturated flow: criteria Po = Pbp and Sg > 0.


dependencies Bo = f ( Po ) and Rso = f ( Po ) .
c) Po, Sw and Sg
d) Undersaturated flow: criteria Po > Pbp and Sg = 0.
dependencies Bo = f ( Po ,Pbp ) and Rso = f ( Pbp ) .
e) Po, Sw and Pbp

∂ ⎛ kk ro ∂Po ⎞ ∂ ⎛ φSo ⎞
f) ⎜ ⎟ − q′o = ⎜ ⎟
∂x ⎝ µo Bo ∂x ⎠ ∂t ⎝ Bo ⎠
Final Exam page 16 of 17
TPG4160 Reservoir Simulation, June 4, 2013

∂⎛ kkro ∂Po ⎞ ∂ ⎛ φSo ⎞


⎜ Rso ⎟ − q′g − Rsoq′o = ⎜ Rso ⎟,
∂x ⎝ µo Bo ∂x ⎠ ∂t ⎝ Bo ⎠
∂ ⎛ kk rw ∂Pw ⎞ ∂ ⎛ φSw ⎞
⎜ ⎟ − q′w = ⎜ ⎟
∂x ⎝ µw Bw ∂x ⎠ ∂t ⎝ Bw ⎠
Final Exam page 17 of 17
TPG4160 Reservoir Simulation, June 4, 2013

Question 8 (2+5+2 points)

For a one-dimensional, vertical (z), 3 phase oil, water, gas system, outline how initial pressures
and saturations may be computed in a simulation model, assuming that equilibrium conditions
apply:
a) Sketch the reservoir, with a grid superimposed, including gas-oil-contact (GOC) and water-
oil-contact (WOC).
b) Sketch the oil-gas and oil-water capillary pressure curves, and show the how the initial
equilibrium pressures and saturations are determined in the continuous system.
c) Sketch the initial saturations as they are applied to the grid blocks.

Solution

You might also like