Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
PETROLEUM ENGINEERING DEPARTMENT
TEXTBOOK SERIES
VOLUME 1
FLUID FLOW IN POROUS MEDIA
by
Zoltán E. HEINEMANN
Professor for Reservoir Engineering
Leoben, January 2003
© No part of this publication may be reproduced in any form.
Only students of the University of Leoben may copy for studying purposes.
Edited by DI Johannes Pichelbauer  January 2003
Table of Contents 3
Table of Contents
1 Fundamental Properties of Porous Media ................................................. 11
1.1 Porosity ..................................................................................................................... 12
1.1.1 General Aspects and Definition ...................................................................... 12
1.1.2 Determination of Porosity ............................................................................... 13
1.1.3 Compaction ..................................................................................................... 15
1.1.4 Compressibility of Porous Media ................................................................... 16
1.2 Capillary Properties ................................................................................................... 19
1.2.1 Saturation ........................................................................................................ 19
1.2.2 Wettability ...................................................................................................... 19
1.2.3 Capillary Pressure ........................................................................................... 20
1.2.4 Measurements of Capillary Pressure in a Porous Medium ............................. 23
1.2.5 Conversion of Laboratory Data ...................................................................... 28
1.2.6 The Leverett Function ..................................................................................... 30
1.2.7 Pore Size Distribution ..................................................................................... 31
1.3 Permeability .............................................................................................................. 33
1.3.1 Darcy’s Law ................................................................................................... 33
1.3.2 Definition and Units of Permeability .............................................................. 34
1.3.3 Measurements of Permeability ....................................................................... 36
1.3.4 Klinkenberg Effect .......................................................................................... 41
1.3.5 Analogies between the Laws of Darcy, Ohm and Fourier ............................. 42
1.3.6 Filtration Velocity ........................................................................................... 43
1.3.7 Quadratic Equation of Filtration ..................................................................... 44
1.4 Relative Permeabilities ............................................................................................. 45
1.4.1 Definition of Relative Permeability ................................................................ 46
1.4.2 Relative Permeability Measurements ............................................................. 47
1.4.2.1 The HASSLER method...................................................................... 47
1.4.2.2 PENNSTATEMethod...................................................................... 51
1.4.2.3 WelgeMethod.................................................................................... 52
1.4.3 Saturation Distribution and Relative Permeability ......................................... 52
1.5 References ................................................................................................................. 55
2 Equations of SinglePhase Filtration .......................................................... 57
2.1 Fundamental Equation of Filtration. ......................................................................... 59
2.1.1 Differential Form of the DarcyLaw .............................................................. 59
2.1.2 Anisotropic Porous Media .............................................................................. 62
2.2 Equation of State ....................................................................................................... 65
2.2.1 Incompressible Fluids ..................................................................................... 65
2.2.2 Low Compressibility Fluids ........................................................................... 65
2.2.3 Formation Volume Factor ............................................................................... 66
2.2.4 Ideal and Real Gases ....................................................................................... 67
2.2.5 Equation of continuity .................................................................................... 68
2.3 Special Forms of the Equation of Filtration .............................................................. 71
2.3.1 Incompressible Fluids ..................................................................................... 71
Table of Contents 4
2.3.2 Low Compressibility Fluids ........................................................................... 71
2.3.2.1 Elastic Porous Media.......................................................................... 72
2.4 Real and Ideal Gases ................................................................................................. 74
2.5 Boundary and Initial Conditions ............................................................................... 76
2.5.1 Boundary Conditions ...................................................................................... 76
2.5.2 Initial Conditions ............................................................................................ 77
2.5.3 Discontinuities in Porous Media ..................................................................... 78
2.6 Schematic of the Filtration Equations ....................................................................... 79
3 Solutions of the SinglePhase Equation of Filtration ................................ 83
3.1 Steady State Filtration ............................................................................................... 84
3.1.1 Steady State Filtration of Low Compressibility Fluid .................................... 84
3.1.2 Steady State Filtration in a Radial System ..................................................... 85
3.1.3 Steady State Gas Filtration ............................................................................. 88
3.2 NonSteady State Filtration in Infinite Acting Systems ............................................ 90
3.2.1 Radial Systems with Constant Production Rate ............................................. 90
3.2.2 Properties of the EiFunction .......................................................................... 94
3.2.3 Pressure Drop in Space and Time ................................................................... 95
3.2.4 The Spatial Distribution of Flow .................................................................... 99
3.3 Dimensionless Variables ......................................................................................... 100
3.4 The Infinite Radial System with Constant Pressure at the Interior Boundary ........ 105
3.5 NonSteady State Filtration in a Finite System ....................................................... 110
3.5.1 Constant Production Rate ............................................................................. 110
3.5.1.1 Closed Exterior Boundary................................................................ 110
3.5.1.2 Boundary with Constant Pressure .................................................... 112
3.5.2 Constant Pressure at the Interior Boundary and Closed Exterior Boundary 113
3.6 NonSteady State Filtration in Linear System ........................................................ 114
3.6.1 Linear Flow with Constant Production Rate ................................................ 114
3.7 The Principle of Superposition ............................................................................... 120
3.7.1 The First Law of Superposition .................................................................... 120
3.7.2 The Second Law of Superposition ................................................................ 123
3.7.3 Calculation of MultiWell Problems ............................................................ 125
3.7.4 Single Well with Variable Production Rates ................................................ 125
3.7.5 Pressure Buildup of ShutIn Well ............................................................... 127
3.7.6 Method of Image ........................................................................................... 128
3.7.6.1 Pressure Buildup Test Near No Flow Boundary.............................. 130
3.7.6.2 Constant Pressure Boundary ............................................................ 132
3.8 References ............................................................................................................... 133
4 TwoPhase Filtration.................................................................................. 137
4.1 The Equation of TwoPhase Filtration .................................................................... 137
4.2 Vertical TwoPhase Filtration of Incompressible Fluids ........................................ 139
4.3 The BUCKLEYLEVERETT Solution .................................................................. 142
4.3.1 The WelgeMethod ....................................................................................... 145
4.4 Influence of Gravity and Capillary Force ............................................................... 151
4.4.1 Influence of Gravity ...................................................................................... 151
Table of Contents 5
4.4.2 Influence of the Capillary Force ................................................................... 152
4.4.3 The Capillary EndEffect .............................................................................. 154
4.4.4 Imbibition ..................................................................................................... 155
4.5 References ............................................................................................................... 159
5 PistonLike Displacement .......................................................................... 163
5.1 The Mobility Ratio .................................................................................................. 163
5.2 Propagation of a Displacement Front ...................................................................... 164
5.2.1 Linear Displacement ..................................................................................... 166
5.2.2 Displacement in an Inclined Layer ............................................................... 168
5.2.3 Supercritical Displacement ........................................................................... 175
5.3 References ............................................................................................................... 179
6 Examination Outline .................................................................................. 181
6.1 A  Questions .......................................................................................................... 182
6.1.1 Part 1 ............................................................................................................. 182
6.1.1.1 Porous Material ................................................................................ 182
6.1.1.2 Wettability, Saturation, Capillary Pressure ...................................... 182
6.1.1.3 Darcy Equation, Liquid and Gas Permeability, Anisotropy............. 182
6.1.1.4 Relative Permeability ....................................................................... 183
6.1.2 Part 2 ............................................................................................................. 183
6.1.2.1 Continuity Equation, Darcy´s Law, Equation of State..................... 183
6.1.2.2 Equation of Filtration for Low Compressible Fluids,
Boundary Conditions........................................................................ 183
6.1.2.3 Equation of Filtration for Real Gas .................................................. 183
6.1.2.4 SteadyState Filtration through Linear Systems for
Liquids and Real Gases .................................................................... 183
6.1.2.5 SteadyState Filration through Radial Systems................................ 184
6.1.2.6 NonSteadyState Filtration with Constant Production Rate
through Infinite Radial Systems ....................................................... 184
6.1.2.7 Calculation of the Cumulative Influx............................................... 184
6.1.2.8 Finite Radial Systems....................................................................... 184
6.1.2.9 Physical Meaning of Superposition.................................................. 185
6.1.2.10PressureBuild Up of a ShutIn Well................................................ 185
6.1.3 Part 3 ............................................................................................................. 185
6.1.3.1 BuckleyLeverett Theory.................................................................. 185
6.1.3.2 Effect of Gravity and Viscosity on the fCurve, Influence of
Capillary Pressure on the Saturation Profile .................................... 185
6.1.3.3 PistonLike Displacement in Linear Systems .................................. 186
6.1.3.4 Displacement in Inclined Layers...................................................... 186
6.2 B  Questions ........................................................................................................... 187
6.2.1 Part 1 ............................................................................................................. 187
6.2.1.1 Porosity Measurements .................................................................... 187
6.2.1.2 Evaluation of Capillary Pressure Curves.......................................... 187
6.2.1.3 Permeabilty Measurements .............................................................. 187
6.2.1.4 Evaluation of Relative Permeability Functions................................ 187
Table of Contents 6
6.2.2 Part 2 ............................................................................................................. 187
6.2.2.1 Derivation of Special Forms of the Equation of Filtration 1............ 187
6.2.2.2 Derivation of Special Forms of the Equation of Filtration 2............ 188
6.2.2.3 NonSteadyState Filtration through Infinite Systems ..................... 188
6.2.2.4 Filtration of Low Compressible Fluids through Radial Systems
with Closed Boundaries.................................................................... 188
6.2.2.5 Second Law of Superposition........................................................... 188
6.2.2.6 Methods of Image............................................................................. 188
6.2.3 Part 3 ............................................................................................................. 189
6.2.3.1 Derivation of the BuckleyLeverett Theory ..................................... 189
6.2.3.2 Welge Method .................................................................................. 189
6.2.3.3 PistonLike Displacement in Linear Systems .................................. 189
6.2.3.4 Supercritical Displacement............................................................... 189
List of Figures 7
List of Figures
Figure 1.1: Definition of representative control volume for porosity measurements .............. 13
Figure 1.2: Packing of spheres and porosity (after P.K.Link) .................................................. 15
Figure 1.3: Sediment compaction and porosity (from Krumberlain and Sloss) ...................... 16
Figure 1.4: Pore compressibilities of rocks (after H.N.Hall) ................................................... 17
Figure 1.5: Pore volume compressibility factor in terms of overburden pressure
(after I.Fatt) ........................................................................................................... 18
Figure 1.6: Comparison of wetting to nonwetting fluid ......................................................... 19
Figure 1.7: Definition of the contact angle by Young .............................................................. 20
Figure 1.8: Illustration of the principal radii of the curvatures ................................................ 20
Figure 1.9: Modeling the porous medium as a bundle of cylindrical rods .............................. 22
Figure 1.10: Capillary pressure versus saturation of the wetting phase for the
model in Figure 1.9................................................................................................ 23
Figure 1.11: Schematic diagram of a diaphragm device for capillary pressure determination
by drainage (after Welge and Bruce)...................................................................... 24
Figure 1.12: Assembly with mercury pump for capillary pressure measurement
(Purcellmethod).................................................................................................... 25
Figure 1.13: Hysteresis of the capillary pressure curve ............................................................. 26
Figure 1.14: Drainage and imbibition in capillary tubes............................................................ 27
Figure 1.15: Equilibrium between gravity and capillary forces................................................. 27
Figure 1.16: The dimensionless capillary J function curve (after Leverett) .............................. 30
Figure 1.17: Non wetting fluid saturation versus the effective pore size distribution ............... 31
Figure 1.18: Bivariant pore radii distribution (from Dullien and Mehta) .................................. 32
Figure 1.19: Schematic diagram of DARCY’s experiment ....................................................... 34
Figure 1.20: Air permeameter: Schematic Flow diagram (after Monicard) .............................. 37
Figure 1.21: Schema of permeability measurement for unconsolidated media
(from Monicard) .................................................................................................... 39
Figure 1.22: Hassler type core holder (from Monicard)............................................................ 40
Figure 1.23: Variation in gas permeability with mean pressure and type of gas
(from Klinkenberg) ................................................................................................ 41
Figure 1.24: Schematic diagram of a device for measuring relative permeabilities of the
nonwetting phase (after Leas, Jenks, and Rassel) ................................................ 48
Figure 1.25: Schematic diagram of a device for measuring relative permeabilities of the
wetting phase (after Rappoport and Leas)............................................................. 49
Figure 1.26: Schematic diagram of a device for relative permeability measurements
(after Osoba).......................................................................................................... 50
Figure 1.27: Schematic diagram of a device for relative permeability measurements
(after Hafford)........................................................................................................ 51
Figure 1.28: Schematic diagram of the PENNSTATE device for relative permeability
measurements (after Morse, Terwilliger, and Yuster)............................................ 51
Figure 1.29: Schematic diagram of water invasion into porous media permeabilities of the
wetting phase (after Craig).................................................................................... 53
Figure 1.30: Drainage and imbibition relative permeability characteristic (after Craig) .......... 54
Figure 1.31: Typical water/oil relative permeability characteristic (after Craig) ...................... 54
Figure 2.1: Schematic diagram of a field segment................................................................... 59
List of Figures 8
Figure 2.2: Transformation of the coordinate system.............................................................. 63
Figure 2.3: Volume element in a cartesian coordinate system................................................ 68
Figure 2.4: Illustration of the boundary conditions.................................................................. 77
Figure 3.1: The Radial Coordinate System.............................................................................. 86
Figure 3.2: Illustration of steadystate filtration in a radial system......................................... 87
Figure 3.3: Plots of production equation for gas wells ............................................................ 89
Figure 3.4: Plots of the Ei(z) function (after Chaumet)......................................................... 94
Figure 3.5: Plots of pressure drop in the vicinity of a well (infinite reservoir,
compressible fluid) ................................................................................................ 96
Figure 3.6: Plots of pressure drop in the vicinity of a well (infinite reservoir,
compressible fluid) ................................................................................................ 96
Figure 3.7: The flow rate in function of the dimensionless variable Kt/r2 (after Chaumet).... 99
Figure 3.8: Solution for the infinitive and finite radial filtration problem with
closed boundary and constant pressure drop (after Van Everdingen and Hurst) . 102
Figure 3.9: Solution for the infinitive and finite radial filtration problem with
closed boundary and constant bottom hole pressure (after Silder)...................... 107
Figure 3.10: Solution for the infinitive and finite radial filtration problem with
closed boundary and constant bottom hole pressure (after Silder)...................... 108
Figure 3.11: Solution for the infinitive and finite radial filtration problem with
closed boundary and constant bottom hole pressure (after Silder)...................... 108
Figure 3.12: Solution for the infinitive and finite radial filtration problem with
closed boundary and constant bottom hole pressure (after Silder)...................... 109
Figure 3.13: Dimensionless pressure for single fractured well in an infinite acting
system (after Gringarten, Ramey, and Ragavan) ................................................ 117
Figure 3.14: Variable production rate in case of a ideal reservoir (after Hurst) ...................... 122
Figure 3.15: Pressure change at point R in infinite reservoir, with two production wells....... 124
Figure 3.16: Superposition of several wells in a infinite reservoir .......................................... 126
Figure 3.17: Application of the second law of superposition on a well with a
variable production.............................................................................................. 126
Figure 3.18: Pressure buildup analysis plot (after Horner) .................................................... 128
Figure 3.19: Production from a well near impermeable boundary (after Bear)....................... 129
Figure 3.20: Pressure buildup curve near a discontinuity....................................................... 131
Figure 3.21: Production in the vicinity of a boundary with a constant potential (after Bear) . 132
Figure 4.1: Calculation of fractional curve (after Marle) ...................................................... 142
Figure 4.2: Propagation of saturation profile (after Marle) ................................................... 144
Figure 4.3: The displacement front as discontinuity of saturation (after Marle) ................... 145
Figure 4.4: Determination of average saturation of the wetting phase after breakthrough
(after Welge)......................................................................................................... 146
Figure 4.5: Cumulative production by linear displacement (after Marle) ............................. 148
Figure 4.6: The influence of gravity on the fractional curve (after Marle)............................ 152
Figure 4.7: Influence of the velocity of displacement on the distribution of saturation
regarding the capillary force (by Douglas et al 1958)......................................... 153
Figure 4.8: The displacing efficiency as a function of velocity (by Kyte, Rappoport 1958). 153
Figure 4.9: "Endeffect” in case of a wetting displacing phase (after Marle)......................... 154
Figure 4.10: Endeffect” in case of a nonwetting displacing phase (after Marle) .................... 155
Figure 4.11: Displacement in countercurrent........................................................................... 155
Figure 4.12: Capillary pressure and relative permeability functions used in the
List of Figures 9
calculation by Blair ............................................................................................. 157
Figure 4.13: Distribution of pressure and saturation in case of linear (counterflowing)
imbibition (by Blair)............................................................................................ 157
Figure 4.14: Recovery in case of linear counterflowing imbibition and the experimental
determination of the influence of a certain in corelength.
(by Graham and Richardson) .............................................................................. 158
Figure 5.1: Comparison of saturation profiles according to different mathematical models. 164
Figure 5.2: Schematic diagram of pistonlike displacement .................................................. 166
Figure 5.3: Influence of the mobility ratio on front propagations in case of a linear
displacement ........................................................................................................ 169
Figure 5.4: Possible positions of the displacing front in inclined layer. ................................ 170
Figure 5.5: Forces acting on the displacing front................................................................... 171
Figure 5.6: Position of the displacing front by favorite mobility ratio (after Marle)............. 172
Figure 5.7: Position of the displacing front by unfavorable mobility ratio............................ 173
Figure 5.8: Supercritical displacement in inclined layer (after Marle) .................................. 175
List of Figures 10
111
1 Fundamental Properties of Porous
Media
A porous medium is a solid containing void spaces (pores), either connected or
unconnected, dispersed within it in either a regular or random manner. These so called
pores may contain a variety of fluids such as air, water, oil etc. If the pores represent a
certain portion of the bulk volume, a complex network can be formed which is able to
carry fluids. Only these permeable and porous media are taken into consideration in this
volume.
Various examples can be named where porous media play an important role or where the
technology requires them as a tool.
• In Soil Science:
The porous medium (soil) contains and transports water and nutrients to plants.
• In Hydrology:
The porous medium is a water bearing and sealing layer.
• In Chemical Engineering:
Porous medium is applied as filter or catalyst bed.
• In Petroleum Engineering:
Porous medium (reservoir rock) stores crude oil and natural gas
112 1: Fundamental Properties of Porous Media
1.1 Porosity
1.1.1 General Aspects and Definition
Basically one must distinguish between two groups of porous media:
• intergranular
• fractured.
Materials having both fractured and intergranular porosity are called dual (double) porous
media. On the other hand, concerning the mechanical properties, one should distinguish
between:
• consolidated
• unconsolidated
porous media. In a consolidated porous medium the particles (grains) are held together by
a cementing material, in the other type the grains are loose. A typical characteristic of a
consolidated medium is the possibility to form shaperetaining samples.
Definition of Porosity
The porosity of porous media is defined as the ratio of the volume of the pores to the total
bulk volume of the media (usually expressed as fraction or percent). Let us select any
point of the porous media and its environment with a sufficiently large volume V
T
, where:
, (1.1)
where
V
p
is the void volume (pore volume) and
V
s
is the volume of the solid material.
Porosity is defined as the ratio of pore volume to total volume, which can be expressed as:
. (1.2)
V
T
V
p
V
s
+ =
φ
V
p
V
T

V
T
V
s
–
V
T
 = =
1: Fundamental Properties of Porous Media 113
Basically one must distinguish between two kinds of porosities:
• Total porosity (isolated pores are considered also) and
• Effective porosity (effective in the sense of fluid transport).
The storage capacity of a reservoir rock always depends on the effective porosity, since it
contains the reservoir fluids.
Figure 1.1: Definition of representative control volume for porosity measurements
Porosity is a statistical quantity which depends on the magnitude of the total volume taken
into consideration (see Figure 1.1). If the selected volume is too small (e.g.: V
T
= 10
9
m
3
)
the calculated porosity can deviate greatly from the true value, therefore the volume V
T
should be large enough in order to obtain the statistical average of porosity.
On the other side if the volume is too large the porosity may deviate from the real value
due to the influence of heterogeneity.
1.1.2 Determination of Porosity
The determination of the porosity with sufficient accuracy is not a trivial issue, especially
for small samples. If the errors in measuring V
T
and V
P
are ∆V
T
and ∆V
P
then Eq. 1.2
will lead to:
(1.3)
where is the error in calculating the porosity.
φ
ef f
∆φ
φ

∆V
p
V
p

∆V
T
V
T
 + =
∆φ
114 1: Fundamental Properties of Porous Media
Assuming that then Eq. 1.3 can be written as:
. (1.4)
According to Eq. 1.4, the relative error of the porosity measurement depends on V
T
and
φ. Assuming an error ∆V=10
9
m
3
for the values of V, this error can be up to 50% as shown
in Table 1.1.
Table 1.1: Relative errors in measuring porosity
The following quantities are necessary in order to calculate the porosity based on Eq. 1.2:
• the total volume (V
T
),
• the solid volume (V
s
),
• the void volume (V
p
).
Example 1.1:
A core plug has a radius of 1.25 10
2
m, and 5.0 10
2
m in length. It is completely saturated
with brine having a density of 1200 kg/m
3
.
The dry core plug weighted 5.1 10
3
kg, and
10.4 10
3
kg when it was saturated with brine. Calculate the effective porosity of the core
plug.
Solution:
Weight of brine in the plug
φ[%]
1 4 12 20
V
T
[m
3
]
2.10
9
50.50 13.00 4.25 3.00
4 25.20 6.50 2.10 1.50
8 12.60 3.25 1.10 0.70
16 6.30 1.60 0.50 0.37
32 0.30 0.80 0.26 0.18
64 0.16 0.40 0.13 0.09
∆V
p
∆V
T
∆V = =
∆φ
φ

∆V
V
T
 1
1
φ

J
`
+
(

=
V
T
∆ 10
9 –
± m
3
] [ = ( )
w 10.4
3 –
×10 5.1
3 –
×10 – 5.3
3 –
×10 kg = =
1: Fundamental Properties of Porous Media 115
Volume of brine (pore volume)
Bulk Volume of plug
Porosity of plug
1.1.3 Compaction
Figure 1.2 shows porous media built with spheres of equal size. The spheres are ordered
in three different ways to illustrate the effect of compaction on the porosity of a pack.
However, no characteristic factor has been introduced yet to describe the compaction as a
property.
Compaction  and thus porosity  of a sediment depends on the greatest depth a rock
reached during its genesis. Figure 1.3 shows the porosity of clay and sandstone as a
function of depth. The compaction  in contrast to the compressibility  is irreversible.
Figure 1.2: Packing of spheres and porosity (after P.K.Link)
V
p
w
ρ
w

5.3
3 –
×10
1.2
3
×10
 4.42
6 –
×10 = = =
V
T
r
2
πh 24.5
6 –
×10 = =
φ
V
p
V
T
 0.18 or 18% = =
116 1: Fundamental Properties of Porous Media
1.1.4 Compressibility of Porous Media
Reservoir rock is not considered to be a rigid system but as a  only minor though  elastic
and thus compressible medium. Change of pressure inside the pore space during
production also affects the porosity.
The isothermal compressibility of porosity is defined as:
. (1.5)
Integration of the preceding equation leads to:
. (1.6)
where φ
0
is the porosity at the pressure p
0.
Figure 1.3: Sediment compaction and porosity (from Krumberlain and Sloss)
However, c
φ
is small and normally regarded as a constant. The pore volume alteration
during the pressure drop in the reservoir has its source in the elasticity of the solid.
Therefore c
φ
will be a function of porosity. Figure 1.4 illustrates this relation.
c
φ
1
φ

∂φ
∂p

( J
 `
T
=
φ φ
0
e
c
φ
p p
0
– ( )
φ
0
1 c
φ
p p
0
– ( ) + [ ] ≈ =
1: Fundamental Properties of Porous Media 117
Figure 1.4: Pore compressibilities of rocks (after H.N.Hall)
The compressibility of the pore space is influenced by overburden pressure too, which is
illustrated in Figure 1.5.
118 1: Fundamental Properties of Porous Media
Figure 1.5: Pore volume compressibility factor in terms of overburden pressure (after I.Fatt)
1: Fundamental Properties of Porous Media 119
1.2 Capillary Properties
1.2.1 Saturation
Basically pore space may contain several phases. The saturation of a certain phase is
defined as:
. (1.7)
Summing the saturations results in:
. (1.8)
1.2.2 Wettability
A fluid drop on a plane solid surface can take various shapes. The respective shape (either
flat or shaped like a pearl) depends on the wettability of the considered solid. Figure 1.6
illustrates that property. In case of air and water the water is the wetting phase, for air and
mercury the air.
Figure 1.6: Comparison of wetting to nonwetting fluid
The contact angle is used as a measure of Wettability. In the case of a wetting fluid, the
contact angle is smaller than 90°. If the contact angle is larger than 90°, then the fluid is
referred to as nonwetting.
Young, 150 years ago, defined the contact angle as a consequence of the static equilibrium,
between a drop of liquid and a plane of a solid surface. The drop of liquid will take a
certain shape due to the interfacial tensions acting on it, which are:
σ
12
the interfacial tension between fluid 1 and 2,
σ
s1
and σ
s2
the interfacial tensions between solid and fluids.
S
i
Volume of phase i in the porous media
Effective pore volume of the porous media
 =
S
i
1 =
i
∑
120 1: Fundamental Properties of Porous Media
Figure 1.7: Definition of the contact angle by Young
From Figure 1.7, Eq. 1.9 and Eq. 1.10 will result:
(1.9)
(1.10)
Interfacial tensions and thus θ are regarded as temperaturedependent. At room
temperature the interfacial tension between water and air is 0.073 N/m and between oil
and water about 0.03 N/m. When regarding oil reservoirs it is necessary to consider the
specific rock and fluid properties in order to determine whether the reservoir rock is water
or oil wet.
1.2.3 Capillary Pressure
For any two immiscible fluids (e.g. oil and water), the pressures at both sides of the fluid
interface are not equal. This pressure difference, called capillary pressure (P
c
), can be
calculate by the Laplace equation:
. (1.11)
Figure 1.8: Illustration of the principal radii of the curvatures
σ
S1
σ
12
Fluid 1
Fluid 2
Solid
σ
S2
θ cos
σ
s2
σ
s1
–
σ
12
 =
σ
s1
σ
12
θ cos + σ
s2
=
P
c
σ
12
1
r
1

1
r
2
 +
( J
 `
p
nw
p
w
– = =
1: Fundamental Properties of Porous Media 121
Figure 1.8 illustrates the principle curvatures radii r
1
and r
2
which are elements of the
Laplace equation.
If a porous medium is completely (100%) saturated with the nonwetting Fluid 2 and
contacted by the wetting Fluid 1, then Fluid 1 spontaneously intrudes into the pore space.
Fluid 1 will displace Fluid 2, until equilibrium is achieved. This is called imbibition.
Equilibrium is achieved when Fluid 1 uniformly occupies pore spaces with greatest
possible interfacial curvature.
Thus in the case of imbibition, Fluid 1 will occupy the smaller pore spaces first. In order
to illustrate this capillary equilibrium in a simplified manner, one could imagine the
porous medium as a bundle of infinitely long glass rods with uniform radius R (see Figure
1.9).
If air is the nonwetting and water is the wetting fluid, then and thus
and .
Since one of the two principle radii of curvature (r
2
) is infinite, it is easy to calculate the
porosity and water saturation of the wetting fluid as follows:
(1.12)
and the water saturation can be calculated from:
. (1.13)
Based on the geometry of the cylindrical rods Figure 1.9 the capillary pressure can be
calculated from Eq. 1.11 as:
(1.14)
where which is the radius of the water air interface.
σ
s2
0 = σ
s1
σ
12
=
θ cos 1 =
φ 1
π
4
 – =
S
1
4
4 π –

r
R

( J
 `
2
2
r
R

( J
 `
+
R
r R +

J
`
r
R

J
`
2
R
r R +

J
`
(

asin
(

–
(

acos – =
P
c
σ
12
r
 =
r r
1
=
122 1: Fundamental Properties of Porous Media
Figure 1.9: Modeling the porous medium as a bundle of cylindrical rods
Since S
1
and P
c
are functions of r, the following is also valid:
(1.15)
For a bundle of rods having a radius of R = 7.3 10
5
m and σ
12
= 0.037 N/m the
relationship between the capillary pressure and S
1
is shown in Figure 1.10.
Let p
w
be the pressure in the wetting phase and p
nw
in the nonwetting phase, then the
capillary pressure is defined as follows:
(1.16)
P
c
P
c
S
1
( ) =
p
nw
p
w
– P
c
S
w
( ) =
1: Fundamental Properties of Porous Media 123
Figure 1.10: Capillary pressure versus saturation of the wetting phase for the model in Figure 1.9
1.2.4 Measurements of Capillary Pressure in a
Porous Medium
The various kinds of measurement methods are all based on the same principle: A
constant pressure is exerted on the porous medium, until capillary equilibrium has
generated a constant saturation. Consequently this applied pressure equals capillary
pressure which corresponds to the given saturations. The saturation can be calculated with
the help of material balance.
One of the devices used for measurement of the capillary pressure curve is illustrated in
Figure 1.11. The upper and lower cells are separated by a diaphragm, which is
impermeable to the nonwetting fluid. The sample (core) and the diaphragm are both
saturated with the wetting fluid. The porous medium is placed into the upper chamber.
The nonwetting fluid surrounding the core, e.g. air, is then set under constant pressure
and is thus pressed into the porespace of the core.
124 1: Fundamental Properties of Porous Media
Figure 1.11: Schematic diagram of a diaphragm device for capillary pressure determination by
drainage (after Welge and Bruce)
The wetting fluid, displaced by the non wetting fluid, flows through the diaphragm into a
graduated pipette to be measured. This procedure is then repeated several times at higher
pressures. It is essential to pay attention to the fact that capillary equilibrium should be
achieved at every pressure step so the displaced volume is no longer a function of time.
However, it will never come to a complete displacement of the wetting fluid. The so called
connate water saturation always remains. After completing the described drainage and
measurement procedure, it is possible to establish the capillary pressure curve for
imbibition by reducing the pressure stepwise inside the upper chamber.
It is easier to measure the capillary pressure using mercury as a nonwetting fluid. The
original devise used by Purcell is shown in Figure 1.12 with the following procedure:
The core is evacuated inside the cell, then the cell is filled with Hg up to the level indicator
(nonwetting fluid Hg will not be imbibed into the pore space of the core), then the volume
indicator of the pump is set to zero.
Nitrogen is used for applying a constant pressure to force Hg into the core. A constant Hg
level is held by the pump. At each pressure step the Hgvolume is measured which has to
be pumped into the cell in order to maintain the level.
This procedure enables the capillary pressure to be determined as a function of the
Hgsaturation.
1: Fundamental Properties of Porous Media 125
Calibration of the device is accomplished with the empty cell, applying a wide range of
pressure steps. Final pressures may range from 1 to 100 MPa. Due to the necessity of
capillary equilibrium for a correct reading of the injected volume, a measurement may
take from hours to several days.
Another method uses a centrifuge to determine the capillary pressure curve. Initially the
core is saturated with the wetting fluid and the volume extruded is dependent on the
rotational speed (up to 20 000 rpm).
Figure 1.13 shows a typical capillary pressure curve. If the core is saturated completely
with the wetting phase (e.g. water) at the beginning of the measurement, then a certain
pressure must be applied to enable the nonwetting phase to intrude the pore space. This
pressure is the threshold pressure which depends on the largest pore diameter.
Figure 1.12: Assembly with mercury pump for capillary pressure measurement (Purcellmethod)
126 1: Fundamental Properties of Porous Media
Figure 1.13: Hysteresis of the capillary pressure curve
The capillary pressure curve will be asymptotic to a certain limit value of saturation  the
so called connate water saturation. This procedure of a wetting fluid being displaced by
a nonwetting fluid is called drainage.
On the other hand, in case of imbibition, the nonwetting fluid with which the core is
initially saturated will be displaced spontaneously by the wetting fluid. This means that
the porous medium is imbibing the wetting fluid immediately. Thus the capillary pressure
curve appears as a hysteresis (see Figure 1.13).
This appearance (hysteresis) may also be explained using capillary tubes. In the case of
an uniform radius of the capillary tube (see Figure 1.14), the height of the meniscus of a
wetting phase above the contact level will be independent of the displacement process. It
makes no difference in the measurement of P
c
if (i) the capillary tube is filled first with
the wetting fluid and placed in the tub (drainage), or (ii) an empty tube is placed in the tub
and the wetting fluid enters the tube (imbibition).
If this procedure is repeated with capillary tubes having sequentially different diameters,
then the height of the meniscus will depend on the saturation process (see Figure 1.15). If
 as graphically illustrated in Figure 1.15  empty capillary tubes are placed into a tub
filled with a wetting fluid, the menisci will rise against gravity until equilibrium between
gravity and capillary force is reached.
1: Fundamental Properties of Porous Media 127
Figure 1.14: Drainage and imbibition in capillary tubes
Figure 1.15: Equilibrium between gravity and capillary forces
Thus at equilibrium the following is valid (for a single capillary tube):
. (1.17)
In addition, Figure 1.15 illustrates that S
w
is a function of height h if the porous medium
is regarded as a bundle of capillary tubes with equal length, but different diameters.
Correspondingly the capillary pressure function may be expressed in terms of h = h(S
w
).
Therefore the saturation may be determined as a function of the vertical distance from the
P
c
= 0 plane.
, (1.18)
, (1.19)
P
c
h ρ
w
ρ
nw
– ( )g
2σ θ cos
r
 = =
P
c
S
w
( ) h S
w
( ) ρ
w
ρ
nw
– ( )g =
h S
w
( )
P
c
S
w
( )
ρ
w
ρ
nw
– ( )g
 =
128 1: Fundamental Properties of Porous Media
where
ρ
w
is the density of the wetting phase,
ρ
nw
is the density of the nonwetting phase,
P
c
(S
w
) is the capillary pressure of the wetting phase (e.g. water),
g is the gravitational constant.
1.2.5 Conversion of Laboratory Data
In Order to use capillary pressure data measured in the laboratory for capillary pressure
determination under reservoir conditions, a proper conversion of this data should be done
at first. The conversion is based on Eq. 1.20
(1.20)
where
P
cR
is the capillary pressure under reservoir conditions,
P
cL
is the capillary pressure measured under laboratory conditions,
σ
R
is the interfacial tension under reservoir conditions,
σ
L
is the interfacial tension measured under laboratory conditions,
θ
R
is the contact angle measured under reservoir conditions,
θ
L
is the contact angle measured under laboratory conditions.
Table 1.2 shows some interfacial tension values measured in the laboratory and estimated
values in reservoir.
Table 1.2: Interfacial tension and contact analyses
In the Laboratory In the Reservoir
σ
w/a
= 0.07 N/m σ
w/o
= 0.028N/m
θ
w/a
= 0 θ
w/o
= 33° to 55°
σ
Hg/a
= 0.48
θ
Hg/a
= 140°
P
cR
θ cos
R
σ
R
θ cos
L
σ
L
 P
cL
⋅ =
1: Fundamental Properties of Porous Media 129
Exampl e 1. 2:
Calculate the height of a transition zone at Sw = 0.7 and above the level of Pc = 0, if the
laboratory measurement of Pc using air/water system is 20600 N/m2 at Sw = 0.7. Use
Table 1.2 and the following data:
ρ
o
= 800 [kgm
3
]
ρ
w
= 1100 [kgm
3
]
θ
R
= 45°
Solution:
P
cR
θ
R
σ
R
cos
θ
L
σ
L
cos
 P
cL
⋅
0.71 0.028 ⋅
0.07
 20600 ⋅ 5850Nm
2
= = =
h
P
cR
ρ
w
ρ
o
– ( )g

5850
1100 800 – ( ) 9.81 ( )
 2m = = =
130 1: Fundamental Properties of Porous Media
1.2.6 The Leverett Function
According to Leverett the capillary pressure curves of core plugs  obtained from samples
of the same formation  may be correlated with other properties. These investigations
resulted in the dimensionless Jfunction (see Figure 1.16) which is given by:
, (1.21)
where
Pc is the capillary pressure,
σ
12
is the interfacial tension between fluid 1 and 2,
θ is the contact angle,
k is the permeability,
φ is the porosity.
Figure 1.16: The dimensionless capillary J function curve (after Leverett)
J S
w
( )
P
c
σ
12
θ cos

k
φ
 =
1: Fundamental Properties of Porous Media 131
1.2.7 Pore Size Distribution
Eq. 1.17 indicates that each capillary pressure value can be related to a certain radius,
which corresponds to a specific saturation as indicated by Eq. 1.15. Again consider a
bundle of equally long capillary tubes, so the circular crosssections of the capillary tubes
correspond to a certain partition function, then it is possible to set up a V
i
= V
i
(r
e
)function
using the Purcellmethod:
(1.22)
Where r is defined as the smallest radius which has to be filled with the nonwetting fluid
in order to reach a certain part of the pore space. It is customary to use the terms r
e
(pore
entry radius) and S
e
(effective saturation). This function is a distinctive property of
reservoir rocks and is therefore very often used to characterize a porous medium. Figure
1.17 shows the measured non wetting fluid saturation S
e
(P
c
) versus the effective pore size
r
e
(P
c
) calculated from Eq. 1.22. The function shown is similar to the function of
cumulative frequency distribution in statistics.
Figure 1.17: Non wetting fluid saturation versus the effective pore size distribution
Another way to determine the poreradii distribution is by means of statistical methods
(e.g. from thin ground sections). Characteristic for this method is the fact that the pore
radius is measured directly. The extent to which the distribution curves of the two methods
diverge is illustrated in Figure 1.18, which shows the pore volume distribution function
determined by the centrifuge method (the curve with sharp peak), and the other one
determined by photography of thinsections.
P
c
r ( )
2σ θ cos
r
 P
c
S
Hg
( ) = =
132 1: Fundamental Properties of Porous Media
Figure 1.18: Bivariant pore radii distribution (from Dullien and Mehta)
1: Fundamental Properties of Porous Media 133
1.3 Permeability
It is important to distinguish between mass flow and filtration:
In the case of mass flow all particles in the field of flow are in motion, whereas in the case
of filtration, only a portion of the mass particles flows and the remaining part forms the
flooded framework.
First investigations regarding filtration date back to the year 1825. Chaumont had the idea
of diging a trench parallel to the river Garonne (in Southern France) and then to dewater
this trench using an Archimedian screw. He measured the time elapsed until the water
level had resumed a certain height. Results of this experiment were unfortunately lost.
1.3.1 Darcy’s Law
In the year 1854 Dupuit made experiments with urban water filters in London. Velocity
of filtration had been 5 [m
3
m
2
] per day. From the results he deduced that the pressure
drop caused by the filter was proportional to the velocity of filtration.
Henry Darcy proved this hypothesis in the year 1856 using the equipment illustrated in
Figure 1.19. Investigations were made with various sand gravel packs. Darcy found that
the flow through the sand filter corresponded to the following formula:
, (1.23)
where
K is the permeability coefficient,
S is the crosssection of the packing,
L is the length of the packing,
h is the difference in piezometric head.
At the first World Oil Congress in 1933 the permeability was defined by Fancher, Lewis,
and Barnes. In the same year Wyckoff, Botset, Muskat and Reed suggested to give the unit
of permeability the name Darcy.
From that time on the equation is called Darcy’s Law:
, (1.24)
where:
A is the cross section of the porous medium perpendicular to the direction of flow
Q K
S
L
 h ⋅ ⋅ =
q A
k
µ

∆p
L
 ⋅ ⋅ =
134 1: Fundamental Properties of Porous Media
k is the permeability as a material property of the porous medium
L is the length of the porous media in direction of flow
∆p is the pressure difference along the porous medium
µ is the viscosity of the flowing fluid
The Darcy’slaw in Eq. 1.17 is valid for a laminar and steadystate onphase flow through
a porous medium only. Moreover, the fluid has to be largely incompressible.
Figure 1.19: Schematic diagram of DARCY’s experiment
1.3.2 Definition and Units of Permeability
Even today it is customary to use DARCY [D] as a unit of permeability. The permeability
of a porous media will be 1 Darcy, if at a 1 cm
2
crosssection a fluid with 1 cP viscosity
flowing with a rate of 1 cm
3
/s will cause a pressure drop of 1 atm/cm:
. (1.25)
By using the SIunits
k
µq
A
/
∆P
L

cp [ ] cm /
3
s [ ]
cm
2
[ ]
/
atm [ ]
cm [ ]
 1Darcy = = =
1: Fundamental Properties of Porous Media 135
The relationship between the two units is:
1 Darcy = 0.987 x 10
12
[m
2
]
Examples 1.3:
A cylindrical core having a radius 2.54 10
2
[m] and length of 0.3 [m], was flooded with
brine at a steady rate of 1.10
6
[m
3
s
1
], the differential pressure across the core was 10
[bar]. Calculate the absolute permeability of the core. Assume brine viscosity 0.001
[Pa.s].
Solution:
From Darcy’s law
Exercise:
Calculate the permeability of a core plug from the following test:
• Flow rate = 2.10
6
[m
3
s
1
]
• Inlet pressure = 5 [bar]
• Outlet pressure = 1 [bar]
• Length of core = 0.1 [m]
• Area = 1.10
4
[m
2
]
• Viscosity = 0.002 [Pa.s]
Nsm
2 –
[ ] m
3
s
1 –
[ ]
m
2
[ ]
/
Nm
2 –
[ ]
m [ ]
 m
2
[ ] =
q A
k
µ

∆p
L
 =
k q
µ
A

L
∆p
 =
k
1
6 –
×10 m
3
[ ]
s

0.001 Pas [ ]
20.3
4 –
×10 m
2
[ ]
 ×
0.3 m [ ]
10 bar [ ]
 × =
k 1.48
13 –
×10 m
2
[ ] 0.148 D [ ] = =
136 1: Fundamental Properties of Porous Media
1.3.3 Measurements of Permeability
The method depends on the following factors:
• consolidation of the medium
• core size
• fluid properties
• the applied pressure.
Samples from a consolidated media can be shaped as regular geometrical forms:
• cylinders with a diameter ranging from 0.02 to 0.05 [m]
• cubes with 0.02 [m] length of the side.
The rock sample  mostly formed cylindrically (core)  is fixed in the device with a sealing
rubber gaiter. Usually gas is used to measure permeability of core samples instead of
liquids. Since gas is the nonwetting fluid it does not alter the original state of the core and
the measurements can be repeated. It can also be used for low permeability cores where a
higher pressure difference is required.
By applying constant pressure, the gas commences to flow through the sample, and thus
permeability may be calculated using Darcy’s law for ideal gas as follows:
, (1.26)
where:
p
1
is the inlet pressure,
p
2
is the outlet pressure,
k is the permeability of the core,
q is the flow rate of gas at average pressure ,
µ is the dynamic viscosity of gas at the average pressure
To convert the flow rate q to a flow rate measured under atmospheric pressure (p
a
), the
ideal gas behavior is assumed, then:
, (1.27)
q
kA
µL
 p
1
p
2
– ( ) – =
p
p
1
p
2
+
2
 =
p
qp q
a
p
a
=
1: Fundamental Properties of Porous Media 137
where:
q
a
is the flow rate of gas at atmospheric pressure p
a
.
From Eq. 1.27 the flow rate q can be written as:
. (1.28)
Substituting Eq. 1.28 into yields:
. (1.29)
From Eq. 1.29 the permeability k
a
can be written as:
, (1.30)
where
q
a
is the flow rate at atmospheric pressure p
a
.
The prearrangement for unconsolidated media is very similar to the equipment of Darcy
(see Figure 1.19). The device may be described as a cylinder inside which a porous
medium is positioned between two lattices see Figure 1.21.
Figure 1.20: Air permeameter: Schematic Flow diagram (after Monicard)
q
q
a
p
a
p
 =
q
a
k
a
A –
µLp
a
 p
1
p
2
– ( )
p
1
p
2
+
2

( J
 `
=
k
a
2q
a
µL –
A

p
a
p
1
2
p
2
2
– ( )
 =
138 1: Fundamental Properties of Porous Media
In order to measure the permeability of very compact media using a liquid, high pressure
gradients must be applied. The liquid is either injected with constant gas pressure or by
use of a micropump. The Hassler core holder is a common device. Hereby the
permeability measurement is possible both in horizontal and vertical direction (see Figure
1.22). Permeability measurement in a vertical direction presupposes the casing of the core
by an impermeable rubber gaiter. In doing so air, water or oil can be pressed through the
core in an axial direction.
Horizontal measurement makes a sealing of the top surfaces of the core necessary. Then
the area of the cylindrical surface is covered both at the inflow and outflow opening to one
quarter with a lattice. This enables the core to remain permeable in these areas when cased
by the rubber gaiter. The fluid is then pressed horizontally through the lattice and the core.
Advantages of the Hassler core holder are:
• excellent sealing,
• optimal selection of core length,
• compatibility of large pressure gradients,
• permeability measurement in two directions,
• independency from the fluid used.
1: Fundamental Properties of Porous Media 139
Figure 1.21: Schema of permeability measurement for unconsolidated media (from Monicard)
140 1: Fundamental Properties of Porous Media
Figure 1.22: Hassler type core holder (from Monicard)
1: Fundamental Properties of Porous Media 141
1.3.4 Klinkenberg Effect
Klinkenberg noticed that the permeability of gas is not the same as for liquids and in
addition the gas permeability depends on pressure. The correlation between gas
permeability k
a
, liquid permeability k
1
and mean pressure inside the core, p
m
is given by
Eq. 1.31. The parameter b depends on the gas used.
(1.31)
(1.32)
Figure 1.23: Variation in gas permeability with mean pressure and type of gas (from Klinkenberg)
Physically the Klinkenbergeffect may be explained by the phenomenon of surface
slipping of gases caused by the Brown motion. This surface slipping decreases with
increasing pressure. However, at low pressure this effect is responsible for the deviation
of the gas permeabilities.
k
a
k
l
1
b
p
m
 +
( J
 `
=
142 1: Fundamental Properties of Porous Media
1.3.5 Analogies between the Laws of Darcy, Ohm and
Fourier
The following is a comparison of different laws which have a similar form to DARCY’s
law:
Darcy’s law
(1.33)
Ohm’s law of electrical current:
(1.34)
Fourier’s law of heat conduction:
, (1.35)
where:
A is the cross section [m
2
]
J is the amperage [A]
k is the permeability [m
2
]
L is the length [m]
Q is the rate of heat [Js
1
]
q is the flow rate [m
3
s
1
]
U is the voltage [V]
∆p is the difference in pressure [Pa]
∆Τ is the difference in temperature [°C]
µ is the dynamic viscosity [Pa.s]
λ is the thermal conductivity [Wm
1
°K]
ρ is the electrical resistance [Ωm]
The form of Eq. 1.33, Eq. 1.34 and Eq. 1.35 is similar.
These analogies enable a simulation of filtration processes with the help of electrical
models and the adoption of mathematical solutions obtained from heat flow problems.
q A
k
µ

∆p
L
 =
J A
1
ρ

U
L
 =
Q Aλ
∆T
L
 =
1: Fundamental Properties of Porous Media 143
1.3.6 Filtration Velocity
The velocity of filtration is defined as a fluid volume q flowing through the surface A of
a porous medium within unit time:
. (1.36)
u is an arithmetical quantity. In comparison to the actual velocity of flow in pore channels,
great differences can be recognized. However a statistical average is easily calculated as
follows:
, (1.37)
where φ is the porosity and v the displacement velocity.
If a fluid at a velocity of u = 1 [mday
1
] is injected into a porous medium with a porosity
of 0.1, a specific fluid particle will be transported within a distance of 10 [m] in one day.
Table 1.3: Comparison between the laws of Darcy, Ohm, and Fourier
Darcy Ohm Fourier
Flow Rate Amperage Heat conduction rate
Permeability
coefficent
Electrical
conductivity
Thermal
conductivity
Pressure difference Voltage Temperature
difference
u m
3
s
1 –
[ ]
J A [ ]
Q Js
1 –
[ ]
K
k
µ
 =
m
2
Pa
1 –
s
1 –
[ ]
γ
1
ρ
 =
1
Ωm

λ
W
mK

∆p Pa [ ] U V [ ] ∆T K [ ]
u
q
A
 ms
1 –
[ ] =
v
q
Aφ

u
φ
 = =
144 1: Fundamental Properties of Porous Media
1.3.7 Quadratic Equation of Filtration
Darcy’s law is only an approximation of the general equation of filtration given by:
. (1.38)
If
is small  in comparison to the term of first order  then Eq. 1.38 will be reduced
to the Darcy formula and α can be calculated by:
. (1.39)
In consequence of experiments, b can be defined as:
, (1.40)
where b, respectively ß, is independent of the fluid properties. The unit of ß is [m] and is
considered as a characteristic length of the porous medium. Further the velocity of mass
filtration Q
m
may be introduced as the product of filtration velocity and density:
, (1.41)
where A is the crosssection.
Then Eq. 1.38 can be written as:
(1.42)
The order of magnitude of ß ranges at 10
5
[m]. Therefore the correction factor:
(1.43)
becomes negligible in oil reservoirs, where Q
m
is not too large. However, in gas reservoirs
Q
m
is large in the vincinity of wells. Therefore the correction factor may often be
significant.
dp
dx
 αu bu
2
+ =
bu
2
α
µ
k
 – =
b b’ρ
βρ
k
 – = – =
ρu Q
m
A ⁄ =
ρdp
µQ
m
Ak
 1
βQ
m
µA
 +
( J
 `
dx – =
βQ
m
µA

1: Fundamental Properties of Porous Media 145
1.4 Relative Permeabilities
It was not until 1917 that Lewis discovered the coexistence of water, oil and gas in a
hydrocarbon reservoir. In 1926 Becker and Parkhurst had the idea to consider oil and gas
in a reservoir as liquid and gaseous phases of the hydrocarbon system being under
thermodynamic equilibrium.
The knowledge that finally led to a systematic research of multiphase filtration in porous
media originally referred to the gas/oil ratio. In the year 1927 Uren conducted
experiments, in which he simulated an oil reservoir with gas drive by using a tank filled
with sand and gassaturated oil at a pressure of several bars. From the results of this
experiment, Uren recognized that on one side the detached gas reduces the permeability,
but on the other side provides the energy needed to mobilize the oil.
In the same year Power made experiments with a sandfilled pipe 2.20 [m] long and 0.06
[m] in diameter. He injected saturated oil slightly above bubble point pressure into the
pipe. In doing so, he observed that at a constant rate the pressure gradient is larger in cases
of gas liberation. Thus a two phase flow causes an additional pressure drop for the flowing
system. Power concluded that the additional pressure loss was due to separated gas
bubbles caused by the JAMINeffect.
An overestimation of the JAMINeffect led Herold to the opinion that a given pressure
gradient can only set a certain number of oilgas interfaces (gas bubbles) in motion.
However, this would mean that the drainage area of a well was limited. Though this theory
was greatly contested  especially by Versluy’s and Lewis; the slogan “more wells, more
oil” was accepted in practice. It lasted until the year 1945 when this phenomenon was
understood. Hassler, Brunner and Deahl were convicted that a gas bubble was not able to
plug a porous medium due to the existence of various cross linkings between the pore
channels and the possibility of the gas molecules diffusing through the oil.
In the year 1936 Hassler made experiments which contributed greatly to the solution of
multiphase problems. The so called Hassler core holder was already described in Figure
1.22. First the cores were saturated with oil derivates which were then displaced by air.
Hassler measured the permeability of air and determined the saturation by weight control
of the core. This was possible due to the advantage of a quick demount and mount of the
core.
Hassler observed that it is essential to determine the permeability not only for dry cores
but for cores with any saturation. The retention of the wetting phase at the outlet is called
end effect and was also discovered by Hassler.
Wyckoff and Botset conducted research work with brine and carbon dioxide in the same
year as Hassler made his investigations. Brine was used to ensure measurability of the
electrical conductivity as shown in Figure 1.25.
The model used by Wyckoff and Botset was 3 [m] long and 0.05 [m] in diameter. They
146 1: Fundamental Properties of Porous Media
measured the pressure at 10 equidistant points determining the permeability both of the
wetting (k
w
) and nonwetting (k
nw
) phase as a function of saturation. The variation of
viscosity and saturation of the samples, at steadystate flow, is determined by the gas/oil
ratio. They made the observation that saturation may be independent of the absolute
permeability at a given gas/oil ratio.
1.4.1 Definition of Relative Permeability
As the historical review showed, all efforts were made to extend the validity of the
Darcylaw to multiphase filtration. If this is possible, then the following formulas may be
set up:
(1.44)
(1.45)
where:
k
rw
is the relative permeability of the wetting phase
k
rnw
is the relative permeability of the nonwetting phase.
The same indices are also valid for the flow rate q and the viscosity µ. The relative
permeability of the wetting phase is defined as:
, (1.46)
where k
w
is the effective permeability of the wetting phase. This definition is valid for
k
rnw
, also.
Finally, Leverett had conducted several experiments concerning the relative permeability
in case of a fluid/fluid system and then extended these to threephase
wateroilgassystems. He introduced not only the capillary pressure into the equations
of multiphase filtration, but together with Buckley he also set up the theory of frontal
displacement which will be covered in Chapter 4.
q
w
A
kk
rw
µ
w

∆p
L
 – =
q
nw
A
kk
rnw
µ
nw

∆p
L
 – =
k
rw
S
w
( )
k
w
S
w
( )
k
; = 0 k
rw
S
w
( ) 1 ≤ ≤
1: Fundamental Properties of Porous Media 147
1.4.2 Relative Permeability Measurements
The following three methods can be applied:
• The Hasslermethod: The principle behind this method is the ability to adjust the
phase pressures and the filtration velocity independently by the use of one or two
diaphragms.
• The PENNSTATE method: Two fluids are injected simultaneously at constant rates.
• The Welge method: The principle is the calculation of the relative permeability using
the results received from the displacement experiments.
1.4.2.1 The HASSLER method
Leas, Jenks and Russel used the Hassler device in their instrumentation as shown in
Figure 1.24.
Hasslercore holder contains a core (C) placed on a diaphragm (D). The diaphragm is an
artificial porous medium exhibiting a relatively large capillary threshold pressure. It is
fully saturated with the wetting phase to prevent the flow of the non wetting phase
downstream. Since the diaphragm is completely impermeable to the non wetting phase,
it is perforated in order to inject gas into the system.
The wetting phase can be considered as static under equilibrium conditions. This
equilibrium will be achieved if the pressure drop along the core keeps balance to gravity.
The pressure can be regulated, so the saturation distribution can be regarded as
homogeneous.
The pressure drop is measured along the core, and the wetting phase pressure is regulated
by a mercury flask. The saturations are directly marked by the meniscus m. In doing so
the relative permeability of gases is determined. The wetting phase remains immobile
being in capillary equilibrium with the flowing nonwetting phase.
Using the device by Rappoport and Leas (see Figure 1.25) it is possible to measure the
relative permeability of the wetting phase. A significant distinction from the method
described above is the use of two diaphragms between which the core is positioned. The
desired pressure of the nonwetting phase can be fixed. The wetting phase circulates from
top to bottom. In order to obtain a homogeneous saturation distribution in the core, it is
essential to regulate the flow velocity so that the pressure drop approximately becomes
equal to the hydrostatic pressure difference.
148 1: Fundamental Properties of Porous Media
Figure 1.24: Schematic diagram of a device for measuring relative permeabilities of the nonwetting
phase (after Leas, Jenks, and Rassel)
1: Fundamental Properties of Porous Media 149
Figure 1.25: Schematic diagram of a device for measuring relative permeabilities of the wetting
phase (after Rappoport and Leas)
The Hassler method is considered theoretically adequate and efficiently reproducible.
Summarized, the essence of this method is that only the relative permeability of the
wetting phase will be measured, whereas the nonwetting phase is immobile.
As an example of the efforts of many scientific groups to develop a modification of the
Hasslermethod  in order to measure the relative permeability of both phases some efforts
were made by Osoba as shown in Figure 1.26. The method used by Hafford as shown in
Figure 1.27 which can be considered as a simplification of the instrumentation of Osoba.
150 1: Fundamental Properties of Porous Media
Figure 1.26: Schematic diagram of a device for relative permeability measurements (after Osoba)
If the flow rate is high enough to prevent capillaryendeffects (explained in Chapter 4),
it is possible to run the experiment without a diaphragm on the outlet side.
Although Hafford measured the saturation by weighing, the accuracy of this experiment
is questionable, since demounting the core for weighing (being under high pressure during
the measurement of relative permeability) causes the expansion of the fluids. Therefore it
may be possible that certain fluid quantities will be displaced from the core. Loomis,
Crowell and Richardson investigated this problem and verified the reproducibility of the
Hassler method by neglecting the amount of fluids displaced from the core during
demounting.
1: Fundamental Properties of Porous Media 151
Figure 1.27: Schematic diagram of a device for relative permeability measurements (after Hafford)
1.4.2.2 PENNSTATEMethod
This method was developed by Yuster and colleagues at Pennsylvania State University,
therefore the name PENNSTATE. In many cases it is referred to as the dynamic
steadystate displacement method. Figure 1.28 presents the instrumentation used by
Morse, Terwilliger and Yuster to measure relative permeability. The basic principle behind
this method is the following:
Figure 1.28: Schematic diagram of the PENNSTATE device for relative permeability
measurements (after Morse, Terwilliger, and Yuster)
152 1: Fundamental Properties of Porous Media
At first the core is embedded in a plastic tube in order to achieve the indispensable sealing.
The core placed inside the holder has been fitted between two porous packings which
should provide both satisfactory mixing of fluids and the prevention of capillary end
effects at the inlet and outlet of the core. The fluids are injected at a constant rate by two
micropumps.
Likewise Yuster determined the saturation by weighing. Determination in terms of
measuring the electrical conductivity has not been very successful, since conductivity not
only depends on saturation but also on the fluid distribution.
Problems arise concerning the application of this method in connection with the
development of capillary contact between the core and the two packings. It is also difficult
to guarantee the homogeneous distribution of the phases.
1.4.2.3 WelgeMethod
This method was at first developed by Welge and it is based on the evaluation a continuous
two phase displacement.
1.4.3 Saturation Distribution and Relative
Permeability
The main difference between the Hassler method and the PENNSTATE principle is the
fact that the Hassler method considers the movement of only one fluid. Therefore it is
evident that the mobile phase forms continuous paths, called channels, through the porous
medium. But in case of dynamical methods both phases are in motion. The filtration
processes will only coincide with the results of the Hasslermethod, if each phase
individually forms some channel system of its own and maintains this. In consequence the
fluids should not flow alternately through the same channels as small drops or larger
filaments.
In 1949 and 1950, two experiments were conducted in order to analyse the nature of phase
distribution in porous media. The so called APIresearch project 47B was carried out at
the University of Oklahoma. Between two transparent slices, little spheres were packed
to simulate a porous medium. The water and oil filtrated simultaneously through this
artificial medium. The motion of the fluids had been filmed and magnified.
The film verified that water and oil form their own flow channel systems in which the
wetting fluid occupies the smaller pores. Increasing oil saturation affected a growing
number of oil channels and a decrease of water channels. Further it was noticed that these
channels maintain their position. The flow has been laminar and  in spite of great
tortuosity  free of turbulences.
1: Fundamental Properties of Porous Media 153
Simultaneously the distribution of the residual oil saturation after water displacement has
been investigated. It was noticed that the residual oil is distributed in few oil filaments
occupying the relatively larger pores.
AMOCO also made experiments on sand packing as a porous medium. As the wetting
fluid, Woodmetal was used. The nonwetting fluid was colored synthetic resin. After
transforming the fluids into a solid state, the saturation distribution due to the relative
permeabilities was made observable. The preparation of the sample by grinding off
progressively, a great amount of photographs, enable a journey through the porous
medium. The method also proved the theory of separated flow through the pore channels.
Basically it must be distinguished between two kinds of displacement:
• the wetting phase is the displacing fluid (imbibition)
• the wetting phase is the displaced fluid (drainage)
The phase whose saturation has been increased after displacement is always considered
the displacing phase. Differences between imbibition and drainage are illustrated in
Figure 1.29. Just as supposed the relative permeability is not only a function of saturation,
but also depends on the saturation distribution (in consequence of the structure of the pore
channels and the wettability of the porous medium).
Figure 1.29: Schematic diagram of water invasion into porous media permeabilities of the wetting
phase (after Craig)
154 1: Fundamental Properties of Porous Media
The distribution of the nonwetting phase at drainage differs from the one at imbibition.
Figure 1.30 illustrates this aspect. The deviation of the curve in the direction of
displacement is referred to as the hysteresis of relative permeability. In addition Figure
1.31 shows two oil/water permeability function pairs to indicate the role of wettability.
Figure 1.30: Drainage and imbibition relative permeability characteristic (after Craig)
Figure 1.31: Typical water/oil relative permeability characteristic (after Craig)
1: Fundamental Properties of Porous Media 155
1.5 References
1.1 Craig, F.F. Jr.: "The Reservoir Engineering Aspects of waterflooding" Monograph
Vol. 3 of the Henry L. Doherty series (1971)
1.2 Fatt, I. and Davis, D.H.: "Reduction in Permeability with overburden pressure"
Trans. AIME (1952) p.329
1.3 Hall, H.N.: 5. Petroleum Technology (Jan. 1953)
1.4 Krumbein, W.C. and Sloss, L.L.: "Strategically and Sedimentation" Appleton
century publication, Crofts Inc. New York.
1.5 Leas, W.S., Jenks, W.J. and Russel, Ch.D.: "Relative permeability to Liquid in
Liquid Gas System" Trans. AIME 192 (1951)
1.6 Link,P.K.: "Basic Petroleum Geology" OGCI Publications, Tulsa Oklahoma USA
(1983)
1.7 Monicard, R.P.: "Properties of Reservoir Rocics: Core Analysis" Institut Francais
du Petrole Publications Ed. Technip, Paris (1980)
1.8 Morse, R.A., Terwilliger, P.L. Yuster; S.T.: "Relative Permeability Measurements
on small Core Samples" Producers Monthly II (1947)
1.9 Rappaport, L.A. and Leas, W.J.: "Relative Permeability to Liquid in Liquid Gas
System" Trans AIME 192 (1952)
1.10 Osoba, J.S. et al.: "Laboratory Measurements of Relative Permeability" Trans.
AIME 192 (1951)
156 1: Fundamental Properties of Porous Media
257
2 Equations of SinglePhase Filtration
A quantitative description of a physical process always requires a mathematical
formulation. These mathematics aim at approximating these processes in a more or less
sufficient way, but they will always refer only to the most important aspects of the process.
These mathematics are summarized by the term mathematical model.
Any system properties not included in the mathematical model can’t be taken into
consideration in further calculations.
The description of motion of a continuum are based on the so called constitutive
assumptions or constitutive equations. They usually take the form of relationships
between fluxes and driving forces. We mention as examples of fluxes those of mass,
momentum and energy of various kinds. The conceptional assumption is more
comprehensive than the equation, although outwardly the assumptions take the form of
equations. We do so to emphasize that these equations define the assumed behaviour of
ideal continua. These equations are definitions extracted from physical experiences,
perhaps fortified by experimental evidence. The constitutive equations are often referred
to as phenomenological equations because of their dependence on experimental evidence.
There exist a large number of constitutive equations describing relationships between
fluxes and driving forces. For example: Newton’s law relating shearing force to velocity
gradient; Fourier’s law relating heat flow to temperature gradient; Fick’s law relating flow
of matter of a component to its concentration gradient in a multicomponent system; Ohm’s
law relating electrical current to electrical potential gradient etc.
In all these cases we see a simple, linear dependence of a flow on some conjugated force.
Such a simple relationship does not always hold. There are more or less powerful
approaches for abstracting and simplifying natural phenomena. In each case the
simplification is carried to the point where the model is still amenable to mathematical
treatment, yet is not so simple as to miss those features of the studied phenomena it is
intended to describe. While these equations determine the features of a mathematical
model we refer to this as fundamental equations.
A more general law of the continuum theory is the law of conservation of the extensive
properties as mass, momentum and energy. The resulting equations of continuity are
generally written in the form of partial differential equations. They are commonly referred
to as field equations. However, either the fundamental nor the field equations contain
information regarding the properties of the particular continua under consideration. They
form an under defined mathematical model, insufficient to yield specific answers unless
further equations are supplied.
258 2: Equations of SinglePhase Filtration
Mathematical models of filtration are generally based on Darcy’s law. The field equation
express the conservation of the fluid mass. The mechanical properties of the fluids are
formulated through the equations of state. In order to solve this set of equations, further
equations are required to determine the initial state and the boundary conditions.
Basically one must distinguish between mass flow and filtration. During mass flow all
mass particles of the system are in motion. In case of filtration only a certain part of the
particles are moving, all other mass particles form a solid matrix.
2: Equations of SinglePhase Filtration 259
2.1 Fundamental Equation of Filtration.
2.1.1 Differential Form of the DarcyLaw
Consider a certain volumetric element in space with fixed boundaries. The size of the
volume must be selected in a manner so that random effects may be statistically
eliminated. To simplify but without losing the generality we take a cylindrical element.
The surface of this element is build from stream lines and the base surfaces are
perpendicular to it. Its length is δs and has a cross section δΑ. Equal forces are acting on
the bulk of fluid within the control volume. Two types of forces should be distinguished:
body forces and surface forces. Regarding Newton’s law of motion, the fluid body will
accelerate as long as the resultant acceleration is not zero, but it will keep its velocity; we
seek a value for velocity such that the resultant acceleration equals the null vector.
In filtration process the most important forces are listed in the following:
• force of compression (acting on surface)
• force of gravity (acting on body)
• forces of inertia (acting on body)
• frictional forces (acting on surfaces)
• capillary forces (acting on surfaces)
Figure 2.1: Schematic diagram of a field segment
260 2: Equations of SinglePhase Filtration
Convective and local acceleration are mostly so small in case of filtration, forces of inertia
may be neglected for steady state and nonsteady state filtration as well.
Frictional forces are generally composed from two terms. The first represent the adhesive
forces and is proportional to the velocity inside the pore channels. The second term refers
to the turbulence and is proportional to the squared velocity. If the flow is laminar the
second term disappears.
The frictional force is actually to be regarded as a surface force acting on the surfaces
inside the pore channels. A very complicated labyrinthic structure of the pore channels
makes accurate calculations impossible. Therefore only an average value referring to one
volume unit is considered. Therefore in case of filtration frictional forces become
volumetric forces, but these are only fictious  just as is the velocity of filtration  since it
results only from a mathematical derivation and not from observation and measurement.
Capillary forces are surface forces between two immiscible fluids which are separated by
an interface inside the pore channels. Calculation of these forces is only possible in sum,
in reference to the crosssections of the pore channels. The capillary force is therefore
regarded as a ficitous surface force. In the case of a onephase filtration, capillary forces
are considered only if the waterbearing layer is partially empty.
Such cases are known to exist in hydrology and are called unsaturated filtrations or
unconfined flow.
We assume that only the force of compression , the force of gravity , and the
frictional force , which are proportional to the velocity of filtration, are important. The
forces of inertia and the forces of turbulent friction can be neglected.
The following equation for the equilibrium of forces may be set up:
(2.1)
In reference to a defined elementary volumetric element in Figure 2.1 those forces may
be mathematically specified as follows:
• The force of compression:
, (2.2)
where is the free surface (pore surface) of the element and the pressure
difference along δs.
• The force of gravity:
, (2.3)
F
p
F
g
F
µ
F
µ
F
p
F
g
0 = + +
F
p
φδA∇pδs – =
φδA ∇pδs
F
g
ρ – gφδAi
3
δs =
2: Equations of SinglePhase Filtration 261
where the unit vector is directed upwards.
• The laminar frictional force:
, (2.4)
where B is a coefficient fort a given porous medium.
Substituting Eq. 2.2  Eq. 2.4 into Eq. 2.1 leads to:
, (2.5)
or after reducing:
. (2.6)
In addition, the term is defined as k and interpreted as the permeability. After
transforming Eq. 2.6 the differential form of the law of Darcy is obtained:
(2.7)
We introduce a so called potential function instead of pressure:
(2.8)
where is a reference pressure at .
Then Eq. 2.7 becomes:
, (2.9)
Differentiation of Eq. 2.8, yields:
(2.10)
If Eq. 2.10 is substituted into Eq. 2.9, then Eq. 2.7 is obtained.
i3
F
µ
B – µδAuδs =
∇p µ
B
φ
u i
3
ρg + +
( J
 `
φδAδs ⋅ – 0 =
∇p µ
B
φ
u i
3
ρg 0 = + +
φ B ⁄
u
k
µ
 ∇p i
3
ρg + ( ) – =
ψ gx
3
p d
ρ
 ; ρ
p
o
p
∫
+ ρ p ( ) = =
p
o
x
3
u
kρ
µ
∇ψ – =
ρ∇ψ ∇p ρgi
3
+ =
262 2: Equations of SinglePhase Filtration
2.1.2 Anisotropic Porous Media
The Eq. 2.9 is a linear vectorvector equation. In an isotropic porous medium the
permeability k is a scalar, but in an anisotropic medium is a tensor.
Thus
(2.11)
in detailed form:
. (2.12)
If the matrix is symmetrical (k
ij
= k
ji
) it is possible to transform the coordinate system so
that all values apart from the main diagonal become zero. Directions parallel to these
coordinate axes are called principal directions (axis) of the porous medium. These
principal directions are orthogonal to each other. Writing Eq. 2.12 for a coordinate system
with axis parallel to the principal directions of the porous medium, the Darcy’s law
becomes the following:
u
ρ
µ
k∇ψ – =
u
1
u
2
u
3
ρ
µ

k
11
k
12
k
13
k
21
k
22
k
23
k
31
k
32
k
33
∂ψ
∂x
1

∂ψ
∂x
2

∂ψ
∂x
3

– =
2: Equations of SinglePhase Filtration 263
Figure 2.2: Transformation of the coordinate system
, (2.13)
where:
(2.14)
are defined as the angles between the coordinate axes and vector of gravity. In this
coordinate system Eq. 2.13 becomes a little bit more complicated:
α
2
α
3
α
1
x
1
x’
1
x
3
x’
3
x
2
x’
2
g
i
c
o
s
3
3
α
g
g
i
c
o
s
1
1 α
u’
1
u’
2
u’
3
ρ
µ

k
1
0 0
0 k
2
0
0 0 k
3
∂ψ’
∂x’
1

∂ψ’
∂x’
2

∂ψ’
∂x’
3

– =
ψ’ g x’
i
α
i
cos
i 1 =
3
∑
=
p d
ρ p ( )

p
o
p
∫
+
α
i
264 2: Equations of SinglePhase Filtration
(2.15)
Figure 2.2 Transformation of the coordinate system into . The
potential function ψ in the coordinate system is different to the one in the
coordinate system since gravity is not parallel to the axis.
According to Eq. 2.15 the velocity of filtration is proportional to the potential gradient.
However, the proportions are not the same in different coordinate directions.
u
ρ
µ
k∇ψ’ –
k
µ
 ∇p ρg i
i
α
i
cos
i 1 =
3
∑
+
( J
' J
' J
 `
– = =
x
1
x
2
x
3
( ) x’
1
x’
2
x’
3
, , ( )
x’
1
x’
2
x’
3
, , ( )
x
1
x
2
x
3
( ) x
3
2: Equations of SinglePhase Filtration 265
2.2 Equation of State
2.2.1 Incompressible Fluids
The potential described by Eq. 2.8 includes the density of the fluid. It is a function of
pressure. If this is not the case the fluid is considered incompressible:
(2.16)
or after integration:
. (2.17)
2.2.2 Low Compressibility Fluids
For low compressible fluids one may assume that the fractional change of volume of the
fluid as pressure changes at constant temperature is constant. This constant is called the
coefficient of isothermal compressibility which is defined by:
. (2.18)
By substituting: (2.19)
, (2.20)
where m is the mass of fluid, a constant, into Eq. 2.18 we obtain:
, (2.21)
After integration of Eq. 2.21, yields:
, (2.22)
where: is the density of fluid at any reference pressure .
dρ
dp
 0 =
ρ constant =
c
1
V
f

dV
f
dp

( J
 `
T
– =
V
f
m ρ ⁄ =
c
1
m
ρ


d
m
ρ

( J
 `
dp

1
ρ

dρ
dp
 = – =
ρ ρ
o
e
c p p
o
– ( )
=
ρ
o
p
0
( )
266 2: Equations of SinglePhase Filtration
Applying TAYLOR’s rule and neglecting the terms of higher order we get the following
approximation:
(2.23)
2.2.3 Formation Volume Factor
The volume of produced fluid at surface conditions is usually different than the volume
of fluid entering the wellbore at reservoir conditions. This change of volume is mainly due
to changes in pressure and temperature. Since these fluids are compressible a change of
phase also can take place (gas evolving from oil), but in this text only one phase is
considered. This change of volume should be accounted for by using a conversion factor
(B) which is called Formation Volume Factor. This conversion factor can be defined as:
Formation Volume Factor for oil is the volume of 1 standard m
3
oil (1 m
3
tank oil) under
reservoir conditions p,T and solution gasoil ratio R
s
.
Tthe rate of fluid under reservoir conditions can be calculated by:
(2.24)
ρ ρ
0
1 c p p
o
– ( ) + [ ] ⋅ ≈
B
volume of fluid under reservoir conditions
volume of fluid at standard conditions
 =
q
o
p T R
s
, , ( ) B
o
p T R
s
, , ( )q
os
=
2: Equations of SinglePhase Filtration 267
2.2.4 Ideal and Real Gases
For ideal gases according to the law of BoyleMariott:
, (2.25)
where m is the mass of gas, M the molecular weight and V the volume. Since the density
of a gas is defined as:
. (2.26)
Substituting Eq. 2.25 into Eq. 2.26, yields:
(2.27)
Taking the derivative of Eq. 2.27 yields:
(2.28)
Dividing Eq. 2.28 by Eq. 2.27 we obtain:
(2.29)
For real gases Eq. 2.27 becomes:
(2.30)
where Z(p) is the real gas compressibility factor. Thus the compressibility of a real gas is
defined as:
(2.31)
The formation volume factor for gas is defined by:
(2.32)
Substituting Eq. 2.30 into Eq. 2.32 , yields:
(2.33)
pV
m
M
RT =
ρ
m
V
 =
ρ
M
RT
p =
dρ
dp

M
RT
 =
c
g
1
ρ

dρ
dp

1
p
 = =
ρ
M
RT

p
Z p ( )
, =
c
g
1
ρ

dρ
dp

1
p

1
Z p ( )

∂Z
∂p
 – = =
B
g
ρ
s
ρ
g
 =
B
g
Mp
s
RT
s
⁄
Mp RTZ p ( ) ⁄
 =
268 2: Equations of SinglePhase Filtration
After simplification Eq. 2.33 becomes:
(2.34)
Where T
s
,p
s
are the standard temperature and pressure.
The volume of gas under any pressure and temperature can also be calculated by:
(2.35)
where: V
s
is the volume of gas at standard conditions.
2.2.5 Equation of continuity
The equation of continuity describes the law of mass conservation. We use a rectangular
coordinate system and consider a parallelepiped as a control volume (see Figure 2.3) with
a porosity φ.
Figure 2.3: Volume element in a cartesian coordinate system
The pore volume is then:
. (2.36)
The fluid mass content in the control volume is:
. (2.37)
B
g
p
s
TZ p ( )
pT
s
 C
TZ
p
 ⋅ = =
V p T , ( ) V
s
B
g
=
φdx
1
dx
2
dx
3
ρφdx
1
dx
2
dx
3
2: Equations of SinglePhase Filtration 269
The change of quantity during a time interval dt is:
(2.38)
On the other side, the quantity of fluid flowing through the surface at and is:
(2.39)
and
. (2.40)
The change of flowing quantity results in:
. (2.41)
Also considering the filtration in the direction of other coordinates. Eq. 2.38 and Eq. 2.41
result in:
(2.42)
where the term were already cancelled. After writing it in vector form Eq.
2.42 becomes:
(2.43)
Substituting Eq. 2.15 into Eq. 2.43 yields :
(2.44)
Or in detailed form:
(2.45)
∂ ρφ ( )
∂t
dx
1
dx
2
dx
3
dt.
x
1
x
1
dx
1
+
ρu
1
( )
x
1
dx
2
dx
3
dt
ρu
1
( )
x
1
dx
1
+
dx
2
dx
3
dt
∂
∂x
1
 – ρu
1
( )dx
1
dx
2
dx
3
dt
∂
∂t
 – φρ ( )
∂
∂x
i
 ρu
i
( ).
i 1 =
3
∑
=
dx
1
dx
2
dx
3
∂
∂t
 – φρ ( ) ∇ ρu ( ) =
∇
k
µ
ρ
2
∇ψ
( J
 `
∂
∂t
 φρ ( ) =
∂
∂x
1

k
1
µ
ρ
2
∂ψ
∂x
1

( J
' J
 `
∂
∂x
2

k
2
µ
ρ
2
∂ψ
∂x
2

( J
' J
 `
∂
∂x
3

k
3
µ
ρ
2
∂ψ
∂x
3

( J
' J
 `
+ +
∂
∂t
 φρ ( ) =
270 2: Equations of SinglePhase Filtration
Differentiation of Eq. 2.14 yields:
. (2.46)
If the 3rd coordinate direction is vertical then Eq. 2.46 become more simple:
Substituting Eq. 2.46 into Eq. 2.44 yields:
. (2.47)
If the 3rd coordinate direction is vertical then Eq. 2.47The second term on the left hand
side is often very small compared to the first. For these cases the following equation is
sufficient:
. (2.48)
ρ
2
∇ψ ρ∇p ρ
2
g i
i
α
i
cos
i 1 =
3
∑
+ =
ρ
2
∇ψ ρ∇p ρ
2
gi
3
+ =
∇
k
µ
ρ∇p ∇ ρ
2
k
µ
gi
3
+
∂
∂t
 φρ ( ) =
∇
k
µ
ρ∇p
∂
∂t
 φρ ( ) =
2: Equations of SinglePhase Filtration 271
2.3 Special Forms of the Equation of Filtration
2.3.1 Incompressible Fluids
Assuming an incompressible fluid and porous media, then and are constant and Eq.
2.48 becomes:
. (2.49)
In the case of a homogenous porous media and then Eq. 2.49 becomes:
. (2.50)
Eq. 2.50 can be simplified by introducing a new independent variable:
(2.51)
Then Eq. 2.50 becomes the Laplace equation:
(2.52)
For isotropic porous media Eq. 2.49 becomes:
(2.53)
2.3.2 Low Compressibility Fluids
Based on Eq. 2.22 the following transformation can be made:
(2.54)
Substituting Eq. 2.54 into Eq. 2.48 becomes:
(2.55)
ρ φ , µ
∇ k∇p [ ] 0 =
k const =
∇ k∇p [ ] k
1
∂
2
p
∂x
2
1
 k
2
∂
2
p
∂x
2
2
 k
3
∂
2
p
∂x
2
3
 + + 0 = =
η
i
x
i
k
1
k
i
, i 1 2 3 , , = =
∂
2
p
∂η
1
2

∂
2
p
∂η
2
2

∂
2
p
∂η
3
2
 0 = + +
∇ k∇p [ ] ∇
2
p
∂
2
p
∂x
1
2

∂
2
p
∂x
2
2

∂
2
p
∂x
3
2
 0 = + + = =
ρ∇p ρ
0
e
c p p
0
– ( )
p ∇
1
c
 ρ
0
e
c p p
0
– ( )
( J
 `
∇
1
c
 ∇ρ = = =
∇
k
cµ
∇ρ
∂
∂t
 φρ ( ) =
272 2: Equations of SinglePhase Filtration
Differentiation of Eq. 2.23 gives:
. (2.56)
Substituting Eq. 2.23 and Eq. 2.56 into Eq. 2.55 results in:
(2.57)
In Eq. 2.57 , c and are constant. The viscosity can be regarded as constant as well
hence the fluid compressibility is small.
For a homogeneous isotropic and incompressible porous media and constant fluid
viscosity, Eq. 2.55 becomes:
(2.58)
where:
(2.59)
is defined as the piezometric conductivity. Eq. 2.58 is identical to the
Fourierequation of heat conductivity.
2.3.2.1 Elastic Porous Media
The porosity and the permeability of an elastic porous medium are both functions of
pressure:
(2.60)
The actual functions have to be determined by measurements.
Without serious restictions one may assume that the interdependencies are small and the
alteration of porosity and permeability are proportional to pressure changes. Then these
functions in Eq. 2.60 are similar to Eq. 2.22 and may be written as follows:
(2.61)
Substituting Eq. 2.61 together with Eq. 2.22 into Eq. 2.48 leads to:
∇ρ ρ
o
c∇p =
∇
k
µ
ρ
o
∇p
∂
∂t
 φρ
o
1 c p p
o
– ( ) + [ ] { ¦ =
ρ
0
p
0
µ
∇
2
p
1
K

∂p
∂t
, =
K
k
µcφ
 =
K m
2
s
1 –
[ ]
k k p ( ), φ φ p ( ) = =
k k
o
e
c
k
p p
o
– ( )
=
φ φ
o
e
c
φ
p p
o
– ( )
=
2: Equations of SinglePhase Filtration 273
(2.62)
Eq. 2.62 can be written as:
(2.63)
After differentiation it is quite evident that Eq. 2.63 and Eq. 2.64 are identical. After
simplification:
(2.64)
Since and are very small it can be assumed:
(2.65)
and
In addition:
(2.66)
Substituting Eq. 2.65 and Eq. 2.66 into Eq. 2.64 leads to:
(2.67)
Eq. 2.67 shows that the compressibilities of the fluid and the porous medium are added in
the piezometric conductivity term:
(2.68)
where c
t
is the total or ultimate compressibility of the system.
∇
k
o
ρ
o
µ
e
c c
k
+ ( ) p p
o
– ( )
∇p
∂
∂t
 φ
o
ρ
o
e
c c
φ
+ ( ) p p
o
– ( )
[ ] =
k
o
ρ
o
µ c c
k
+ ( )
∇ ∇ e
c c
k
+ ( ) p p
o
– ( )
( ) [ ]=
φ
o
ρ
o
c c
φ
+ ( )
c c
k
+
e
c
φ
c
k
– ( ) p p
o
– ( ) ∂
∂t
 e
c c
φ
+ ( ) p p
o
– ( )
[ ]
∇
2
e
c c
k
+ ( ) p p
o
– ( )
[ ]
φ
o
µ c c
φ
+ ( )
k
o
e
c
φ
c
k
– ( ) p p
o
– ( ) ∂
∂t
 e
c c
k
+ ( ) p p
o
– ( )
[ ] =
c
φ
c
k
e
c
φ
c
k
– ( ) p p
o
– ( )
1 =
c c
k
+ c c
φ
+ =
e
c c
k
+ ( ) p p
o
– ( )
1 c c
k
+ ( ) p p
o
– ( ) + ≅
∇
2
p
φ
o
µ c c
φ
+ ( )
k
o

∂p
∂t

φ
o
µc
t
k
o

∂p
∂t
. = =
c
t
c c
φ
+ =
274 2: Equations of SinglePhase Filtration
2.4 Real and Ideal Gases
From Eq. 2.30 the real gas density is:
, (2.69)
and the compressibilty is
(2.70)
Substituting Eq. 2.69 into Eq. 2.48 and taking the porosity as constant since the rock
compressibility is several orders of magnitude less than the gas compressibility, the right
side of the Eq. 2.48 can be developed in the following way:
(2.71)
Thus, Eq. 2.48 becomes:
, (2.72)
After simplification of Eq. 2.72, yields:
, (2.73)
AlHussainy, Ramey, Crawford then introduced the following function:
(2.74)
which is called the real gas pseudo pressure. This function enabled the following
derivations:
(2.75)
and
(2.76)
ρ
Mp
RT Z
 =
c
g
1
ρ

dρ
dp
 ⋅ =
∂
∂t
 φρ ( ) φ
∂ρ
∂t
 φ
1
ρ

∂ρ
∂p
 ρ
∂p
∂t

( J
 `
= = =
=φ c
g
Mp
RT Z

∂p
∂t

M
2RT
φc
g
µ
2p
µZ

∂p
∂t
, =
∇
k
µ

Mp
RT Z
∇p
M
2RT
φc
g
µ
2p
µZ

∂p
∂t
 =
∇ k
2p
µZ
∇p φc
g
µ
2p
µZ

∂p
∂t
 =
m p ( ) 2
pdp
µZ
,
p
b
p
∫
=
∇m p ( )
dm p ( )
dp
∇p
2p
µZ
∇p, = =
∂m p ( )
∂t

dm p ( )
dp

∂p
∂t

2p
µZ

∂p
∂t
 = =
2: Equations of SinglePhase Filtration 275
Substitution of Eq. 2.75 and Eq. 2.76 into Eq. 2.73 results in:
, (2.77)
Assuming that the porous medium is isotropic and homogeneous, Eq. 2.77 becomes:
. (2.78)
This equation is identical to Eq. 2.58 for low compressibility fluids. The differences that
c
g
and µ are both functions of pressure and hence Eq. 2.78 will not be linear.
Ideal gases are characterized by the following terms:
Z = 1, and µ = constant,
where is defined as the average pressure of the considered volume and time interval.
For ideal gas Eq. 2.74 becomes:
(2.79)
Substitution of this equation into Eq. 2.78 leads to:
(2.80)
∇ k∇m p ( ) [ ] φc
g
µ
∂m p ( )
∂t
 =
∇
2
m p ( )
φc
g
µ
k

∂m p ( )
∂t
 =
c
g
1
p
 =
p
m p ( )
p
2
µ
. =
∇
2
p
2
φµ
kp

∂p
2
∂t
 =
276 2: Equations of SinglePhase Filtration
2.5 Boundary and Initial Conditions
2.5.1 Boundary Conditions
The space where filtration takes place may be limited or infinite. Infinity is more a
mathematical fiction than a fact but it is very useful for the solution of the equations.
A boundary is considered open if the fluid is able to pass through and closed if it is not.
In the case of a closed boundary the following mathematical formulations may be made:
(2.81)
Where is a unit vector normal to the boundary .
In case of an open boundary it is either the potential or the gradient of the potential given
at the boundary as a function of time:
(2.82)
or
, (2.83)
In both cases the velocity across the boundary is calculated by:
(2.84)
It is possible to combine the boundary conditions Eq. 2.82 and Eq. 2.83. This leads to a
general equation for the boundary conditions:
, (2.85)
a, b and c are functions of time.
The condition in Eq. 2.81 is valid at the top or bottom boundaries of an oil, gas or
waterbearing layer. In addition it is also valid at faults and pinch outs. The condition in
Eq. 2.82 is valid in the case of water intruding from the surface into the reservoir.
u.n
kρ
µ
∇ψn –
kρ
µ

∂ψ
∂n
 –
Γ
0, = = =
∂ψ
∂n
 0, x Γ, ∈ =
n Γ
ψ
Γ
ψ x t , ( ), x Γ ∈ =
ψ′
Γ
∂ψ x t , ( )
∂n
, x Γ ∈ =
u
n
kρ
µ

∂ψ
∂n
 =
a
∂ψ
∂n
 bψ + c, x Γ ∈ =
2: Equations of SinglePhase Filtration 277
Eq. 2.83 must be applied at the surface of a well (inner boundary) producing (or injecting)
at a given rate.
2.5.2 Initial Conditions
In order to solve filtration problems which have time dependent solutions it is necessary
to know the state of the system at a certain date. This is usually the temporal starting point
t = 0. The expression initial conditions refers to this practice. The state of the system at t
= 0 is called initial condition:
, (2.86)
and
(2.87)
By solving problems such as filtration of incompressible fluids or a steady state filtration
no initial conditions are required.
Figure 2.4: Illustration of the boundary conditions
ψ x t
0
, ( ) ψ
0
x ( ) =
p x t
0
, ( ) p
0
x ( ) =
t ∞ → ( )
278 2: Equations of SinglePhase Filtration
2.5.3 Discontinuities in Porous Media
Permeability may change by leaps and bounds at the contact surface of two regions of the
porous medium. In such a case the potential and the normal component of the velocity at
the boundary must be continuous:
(2.88)
and
. (2.89)
Both and are potential functions of the two regions separated by the discontinuity
surface.
ψ
1
ψ
2
=
ρ
k
1
µ

∂ψ
1
∂n
 ρ
k
2
µ

∂ψ
2
∂n
 =
ψ
1
ψ
2
2: Equations of SinglePhase Filtration 279
2.6 Schematic of the Filtration Equations
Equation of Equation of Mass Equation of Flow
State Conservation DarcyLaw
Incompressible Fluid and Porous Medium:
Homogenous and isotropic Porous Medium
Low compressible Fluid and elastic Porous Medium
Homogenous and isotropic Porous Medium
Real Gas:
Homogenous and isotropic Porous Medium
Ideal Gas: ; ; ;
ρ ρ p ( ) = ∇ ρu ( )
∂
∂t
 – φρ ( ) = u
ρ
µ
k∇ψ – =
φ φ p ( ) =
ρ∇ψ ∇p ρgi
3
+ =
∇
ρ
µ
k∇p
∂
∂t
 φρ ( ) =
ρ µ φ , , const =
∇ k∇p ( ) 0 =
∇ ∇p ( ) p
2
∇ 0 = =
ρ ρ
0
1 c
f
p p
0
– ( ) + ( ) =
φ φ
0
1 c
φ
p p
0
– ( ) + ( ) =
c
t
c
f
c
φ
+ =
∇ k∇p ( ) φ
0
c
t
µ
∂p
∂t
 =
p
2
∇
1
K
 
∂
∂t
 p ( ) = K
k
φc
t
µ
 =
ρ
Mp
RTZ
 = m p ( ) 2
p
Zµ
 p d
p
0
p
∫
=
∇ k∇m p ( ) ( ) φc
g
µ
∂m p ( )
∂t
 =
∇
2
m p ( )
φc
g
µ
k

∂m p ( )
∂t
 =
Z 1 = c
g
1
p
 = µ
g
const = m p ( )
p
2
µ
 =
∇
2
p
2 φµ
kp

p
2
∂
t ∂
 =
280 2: Equations of SinglePhase Filtration
Appendix A
Derivation of Laplace Equation in Radial Coordinates
The Laplace equation in the Cartesian coordinate system is:
(A.1)
To convert Eq. A.1 to a radial coordinate system the independent variables r and have
to be considered. The Cartesian coordinates can be expressed:
:
(A.2)
(A.3)
Since:
(A.4)
and
(A.5)
Inserting Eq. A.4 and Eq. A.5 in Eq. A.2 and Eq. A.3 yields following results:
(A.6)
(A.7)
Multiplying Eq. A.6 with leads to:
(A.8)
∂
2
p
∂x
2

∂
2
p
∂y
2
 + 0 =
θ
x r θ cos =
y r θ sin =
∂p x y , ( )
∂r

∂p
∂x

∂x
∂r

∂p
∂y

∂y
∂r
 + =
∂p x y , ( )
∂θ

∂p
∂x

∂x
∂θ

∂p
∂y

∂y
∂θ
 + =
∂x
∂r
 θ;
∂y
∂r
 cos θ sin = =
∂y
∂θ
 r θ;
∂y
∂θ
 sin – r θ cos = =
∂p
∂r

∂p
∂x
 θ cos
∂p
∂y
 θ sin + =
∂p
∂θ

∂p
∂x
r θ sin –
∂p
∂y
 θ cos + =
∂p
∂r

∂p
∂x

∂x
∂r

∂p
∂y

∂y
∂r
 +
∂p
∂x
 θ cos
∂p
∂y
 θ sin + = =
∂p
∂θ

∂p
∂x

∂x
∂θ

∂p
∂y

∂y
∂θ
 +
∂p
∂x
 r θ sin – ( )
∂p
∂y
r θ cos + = =
θ cos
∂p
∂r
 θ cos ⋅
∂p
∂x
 θ
2
cos
∂p
∂y
 θ cos θ sin + =
2: Equations of SinglePhase Filtration 281
Multiplying Eq. A.7 with and some rearrangements lead to:
(A.9)
Adding Eq. A.8 and Eq. A.9 yield:
(A.10)
If Eq. A.6 is multiplyed with and Eq. A.7 with the same calculation leads to:
(A.11)
Since
(A.12)
and
(A.13)
and
(A.14)
Then Eq. A.12 becomes:
(A.15)
Also Eq. A.13 becomes:
(A.16)
θ sin
1
r

∂p
∂θ
 θ sin – ( ) ⋅ ⋅
∂p
∂x
 θ
2
sin
∂p
∂y
 θ cos θ sin – =
∂p
∂r
 θ cos ⋅
1
r

∂p
∂θ
 θ sin – ( ) ⋅ ⋅ +
∂p
∂x
 θ
2
cos θ
2
sin + ( ) ⋅
∂p
∂x
 = =
θ sin θ cos
∂p
∂r
 θ sin ⋅
1
r

∂p
∂θ
 θ cos ⋅ ⋅ +
∂p
∂y
 =
∂
∂x

∂p r θ , ( )
∂x

∂r
∂x

∂
∂r

∂p
∂x

( J
 `
∂θ
∂x

∂
∂θ

∂p
∂x

( J
 `
+ =
∂
∂y

∂p r θ , ( )
∂y

∂r
∂y

∂
∂r

∂p
∂y

( J
 `
∂θ
∂x

∂
∂θ

∂p
∂y

( J
 `
+ =
∂r
∂x

x
r
 θ,
∂θ
∂x
 cos
θ sin –
r
 = = =
∂r
∂y

y
r
 θ,
∂θ
∂y
 sin
θ cos
r
 = = =
∂
2
p
∂x
2
 θ
∂
∂r
 θ
∂p
∂r

1
r
 – cos θ
∂p
∂θ
 sin cos =
θ sin
r

∂
∂θ
 θ
∂p
∂r

1
r
 – cos θ
∂p
∂θ
 sin –
∂
2
p
∂y
2
 θ
∂
∂r
 θ sin
∂p
∂θ

1
r
 + θ cos
∂p
∂r
 sin =
+
θ cos
r

∂
∂θ
 θ sin
∂p
∂θ

1
r
 + θ cos
∂p
∂r

282 2: Equations of SinglePhase Filtration
Substitute Eq. A.15 and Eq. A.16 into Eq. A.1, with simplifications, yields :
(A.17)
Eq. A.17 can be written as:
(A.18)
∂
2
p
∂r
2

1
r

∂p
∂r

1
r
2

∂
2
p
∂θ
2
 + + 0 =
1
r

∂
∂r
 r
∂p
∂r

( J
 `
1
r
2

∂
2
p
∂θ
2
 + 0 =
383
3 Solutions of the SinglePhase
Equation of Filtration
In order to find a suitable solution to the filtration equation, one should know what type
of condition can be used to describe a flow regime. Generally there are two conditions:
NonSteady State and Steady State.
The NonSteady State may be subdivided into the following conditions:
A  Transient condition:
This condition is valid at an early, relatively short time, where the pressure response
in the reservoir is not affected by the presence of an outer boundary, thus the
reservoir appears infinite acting. In well testing this condition is applicable when the
production rate is deliberately changed for a short time, the pressure response is
measured for a few hours and the boundary effects will not be felt and therefore the
reservoir is mathematically infinite.
B  Late Transient condition:
This condition exists in the period when the boundary effects start to show up in the
pressure response. This will occur when the well test period takes a longer time, or
the reservoir is smaller than expected.
C  PseudoSteady State condition:
The Pseudo (or Semi) Steady Sate condition occurs after the late transient condition
and is valid for a reservoir which has been producing for a sufficient period of time
so that the boundary effect has been felt by the pressure response. The outer
boundary could be impermeable to fluids (no flow boundary) or a constant pressure
boundary.
Steady State condition:
The Steady State condition occurs also after the late transient period. This condition is
applicable when the production rate is constant and fluid withdrawal will be exactly
balanced by fluid entry across the open boundary, so that there is no change in pressure
with time in the whole reservoir.
384 3: Solutions of the SinglePhase Equation of Filtration
3.1 Steady State Filtration
3.1.1 Steady State Filtration of Low Compressibility
Fluid
Filtration is steady state if the potential (and so the flow) at any point of the system is
independent of time ( ). For constant fluid viscosity, Eq. 2.57 will be reduced
to:
(3.1)
Eq. 3.1 is similar to Eq. 2.49 for incompressible fluids.
For homogeneous and isotropic porous media (k = k = constant) follows:
(3.2)
A linear one dimensional model is used for the sake of simplicity. Mathematically it can
be described by:
(3.3)
and its solution after integrating Eq. 3.3 twice is given by:
(3.4)
where: a and b are constants of integration. The actual value of the pressure is then
determined by boundary conditions.
A production rate (q) is assumed negative when flowing out of the porous medium and
positive when flowing into it. Thus q becomes negative when produced and positive when
injected.
The boundary conditions are used to determine a and b as follows:
At is .
Then b is determined from Eq. 3.4 as:
(3.5)
At is
∂p/∂t 0 =
k p ∇ ( ) ∇ 0 =
p ∇
2
0 =
x
2
2
∂
∂ p
0 =
p ax b + =
x 0 = p p
i
=
b p
i
=
x L =
kA
µ

dp
dx
 qB =
3: Solutions of the SinglePhase Equation of Filtration 385
where B is the formation volume factor and q is negative. Taking the derivative of Eq. 3.4
then
, (3.6)
and by substituting into the second boundary condition (B.C.), is determined as:
(3.7)
Substituting Eq. 3.5 and Eq. 3.7 into Eq. 3.4 yields:
(3.8)
The value of p  p
i
becomes negative in case of flow taking place in direction of the xaxis
(q is negative) and positive in the inverse case.
3.1.2 Steady State Filtration in a Radial System
The Cartesian coordinates can be transformed into radial coordinates, as shown in
Appendix A, Chapter 2, which will leave Eq. 3.2 in the following form:
(3.9)
If the pressure (or potential) distribution is independent of , then the filtration is radially
symmetrical, and
x d
dp
a =
a
a
µqB
kA
 =
p p
i
–
µqB
kA
x =
1
r

r ∂
∂
r
r ∂
∂p
( J
 `
1
r
2

θ
2
2
∂
∂ p
+ 0 =
θ
θ
2
2
∂
∂ p
0 =
386 3: Solutions of the SinglePhase Equation of Filtration
Figure 3.1: The Radial Coordinate System
This simplifies Eq. 3.9 to:
(3.10)
Eq. 3.10 can be expressed in the following form also:
(3.11)
Integrating Eq. 3.10 twice leads to:
(3.12)
We set a constant pressure at the inner radius r
w
:
at (3.13)
and a constant rate q. From Darcy equation:
at , (3.14)
where is the inner surface and h the thickness of the layer. Note: outfow
rate is negative while inflow rate is positive. Taking the derivative of Eq. 3.12 with
respect to r (dp/dr =a/r) and substituting it into Eq. 3.14 yields: (3.15)
. (3.16)
b can be calculated from Eq. 3.12 as:
(3.17)
1
r

r ∂
∂
r
r ∂
∂p
( J
 `
0 =
r
2
2
∂
∂ p 1
r

r ∂
∂p
+ 0 =
p a lnr b + ⋅ =
p p
wf
= r r
w
=
kA
µ

r d
dp
q – B = r r
w
=
A 2πr
w
h =
a
µqB
2πhk
 – =
b p
wf
a – lnr
w
=
3: Solutions of the SinglePhase Equation of Filtration 387
and
(3.18)
Substituting the values of a and b into Eq. 3.12 leads to:
(3.19)
Figure 3.2: Illustration of steadystate filtration in a radial system
Considering an outer boundary ( ) with constant pressure p
e
Eq. 3.19 leads to:
. (3.20)
This equation is known as Dupuit equation.
If the inner radius r
w
represent a well then p
wf
is called as bottom hole flowing pressure.
b p
wf
µqB
2πhk
lnr
w
+ =
p
wf
p –
µqB
2πhk
ln
r
r
w
 =
r r
e
=
q
2πhk
µB

p
e
p
wf
–
ln r
e
r
w
⁄ ( )
 – =
388 3: Solutions of the SinglePhase Equation of Filtration
3.1.3 Steady State Gas Filtration
From Eq. 1.41 and substituting of by Eq. 2.30 one can set up the following equation:
(3.21)
Then Eq. 3.21 is integrated along a linear segment:
(3.22)
Substituting Eq. 2.74 into Eq. 3.22 we obtain:
(3.23)
In the case of a radial symmetric filtration (see Figure 3.2) the surface area is
and Eq. 3.21 becomes:
(3.24)
or
(3.25)
Assuming that the pressure difference between the two ends of a linear porous body is not
very large, then Z(p) is substituted by Z, for p = p
i
and Eq. 3.25 becomes:
(3.26)
The fraction is generally considered small and so the last term ~ 1.
At a given exterior radius r
e
the pressure is p
i
. Then Eq. 3.26 may be set up as follows:
(3.27)
where:
ρ
Mp
RTZ p ( )
 p d
µQ
m
Ak
 1
βQ
m
µA
 +
( J
 `
x d =
2
p
Z p ( )µ
 p d
p
i
p
∫
2Q
m
RT
MAk
 – 1
βQ
m
µA
 +
( J
 `
x =
m p ( ) m p
i
( ) –
2Q
m
RT
MAk
 – 1
βQ
m
µA
 +
( J
 `
x =
A 2rπh =
2
p
Z p ( )µ
 p d
p
i
p
∫
2Q
m
RT
MAk
 –
1
r

βQ
m
2πr
2
hµ
 +
( J
' J
 `
r d
r
w
r
∫
=
m p ( ) m p
wf
( ) –
Q
m
RT
πhMk
 ln
r
r
w

βQ
m
2πhµ

1
r
w

1
r
  –
( J
 `
+ – =
p
i
2
p
wf
2
–
µQ
m
RTZ
πhMk
 – ln
r
e
r
w

βQ
m
2πhµr
w
 1
r
w
r
e
 –
( J
' J
 `
+ =
r
w
r
e
⁄ 1 r
w
r
e
⁄ – ( )
p
i
2
p
wf
2
– AQ
m
– BQ
m
2
– =
3: Solutions of the SinglePhase Equation of Filtration 389
Q
m
is negative for producing wells and positive for injection wells. Eq. 3.27 can be
transformed into the following form:
(3.28)
This equation is useful to calculate gas well production rates as a function of p
wf
. In
practice the following equation is also used:
Figure 3.3: Plots of production equation for gas wells
(3.29)
The constants C and n have to be determined by fitting of Eq. 3.29 to the measured Q
m
and p
wf
values. In the same way we can evaluate the constants in Eq. 3.27 instead
calculating them with Eq. 3.28. The graphical illustrations are shown in Figure 3.3.. Eq.
3.28 indicates that the slope of the line in Figure 3.3a is equal to B and the intercept is A.
A
µRTZ
πhMk
ln
r
e
r
w
 = B
βRTZ
2π
2
h
2
kMr
w
 =
p
i
2
p
wf
2
–
Q
m
 A B Q
m
+ =
Q
m
C p
i
2
p
wf
2
– ( )
n
=
390 3: Solutions of the SinglePhase Equation of Filtration
3.2 NonSteady State Filtration in Infinite Acting
Systems
3.2.1 Radial Systems with Constant Production
Rate
Eq. 2.58 transformed into the radial yields:
(3.30)
This equation is of essential importance when solving technical problems. It is a base for
calculations of production rates of hydrocarbon wells and deep water wells. Further it is
also applied for the evaluation of pressure buildup curves and for the water inflow into
hydrocarbon reservoirs.
It is assumed that the fluid bearing layer has a thickness h, a porosity φ, a permeability k
and is horizontal, homogeneous, isotropic and infinite. The fluid has a viscosity µ and a
compressibility c.
The radius of the well is r
w
. At the pressure is the same everywhere:
(3.31)
The well should produce continuously at a rate q. According to Darcy’s law we obtaine
the following equation:
(3.32)
The boundary condition in infinity is:
(3.33)
We seek for a solution in the form:
for (3.34)
r
2
2
∂
∂ p 1
r

r ∂
∂p
+
1
K

t ∂
∂p
= K
k
µφc
 =
t 0 =
p p
i
= r r
w
> t 0 =
2πr
w
hk
µ

r ∂
∂p
( J
 `
r r
w
=
qB – =
p p
i
= r ∞ = t 0 >
p p r t ( , ) = r
w
r ∞ < < t 0 >
3: Solutions of the SinglePhase Equation of Filtration 391
To explain the solution method used below we take Pythagoras theorem as an example.
To calculate the hypotenuse of a right triangle we are looking for a solution of:
(3.35)
where x and y are the side lengths and z is a function of two variables, x and y. It is possible
to reduce the number of independent variables by introducing a new one in the form of:
(3.36)
So the variables x and y in Eq. 3.35 can be replaced by
(3.37)
In a similar manner we assume that p is a function of only one variable z and we introduce
the following relationship:
(3.38)
Based on this assumption we get:
(3.39)
(3.40)
(3.41)
Substituting Eq. 3.39  Eq. 3.41 into Eq. 3.30 yields:
(3.42)
To solve Eq. 3.42 the boundary conditions must also be transformed. Using Eq. 3.38 the
B.C. in Eq. 3.33 can be transferred into:
(3.43)
Using Eq. 3.38 and Eq. 3.40, the first B.C. in Eq. 3.32 can be transformed as well to:
(3.44)
z z x y , ( ) =
w x
2
y
2
+ =
z z w ( ) =
z
r
2
K t ⋅
 =
1
K

t ∂
∂p 1
K

z ∂
∂p
t ∂
∂z r
2
K
2
t
2

z d
dp z
Kt

z d
dp
– = = =
r ∂
∂p
z d
dp
r ∂
∂z 2r
Kt

z d
dp 2r
2
Kt

1
r
 
z d
dp 2z
r

z d
dp
= = = =
r
2
2
∂
∂ p 2
Kt

z d
dp 4r
2
K
2
t
2

z
2
2
d
d p
+
2
Kt

z d
dp 4z
Kt
 
z
2
2
d
d p
+ = =
4z
z
2
2
d
d p
4 z + ( )
z d
dp
+ 0 =
p p
i
= z ∞ = t 0 >
z
z d
dp µqB
4πhk
 – =
392 3: Solutions of the SinglePhase Equation of Filtration
Eq. 3.42 can be written as:
(3.45)
or:
(3.46)
Let
then Eq. 3.46 becomes:
(3.47)
By separation of variables:
(3.48)
By integrating Eq. 3.48 yields:
(3.49)
where A
1
is a constant of integration, taking the exponential of Eq. 3.49
(3.50)
or
(3.51)
and the second substitution with the notation and by separation of variables, Eq.
3.51 becomes:
(3.52)
z
z
2
2
d
d p
z d
dp
+
z
4
 –
z d
dp
=
z d
d
z
dp
dz

( J
 `
1
4
z
dp
dz
 – =
y z
z d
dp
=
z d
d
y y/4 – =
z d
d
y
y
4
 – =
lny lnA
1
–
z
4
 – =
y A
1
e
z 4 ⁄ –
=
z
z d
dp
A
1
e
z 4 ⁄ –
=
ξ z 4 ⁄ =
dp A
1
e
ξ –
ξ
dξ =
3: Solutions of the SinglePhase Equation of Filtration 393
Integrating Eq. 3.52 yields:
(3.53)
Using Eq. 3.51 and the boundary condition in Eq. 3.44 and assuming that r
w
is very small
so that the condition is valid, can be determined as:
(3.54)
The integral in Eq. 3.53 cannot be solved in a closed form. This integral is defined as the
so called exponential integral and its numerical solution can be found in any mathematical
handbook:
(3.55)
Finally Eq. 3.53 becomes:
(3.56)
The calculation of the pressure drop at the well bottom is made by substituting
and into Eq. 3.56:
(3.57)
p
i
p – A
1
e
ξ –
ξ
dξ
z 4 ⁄
∞
∫
=
r
w
2
4Kt
 0 ≈ A
1
A
1
µqB
4πhk
 – =
Ei x – ( )
e
ξ –
ξ
dξ
x
∞
∫
– =
p
i
p r t , ( ) –
µqB
4πhk
Ei
r
2
4Kt
 –
( J
 `
=
r r
w
=
p r t ( , ) p
wf
t ( ) =
p
i
p
wf
–
µqB
4πhk
Ei
r
w
2
4Kt
 –
( J
' J
 `
=
394 3: Solutions of the SinglePhase Equation of Filtration
Figure 3.4: Plots of the Ei(z) function (after Chaumet)
3.2.2 Properties of the EiFunction
The function is illustrated in Figure 3.4. In the vicinity of the
TAYLORSeries of is defined as:
(3.58)
where is the EULERConstant.
If the series in Eq. 3.58 will have very small values in terms higher than the third
term, which makes the following approximation valid:
(3.59)
then:
(3.60)
Ei z – ( ) – z 0 =
Ei z – ( ) –
Ei z – ( ) – γ – ln z z
z
2
4
 .... + – + – =
γ 0 57722 , =
z 1 «
Ei z – ( ) – 0 57722 , ln z – – =
Ei
r
2
4Kt
 –
( J
 `
– 0 57722 , ln
4Kt
r
2
 – – 0 80907 , ln
Kt
r
2
  + = =
3: Solutions of the SinglePhase Equation of Filtration 395
3.2.3 Pressure Drop in Space and Time
Substitution of Eq. 3.60 into Eq. 3.56 leads to:
(3.61)
The error of this approximation becomes less than 1 % if the following condition is
satisfied:
(3.62)
Eq. 3.61 enables the calculation of the pressure drop between two selected radii. The
pressure is evaluated at the the two radii seperately using Eq. 3.61 and the difference gives
the pressure drop between the two points. A graphical illustration of these values indicates
a parallelism of the pressure drop curves. This means that the pressure difference between
two radii becomes a constant for a given production time interval.
This fact can be verified mathematically with Eq. 3.61 where r
w
and r
e
are the two
considered radii:
(3.63)
(3.64)
By substracting Eq. 3.64 from Eq. 3.63 we get:
(3.65)
The pressure difference is therefore independent of time but Eq. 3.65 is only valid if
.
Further it is possible to illustrate the spatial distribution of pressure according to Eq. 3.61
as shown in Figure 3.5.
p
i
p r t , ( ) –
µqB
4πhk
 – 0 80907 , ln
Kt
r
2
  +
( J
 `
=
Kt
r
2
 10 >
p
i
p
wf
–
µqB
4πhk
 – 0 80907 , ln
Kt
r
w
2
 +
( J
' J
 `
=
p
i
p
e
–
µqB
4πhk
 – 0 80907 , ln
Kt
r
e
2
 +
( J
' J
 `
=
p
e
p
wf
–
µqB
4πhk
ln
r
e
2
r
w
2

µqB
2πhk
 ln
r
e
r
w
 – = – =
Kt
r
2
 10 >
396 3: Solutions of the SinglePhase Equation of Filtration
Figure 3.5: Plots of pressure drop in the vicinity of a well (infinite reservoir, compressible fluid)
Figure 3.6: Plots of pressure drop in the vicinity of a well (infinite reservoir, compressible fluid)
Figure 3.6 shows the pressure distribution as a function of log r for various t. The
continuous line represents calculations by Eq. 3.61, the staggered line calculations by Eq.
3.56. The linear section of the plot is shifted parallel in time.
3: Solutions of the SinglePhase Equation of Filtration 397
Example 3.1:
The following data of an oil bearing layer are known:
Permeability: k = 10 [mD] = 20.10
15
[m
2
]
Porosity: =0.2[]
Thickness: h = 5 [m]
Well Radius: = 0.1 [m]
Oil Viscosity: = 10
3
[Pa.s]
Oil Compressibility: c = 10
9
[Pa
1
(10
4
bar
1
)]
Production rate: q = 10 [m
3
/day] (Bo=1)
For how long should the well be produced to allow the use the Eq. 3.61 in the following
cases:
a) For calculation of the bottom hole flowing pressure
b) For calculation of the pressure at 100 [m] distance from the well
Solution:
At the well radius 3.61 is valid if :
At a 100 m distance from the well the time limit for application of Eq. 3.61 instead of Eq.
3.57 will be much higher:
φ
r
w
µ
t
1
10r
w
2
K
 >
10 10
2 –
⋅
0 1 ,
 1 [sec] = =
t
2
10r
w
2
K
 >
10 10
4
⋅
0 1 ,
 10
6
[sec] 11 6 [days] , = = =
398 3: Solutions of the SinglePhase Equation of Filtration
Example 3.2:
The task is to determine the pressure at the well radius and at r = 100 [m] after 30 days
of production. The reservoir data are to be taken from Example 3.1.
Solution:
Since:
Time is large enough to apply Eq. 3.61. The bottom hole flowing pressure is calculated as
follows:
µqB
4πhk
 
10
3 – 10 –
86400

4 π 5 0 02 , 10
12 –
⋅ ⋅ ⋅ ⋅
 0 92 , 10
5
[Pa] ⋅ – = =
k
µcφ

0 02 10
12 –
⋅ ,
10
3 –
10
9 –
0 2 , ⋅ ⋅
 0 1 [m
2
sec
1 –
] , = = =
p
wf
p
i
µqB
4πhk
 + 0 80907 , ln
Kt
r
w
2
  +
( J
' J
 `
10 = 10
6
0 92 10
5
0 80907 ln
0 1 30 86400 ⋅ ⋅ ,
0 1 ,
2
  + ,
( J
 `
⋅ , – ⋅ 8 35 [MPa] , =
=
p
r 100 = ( )
p
i
µqB
4πhk
 + 0 80907 , ln
Kt
r
w
2
 +
( J
' J
 `
10 = 10
6
0 92 10
5
0 80907 ln
0 1 30 86400 ⋅ ⋅ ,
100
2
  + ,
( J
 `
⋅ , – ⋅ 9 36 [MPa] , =
=
3: Solutions of the SinglePhase Equation of Filtration 399
3.2.4 The Spatial Distribution of Flow
We assume a cylinder with radius r, centered by the well inside an infinite, homogenous
and isotropic layer. The quantity of fluid flowing through the cylindrical surface is a
function of time and may be calculated from Darcy’s law:
(3.66)
Eq. 3.40, Eq. 3.38 and Eq. 3.51 yield:
(3.67)
respectively from Eq. 3.67 and Eq. 3.38
(3.68)
Substituting Eq. 3.68 into Eq. 3.66 gives:
(3.69)
In Figure 3.7 the relation q(r,t)/q is illustrated as a function of Kt/r
2
.
Figure 3.7: The flow rate in function of the dimensionless variable Kt/r
2
(after Chaumet)
q r t , ( )
2πrhk
Bµ

r ∂
∂p
– =
r ∂
∂p 2r
Kt

z d
dp 2r
Kt
A
1
e
z
4
 –
z
 = =
r ∂
∂p 2r
Kt

µqB
4πhk

( J
 `
Kt
r
2
e
r
2
4Kt
 –
µqB
2πrhk
e
r
2
4Kt
 –
= =
q r t , ( ) qe
r
2
4Kt
 –
– =
3100 3: Solutions of the SinglePhase Equation of Filtration
3.3 Dimensionless Variables
Simplifications of the calculation may be achieved by introducing the so called
dimensionless variables.
radius:
time:
pressure:
Thus
(3.70)
(3.71)
(3.72)
Substitution of the variables given in Eq. 3.70  Eq. 3.72 into Eq. 3.30 and Eq. 3.31 Eq.
3.33 leads to the following formulas:
(3.73)
(3.74)
(3.75)
(3.76)
A similar transformation of the variables in Eq. 3.30 can be done and as r
w
, c, p
i
, K are
constants and therefore can be cancelled. Thus Eq. 3.73  Eq. 3.76 become:
r r
D
→
r
r
w
 =
t t
D
→
Kt
r
w
2

kt
φµcr
w
2
 = =
p P
D
→ p p
i
– ( )
2πhk
qBµ
 =
r r
D
r
w
=
t t
D
r
w
2
K
 =
p P
D
qbµ
2πhk

( J
 `
= p
i
P
D
C p
i
+ = +
r
D
r
w
( )
2
2
∂
∂
P
D
C p
i
+ [ ]
1
r
D
r
w

r
D
r
w
( ) ∂
∂
P
D
C p
i
+ [ ] +
1
K

t
D
r
w
2
K

( J
' J
 `
∂
∂
P
D
C p
i
+ [ ] =
P
D
C p
i
+ p
i
= r
D
r
w
r
w
>
t
D
r
w
2
K
 0 =
2πr
w
hk
Bµ

r
D
r
w
( ) ∂
∂
P
D
C p
i
+ [ ] q – =
P
D
C p
i
+ p
i
= r
D
r
w
∞ =
t
D
r
w
2
K
 0 ≥
3: Solutions of the SinglePhase Equation of Filtration 3101
(3.77)
(3.78)
(3.79)
(3.80)
Solutions of this boundary value problem are obtained by using Eq. 3.56. Rearranging of
this equation leads to:
(3.81)
Substituting the dimensionless variables into Eq. 3.81 yields:
(3.82)
Thus P
D
may be regarded as a function of :
(3.83)
Figure 3.8 shows the function P
D
according to Eq. 3.82 with .
The logaritmic approximation (for t
D
/r
D
2
>10) of Eq. 3.82 becomes:
(3.84)
At the wellbore , so Eq. 3.84 becomes more simple:
(3.85)
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
P
D
0 = r
D
1 > t
D
0 =
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
1 – = t
D
0 >
P
D
0 = r
D
∞ = t
D
0 >
2πhk
Bqµ
 p
i
p – ( )
1
2
Ei
r
2
r
w
2
⁄
4Kt r
w
2
⁄
 –
( J
' J
' J
 `
=
P
D
1
2
Ei
r
D
2
4t
D
 –
( J
' J
 `
– =
t
D
r
D
2
⁄
P
D
t
D
r
D
2
⁄ ( )
1
2
Ei
1
4
 
1
t
D
r
D
2
⁄ ( )
 –
( J
' J
 `
– =
r
e
r
w
⁄ ∞ =
P
D
1
2
 0
·
80907 t
D
/r
D
2
( ) ln + [ ] =
r
D
1 =
P
Dw
1
2
 0 80907 ln t
D
+ , [ ] =
3102 3: Solutions of the SinglePhase Equation of Filtration
Figure 3.8: Solution for the infinitive and finite radial filtration problem with closed boundary and
constant pressure drop (after Van Everdingen and Hurst)
3: Solutions of the SinglePhase Equation of Filtration 3103
Example 3.3:
Convert the following variables to dimensionless variables using the data in Example 3.1.
r = 100[m], t = 30[days] p = 9.63[MPa]
Solution:
Example 3.4:
Calculate the pressure at a distance of 100 [m] from the well after 5 days of production.
Use the data given in Example 3.1  Example 3.3.
Solution:
At this time the approximations used in Eq. 3.61 and Eq. 3.84 are not applicable, so one
must use Eq. 3.83 or the graphical illustration of the Ei function in Figure 3.8.
The first task is to calculate the dimensionless variables:
From the ( ) curve in Figure 3.8, P
D
=1.15.
r
D
r
r
w
 1000 = =
t
D
Kt
r
w
2

0 1 30 86400 ⋅ ⋅ ,
0 1
2
,
 2 59 10
7
⋅ , = = =
D
p p
i
– ( )
2πhk
qµ
 9 63 10
6
10 10
6
⋅ – ⋅ , ( )
2 π 5 0 02 10
12 –
⋅ , ⋅ ⋅ ⋅
10
3 – 10 –
86400
 

37 10
6
⋅ , 0 54286 10
5 –
⋅ , ⋅ 2 01 , –
=
=
= =
t
D
r
D
2
⁄ ( )
Kt
r
w
2

( J
' J
 `
r
2
r
w
2

( J
' J
 `
Kt
r
2

0 1 5 86400 ⋅ ⋅ ,
100
2
  4 32 , = = = =
K 0 1 [m
2
sec
1 –
] , =
Bqµ
2πhk
 1 84 10
5
[Pa] ⋅ , – =
r
e
r
w
⁄ ∞ =
3104 3: Solutions of the SinglePhase Equation of Filtration
Since:
From Eq. 3.72 follows:
(3.86)
P
D
1
2
Ei
r
D
2
4t
D
 –
( J
' J
 `
1 15 , = – =
p p
i
p
D
qBµ
2πhk
 + 10 10
6
1 84 , 10
5
1 15 , ⋅ ⋅ – ⋅ 9 79 [MPa] , = = =
3: Solutions of the SinglePhase Equation of Filtration 3105
3.4 The Infinite Radial System with Constant
Pressure at the Interior Boundary
Sometimes the well is produced under a constant bottom hole flowing pressure (p
wf
),
rather than under a constant production rate, for example to control water conning
problems.
The function p(r,t) is no longer of interest for us in this case. Because the bottom hole
pressure p
wf
and the pressure in infinity are both given and thus pressure at r will range
between these two values. The main interest though appears to lie in the determination of
the flow rate q or the cumulative production Q at the well radius r
w
. These two variables
are both regarded as functions of time.
Eq. 3.30 and the conditions in Eq. 3.31 and Eq. 3.33 are still valid. The boundary
condition in Eq. 3.32 though is substituted by:
(3.87)
The flow rate is calculated by Darcylow (see Eq. 3.32):
(3.88)
and the cumulative outflow by integrating Eq. 3.88:
. (3.89)
Again dimensionless variables are applied:
(3.90)
(3.91)
(3.92)
r
D
and t
D
are similar to those used in Eq. 3.70 and Eq. 3.71, only P
D
is different. Also note
that Eq. 3.56 is not applicable in this case since the production q is not constant.
p p
wf
= r r
w
= t 0 >
q t ( )
2πr
w
hk
Bµ
 – =
r ∂
∂p
( J
 `
r r
w
=
Q t ( ) q t ( ) t
2πr
w
hk
Bµ

r ∂
∂p
( J
 `
r r
w
=
t d
0
t
∫
– = d
0
t
∫
=
r
D
r
r
w
 =
t
D
Kt
r
w
2

k
µφcr
w
2
t = =
D
p
i
p –
p
i
p
wf
–

p
i
p –
∆p
wf
 = =
3106 3: Solutions of the SinglePhase Equation of Filtration
Eq. 3.90  Eq. 3.92 are then substituted into Eq. 3.30, Eq. 3.31 and Eq. 3.33. The B.C. in
Eq. 3.87 can be converted to dimensionless B.C. using Eq. 3.92. The following equations
are obtained:
(3.93)
(3.94)
(3.95)
(3.96)
The above boundary value problem (B.V.P.) described by Eq. 3.93  Eq. 3.96 is similar to
the B.V.P. described by Eq. 3.77  Eq. 3.80, except that the B.C. in Eq. 3.79 is replaced by
the B.C. in Eq. 3.95. Substituting the dimensionless variables in Eq. 3.89, leads to:
(3.97)
Let:
(3.98)
Figure 3.9 to Figure 3.12 show the graphical solution of Eq. 3.98.
Since Q
D
(t
D
) can be obtained from Table 3.3 at the end of Chapter 3 or from Figure 3.9 
Figure 3.12 (for ) as a function of t
D
then:
(3.99)
Where
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
P
D
0 = r
D
1 > t
D
0 =
P
D
1 = r
D
1 = t
D
0 >
P
D
0 = r
D
∞ = t
D
0 >
Q t ( )
2πr
w
hk
Bµ
 –
( J
 `
r
D
r
w
( ) ∂
∂
p
i
∆p
wf
P
D
– [ ]
( J
 `
r
D
1 =
Bµφcr
w
2
k
 t
D
2πhφr
w
2
c∆p
wf
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
t
D
d
0
t
D
∫
= d
0
t
D
∫
=
Q
D
t
D
( )
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
t
D
d
0
t
D
∫
=
r
e
r
w
⁄ ∞ =
t ( ) 2πhφr
w
2
c∆p
wf
Q
D
t
D
( ) =
∆p
wf
p
i
p
wf
– =
3: Solutions of the SinglePhase Equation of Filtration 3107
Example 3.5:
Consider the well data from Example 3.1. The bottom hole flowing pressure is fixed at:
Calculate the cumulative production after 100 days.
Solution:
The dimensionless time results in:
From the Van Everdinger and Hurst solution (Table 3.3).
The cumulative oil production after 100 days will be:
Figure 3.9: Solution for the infinitive and finite radial filtration problem with closed boundary and
constant bottom hole pressure (after Silder)
p
wf
8 [MPa] =
t
D
Kt
r
w
2

0 1 ,
0 1
2
,
100 86400 ⋅ 8 64 10
7
⋅ , = = =
Q
D
t
D
( ) 8 895 10
6
⋅ , =
2πhφr
w
2
c∆p
wf
Q
D
t
D
( )
π 5 0 2 0 1
2
10
9 –
2 10
6
8 895 10
6
1117 2 [m
3
] , = ⋅ , ⋅ ⋅ ⋅ ⋅ , ⋅ , ⋅ ⋅ ⋅
= =
3108 3: Solutions of the SinglePhase Equation of Filtration
Figure 3.10: Solution for the infinitive and finite radial filtration problem with closed boundary and
constant bottom hole pressure (after Silder)
Figure 3.11: Solution for the infinitive and finite radial filtration problem with closed boundary and
constant bottom hole pressure (after Silder).
3: Solutions of the SinglePhase Equation of Filtration 3109
Figure 3.12: Solution for the infinitive and finite radial filtration problem with closed boundary and
constant bottom hole pressure (after Silder).
3110 3: Solutions of the SinglePhase Equation of Filtration
3.5 NonSteady State Filtration in a Finite System
3.5.1 Constant Production Rate
3.5.1.1 Closed Exterior Boundary
The dimensionless variables of Eq. 3.70  Eq. 3.72 are again applied. The differential
equation given by Eq. 3.77, the initial condition in Eq. 3.78 and the boundary condition
in Eq. 3.79 are further valid.
The only difference can be found at the exterior boundary. In this case the boundary is not
considered to be located in infinity but at a finite distance with radius r
eD
:
(3.100)
(3.101)
(3.102)
(3.103)
We now seek a solution for the boundary value problem described by Eq. 3.100  Eq.
3.103 in the form of:
(3.104)
This solution can be achieved by transforming Eq. 3.100  Eq. 3.103 into the complex
plane, followed by integration and retransformation.
This operation is called LAPLACE transformation and is a little bit complicated. Only the
results are given here:
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
P
D
0 = r
D
1 > t
D
0 =
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
1 – = t
D
0 >
r
D
∂
∂P
D
0 = r
D
r
eD
= t
D
0 >
D
P
D
r
D
t
D
, ( ) =
3: Solutions of the SinglePhase Equation of Filtration 3111
(3.105)
are the roots of the equation:
(3.106)
where J
i
and Y
i
are the Bessel functions.
Eq. 3.105 is expressed graphically in Figure 3.8 for different values of r
eD
. If it is
possible to write the well flowing pressure formula in a more simple form:
(3.107)
The roots increase monotonous, if n increases. This means:
This is the cause for a monotonous decrease of the exponential factor. From this one may
conclude that if t
D
is large enough this sum will become very small and thus can be
neglected.
The approximation for large t
D
is:
(3.108)
If the pressure disturbance has not yet reached the exterior boundary, Eq. 3.77  Eq. 3.80
and Eq. 3.100  Eq. 3.103 will have an identical solution which is:
(3.109)
P
D
2
r
eD
2
1 –

r
D
2
4
 t
D
+
( J
' J
 `
r
eD
2
ln r
D
r
eD
2
1 –

3r
eD
2
4r
eD
2
ln r
eD
2r
eD
2
1 – – –
4 r
eD
2
1 – ( )
2

π
e
α
n
2
t
D
–
J
1
2
α
n
r
eD
( ) J
1
α
n
( )Y
0
α
n
r
D
( ) Y
1
α
n
( )J
0
α
n
r
D
( ) – [ ]
α
n
J
1
2
α
n
r
eD
( ) J
1
2
α
n
( ) – [ ]

n 1 =
∞
∑
+
– –
=
α
n
J
1
α
n
r
eD
( )Y
1
α
n
( ) J
1
α
n
( )Y
1
α
n
r
eD
( ) – 0 =
r
w
r
e
«
P
Dw
2t
D
r
eD
2
 ln r
eD
3
4
 – 2
e
α
n
2
t
D
–
J
1
2
α
n
r
eD
( )
α
n
2
J
1
2
α
n
r
eD
( ) J
1
2
α
n
( ) – [ ]

n 1 =
∞
∑
+ + =
e
α
1
2
t
D
–
e
α
2
2
t
D
–
e
α
3
2
t
D
–
> >
P
Dw
2t
D
r
eD
2
 ln r
eD
3
4
 – + =
P
Dw
1
2
 ln t
D
0 80907 Y
D
t
D
( ) + , + [ ] =
3112 3: Solutions of the SinglePhase Equation of Filtration
where:
(3.110)
Eq. 3.110 is the combination of the equations Eq. 3.85 and Eq. 3.107. It is evident that the
function Y
D
(t
D
) will become zero, if the exterior boundary is not yet reached. The
solutions of the finite and the infinite systems are therefore identical for a given time t .
With dimensioned variables Eq. 3.109 becomes:
(3.111)
3.5.1.2 Boundary with Constant Pressure
If the system has an open boundary, but a constant pressure at the exterior boundary.
Eq. 3.100  Eq. 3.102 remain the same, but the boundary condition in Eq. 3.103 is
substituted by:
(3.112)
The solution of the boundary value problem is described by Eq. 3.100  Eq. 3.102 and Eq.
3.112 considering again that is given by:
(3.113)
are defined as the roots of this equation:
It t
D
is large enough it is possible to neglect the summation term in Eq. 3.113, then Eq.
3.113 becomes:
Y
D
t
D
( ) ln t
D
0 80907 , + ( ) –
4t
D
r
eD
2
 2 ln r
eD
3
4
 –
( J
 `
4
e
α
n
2
t
D
–
J
1
2
α
n
r
eD
( )
α
n
2
J
1
2
α
n
r
eD
( ) J
1
2
α
n
( ) – [ ]

n 1 =
∞
∑
+ + + =
p
wf
p
i
µqB
4πhk
 + 0 80907 , ln
Kt
r
w
2
  Y t ( ) + +
( J
' J
 `
=
P
D
0 = t
D
0 > r
D
r
eD
=
r
w
r
e
«
P
D
ln r
eD
2
e
β
n
2
t
D
–
J
0
2
β
n
r
eD
( )
β
n
2
J
1
2
β
n
( ) J
0
2
β
n
r
eD
( ) – [ ]

n 1 =
∞
∑
– =
β
n
J
1
β
n
( )Y
0
β
n
r
eD
( ) Y
1
β
n
( )J
1
β
n
r
eD
( ) – 0 =
3: Solutions of the SinglePhase Equation of Filtration 3113
(3.114)
which is identical with Eq. 3.20.
3.5.2 Constant Pressure at the Interior Boundary and
Closed Exterior Boundary
Just as in Section 3.4: The rate at the interior boundary (e.g. at the well radius) is not fixed,
but the pressure p
wf
is set at a certain value. Again the dimensionless pressure form is
used. The differential equation and the boundary conditions may then be written as
follows:
(3.115)
(3.116)
(3.117)
(3.118)
The solution of the above B.V.P. is given by Eq. 3.99, but the dimensionless cumulative
inflow Q
D
(t
D
). is determined from Figure 3.9  Figure 3.12 for different ratios of
.
Example 3.6:
Consider the data given in example 3.5 and the exterior radius (r
e
) of the reservoir was
estimated as 320 [m]. Calculate the cumulative production after 100 days.
Solution:
From the solution of example 3.5, t
D
is 8.64 10
7
From Figure 3.9a: Q
D
= 3.5 10
6
The cumulative production will be:
P
D
ln
r
e
r
w
 =
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
P
D
0 = 1 r
D
r
eD
≤ ≤ t
D
0 =
P
D
1 = r
D
1 = t
D
0 >
r
D
∂
∂P
D
0 = r
D
r
eD
= t
D
0 >
r
eD
r
e
r
w
⁄ =
Q 2πhr
w
2
p ∆
wf
Q
D
t
D
( ) 440 [m
3
] = =
3114 3: Solutions of the SinglePhase Equation of Filtration
3.6 NonSteady State Filtration in Linear System
Linear flow systems are applicable in the case of a massive fracture or a horizontal well,
where the fluids flow linearly to the fracture or the wellbore of the horizontal well.
3.6.1 Linear Flow with Constant Production Rate
For a vertically fractured wells, the fracture intersects the wellbore perpendicularly, or
horizontal wells in an infinite acting reservoir, the fluids will flow linearly to the vertical
fracture or the horizontal well especially at the beginning of the flow and after a relatively
short time period.
Consider a one dimensional flow problem described by:
(3.119)
(3.120)
(3.121)
(3.122)
Gringarten, Ramey and Raghavan used a flow model "Uniform Flux Fracture" which is
the first approximation to the behavior of a vertically fractured well. Fluid enter the
fracture at a uniform rate per unit surface area of the fracture, so that there is a pressure
drop created by the fracture. This pressure difference can be calculated from the following
equation:
, (3.123)
where:
x
f
is half length of the vertical fracture and
A is the surface area of the flow
x
2
2
∂
∂ p 1
K

t ∂
∂p
=
x ∂
∂p
q –
Bµ
Ak
 = x 0 = t 0 >
p ∞ t , ( ) p
i
= x ∞ = t 0 >
p x 0 , ( ) p
i
= x 0 > t 0 =
p p
i
–
qBµ
Ak

2πKt
x
f
erf
x
f
2 Kt

( J
 `
Ei
x
f
2
–
4Kt

( J
' J
 `
–
 
¦ 
 
=
3: Solutions of the SinglePhase Equation of Filtration 3115
In order to convert Eq. 3.123 to a dimensionless equation the following dimensionless
variable is introduced:
(3.124)
using the definition of p
D
and t
D
the dimensionless pressure is obtained by:
(3.125)
For Eq. 3.125 can be approximated by
(3.126)
For Eq. 3.125 becomes:
(3.127)
which indicates that during a short time the flow into the fracture is linear.
Figure 3.13 shows the relation between p
D
and t
Dxf
. For a half slope straight
line is obtained which indicates linear flow behaviour.
Example 3.7:
A) Calculate the pressure drop created by a vertical fracture having a length of 5 [m] after
5.2 and 15.6 min. Use the data in example 3.1.
B) Compare the pressure drop in part A with the pressure drop, created by the unfractured
well in example 3.1 after 15.6 min.
Solution:
A) After 5.2 min.
t
Dx
f
t
D
r
w
x
f

( J
' J
 `
2
=
p
D
πt
Dx
f
erf
1
2 t
Dx
f

( J
' J
 `
1
2
 Ei
1 –
4t
Dx
f

( J
' J
 `
– =
t
Dxf
10 >
p
D
1
2
 ln t
Dx
f
2.80907 + [ ] =
t
Dxf
0.1 >
p
D
πt
Dx
f
=
t
Dxf
0.1 >
k
φµc

0 02 10
12 –
⋅ ,
0 2 10
3 –
10
9 –
⋅ ⋅ ,
 0 1 [m
2
sec
1 –
] , = = =
t
D1
K
t
1
r
w
2
 0 1
5 2 60 ⋅ ,
0 1
2
,
 ⋅ , 3 12 10
3
⋅ , = = =
3116 3: Solutions of the SinglePhase Equation of Filtration
From Figure 3.13, P
D1
= 2.25
The pressure drop after 5.2 min. is:
After 15.6 min.:
Since t
Dxf
> 10 , then p
D2
can be calculated from:
Then the pressure drop after 15.6 min. is:
B) For the unfractured well in example 3.1 and after 15.6 min.:
, then p
D
can be calculated from:
t
Dx
f 1
t
D1
r
w
x
f

( J
' J
 `
2
3 12 10
3 0 1 ,
2 5 ,

( J
 `
2
⋅ ⋅ , 5 0 , = = =
∆p
1
µBq
2πhk
p
D1
10 184300 ⁄ – ( )10
3 –
2 3 14 5 0 02 10
12 –
⋅ , ⋅ ⋅ , ⋅
 2 25 , ( ) 415 [kPa] = = =
t
D2
K
t
2
r
w
2
 0 1
15 6 60 ⋅ ,
0 1
2
,
 ⋅ , 9 36 10
3
⋅ , = = =
t
Dx
f 2
t
D2
r
w
x
f

( J
' J
 `
2
15 = =
p
D2
1
2
 ln t
Dx
f
0 80907 , + [ ] 2 76 , = =
∆p
µBq
2πhk
p
D2
509 [kPa] = =
t
D
r
w
2
 10 >
p
Dw
1
2
 ln t
D
0 80907 , + [ ] 4 43 , = =
3: Solutions of the SinglePhase Equation of Filtration 3117
The pressure drop created by the unfractured well, after 15.6 min., is 1.6 times higher than
the one created by the same well with a single vertical fracture.
Figure 3.13: Dimensionless pressure for single fractured well in an infinite acting system (after
Gringarten, Ramey, and Ragavan)
∆p
unf
∆p
f

p
unf
p
Df

4 43 ,
2 76 ,
 1 61 , = = =
3118 3: Solutions of the SinglePhase Equation of Filtration
Table 3.1: Solution of Transient Filtration in the Case of Infinite Radial System
Mathematical
Formulation
Dimensioned Dimensionless
Distance
Time
Pressure
r
t
p
Differential Equation
Initial Condition
; ; ; ;
Boundary Conditions
1.) Well Radius
2.) Outer Boundary
;
;
Solution
Solution in Case of
r
D
r
r
w
 =
t
D
Kt
r
w
2
  =
P
D
p p
i
– ( )
2πhk
qBµ
 =
r
2
2
∂
∂ p 1
r

r ∂
∂p
+
1
K
 
t ∂
∂p
=
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
p p
i
= r r
w
≥ t 0 = P
D
0 = r
D
1 ≥ t
D
0 =
2πr
w
hk
µ

r ∂
∂p
( J
 `
r r
w
=
qB – =
p p
i
= r ∞ =
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
1 – =
P
D
0 = r
D
∞ =
Kt
r
2
 10 ≥
p
i
p r t , ( ) –
µqB
4πhk
 Ei
r
2
4Kt
 –
( J
 `
=
p
i
p r t , ( ) –
µqB
4πhk
 – 0 80907 , ln
Kt
r
2
  +
( J
 `
=
P
D
t
D
r
D
2

( J
' J
 `
1
2
Ei
r
D
2
4t
D
 –
( J
' J
 `
– =
P
D
1
2
 0 80907 ln
t
D
r
D
2
 + , =
3: Solutions of the SinglePhase Equation of Filtration 3119
Table 3.2: Solution of Transient Filtration in the Case of Infinite Radial System
Conditions Constant Production Constant Borehole Pressure
Dimensionless Variables:
Distance
Time
Pressure
Differential Equation
Initial Condition
; ; ; ;
Boundary Conditions
1.) Well Radius
2.) Outer Boundary
;
; ;
; ; ;
; ; ;
Borehole Pressure
Cumulative Influx
;
Trivial [ ]
Trivial [ ; ]
r
D
r
r
w
 =
t
D
Kt
r
w
2
  =
P
D
p p
i
– ( )
2πhk
qBµ
 =
r
D
r
r
w
 =
t
D
Kt
r
w
2
  =
P
D
p
i
p –
p
i
p
wf
–
 =
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
r
D
2
2
∂
∂ P
D 1
r
D

r
D
∂
∂P
D
+
t
D
∂
∂P
D
=
P
D
0 = r
D
1 ≥ t
D
0 = P
D
0 = r
D
1 ≥ t
D
0 =
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
1 – = t
D
0 >
P
D
0 = r
D
∞ = t
D
0 >
P
D
1 = r
D
1 = t
D
0 >
P
D
0 = r
D
∞ = t
D
0 >
r
D
1 = t
D
0 >
P
D
1
2
 0 80907 ln t
D
+ , [ ] =
Q t ( ) qt =
P
D
1 = r
D
1 =
Q
D
t
D
( )
r
D
∂
∂P
D
( J
' J
 `
r
D
1 =
t
D
d
0
t
D
∫
=
Q t ( ) 2πhφr
w
2
c∆p
wf
Q
D
t
D
( ) =
3120 3: Solutions of the SinglePhase Equation of Filtration
3.7 The Principle of Superposition
In mathematics the superposition theorem states that any sum of individual solutions of a
linear differential equation is also a solution of this differential equation, but for different
boundary conditions. In practice, the superposition theorem is considered to be one of the
most powerful tools to get the solutions of complex flow problems without solving the
differential equation for different boundary conditions over and over again.
3.7.1 The First Law of Superposition
Consider a well producing with a continuously changing bottom hole pressure as it is
shown in Figure 3.14. The pressure can then be approximated by a step function. For all
single pressure changes the cummulative production (influx) can be calculated by Eq.
3.99. According to the theorem of superposition the overall influx caused by the n
subsequent finite pressure drops is the sum of the elementary solutions:
(3.128)
where C is a constant. Eq. 3.128 is the original form of the van EverdingenHurst solution.
Vogt and Wang improved this model assuming piecewise linear pressure change instead
of a stepping one.
Eq. 3.128 can be written in the following form too:
(3.129)
or after replacing the summation by integral:
(3.130)
is the integration variable. The pressure derivative can be approximated piecewise by
finite differences:
(3.131)
W
e
t ( ) C ∆p
wf
Q
D
t
D
t
Dj
– ( )
j 0 =
n
∑
=
W
e
t ( ) C
∆p
wf
∆t
D
Q
D
t
D
t
Dj
– ( )∆t
D
j 0 =
n
∑
=
W
e
t ( ) C
τ d
dp
wf
Q
D
t
D
τ – ( ) τ d
0
t
D
∫
=
τ
τ d
dp
wf
p
j
p
j 1 –
–
t
Dj
t
Dj 1 –
–
 = t
Dj 1 –
τ t
Dj
≤ ≤
3: Solutions of the SinglePhase Equation of Filtration 3121
and then the integral in Eq. 3.131 can be splitted into n terms:
(3.132)
Let , then , and the integrals in Eq. 3.132 can be rewritten as follows:
(3.133)
Introducing the integral of the Q
D
(t
D
) function
(3.134)
Eq. 3.132 yields:
(3.135)
W
e
t
n
( ) C
p
0
p
1
–
t
D1
 Q
D
t
Dn
τ – ( ) τ
p
1
p
2
–
t
D2
t
D1
–
 Q
D
t
Dn
τ – ( ) τ
......+
p
n 1 –
p
n
–
t
Dn
t
Dn 1 –
–
 Q
D
t
Dn
τ – ( ) τ d
t
Dn 1 –
t
Dn
∫
+ d
t
D1
t
D2
∫
+ d
0
t
D1
∫




¦





– =
u t
Dn
τ – = du dτ – =
Q
D
t
Dn
τ – ( ) τ d
t
Dj 1 –
t
Dj
∫
Q
D
u ( ) u d
t
Dn
t –
Dj 1 –
t
Dn
t –
Dj
∫
Q
D
u ( ) u d
0
t
Dn
t –
Dj
∫
– Q
D
u ( ) u d
0
t
Dn
t –
Dj 1 –
∫
+
= =
Q
D
∗
t
D
( ) Q
D
u ( ) u d
0
t
D
∫
=
W
e
t
n
( ) C
p
0
p
1
–
t
D1
Q
D
∗
t
Dn
( )
p
1
p
2
–
t
D2
t –
D1

p
0
p
1
–
t
D1
 –
( J
' J
 `
Q
D
∗
t
Dn
t
D1
– ( )
.....
p
n 1 –
p
n
–
t
Dn 1 –
t –
Dn

p
n 2 –
p
n 1 –
–
t
Dn 1 –
t
Dn 2 –
–
 –
( J
' J
 `
Q
D
∗
t
Dn
t
Dn 1 –
– ( )
+ +
+


¦



– =
3122 3: Solutions of the SinglePhase Equation of Filtration
Let be and then Eq. 3.135 becomes:
(3.136)
where
(3.137)
Comparing Eq. 3.136 with Eq. 3.128 it is evident that:
if (3.138)
Figure 3.14: Variable production rate in case of a ideal reservoir (after Hurst)
p
j 1 +
∆ p
j
p
j 1 +
– = t
j 1 +
∆ t
j
t
j 1 +
– =
W
e
t
n
( ) C
p
1
∆
t
D1
Q
D
∗
t
Dn
( )
p
2
∆
t
D2
∆

p
1
∆
t
D1
 –
( J
' J
 `
Q
D
∗
t
Dn
t
D1
– ( )
.....
p
n
∆
t
Dn
∆

p
n 1 –
∆
t
Dn 1 –
∆
 –
( J
' J
 `
Q
D
∗
t
Dn
t
Dn 1 –
– ( )
+ +
+


¦



–
C
p
j
∆
t
Dj
∆

( J
' J
 `
Q
D
∗
t
Dn
t
Dj
– ( ) ∆
j 0 =
n 1 –
∑
–
=
=
∆
∆p
j
∆t
Dj

( J
' J
 `
∆p
j
∆t
Dj

∆p
j 1 –
∆t
Dj 1 –
 – =
∆p
j
Q
D
t
Dn
t
Dj
– ( )
j 0 =
n
∑
∆
∆p
j
∆t
Dj

( J
' J
 `
Q
D
∗
t
Dn
t
Dj
– ( )
j 0 =
n 1 –
∑
= n ∞ →
3: Solutions of the SinglePhase Equation of Filtration 3123
3.7.2 The Second Law of Superposition
Figure 3.15 illustrates the pressure change of two wells inside an infinite reservoir. At first
well 1 is put on production, with rate q
1
, at a time t
1
. The pressure change at any time
at point R can be calculated using Eq. 3.56:
(3.139)
If the rate of well 1 is set to then the pressure drop is defined as . Eq. 3.56
yields again:
(3.140)
or dimensionless:
(3.141)
Using Eq. 3.139 and Eq. 3.140 it is trivial to state that if well 1 producec at a constant rate
the pressue change at point R will be proportinal to :
(3.142)
and
(3.143)
where is equal to q
1
but it is dimensionless. Analogous the formulas for well 2
producing with a rate and starting at a time
may be set up:
(3.144)
(3.145)
(3.146)
t t
1
>
∆p
1
q
1
µB
4πhk
Ei
r
1
2
4K t t
1
– ( )
 –
( J
' J
 `
=
q
1
1 = p
1
∗
∆
∆p
1
∗
µB
4πhk
Ei
r
1
2
4K t t
1
– ( )
 –
( J
' J
 `
=
P
D1
∗
2πhk
µB
∆p
1
∗
1
2
Ei
r
D1
2
4 t
D
t
D1
– ( )
 –
( J
' J
 `
= =
q
1
1 ≠ q
1
∆p
1
q
1
∆p
1
∗
=
P
D1
q
ˆ
1
P
D1
∗
=
q
ˆ
1
q
2
1 = t
2
∆p
2
∗
µB
4πhk
Ei
r
2
2
4K t t
2
– ( )
 –
( J
' J
 `
=
P
D2
∗
2πhk
µB
∆p
2
∗
1
2
Ei
r
D2
2
4 t
D
t
D2
– ( )
 –
( J
' J
 `
= =
∆p
2
q
2
∆p
2
∗
=
3124 3: Solutions of the SinglePhase Equation of Filtration
and
(3.147)
If both wells are under production the pressure changes at point R may be added:
(3.148)
(3.149)
Figure 3.15: Pressure change at point R in infinite reservoir, with two production wells
P
D2
q
ˆ
2
P
D2
∗
=
P
D
P
D1
P
D2
+ q
ˆ
1
P
D1
∗
q
ˆ
2
P
D2
∗
+ = =
∆p p
i
p
r
t ( ) – q
1
∆p
1
∗
q
2
∆p
2
∗
+
µB
4πhk
 q
1
Ei
r
1
2
4K t t
1
– ( )
 –
( J
' J
 `
q
2
Ei
r
2
2
4K t t
2
– ( )
 –
( J
' J
 `
+
 
¦ 
 
= = =
3: Solutions of the SinglePhase Equation of Filtration 3125
3.7.3 Calculation of MultiWell Problems
Fig. 3.16 shows various wells put on production with constant rates inside an infinite
acting reservoir. Wells started with production successively at the times t
1
,t
2
,t
3
.
The diagram includes the overall production rate. Bottom hole flowing pressure of well 1
is determined by the second law of superposition as follows:
(3.150)
or from Eq. 3.148  Eq. 3.149:
(3.151)
r
1
is the radius of well 1. Summation is only made for .
3.7.4 Single Well with Variable Production Rates
When applying the second law of superposition the distances of the wells from well
1 are not considered. As shown in Figure 3.17 the wells are projected imaginatively
into well 1. Then every r
j
is substituted by r
w
:
(3.152)
The value of the bottom hole pressure is the result of this equation where the q
j
represent
the rate changes at the times t
j
.
For the calculation of the flowing pressure it is permitted to use the logarithmic
approximation formula Eq. 3.61. Thus Eq. 3.152 becomes:
P
D1
q
ˆ
j
j 1 =
n
∑
1
P
Dj
∗
=
p
i
p
wf1
t ( ) –
µB
4πhk
 q
j
E
j 1 =
n
∑
i
r
j
2
4K t t
j
– ( )
 –
( J
' J
 `
=
t t
j
>
2 n →
2 n →
p
i
p
wf1
t ( ) –
µB
4πhk
 q
j
E
j 1 =
n
∑
i
r
w
2
4K t t
j
– ( )
 –
( J
' J
 `
=
3126 3: Solutions of the SinglePhase Equation of Filtration
(3.153)
Figure 3.16: Superposition of several wells in a infinite reservoir
Figure 3.17: Application of the second law of superposition on a well with a variable production
p
i
p
wf1
t ( ) –
µB
4πhk
 – q
j
0 80907 , ln
K t t
j
– ( )
r
w
2
 +
( J
' J
 `
j 1 =
n
∑
µB
4πhk
 – q
j
ln t t
j
– ( ) 0 80907 , ln
K
r
w
2
 +
( J
' J
 `
q
j
j 1 =
n
∑
+
j 1 =
n
∑
 
 
¦ 
 
 
= =
3: Solutions of the SinglePhase Equation of Filtration 3127
3.7.5 Pressure Buildup of ShutIn Well
At first the well is put on production during time t
1
with a constant rate q
1
. Then at the end
of t
1
the well is shut in ( ) for the time . The pressure drop after the time
may then be calculated by Eq. 3.153 as follows:
(3.154)
Eq. 3.154 considers the well is on production during the time , but the well was
shut in during the time . To make up for this discrepancy, we assume that the well is
producing with a rate (as injector) during the shutin time . Then:
So Eq. 3.154 is simplified to:
(3.155)
Instead of p
wf
(flowing pressure) p
ws
(shut in pressure) was used which states that the well
is shut in. Usage of the common logarithm changes Eq. 3.155 to:
(3.156)
This equation calculates the pressure buildup at the well bottom at time t
1
until the end
of the buildup test. This curve is of great importance since the determination of
permeability and static reservoir pressure is made possible. Figure 3.18 illustrates a
pressure buildup curve. When the shutin pressure (p
ws
) is plotted versus Horner time in
a Semilog plot, Eq. 3.156 should give a straight line portion and the slope of this portion
is m and according to Eq. 3.156:
(3.157)
q
2
0 = t ∆
t t
1
t ∆ + =
p
i
p
wf1
t
1
∆t + ( ) –
µB
4πhk
 –
q
1
ln t
1
∆t + ( ) q
2
ln ∆t ( ) 0 80907 , ln
K
r
w
2
 +
( J
' J
 `
q
1
q
2
– ( ) + +
 
¦ 
 
=
t t
1
t ∆ + =
t ∆
q
2
q
1
– = t ∆
q
1
q
2
0 = +
p
ws
t
1
∆t + ( ) p
i
Bq
1
µ
4πhk
 ln
t
1
∆t + ( )
∆t
 + =
p
ws
p
i
0 1832Bq
1
µ ,
hk

t
1
∆t + ( )
∆t
 log + =
hk
01832µq
1
B
m –
 =
3128 3: Solutions of the SinglePhase Equation of Filtration
3.7.6 Method of Image
The second law of superposition may also be useful in the case boundaries exist inside or
at the edges of the reservoir. This boundary may be a fault as well as a pinch out. Such a
case is illustrated in Figure 3.19.
If an image of well 1, mirrored to the other side of the fault, will be produced with the
same rate, the problem is reduced to a dualwell problem in an infinite reservoir as
discussed before.
It can easily be taken for granted that for symmetry reasons no flow will occur through
the axis of symmetry (fault). Therefore the potential distribution inside the real region will
be identical in both cases. In the cartesian coordinate system the distance between point
B(x
1
,x
2
) and the well (x
10
,0) is:
(3.158)
and from the image well (x
10
,0):
(3.159)
According to Eq. 3.149 the pressure drop is given by:
(3.160)
Figure 3.18: Pressure buildup analysis plot (after Horner)
r
1
2
x
1
x
10
– ( )
2
x
2
2
+ =
r
2
2
x
1
x
10
+ ( )
2
x
2
2
+ =
p x t , ( ) p
i
–
Bµq
4πhk
 Ei
r
1
2
4Kt
 –
( J
' J
 `
Ei
r
2
2
4Kt
 –
( J
' J
 `
+ =
3: Solutions of the SinglePhase Equation of Filtration 3129
Figure 3.19: Production from a well near impermeable boundary (after Bear)
By applying the approximation Eq. 3.61:
(3.161)
or with cartesian coordinates
(3.162)
The pressure drop is calculated by setting and . Eq. 3.160 yields:
(3.163)
p x t , ( ) p
i
–
µqB
2πhk
 0 80907 ln
Kt
r
1
2
 ln
r
1
r
2
 + + ,
 
¦ 
 
=
p x t , ( ) p
i
–
µqB
2πhk
 0 80907 ln
Kt
x
1
x
10
– ( )
2
x
2
2
+

1
2
ln
x
1
x
10
– ( )
2
x
2
2
+
x
1
x
10
+ ( )
2
x
2
2
+
 + + ,
 
 
¦ 
 
 
=
r
1
r
w
= r
2
2x
10
=
p x t , ( ) p
i
–
µqB
4πhk
 Ei
r
w
2
4Kt
 –
( J
' J
 `
Ei
x
10
2
Kt
 –
( J
' J
 `
+
 
¦ 
 
=
3130 3: Solutions of the SinglePhase Equation of Filtration
3.7.6.1 Pressure Buildup Test Near No Flow Boundary
The well is put on production with a constant rate q, until the time t
1
and then it is shut in.
The pressure buildup curve can be calculated in the same manner as demonstrated before
(by using the second law of superposition).
(3.164)
where the shutin time is .
For the first two Eifunctions it is possible to use the logarithmic approximation formula
without any restriction. Thus Eq. 3.164 according to Eq. 3.154 may be written as follows:
(3.165)
since:
,
If is small the last Eiterm becomes zero and the preceding one is practically constant,
then:
(3.166)
Then 3.165 becomes:
(3.167)
If is large it is possible to apply the logarithmic approximation to 3.165 for all terms
and thus:
, (3.168)
and
p
ws
p
i
–
µqB
4πhk
 Ei
r
w
2
4Kt
 –
( J
' J
 `
Ei
r
w
2
4K t t
1
– ( )
 –
( J
' J
 `
Ei
x
10
2
Kt
 –
( J
' J
 `
Ei
x
10
2
K t t
1
– ( )
 –
( J
' J
 `
–
+ –


¦



=
t ∆ t t
1
– =
p
ws
p
i
–
µqB
4πhk
 ln
t
1
∆t + ( )
∆t
 Ei
x
10
2
K t
1
∆t + ( )
 –
( J
' J
 `
Ei
x
10
2
K∆t
 –
( J
' J
 `
– +
 
¦ 
 
=
t ∆ t t
1
– =
t ∆
Ei
x
10
2
K t
1
∆t + ( )
 –
( J
' J
 `
Ei
x
10
2
Kt
1
 –
( J
' J
 `
≈ b =
p
ws
p
i
Bqµ
4πhk
 ln
t
1
∆t + ( )
∆t
 b –
 
¦ 
 
+ =
t ∆
p
ws
p
i
–
µqB
4πhk
 ln
t
1
∆t + ( )
∆t
 ln
K t
1
∆t + ( )
x
10
2
 ln –
K∆t
x
10
2
 +
 
¦ 
 
=
3: Solutions of the SinglePhase Equation of Filtration 3131
(3.169)
Figure 3.20 shows the pressure buildup curve. The first section of the curve is described
by Eq. 3.167 (after a short time no boundary effects) and the second by Eq. 3.169 (after
the pressure disturbance reached the boundary).
The slope of the second straight line (displayed in the semilogarithmic coordinate sheet)
is exactly the double of the slope of the first. (m
2
=2m
1
).
Figure 3.20: Pressure buildup curve near a discontinuity
p
ws
p
i
Bqµ
2πhk
 ln
t
1
∆t + ( )
∆t
 + =
3132 3: Solutions of the SinglePhase Equation of Filtration
3.7.6.2 Constant Pressure Boundary
Now let us consider the case of a constant potential at the outer boundary as shown in
Figure 3.21. The specific boundary condition given by Eq. 2.81 is used. If the real region
is again mirrored, the image well will now be regarded as an injection well, It is evident
that the distribution of the potential in the real region will again be the same for both cases.
We ignore the boundary, but we consider the image injection well. The corresponding
equation is given by Eq. 3.151. The rate for an injection well is :
(3.170)
Using the logarithmic approximation and cartesian coordinates yields:
(3.171)
Figure 3.21: Production in the vicinity of a boundary with a constant potential (after Bear)
The pressure change at well bottom can be calculated by a more simple equation. For this
case:
and
therefore:
(3.172)
q
2
q
1
– =
p
i
p x t , ( ) –
µqB
4πhk
 Ei
r
1
2
4Kt
 –
( J
' J
 `
E – i
r
2
2
4Kt
 –
( J
' J
 `
 
¦ 
 
=
p
i
p x t , ( ) –
µqB
4πhk
ln
r
2
2
r
1
2
 –
µqB
4πhk
 ln
x
1
x
10
+ ( )
2
x
2
2
+
x
1
x
10
– ( )
2
x
2
2
+
 – = =
x x
10
– r
w
= x x
10
+ 2x
10
≅
p
i
p
wf
–
µqB
2πhk
 – ln
2x
10
r
w
 =
3: Solutions of the SinglePhase Equation of Filtration 3133
3.8 References
3.1 Bear, J.: "Dynamics of fluids in Porous Media" Elsevier New York (1972)
3.2 Chaumet, P.: "Cours de production" Thome III Econlement monophasique de
Fluides dans les millieux poreux. I.F.P. Publications, Editions Technip, Paris (1964).
3.3 Van Everdingen, A.F. and Hurst, W.: "The application of the Laplace
transformations to flow problems in Reservoir" Trans. AIME (1949)
3.4 Gringarten, A.C., Ramey, H. Jr., and Raghavan, R. "Applied pressure analysis for
fractured wells" J. Pet. Tech. (July 1975) 887892; Trans. AIME 259.
3.5 Slider, H.C.: "Practical Pertroleum Reservoir Engineering Methods". Petroleum
Publishing Company Tulsa O.K., U.S.A.(1976)
3134 3: Solutions of the SinglePhase Equation of Filtration
Table 3.3: Hurst  Van Everdingen: Constant Pressure Q
tD
Functions for Infinite Acting
Radial Reservoirs
t
D
Q
tD
t
D
Q
tD
t
D
Q
tD
t
D
Q
tD
0.00 0.000 41 21.298 96 41.735 355 121.966
0.01 0.112 42 21.701 97 42.084 360 123.403
0.05 0.278 43 22.101 98 42.433 365 124.838
0.10 0.404 44 22.500 99 42.781 370 126.270
0.15 0.520 45 22.897 100 43.129 375 127.699
0.20 0.606 46 23.291 105 44.858 380 129.126
0.25 0.689 47 23.684 110 46.574 385 130.550
0.30 0.758 48 24.076 115 48.277 390 131.972
0.40 0.898 49 24.466 120 49.968 395 133.391
0.50 1.020 50 24.855 125 51.648 400 134.808
0.60 1.140 51 25.244 130 53.317 405 136.223
0.70 1.251 52 25.633 135 54.976 410 137.635
0.80 1.359 53 26.020 140 56.625 415 139.045
0.90 1.469 54 26.406 145 58.265 420 140.453
55 26.791 150 59.895 425 141.859
1.0 1.569 56 27.174 155 61.517 430 143.262
2.0 2.447 57 27.555 160 63.131 435 144.664
3.0 3.202 58 27.935 165 64.737 440 146.064
5.0 4.539 60 28.691 175 67.928 450 148.856
6.0 5.153 61 29.068 180 69.512 455 150.249
7.0 5.743 62 29.443 185 71.090 460 151.640
8.0 6.314 63 29.818 190 72.661 465 153.029
9.0 6.869 64 30.192 195 74.226 470 154.416
10 7.411 65 30.565 200 75.785 475 155.801
11 7.940 66 30.937 205 77.338 480 157.184
12 8.457 67 31.308 210 78.886 485 158.565
13 8.964 68 31.679 215 80.428 490 159.945
14 9.461 69 32.048 220 81.965 495 161.322
15 9.949 70 32.417 225 83.497 500 162.698
16 10.434 71 32.785 230 85.023 510 165.444
17 10.913 72 33.151 235 86.545 520 168.183
18 11.386 73 33.517 240 88.062 525 169.549
19 11.855 74 33.883 245 89.575 530 170.914
20 12.319 75 34.247 250 91.084 540 173.639
21 12.778 76 34.611 255 92.589 550 176.357
22 13.233 77 34.974 260 94.090 560 179.069
23 13.684 78 35.336 265 95.588 570 181.774
24 14.131 79 35.697 270 97.081 575 183.124
25 14.573 80 36.058 275 98.571 580 184.473
26 15.013 81 36.418 280 100.057 590 187.166
27 15.450 82 36.777 285 101.540 600 189.852
28 15.883 83 37.136 290 103.019 610 192.533
29 16.313 84 37.494 295 104.495 620 195.208
30 16.742 85 37.851 300 105.968 625 196.544
31 17.167 86 38.207 305 107.437 630 197.878
32 17.590 87 38.563 310 108.904 640 200.542
33 18.011 88 38.919 315 110.367 650 203.201
34 18.429 89 39.272 320 111.827 660 205.854
35 18.845 90 39.626 325 113.284 670 208.502
36 19.259 91 39.979 330 114.738 675 209.825
37 19.671 92 40.331 335 116.189 680 211.145
38 20.080 93 40.684 340 117.638 690 213.784
39 20.488 94 41.034 345 119.083 700 216.417
40 20.894 95 41.385 350 120.526 710 219.046
3: Solutions of the SinglePhase Equation of Filtration 3135
Table 3.4: Continuation
720 221.670 1,175 337.142 1,900 510.861 4,050 990.108
725 222.980 1,180 338.376 1,925 516.695 4,100 1,000.858
730 224.289 1,190 340.843 1,950 522.520 4,150 1,011.595
740 226.904 1,200 343.308 1,975 825.337 4,200 1,022.318
750 229.514 1,210 345.770 2,000 534.145 4,250 1,033.028
760 232.120 1,220 348.230 2,025 539.945 4,300 1,043.724
770 234.721 1,225 349.460 2,050 545.737 4,350 1,054.409
775 236.020 1,230 350.688 2,075 551.522 4,400 1,065.082
780 237.318 1,240 353.144 2,100 557.299 4,450 1,075.743
790 239.912 1,250 355.597 2,125 563.068 4,500 1,086.390
800 242.501 1,260 358.048 2,150 568.830 4,550 1,097.024
810 245.086 1,270 360.496 2,175 574.585 4,600 1,107.646
820 247.668 1,275 361.720 2,200 580.332 4,650 1,118.257
825 248.957 1,280 362.942 2,225 586.072 4,700 1,128.854
830 250.245 1,290 365.386 2,250 591.806 4,750 1,139.439
840 252.819 1,300 367.828 2,275 597.532 4,800 1,150.012
850 255.388 1,310 370.267 2,300 603.252 4,850 1,160.574
860 257.953 1,320 372.704 2,325 608.965 4,900 1,171.125
870 260.515 1,325 373.922 2,350 614.672 4,950 1,181.666
875 261.795 1,330 375.139 2,375 620.372 5,000 1,192.198
880 263.073 1,340 377.572 2,400 626.066 5,100 1,213.222
890 265.629 1,350 380.003 2,425 631.755 5,200 1,234.203
900 268.181 1,360 382.432 2,450 637.437 5,300 1,255.141
910 270.729 1,370 384.859 2,475 643.113 5,400 1,276.037
920 273.274 1,375 386.070 2,500 648.781 5,500 1,296.893
925 274.545 1,380 387.283 2,550 660.093 5,600 1,317.709
930 275.815 1,390 389.705 2,600 671.379 5,700 1,338.486
940 278.353 1,400 392.125 2,650 682.640 5,800 1,359.225
950 280.888 1,410 394.543 2,700 693.877 5,900 1,379.927
960 283.420 1,420 396.959 2,750 705.090 6,000 1,400.593
970 285.948 1,425 398.167 2,800 716.280 6,100 1,421.224
975 287.211 1,430 399.373 2,850 727.449 6,200 1,441.820
980 288.473 1,440 401.786 2,900 738.598 6,300 1,462.383
990 290.995 1,450 404.197 2,950 749.725 6,400 1,482.912
1,000 293.514 1,460 406.608 3,000 760.833 6,500 1,503.408
1,010 296.030 1,470 409.013 3,050 771.922 6,600 1,523.872
1,020 298.543 1,475 410.214 3,100 782.992 6,700 1,544.305
1,025 299.799 1,480 411.418 3,150 794.042 6,800 1,564.706
1,030 301.053 1,490 413.820 3,200 805.075 6,900 1,585.077
1,040 303.560 1,500 416.220 3,250 816.090 7,000 1,605.418
1,050 306.065 1,525 422.214 3,300 827.088 7,100 1,625.729
1,060 308.567 1,550 428.196 3,350 838.067 7,200 1,646.011
1,070 311.066 1,575 434.168 3,400 849.028 7,300 1,666.265
1,075 312.314 1,600 440.128 3,450 859.974 7,400 1,686.490
1,080 313.562 1,625 446.077 3,500 870.903 7,500 1,706.688
1,090 316.055 1,650 452.016 3,550 881.816 7,600 1,726.859
1,100 318.545 1,675 457.945 3,600 892.712 7,700 1,747.002
1,110 321.032 1,700 463.863 3,650 903.594 7,800 1,767.120
1,120 323.517 1,725 469.771 3,700 914.459 7,900 1,787.212
1,125 324.760 1,750 475.669 3,750 925.309 8,000 1,807.278
1,130 326.000 1,775 481.558 3,800 936.144 8,100 1,827.319
1,140 328.480 1,800 487.437 3,850 946.966 8,200 1,847.336
1,150 330.958 1,825 493.307 3,900 957.773 8,300 1,867.329
1,160 333.433 1,850 499.167 3,950 968.566 8,400 1,887.298
1,170 335.906 1,875 505.019 4,000 979.344 8,500 1,907.243
3136 3: Solutions of the SinglePhase Equation of Filtration
Table 3.5: Continuation
8,600 1,927.166 2.510
7
2.96110
6
8,700 1,947.065 3.010
7
3.51710
6
8,800 1,966.942 4.010
7
4.61010
6
8,900 1,986.796 5.010
7
5.68910
6
9,000 2,006.628 6.010
7
6.75810
6
9,100 2,026.438 7.010
7
7.81610
6
9,200 2,046.227 8.010
7
8.86610
6
9,300 2,065.996 9.010
7
9.91110
6
9,400 2,085.744 1.010
8
1.09510
7
9,500 2,105.473 1.510
8
1.60410
7
9,600 2,125.184 2.010
8
2.10810
7
9,700 2,144.878 2.510
8
2.60710
7
9,800 2,164.555 3.010
8
3.10010
7
9,900 2,184.216 4.010
8
4.07110
7
10,000 2,203.861 5.010
8
5.03210
7
12,500 2,688.967 6.010
8
5.98410
7
15,000 3,164.780 7.010
8
6.92810
7
17,500 3,633.368 8.010
8
7.86510
7
20,000 4,095.800 9.010
8
8.79710
7
25,000 5,005.726 1.010
9
9.72510
7
30,000 5,899.508 1.510
9
1.42910
8
35,000 6,780.247 2.010
9
1.88010
8
40,000 7,650.096 2.510
9
2.32810
8
50,000 9,363.099 3.010
9
2.77110
8
60,000 11,047.299 4.010
9
3.64510
8
70,000 12,708.358 5.010
9
4.51010
8
75,000 13,531.457 6.010
9
5.36810
8
80,000 14,350.121 7.010
9
6.22010
8
90,000 15,975.389 8.010
9
7.06610
8
100,000 17,586.284 9.010
9
7.90910
8
125,000 21,560.732 1.010
10
8.74710
8
1.510
5
2.53810
4
1.510
10
1.28810
9
2.010
5
3.30810
4
2.010
10
1.69710
9
2.510
5
4.06610
4
2.510
10
2.10310
9
3.010
5
4.81710
4
3.010
10
2.50510
9
4.010
5
6.26710
4
4.010
10
3.29910
9
5.010
5
7.69910
4
5.010
10
4.08710
9
6.010
5
9.11310
4
6.010
10
4.86810
9
7.010
5
1.05110
5
7.010
10
5.64310
9
8.010
5
1.18910
5
8.010
10
6.41410
9
9.010
5
1.32610
5
9.010
10
7.18310
9
1.010
6
1.46210
5
1.010
11
7.94810
9
1.510
6
2.12610
5
1.510
11
1.1710
10
2.010
6
2.78110
5
2.010
11
1.5510
10
2.510
6
3.42710
5
2.510
11
1.9210
10
3.010
6
4.06410
5
3.010
11
2.2910
10
4.010
6
5.31310
5
4.010
11
3.0210
10
5.010
6
6.54410
5
5.010
11
3.7510
10
6.010
6
7.76110
5
6.010
11
4.4710
10
7.010
6
8.96510
5
7.010
11
5.1910
10
8.010
6
1.01610
6
8.010
11
5.8910
10
9.010
6
1.13410
6
9.010
11
6.5810
10
1.010
7
1.25210
6
1.010
12
7.2810
10
1.510
7
1.82810
6
1.510
12
1.0810
11
2.010
7
2.39810
6
2.010
12
1.4210
11
4137
4 TwoPhase Filtration
4.1 The Equation of TwoPhase Filtration
This chapter deals with the physical phenomenon of two immiscible fluids flowing
simultaneously through porous media. Assuming that there is no mass transfer between
these two phases at the interface separating them and that phase equilibrium has been
achieved between the two phases. The concept of relative permeability as described in
Chapter 1 actually based on the mathematical description of this phenomenon. In this
chapter a general mathematical approach was developed to describe the movement of
these two fluids. Using the same assumption which has been previously discussed, the
fundamental equations of a twophase filtration may be set up in the same form as in Eq.
2.7.
(4.1)
(4.2)
Index 1 refers to the displacing phase index 2 to the displaced phase.
The displacing phase can either be the wetting phase or the nonwetting phase. As in a
onephase filtration it is assumed that the process is isothermal and thus density and
viscosity will be functions of pressure only:
(4.3)
(4.4)
The relation between density and pressure is given by Eq. 2.22 and Eq. 2.23. The
difference between the two phase pressures is specified as the capillary pressure which is
a function of saturation:
(4.5)
u
1
kk
r1
µ
1
 p
1
∇ ρ
1
gi
3
+ ( ) – =
u
2
kk
r2
µ
2
 p
2
∇ ρ
2
gi
3
+ ( ) – =
ρ
1
ρ
1
p
1
( ) = ρ
2
ρ
2
p
2
( ) =
µ
1
µ
1
p
1
( ) = µ
2
µ
2
p
2
( ) =
p
2
p
1
– P
c
S
1
( ) =
4138 4: TwoPhase Filtration
The equation of continuity only differs from Eq. 2.43 in the fact that fluid 1 only takes the
portion S
1
of the pore space φ and fluid 2 the portion S
2
. S
1
and S
2
are defined as the
saturations of the two phases.
It is evident that
(4.6)
and so
(4.7)
(4.8)
S
1
S
2
+ 1 =
ρ
1
u
1
( ) ∇
∂ φS
1
ρ
1
( )
∂t
 – =
ρ
2
u
2
( ) ∇
∂ φS
2
ρ
2
( )
∂t
 – =
4: TwoPhase Filtration 4139
4.2 Vertical TwoPhase Filtration of Incompressible
Fluids
For this case we assume a vertical one dimensional filtration of incompressible fluids in
a porous medium. Therefore Eq. 4.1  Eq. 4.8 may be written as follows:
(4.9)
(4.10)
(4.11)
(4.12)
(4.13)
(4.14)
Instead of coordinate x
3
simple x was written. Now it is useful to introduce the total
velocity of filtration as a new variable:
(4.15)
When adding Eq. 4.13 and Eq. 4.14:
(4.16)
it is evident that u is independent of x and therefore:
(4.17)
Then a new function is introduced:
(4.18)
and
u
1
kk
r1
µ
1

∂p
1
∂x
 ρ
1
g +
( J
 `
– =
u
2
kk
r2
µ
2

∂p
2
∂x
 ρ
2
g +
( J
 `
– =
p
2
p
1
– P
c
S
1
( ) =
S
1
S
2
+ 1 =
∂u
1
∂x
 φ
∂S
1
∂t
 + 0 =
∂u
2
∂x
 φ
∂S
2
∂t
 + 0 =
u u
1
u
2
+ =
∂ u
1
u
2
+ ( )
∂x
 φ
S
1
S
2
+ ( )
∂t
 + 0 =
∂u
∂x
 0 =
F
1
u
1
u
 =
4140 4: TwoPhase Filtration
(4.19)
where F
1
is the portion of fluid 1 in reference to the total flow. F
1
and F
2
are called
fractional flow functions. The usual symbol of fractional flow value or function is f
1
. We
use both F
1
and f
1
to distinguish between two cases. We use f
1
if the capillary pressure is
neglegted and F
1
if not.
Substitute Eq. 4.18 and Eq. 4.19 into Eq. 4.9 and Eq. 4.10 leads to:
(4.20)
(4.21)
Taking the derivative of Eq. 4.11 yields:
(4.22)
Then Eq. 4.21 is subtracted from Eq. 4.20 and equating the result with Eq. 4.22 we obtain:
(4.23)
F
1
is then obtained in the following form:
(4.24)
Also F
1
can be expressed as:
(4.25)
F
2
u
2
u

u u
1
–
u
 1 F
1
– = = =
∂p
1
∂x
  ρ
1
g –
µ
1
F
1
u
kk
r1
 – =
∂p
2
∂x
  ρ
2
g –
µ
2
1 F
1
– ( )u
kk
r2
 – =
∂p
2
∂x
 
∂p
1
∂x
 –
∂P
c
S
1
( )
∂x

dP
c
dS
1

∂S
1
∂x
 ⋅ = =
dP
c
dS
1

∂S
1
∂x
 ⋅ ρ
1
ρ
2
– ( )g
µ
1
F
1
u
kk
r1

µ
2
1 F
1
– ( )u
kk
r2
 – + =
F
1
µ
2
k
r2

k ρ
1
ρ
2
– ( )g
u
 –
µ
1
k
r1

µ
2
k
r2
 +

k
u

µ
1
k
r1

µ
2
k
r2
 +

dP
c
dS
1

∂S
1
∂x
  ⋅ ⋅ + =
F
1
1
kk
r2
uµ
2

dP
c
dS
1

∂S
1
∂x
 ⋅ ρg ∆ –
( J
' J
 `
+
1
µ
1
k
r2
µ
2
k
r1
 +
 =
4: TwoPhase Filtration 4141
If
(4.26)
and
(4.27)
then
(4.28)
where f
1
and ψ
1
are functions of S
1
and u.
Substituting into Eq. 4.13 and Eq. 4.16 leads to:
(4.29)
or after using Eq. 4.28:
(4.30)
f
1
µ
2
k
r2

k ρ
1
ρ
2
– ( )g
u
 –
µ
1
k
r1

µ
2
k
r2
 +
 =
ψ
1
k
u

µ
1
k
r1

µ
2
k
r2
 +

dP
c
dS
1
 ⋅ =
F
1
f
1
ψ
1
∂S
1
∂x
 ⋅ + =
u
1
u F
1
⋅ =
u
φ

∂F
1
∂x
 ⋅
∂S
1
∂t
 + 0 =
u
φ

df
1
dS
1

∂S
1
∂x
 ⋅
∂
∂x
 ψ
1
∂S
1
∂x
 ⋅
( J
 `
+
∂S
1
∂t
 + 0 =
4142 4: TwoPhase Filtration
4.3 The BUCKLEYLEVERETT Solution
Eq. 4.30 is not linear, thus a solution can only be achieved numerically.
In this chapter though the problem is discussed in a simplified manner.
In the year 1942 Buckley and Leverett published their theory which enabled great progress
on behalf of multiphase filtration.
This theory neglects the capillary force also considers the following assumptions:
• Incompressible fluids and porous media,
• Immiscible fluids,
• Darcy’s law of two phase filtration is valid and
• Linear displacement.
Eq. 4.11 and Eq. 4.27 are no longer required. According to Eq. 4.28:
. (4.31)
Eq. 4.30 becomes:
(4.32)
Figure 4.1: Calculation of fractional curve (after Marle)
The task is to compute the velocity w of a given saturation at any given point i.e. to ”track”
F
1
u
1
u
 f
1
= =
u
φ

df
1
dS
1

∂S
1
∂x
 ⋅ ⋅
∂S
1
∂t
 + 0 =
0 1 S
1m
S
1M
0
f
1
1
S
1
df
1
dS
1
f
1
4: TwoPhase Filtration 4143
the displacing front. Any given value of the saturation S
1
is valid for a specified point x
and time t. Mathematically, this means that for every constant S
1
value there is a function:
(4.33)
which gives the location x as a function of time.
Taking the derivative of Eq. 4.33:
(4.34)
and since:
(4.35)
this leads to:
(4.36)
Comparing Eq. 4.36 with Eq. 4.32 leads to:
(4.37)
where
q
1
is the injection rate of phase 1 and
A is the cross section area of the porous media.
This term (w) is only a function of S
1
and u. Therefore it is sufficient to know the
distribution of initial saturation and the velocity of displacement in order to calculate the
saturation distribution.
It is assumed that the initial distribution of saturation in a vertical porous medium at t = 0
corresponds to the curve shown in Figure 4.2. On one side at x = 0 the velocity u
1
= u is
constant. This means that the displacing phase is injected at a constant rate. S
1m
and S
1M
are the possible minimum and maximum values of S
1
.
If f
1
is plotted versus S
1
as shown in Figure 4.1 which illustrates the so called Sshaped
curve which is characteristical for most porous media. This curve has an inflection point
at which the differential of f
1
is a maximum.
S
1
x t , ( ) constant =
∂S
1
∂x
dx
∂S
1
∂t
dt + 0 =
dx wdt =
w
∂S
1
∂x

∂S
1
∂t
 + 0 =
w
u
φ

df
1
dS
1
 ⋅
q
1
Aφ

df
1
dS
1
 ⋅ = =
4144 4: TwoPhase Filtration
If porosity φ is also regarded as a constant then the speed of propagation for every value
of saturation will, as a result of Eq. 4.37, be proportional to:
(4.38)
Figure 4.2: Propagation of saturation profile (after Marle)
As time elapses the distance travelled for all saturations is plotted. As expected the points
with small or large saturation values progress at a lower speed than areas with a middle
saturation value as shown in Figure 4.2.
It means that after a certain time period we will have a saturation profile without physical
sense since several xvalues correspond to two different saturation values. This difficulty
only appears after the displacement has travelled a certain distance since the initial
saturation distribution was assumed to be continuously decreasing. If at t = 0 the
saturation S
1
equals S
1m
everywhere then these problems will arise for every time t > 0.
These difficulties are found also in other fields of physics for example in case of
supersonic streaming of gas.
The solution achieved for the saturation distribution may only be interpreted if the profile
is not continuous. (Figure 4.3 displays this discontinuity) As a consequence of the
conservation of mass the location of the discontinuity must be fixed in a way so that the
areas on both sides of the discontinuity are equal in size. (In Figure 4.3 the crosshatched
area).
df
1
dS
1

4: TwoPhase Filtration 4145
Figure 4.3: The displacement front as discontinuity of saturation (after Marle)
4.3.1 The WelgeMethod
Let the initial saturation in a porous medium be S
1a
which can either be smaller or larger
than S
1m
. Then the profile of saturation is plotted at sequential dates t
1
<t
2
<t
3
...
The discontinuity in saturation and the front saturation S
1f
should be determined for every
profile. It will be proven that this value is the same for all t.
As already mentioned it is essential that the discontinuity may not contradict the law of
conservation of mass. This means that the area below the analytical curve must be equal
in value to the area below the profile corrected by the discontinuity as shown in Figure 4.3.
The area below the analytical curve is calculated as follows:
(4.39)
w S
1
( )t S
1
d
S
1a
S
1M
∫
ut
φ

df
1
dS
1
 S
1
d
S
1a
S
1M
∫
ut
φ
  f
1
S
1M
( ) f
1
S
1a
( ) – [ ] = =
4146 4: TwoPhase Filtration
The area below the corrected profile (Welge approximation) will be:
(4.40)
Eq. 4.39 and Eq. 4.40 though must result in:
(4.41)
There is only one point that can satisfy Eq. 4.41. This point is the tangency point of the
line drawn from point S
1a
to the curve f
1
, which is independent of time. The meaning of
Eq. 4.41 is illustrated in Figure 4.4. The tangency point also gives the value of the
saturation at the front (S
1f
).
Figure 4.4: Determination of average saturation of the wetting phase after breakthrough (after
Welge)
S
1f
( )t S
1f
S
1a
– ( ) w S
1
( )t S
1
d
S
1f
S
1M
∫
+
ut
φ

df
1
dS
1

( J
' J
 `
S
1
S
1f
=
S
1f
S
1a
– ( ) f
1
S
1M
( ) f
1
S
1f
( ) – + ⋅ =
df
1
dS
1

( J
' J
 `
S
1
S
1f
=
f
1
S
1f
( ) f
1
S
1a
( ) –
S
1f
S
1a
–
 =
4: TwoPhase Filtration 4147
In order to calculate the average saturation of phase 1 in the swept zone behind the front
, one can write the equation of the tangent of the fractional flow curve as follows:
(4.42)
where is the slope of the tangent and C is the interception of f
1
axis.
At f
1
= f
1f
; S
1
= S
1f
(as shown in Figure 4.4) and Eq. 4.42 becomes:
(4.43)
Substituting Eq. 4.43 into Eq. 4.42 yields:
(4.44)
At f
1
= 1.0 ; and Eq. 4.44 becomes:
(4.45)
From Eq. 4.37, can be expresses as:
(4.46)
After integrating, Eq. 4.46 can be written as:
(4.47)
where Q is the cumulative amount of injected displacing phase (1). Substituting Eq. 4.47
into Eq. 4.45 and solving for S
1bf
yields:
(4.48)
S
1bf
( )
f
1
df
1
dS
1

( J
' J
 `
S
1
C + =
df
1
dS
1

C f
1f
df
1
dS
1

( J
' J
 `
S
1f
– =
f
1
S
1
S
1f
– ( )
df
1
dS
1
 f
1f
+ ⋅ =
S
1
S
1bf
=
1 S
1bf
S
1f
– ( )
df
1
dS
1
 f
1f
+ ⋅ =
df
1
dS
1

df
1
dS
1

φ
u
w
φ
u

dx
dt
 ⋅
φA
q

dx
dt
  ⋅ = = =
df
1
dS
1

φAL
Q
 =
S
1bf
S
1f
Q
φAL
 1 f
1f
– ( ) + =
4148 4: TwoPhase Filtration
There also is an alternative derivation of Eq. 4.48, based on material balance, is given in
Reference 4.3.
Figure 4.5 shows the cumulative production by a linear displacement. It illustrates that
after breakthrough the cumulative production of the displacing fluid (Q
1
) will increase
rapidly and the cumulative production of the displaced fluid (Q
2
) will decrease by the
same amount so the total cumulative production (Q) is a linear function.
Figure 4.5: Cumulative production by linear displacement (after Marle)
4: TwoPhase Filtration 4149
Example 4.1:
A cylindrical sandstone core having diameter of 0.05 [m], length of 1 [m] and porosity of
20%. Saturated with oil (35 API) and irreducable water saturation (S
1m
= 0.2). Water was
injected into the core with a steady rate of 0.1 [cm
3
s
1
] to displace the oil. The fractional
flow was calculated and listed below:
Sw:
Calculate the following:
a.) The water saturation at the water front (S
1f)
.
b.) The average oil saturation in the swept area of the core (behind the water front).
c.) The cumulative water injected to reach the average water saturation behind the
front determined in part (B).
d.) The velocity of the water front.
e.) Estimate the time of the water breakthrough.
Solution:
a.) From the plot of the fractional flow curve (f
w
verses S
w
):
b.) The point of tangency represents the water saturation at the front is S
1f
= 0.55.
c.) Extension of the tangent to f
w
=1.0, the average water saturation behind the front is
S
wbf
= 0.65, Then the average oil saturation behind the front = 1.00 – 0.65 = 0.35.
From Eq. 4.48, the cumulative water injected can be determined from:
S
w
f
w
0.20 0.00
0.40 0.30
0.50 0.62
0.55 0.75
0.60 0.85
0.65 0.92
0.70 0.96
0.75 0.99
4150 4: TwoPhase Filtration
d.) The slope of f
w
curve at (S
w
= 0.55) = 2.17
Then the velocity of the front calculated from:
e.) The time of water breakthrough is:
Q φAL
S
1bf
S
1f
–
1 f
1f
–
 0.2
3.14 25 ⋅
4
  100
0.65 0.55 –
1 0.775 –
 
( J
 `
⋅ ⋅ ⋅ 175 cm
3
= = =
w
q
1
Aφ

df
1
dS
1
 ⋅
0.1
π 2.5 ( )
2
0.2 ⋅
 2.17 ⋅ 0.055 cm s ⁄ = = =
t
L
w
f

100
0.055
 30 min = = =
4: TwoPhase Filtration 4151
4.4 Influence of Gravity and Capillary Force
4.4.1 Influence of Gravity
Eq. 4.26 may be written in the following form:
(4.49)
Displacement proceeds in vertical direction from bottom to top. At first the curve f
1
for
ρ
1
= ρ
2
is drawn. The second term of Eq. 4.49 vanishes and gravity has no more influence.
If ρ1 > ρ2 the term inside the parenthesis becomes smaller than 1 which means that the
fractional curve shifts to the right. The tangent drawn from the initial point [S
1i
, f
1
(S
1i
)]
is not so steep but the saturation at the front becomes larger and displacement more
effective. The limiting points of the curve f
1
remain unchanged since they only depend on
the k
r
functions (see Figure 4.6).
The relation:
(4.50)
is simply the relation between the frictional force and gravity. If the velocity of filtration
increases the fraction will become smaller and the fractional curve shifts back to the case
ρ
1
= ρ
2
.
The theoretical case of ρ
1
< ρ
2
can’t be handled by the theory previously discussed.
Because this state is not stable and the fluids will exchange their positions in counterflow.
A stable frontal displacement is only possible if the heavier fluid is maintained below the
lighter fluid.
f
1
1
1
k
r2
k
r1

µ
1
µ
2
 ⋅ +
 1
kk
r2
µ
2
 –
ρ
1
ρ
2
– ( )g
u
 ⋅ ⋅ =
kk
r2
µ
2

ρ
1
ρ
2
– ( )g
u
 ⋅
4152 4: TwoPhase Filtration
Figure 4.6: The influence of gravity on the fractional curve (after Marle)
4.4.2 Influence of the Capillary Force
When considering the capillary force one must regard the fact that Eq. 4.30 is not linear.
The solution can only be achieved with help of numerical methods, for example, by the
method of finite differences. Discussion of such methods of solution would surpass the
objective of this textbook.
The BuckleyLeverett solution neglecting the capillary force is illustrated in Figure 4.7
which has been discussed previously. The other profiles were calculated by applying Eq.
4.30 for various rates of filtration with the help of the method of finite differences.
In case of a slow displacement the capillary force is larger than the viscous forces. This is
expressed in a rather flat saturation profile. In the case of a fast displacement the profile
becomes steeper and tends to the BuckleyLeverett solution. It can be observed that when
the displacing phase reaches the end of the medium the displacing efficiency is larger at
a fast displacement than at a slow displacement.
4: TwoPhase Filtration 4153
Figure 4.7: Influence of the velocity of displacement on the distribution of saturation regarding the
capillary force (by Douglas et al 1958)
Figure 4.8: The displacing efficiency as a function of velocity (by Kyte, Rappoport 1958)
Figure 4.8 shows the oil recovery versus the rate factor for different core lengths and for
a strong waterwet system. Two things are of importance: First the efficiency of
displacement is at certain values independent of velocity. Second the time period between
4154 4: TwoPhase Filtration
the arrival and breakthrough of the displacing phase at a small displacing speed is large.
The point of breakthrough is defined as the moment of first outflow of the displacing
phase. The deviation is effected by the capillary endeffect.
4.4.3 The Capillary EndEffect
The endeffect is a phenomenon at which the wetting phase is held back by the capillary
force at the boundary of the medium. Until saturation at the boundary has not yet reached
the value S
1M
the capillary force will be larger than zero and the capillary gradient is
infinitely large.
In Figure 4.9 the fluid arriving at the boundary of the medium accumulates and causes a
peculiar deformation of the saturation profile. Breakthrough takes place at S
1M
and in the
following the saturation profile will tend towards the line of the saturation S
1M
.
On the other side the displacing phase may be nonwetting. Then the endeffect must be
considered in opposite. The saturation of the displacing phase remains S
1m
and the profile
developes as shown in Figure 4.10.
Figure 4.9: "Endeffect” in case of a wetting displacing phase (after Marle)
4: TwoPhase Filtration 4155
Figure 4.10: Endeffect” in case of a nonwetting displacing phase (after Marle)
4.4.4 Imbibition
Let us consider a porous medium contacted with the wetting phase at its bottom surface
and all other sides are covered by a impermeable layer. At the initial time t = 0 the
saturation of the wetting phase is S
1m
. Due to capillary forces the wetting phase tends to
intrude at the bottom side and thus displaces the nonwetting phase in counter flow. It is
assumed that the fluids are incompressible.
Figure 4.11: Displacement in countercurrent
Therefore:
4156 4: TwoPhase Filtration
(4.51)
Eq. 4.51 along with Eq. 4.15, Eq. 4.2, Eq. 4.9 – Eq. 4.14 remain further valid. Then Eq.
4.9 is divided by k
r1
/µ
1
and Eq. 4.10 by k
r2
/µ
2
. Afterwards they are subtracted one from
another yields:
(4.52)
From Eq. 4.51 and Eq. 4.12. Therefore:
(4.53)
where:
(4.54)
(4.55)
Substituting of Eq. 4.39 into Eq. 4.13 leads to:
(4.56)
or
(4.57)
Eq. 4.57 is in its form identical with Eq. 4.30. The boundary conditions are:
At the outlet x = L:
(4.58)
At the inlet x = 0, the capillary pressure is zero, thus:
u
1
u
2
+ 0 =
u
1
µ
1
k
r1

µ
2
k
r2
 +
( J
' J
 `
k ρ
1
ρ
2
– ( )g k
dP
c
dS
1

∂S
1
∂x
  ⋅ + =
u
1
ϕ
1
+
ψ
1
+
∂S
1
∂x
 ⋅ + =
ϕ
1
+
k ρ
1
ρ
2
– ( )g
µ
1
k
r1

µ
2
k
r2
 ⋅
 =
ψ
1
+ k
µ
1
k
r1

µ
2
k
r2
 +

dP
c
dS
1
 ⋅ =
∂
∂x
 ϕ
1
+
ψ
1
+
∂S
1
∂x
 + φ
∂S
1
∂t
 + 0 =
1
φ

∂ϕ
1
+
dS
1

∂S
1
∂x
 ⋅
∂
∂x
 ψ
1
+
∂S
1
∂x

( J
 `
+
∂S
1
∂t
  + 0 =
u
1
ϕ
1
+
ψ
1
+
∂S
1
∂x
 +
x L =
0 = =
4: TwoPhase Filtration 4157
(4.59)
The boundary value problem Eq. 4.57 – Eq. 4.59 can only be solved numerically.
To illustrate this it is of use to regard the calculations by Blair (1960). The result is shown
in Figure 4.12 and Figure 4.13. It is appropriate to mention that the pressure gradients of
the phases are corresponding to the counterflow opposed to one another. These numerical
results were verified by experiments of Graham and Richardson shown in Figure 4.14.
Figure 4.12: Capillary pressure and relative permeability functions used in the calculation by Blair
Figure 4.13: Distribution of pressure and saturation in case of linear (counterflowing) imbibition (by
Blair)
S
1
( )
x 0 =
S
1M
=
4158 4: TwoPhase Filtration
Figure 4.14: Recovery in case of linear counterflowing imbibition and the experimental
determination of the influence of a certain in corelength. (by Graham and Richardson)
4: TwoPhase Filtration 4159
4.5 References
4.1 Blair, P.M. paper no. 1475G, Trans. AIME secondary Rec. Symp., Wichita Falls,
Texas, U.S.A. (May 1960)
4.2 Buckley, J.R. and Leverett, M.C., Trans. AIME 146, 107 (1942).
4.3 Craig, F.F., Jr.: "The Reservoir Engineering Aspects of Waterflooding" Monograph
Vol. 3 of the Henry L. Doherty Series N.Y./Dallas (1971).
4.4 Douglas, J.,Jr., Blair, P.M., and Wangner, R.J.: "Calculation of Linear
Waterflooding Behavior Including the Effects of Capillary Pressure" Trans. AIME
215, 96 (1958).
4.5 Kyte, J.R., and Rappoport, L.A.: Trans. AIME 215, 423 (1958).
4.6 Marle, C.M.: "Multiphase Flow in Porous Media" Institut Francais du Petrole, Gulf
Publishing Company (1981).
4.7 Welge, H. J.: "A Simplified Method for Computing Oil Recovery by Gas or Water
Drive" Trans. AIME,195, 91 (1952).
4160 4: TwoPhase Filtration
Table 4.1: Summary of the Equations of One Phase and Two Phase Filtration.
One Phase Two Phase
Basic Equation of Motion
Equation of State
Equation of Continuity
u
k
µ
 – p ∇ ρgi
3
+ ( ) =
ρ ρ p ( ) =
µ µ p ( ) =
ρu ( ) ∇
∂ φρ ( )
∂t
 – =
u
1
kk
r1
µ
1
 – p
1
∇ ρ
1
gi
3
+ ( ) =
u
2
kk
r2
µ
2
 – p
2
∇ ρ
2
gi
3
+ ( ) =
p
2
p
1
– P
c
S
1
( ) =
ρ
1
ρ
1
p
1
( ) = ρ
2
ρ
2
p
2
( ) =
µ
1
µ
1
p
1
( ) = µ
2
µ
2
p
2
( ) =
S
1
S
2
+ 1 =
ρ
1
u
1
( ) ∇
∂ φS
1
ρ
1
( )
∂t
 – =
ρ
2
u
2
( ) ∇
∂ φS
2
ρ
2
( )
∂t
 – =
4: TwoPhase Filtration 4161
Table 4.2: Summary of the Equations of Two Phase Filtration.
Two Phase 1Dimensional
u
1
kk
r1
µ
1

∂p
1
∂x
 ρ
1
g +
( J
 `
– =
u
2
kk
r2
µ
2

∂p
2
∂x
 ρ
2
g +
( J
 `
– =
p
2
p
1
– P
c
S
1
( ) =
ρ
1
constant; = ρ
2
constant =
µ
1
constant; = µ
2
constant =
S
1
S
2
+ 1 =
u u
1
u
2
+ = f
1
u
1
u
 =
∂u
1
∂x
 φ
∂S
1
∂t
 + 0 =
∂p
1
∂x
  ρ
1
g –
µ
1
f
1
u
kk
r1
 – =
∂p
2
∂x
  ρ
2
g –
µ
2
1 f
1
– ( )u
kk
r2
 – =
∂p
2
∂x
 
∂p
1
∂x
 –
∂P
c
S
1
( )
∂x

∂P
c
∂S
1

∂S
1
∂x
 = =
f
1
1
kk
r2
uµ
2

∂P
c
∂S
1

∂S
1
∂x
 ρg ∆ –
( J
' J
 `
+
1
µ
1
µ
2

k
r2
k
r1
 +
 =
u
φ

∂f
1
∂S
1
 
∂S
1
∂x

∂S
1
∂t
 + 0 =
4162 4: TwoPhase Filtration
5163
5 PistonLike Displacement
5.1 The Mobility Ratio
It was observed that in case of neglecting the capillary forces between displacing and
displaced phases a discontinuity in the saturation will develop which may be presented by
a sharp front. It is defined as displacement front. The displaced phase flows ahead of the
front and the displacing phase is predominant behind the front.
If the capillary force is not neglectable then the saturation profile becomes continuous and
a more or less smooth profile. This part is called a transition zone.
If the displacement velocity is not extremely small the BuckeleyLeverett discontinuity
may be considered as a sufficient good approximation.
It is recommended to introduce further simplification by specifying the displacement as
pistonlike. The meaning is illustrated in Figure 5.1: It is assumed that ahead of the front:
(5.1)
and behind the front:
(5.2)
The mobilities of the displacing and displaced fluids are:
(5.3)
and
(5.4)
respectively.
S
1
S
1m
= and u
1
0 =
S
1
S
1M
= and u
2
0 =
λ
1
k
r1M
µ
1
 =
λ
2
k
r2M
µ
2
 =
5164 5: PistonLike Displacement
The mobility ratio (M) is defined as:
(5.5)
The mobility ratio is the ratio between the mobilities of the displaced phase ahead of the
front and the displacing phase behind the front and it is a constant for a given rockfluid
system
Figure 5.1: Comparison of saturation profiles according to different mathematical models
5.2 Propagation of a Displacement Front
It is assumed that the flow regimes behind the front just as ahead of the front are at steady
state and the fluids are incompressible. At time t the position of the front may be described
by the following function:
(5.6)
The propagation velocity of this surface is where is a function of and
t. The new position of the surface after a time period can then be described as follows:
(5.7)
When differentiating with respect to t:
(5.8)
M
λ
1
λ
2

k
r1M
µ
1

k
r2M
µ
2
 ⁄ = =
F x
1
x
2
x
3
t , , , ( ) 0 =
ν ν
1
ν
2
ν
3
, , ( ) = ν x
∂t
F x
1
ν
1
∂t x
2
, ν
2
∂t x
3
, ν
3
∂t t , ∂t + + + + ( ) 0 =
∂F
∂t

∂F
∂x
1
ν
1
∂F
∂x
2
ν
2
∂F
∂x
3
ν
3
+ + + 0 =
5: PistonLike Displacement 5165
or in vector form:
(5.9)
The Darcy’s law in vector form is given by:
(5.10)
Substituting Eq. 5.10 into Eq. 5.9 we obtained:
(5.11)
where the relation between and is:
(5.12)
The potential function for a constant density is:
(5.13)
The problem is now formulated according to Muskat (1934): Determine the distribution
of the potential between the border and the surface and the
distribution of the potential between the surface and the border .
The following boundary conditions are valid:
(5.14)
If neglecting the capillary force the pressure must be equal at the surface :
(5.15)
Then the component of velocity normal to the front must be continuous.
From Eq. 2.89 the equation of motion for the front may be written as follows:
(5.16)
∂F
∂t
 ν F ∇ + 0 =
u
kρ
µ
 ψ ∇ – =
∂F
∂t

kρ
φµ
 ψ ∇ F ∇ – 0 =
u ν
u φν =
ψ gx
3
p
ρ
 + =
ψ
1 Γ
1 ( )
F x t , ( ) 0 =
ψ
2 F x t , ( ) 0 = Γ
2 ( )
ψ
1
ψ
1
1 ( )
= x Γ
1 ( )
ψ
2
ψ
2
2 ( )
= x Γ
2 ( )
∈
∈
F x t , ( )
p
1
p
2
=
k
1
ρ
1
µ
1

∂ψ
1
∂n

k
2
ρ
2
µ
2

∂ψ
2
∂n
 =
5166 5: PistonLike Displacement
or in consequence to Eq. 5.11:
(5.17)
5.2.1 Linear Displacement
Let us look upon a horizontal displacement inside a linear medium with length L. The
front of displacement has proceeded to x
f
during the time period t as shown in Figure 5.2.
Just as in region 1 the filtration in region 2 is also a single phase filtration and according
to Eq. 3.3:
(5.18)
(5.19)
Figure 5.2: Schematic diagram of pistonlike displacement
The boundary conditions at x = 0, and x = L are:
(5.20)
(5.21)
At the displacement front:
φ
∂F
∂t

kρ
µ

( J
 `
1
ψ
1
∇ F ∇ – φ
∂F
∂t

kρ
µ

( J
 `
2
ψ
2
∇ F ∇ – 0 = =
∂
2
p
1
∂x
2
 0 = 0 x x
f
< <
∂
2
p
2
∂x
2
 0 = x
f
x L < <
p
1
p
10
= x 0 =
p
2
p
2L
= x L =
5: PistonLike Displacement 5167
(5.22)
Since , then:
(5.23)
Then the initial condition is given by:
(5.24)
After integrating twice Eq. 5.18 and Eq. 5.19 become:
(5.25)
(5.26)
a
1
, b
1
, a
2
and b
2
are all constants of integration which can be determined by the
Substitution of Eq. 5.20  Eq. 5.23 into Eq. 5.25 and Eq. 5.26 leads to:
(5.27)
(5.28)
(5.29)
(5.30)
The velocity of the front is:
(5.31)
where is the displaced fraction of the pore volume which is determined by:
(5.32)
p
1
p
2
= x x
f
λ
1
∂p
1
∂x
 λ
2
∂p
2
∂x
 =
=
λ
1
λ
2
⁄ M =
M
∂p
1
∂x

∂p
2
∂x
 =
x
f
0 = at t 0 =
p
1
a
1
x b
1
+ =
p
2
a
2
x b
2
+ =
a
1
p
2L
p
10
–
x
f
M L x
f
– ( ) +
 =
b
1
p
10
=
a
2
M p
2L
p
10
– ( )
x
f
M L x
f
– ( ) +
 =
b
2
p
2L
M p
2L
p
10
– ( )
x
f
M L x
f
– ( ) +
 – =
ν
f
dx
f
dt

u
1
φ
D
 = =
φ
D
φ
D
φ S
1M
S
1m
– ( ) =
5168 5: PistonLike Displacement
and
(5.33)
From Eq. 5.27 and Eq. 5.33, the Darcy velocity in Eq. 5.31 can be expressed as:
(5.34)
Eq. 5.34 is then integrated yields:
(5.35)
where t is the time period in which the front proceeds to x
f
. The time t is proportional to
x
f
if M = 1, which means that the two phase mobilities are equal in value.
In order to simplify Eq. 5.35 the dimensionless variables, defined below, were used:
(5.36)
Then Eq. 5.35 becomes:
(5.37)
Eq. 5.37 is illustrated graphically in Fig. (5.3) which shows the relation between t
D
and
x
Df
for different values of M.
5.2.2 Displacement in an Inclined Layer
Let us now regard a non horizontal layer in which both the displaced and displacing phase
are in static equilibrium if and the heavier displacing phase is below. During the
displacement the boundary of the phases become inclined. Figure 5.4 illustrates three
cases: a) the Initial case at t = 0 a static condition is reached. In case b) the displacing
phase pushes forwards at the bottom of the layer and affects an unfavorable efficiency of
displacement. In case c) the displacement is evenly balanced over the whole thickness.
The question arises now is:
u
1
kλ
1
–
∂p
1
∂x
 kλ
1
a
1
– = =
ν
f
dx
f
dt

kλ
1
φ
D

p
10
p
2L
– ( )
x
f
M L x
f
– ( ) +
 = =
t
φ
D
kλ
1
p
10
p
2L
– ( )
 MLx
f
1
2
 1 M – ( )x
f
2
+ =
x
Df
x
f
L
 = t
D
kλ
1
p
10
p
2L
– ( )
φ
D
L
2
 =
t
D
Mx
Df
1
2
 1 M – ( )x
Df
2
+ =
ρ
1
ρ
2
>
5: PistonLike Displacement 5169
When does displacement b) and when does displacement c) takes place?
Same simplifications which has been previously discussed are applied:
• The phases are incompressible.
• Ahead of the front the saturation of the displacing phase is S
1m
= 1  S
2M
, where S
2M
is the initial saturation of the displaced phase. The relative permeability of phase 1 is
zero ahead of the front.
• Behind the front the saturation of the displacing phase is S
1M
= 1  S
2m
and the
relative permeability of phase 2 is zero.
Figure 5.3: Influence of the mobility ratio on front propagations in case of a linear displacement
0
0
0.2
0.4
0.6
0.8
1 2 3 4 5
t
D
1.0
x
D
f
M=10
M=5
M=1
M=0.5
M=0.1
5170 5: PistonLike Displacement
Figure 5.4: Possible positions of the displacing front in inclined layer.
Behind the front:
(5.38)
(5.39)
(5.40)
Ahead of the front:
(5.41)
(5.42)
u
1
kk
r1M
µ
1
 – p
1
∇ ρ
1
gi
3
+ ( ) =
u
2
0 =
u ∇
1
0 =
u
1
0 =
u
2
kk
r2M
µ
2
 – p
2
∇ ρ
2
gi
3
+ ( ) =
5: PistonLike Displacement 5171
(5.43)
The boundary conditions are as follows:
(5.44)
(5.45)
(5.46)
The conditions formulated in Eq. 5.46 neglects the capillary forces and the continuity of
filtration velocity at the front.
The general solution was previously discussed. This though is so complicated that
solutions are only achieved numerically. In order to answer the original question it is not
of importance to know the position and shape of the front at all times. It is satisfactory to
know if a stable shape is formed after a certain distance of displacement. This stable shape
of the front then proceeds translatorial.
Figure 5.5: Forces acting on the displacing front.
Actually it is essential to know if the boundary value problem Eq. 5.38  Eq. 5.46 have a
semi steadystate solution and in addition which kind? The following idea was provided
by Dietz (1953).
In the Figure 5.5 two neighboring points at the front are considered. Those are and
. At both points Eq. 5.46 must be valid:
(5.47)
u ∇
2
0 =
u
1
u = in infinity behind the front
u
2
u = in infinity ahead of the front
u
1
n ⋅ 0 =
u
2
n ⋅ 0 =



at the impermeable boundaries of the layer
p
1
p
2
=
u
1
n ⋅ u
2
n ⋅ =



at the front
r
r dr + ( )
p
1
r ( ) p
2
r ( )
p
1
r dr + ( ) p
2
r dr + ( ) =
=
5172 5: PistonLike Displacement
Since:
(5.48)
(5.49)
from Eq. 5.47:
(5.50)
Substitution of and from Eq. 5.38 and Eq. 5.42 into Eq. 5.50 leads to:
(5.51)
This equation indicates that if a stationary solution for the fluid phase exists it must be
perpendicular to the vector:
(5.52)
where M is the Mobility Ratio:
(5.53)
Figure 5.6: Position of the displacing front by favorite mobility ratio (after Marle)
Figure 5.6 illustrates the position of the front. The heavier fluid 1 displaces the lighter
fluid 2 from bottom to top. The mobility of the displaced fluid is superior. In this case
p
1
r dr + ( ) p
1
r ( ) – p
1
dr ∇ =
p
2
r dr + ( ) p
2
r ( ) – p
2
dr ∇ =
p
1
∇ p
2
∇ – ( )dr 0 =
p
1
∇ p
2
∇
ρ
1
ρ
2
– ( )gi
3
µ
1
kk
r1M

µ
2
kk
r2M
 –
( J
' J
 `
u + dr 0 =
w ρ
1
ρ
2
– ( )gi
3
u
k

µ
1
k
r1M
 1 M – ( ) + =
M
k
r1
µ
1

( J
' J
 `
k
r2
µ
2

( J
' J
 `
⁄ =
5: PistonLike Displacement 5173
. Both terms of the right side of Eq. 5.52 become positive. With increasing
displacement velocity the vector turns to the direction of . This position is semi
steadystate and stable. If the velocity of filtration is reduced to zero the front will turn to
the horizontal position and maintain this position.
Now let us see the case where :
Figure 5.7: Position of the displacing front by unfavorable mobility ratio.
The displacement proceeds from bottom to top. The heavier but more mobile phase
displaces the lighter and less mobile phase. Figure 5.7 states that again a stationary and
stable front will exist. This front though becomes more and more flattered as front
velocity increases.
If the velocity exceeds a certain critical value a stationary position of the front becomes
impossible to maintain. This critical velocity may be calculated using Eq. 5.52. At the
critical velocity the vector is perpendicular to the axis of symmetry of the layer. In
consequence the scalar multiplication with becomes zero:
(5.54)
Since and Eq. 5.54 becomes:
(5.55)
M 1 <
w u
M 1 >
α
w
u
w u ⋅ ρ
1
ρ
2
– ( )gi
3
u
k

µ
1
k
r1M
 1 M – ( ) + u 0 = =
i
3
u ⋅ u α sin ⋅ = u u ⋅ u
2
=
u
k
k ρ
1
ρ
2
– ( )g α sin
µ
1
k
r1M
 1 M – ( )
 – =
5174 5: PistonLike Displacement
u
k
is the critical velocity of filtrations. A filtration with velocity higher then u
k
is called
supercritical.
A case of will not be discussed. In this case though the heavier fluid is on top of
the lighter fluid which makes the front unstable.
ρ
1
ρ
2
<
5: PistonLike Displacement 5175
5.2.3 Supercritical Displacement
If the velocity of displacement is larger than the critical velocity the interface will become
more and more extended Figure 5.8. This displacement is called supercritical. The
theories of Le Fur and Sourieau are used for explanation purposes in this work.
At first it is assumed that displacement is at an advanced stage. The front has travelled a
considerably far distance and the state is to be referred to as practically stable. That means,
the rate of changes is everywhere moderate.
In this case the Dupuitassumption is valid which states that the equipotentials are
perpendicular to the layer. In this case u
1
is the same at every point of the layer which
contains the fluid 1 (section h
1
(x,t)).
The same is valid for u
2
in the layer section h
2
(x,t). The relation between h
1
and h
2
is:
(5.56)
and
(5.57)
Figure 5.8: Supercritical displacement in inclined layer (after Marle)
If the displacing fluid takes the lower part of the layer and h
1
(x,t) is referred to as
the distance between the bottom of the layer and the front. In an abitrary but fixed point
x
a
the value h
1
(x,t) will monotonously increase with time.
h h
1
x t , ( ) h
2
x t , ( ) + =
u
1
h
 h
1
u
1
h
2
u
2
+ [ ] =
ρ
1
ρ
2
>
5176 5: PistonLike Displacement
The velocity of filtration in direction of the xaxis is:
(5.58)
(5.59)
Since filtration is parallel to the axis of the layer at every point (a consequence of the
Dupuitassumption) the following equations are valid:
(5.60)
(5.61)
The equation of continuity is set up for the whole layer as:
(5.62)
and the following average values are defined as:
(5.63)
(5.64)
(5.65)
(5.66)
(5.67)
(5.68)
(5.69)
u
1
kk
r1M
µ
1
 –
∂p
1
∂x
 ρ
1
g α sin +
( J
 `
=
u
2
kk
r2M
µ
2
 –
∂p
2
∂x
 ρ
2
g α sin +
( J
 `
=
∂p
1
∂y
  ρ
1
g α 0 = cos +
∂p
2
∂y
  ρ
2
g α 0 = cos +
∂ u
1
h
1
( )
∂x
 φ S
1M
S
1m
– ( )
∂h
1
∂x
 + 0 =
S
1
h
1
S
1M
h
2
S
1m
+
h
 =
S
2
h
1
S
2m
h
2
S
2M
+
h
 1 S
1
– = =
w
1
h
1
h
u
1
=
w
2
h
2
h
u
2
=
p
1
*
x t , ( ) p
1
x 0 t , , ( ) =
p
2
*
x t , ( ) p
2
x 0 t , , ( ) =
γ g α sin =
5: PistonLike Displacement 5177
(5.70)
(5.71)
As before the capillary force is neglected and the pressures of the phases must be equal at
the front, then:
(5.72)
and from Eq. 5.67 and Eq. 5.68:
(5.73)
Since:
(5.74)
and from Eq. 5.73:
(5.75)
Substitution of Eq. 5.63  Eq. 5.75 into Eq. 5.58, Eq. 5.59 and Eq. 5.62 leads to:
(5.76)
(5.77)
(5.78)
(5.79)
k
r1
*
S
1
( ) k
r1M
S
1
S
1m
–
S
1M
S
1m
–
 =
k
r2
*
S
2
( ) k
r2M
S
1
S
1m
–
S
1M
S
1m
–
 =
p
1
x h
1
t , , ( ) p
2
x h
2
t , , ( ) =
p
1
*
x t , ( ) h
1
ρ
1
g α cos + p
2
*
x t , ( ) h
2
ρ
2
g α cos + =
h
1
h
S
1
S
1m
–
S
1M
S
1m
–
 =
P
c
*
S
1
( ) p
2
*
x t , ( ) p
1
*
x t , ( ) –
h
S
1
S
1m
–
S
1M
S
1m
–
 ρ
1
ρ
2
– ( )g α sin
= =
w
1
kk
r1
*
µ
1
 –
∂p
1
*
∂x
 ρ
1
γ +
( J
' J
 `
=
w
2
kk
r2
*
µ
2
 –
∂p
2
*
∂x
 ρ
2
γ +
( J
' J
 `
=
S
1
S
2
+ 1 =
∂w
1
∂x
 φ
∂S
1
∂t
 + 0 =
5178 5: PistonLike Displacement
(5.80)
(5.81)
Eq. 5.76  Eq. 5.81 may be considerated identical with the equations of Buckley  Leverett
Eq. 4.9  Eq. 4.14. The two dimensional approximation of the supercritical displacement
corresponds with a one dimensional displacement where
is the fictitious gravity,
the average saturations,
are the fictitious phase pressures
are the fictitious fractional velocity,
are the pseudo relative permeabilities
is the pseudo capillary pressure.
All factors, except P
c
, are physically corresponds to the actual factors. Therefore the
fictive saturation is the average saturation the fictive pressure is the pressure at any select
point, etc. Only the function has nothing to do with capillary pressure, because the
capillary forces were neglected due to precondition Eq. 5.72.
w
1
w
2
+ 0 =
p
2
*
p
1
*
– P
c
*
S
1
( ) =
γ
S
1
S
1
,
p
1
*
p
2
*
,
w
1
w
2
,
k
r1
*
S
1
( ) k
r2
*
S
2
( ) ,
P
c
*
S
1
( )
P
c
*
S
1
( )
5: PistonLike Displacement 5179
5.3 References
5.1 Dietz, D.N.: "A theoretical approach to the problem of encroaching and bypassing
edge water" Koninkl, Ned., Akad., Wetenschap, Proc. B56,83 (1953).
5.2 Marle, C.M.: "Multiphase Flow in Porous Media" Institut du Petrole. Gulf
Publishing Company (1981).
5.3 Muskat, M.: "Flow of Homogeneous Fluids through Porous Media". McGraw Hill
Book Co., N.Y. (1937).
5180 5: PistonLike Displacement
6: Examination Outline 6181
6 Examination Outline
The candidate draws three cards of different colours. Every colour marks a question
referring to a distinct part of the subject.
Every card comprises two questions: A and B.
A  Fundamental concepts, the most important formulae and figures supported by short
explanations are asked. Correct and sovereign answers attest that the candidate is familiar
with the subject and understands the topics very well.
B  The theoretical background is asked. The candidate must be able to derivate
formulae. Moreover, the candidate is examined if he/she is skilled to use his/her
knowledge in solving problems which had not been discussed in the lectures.
At first all Aquestions have to be answered orally by adducing short and clear definitions
without any derivations. In doing so, the most important formulae have to be written and
figures have to be sketched. Five minutes are the optimum period of time for answering
the three Aquestions.
As far as the Aquestions are concerned, the candidate is failing in case of only one bad
mistake or one lack of knowledge. In case of optimum answering, the mark fair (3) is
offered. However, the candidate can endeavour to get a better mark by approaching to
answer the Bquestions.
The candidate gets 2030 minutes to prepare the answers to the Bquestions. With the help
of his/her notes, the candidate has to present the answers orally. In doing so, the
demonstration of his/her understanding of physical meanings is more important than the
exactness of derivation. The final mark will range between excellent (1) and fair (3).
6182 6: Examination Outline
6.1 A  Questions
6.1.1 Part 1
6.1.1.1 Porous Material
• definition of porous medium and porosity
• classification of porous material
• What is compressibility?
• What is compaction?
• What is the difference between compaction and compressibility?
6.1.1.2 Wettability, Saturation, Capillary Pressure
• definition of saturation
• explanation of the term wettability
• explain the terms wetting and nonwetting
• What is capillary pressure?
• illustrate the capillary pressure curve
6.1.1.3 Darcy Equation, Liquid and Gas Permeability,
Anisotropy
• Darcy´s law
• definition of permeability
• compare liquid and gas permeability
• explain anisotropy
6: Examination Outline 6183
6.1.1.4 Relative Permeability
• define and explain relative permeability
6.1.2 Part 2
6.1.2.1 Continuity Equation, Darcy´s Law, Equation of State
• equation of continuity
• main forces in the filtration process
• differential form of Darcy´s law
• equation of state
6.1.2.2 Equation of Filtration for Low Compressible Fluids,
Boundary Conditions
• compressibleincompressible fluid
• equation of filtration for low compressible fluid, piezometrical conductivity
• boundary conditions: closedopen boundary
6.1.2.3 Equation of Filtration for Real Gas
• real gas pseudo pressure
• equation of filtration for real gas
• ideal gas
6.1.2.4 SteadyState Filtration through Linear Systems for
Liquids and Real Gases
6184 6: Examination Outline
• explain the term steadystate
• steadystate filtration for liquids
• steadystate filtration for real gases
• What is the major difference between the two equations?
6.1.2.5 SteadyState Filration through Radial Systems
• equation
• explain the term radialsymmetric with respect to potential distribution
• Dupuit equation
6.1.2.6 NonSteadyState Filtration with Constant Production
Rate through Infinite Radial Systems
• differential equation
• inital and boundary conditions
• solution with Eifunction
• solution with lgfunction
6.1.2.7 Calculation of the Cumulative Influx
• infinite radial systems with constant pressure at the inner boundary
• dimensionless variables
6.1.2.8 Finite Radial Systems
• differential equation
• constant boundary conditions (constant production rate, constant pressure)
• plot the pressure changes and the cumulative influx for constant boundary conditions
• asymptotic solutions for large periods of time
6: Examination Outline 6185
6.1.2.9 Physical Meaning of Superposition
• explain the principle of superposition
• first law of superposition
• second law of superposition
6.1.2.10 PressureBuild Up of a ShutIn Well
• illustrate the idea of a pressurebuild up
• Horner plot: What do you plot? What can you determine from the plot?
6.1.3 Part 3
6.1.3.1 BuckleyLeverett Theory
• assumptions
• ffunction
• saturation profile
• front saturation
6.1.3.2 Effect of Gravity and Viscosity on the fCurve, Influence
of Capillary Pressure on the Saturation Profile
• illustrate the effect of gravity on the fcurve
• explain the effect of viscosity on the fcurve
• influence of capillary pressure on the saturation profile
• illustrate the influence of displacement velocity on the saturation profile
• What is the capillary end effect?
6186 6: Examination Outline
6.1.3.3 PistonLike Displacement in Linear Systems
• define the mobility ratio
• illustrate the idea of pistonlike displacement
• front velocity
• explain the effect of different mobility ratios on the pressure distribution
6.1.3.4 Displacement in Inclined Layers
• illustrate possible positions of the displacing front in an inclined layer and the forces
acting at the front
• Dietz stability analysis
• What is supercritical displacement?
6: Examination Outline 6187
6.2 B  Questions
6.2.1 Part 1
6.2.1.1 Porosity Measurements
6.2.1.2 Evaluation of Capillary Pressure Curves
• How can the capillary pressure of a porous medium be measured?
• conversion of laboratory data
• Leverett function
6.2.1.3 Permeabilty Measurements
• Which factors influence the applied method?
• illustrate common permeabilty measurements
• analogies to the Darcy equation
6.2.1.4 Evaluation of Relative Permeability Functions
• How can relative permeability functions be measured?
6.2.2 Part 2
6.2.2.1 Derivation of Special Forms of the Equation of
Filtration 1
• derive the equation of filtration for an incompressible fluid
6188 6: Examination Outline
• derive the equation of filtration for a low compressible fluid
6.2.2.2 Derivation of Special Forms of the Equation of
Filtration 2
• derive the equation of filtration for an ideal gas
• derive the equation of filtration for a real gas
6.2.2.3 NonSteadyState Filtration through Infinite Systems
• derive the equation of filtration for a radial system with constant production rate
• describe the properties of the Eifunction
6.2.2.4 Filtration of Low Compressible Fluids through Radial
Systems with Closed Boundaries
• illustrate the principle of dimensionless variables
• derive equations for boundaries with constant production rate and constant pressure
• give solutions for the early and the late period
6.2.2.5 Second Law of Superposition
• calculation of multiwell problems
• wells with variable production rate
6.2.2.6 Methods of Image
• illustrate the principle of the method
• pressurebuild up test near a no flow boundary for early and late period
• pressurebuild up test near a constant pressure boundary
6: Examination Outline 6189
6.2.3 Part 3
6.2.3.1 Derivation of the BuckleyLeverett Theory
• assumptions
• derive the BuckleyLeverett solution
6.2.3.2 Welge Method
• illustrate the principle of the Welge method
• How do you determine the front saturation?
• How do you determine the average saturation?
6.2.3.3 PistonLike Displacement in Linear Systems
• assumptions
• derive the velocity of the front
• derive the time period in which the front proceeds to x
f
6.2.3.4 Supercritical Displacement
6190 6: Examination Outline
© No part of this publication may be reproduced in any form. Only students of the University of Leoben may copy for studying purposes. Edited by DI Johannes Pichelbauer  January 2003
Table of Contents
3
Table of Contents
1 Fundamental Properties of Porous Media ................................................. 11
1.1 Porosity ..................................................................................................................... 12 1.1.1 General Aspects and Definition ...................................................................... 12 1.1.2 Determination of Porosity ............................................................................... 13 1.1.3 Compaction ..................................................................................................... 15 1.1.4 Compressibility of Porous Media ................................................................... 16 Capillary Properties ................................................................................................... 19 1.2.1 Saturation ........................................................................................................ 19 1.2.2 Wettability ...................................................................................................... 19 1.2.3 Capillary Pressure ........................................................................................... 20 1.2.4 Measurements of Capillary Pressure in a Porous Medium ............................. 23 1.2.5 Conversion of Laboratory Data ...................................................................... 28 1.2.6 The Leverett Function ..................................................................................... 30 1.2.7 Pore Size Distribution ..................................................................................... 31 Permeability .............................................................................................................. 33 1.3.1 Darcy’s Law ................................................................................................... 33 1.3.2 Definition and Units of Permeability .............................................................. 34 1.3.3 Measurements of Permeability ....................................................................... 36 1.3.4 Klinkenberg Effect .......................................................................................... 41 1.3.5 Analogies between the Laws of Darcy, Ohm and Fourier ............................. 42 1.3.6 Filtration Velocity ........................................................................................... 43 1.3.7 Quadratic Equation of Filtration ..................................................................... 44 Relative Permeabilities ............................................................................................. 45 1.4.1 Definition of Relative Permeability ................................................................ 46 1.4.2 Relative Permeability Measurements ............................................................. 47 1.4.2.1 The HASSLER method ...................................................................... 47 1.4.2.2 PENNSTATEMethod ...................................................................... 51 1.4.2.3 WelgeMethod.................................................................................... 52 1.4.3 Saturation Distribution and Relative Permeability ......................................... 52 References ................................................................................................................. 55
1.2
1.3
1.4
1.5
2 Equations of SinglePhase Filtration .......................................................... 57
2.1 Fundamental Equation of Filtration. ......................................................................... 59 2.1.1 Differential Form of the DarcyLaw .............................................................. 59 2.1.2 Anisotropic Porous Media .............................................................................. 62 Equation of State ....................................................................................................... 65 2.2.1 Incompressible Fluids ..................................................................................... 65 2.2.2 Low Compressibility Fluids ........................................................................... 65 2.2.3 Formation Volume Factor ............................................................................... 66 2.2.4 Ideal and Real Gases ....................................................................................... 67 2.2.5 Equation of continuity .................................................................................... 68 Special Forms of the Equation of Filtration .............................................................. 71 2.3.1 Incompressible Fluids ..................................................................................... 71
2.2
2.3
....................8 4 TwoPhase Filtration..5........................................................................5.............5 3.................................................5 Pressure Buildup of ShutIn Well ........................................... 85 3............................1 Linear Flow with Constant Production Rate ..7....5.1 Influence of Gravity ...............................................1 Closed Exterior Boundary . 72 Real and Ideal Gases ...................... 83 3...................................................................................5...................................................7..............5....................4 3.........................................4 Single Well with Variable Production Rates .....7.........................................6....................................3 Steady State Gas Filtration ............................................... 100 The Infinite Radial System with Constant Pressure at the Interior Boundary ........................................................ 79 3 Solutions of the SinglePhase Equation of Filtration ............... 112 3.....6...3 3................. 76 2................................... 137 Vertical TwoPhase Filtration of Incompressible Fluids ...... 110 3............................. 137 4..........................................1 4................................................................................................6............. 133 3.......7................... 110 3.................... 125 3........................7. 130 3...........6 2.............................2 Steady State Filtration in a Radial System .......... 88 NonSteady State Filtration in Infinite Acting Systems .......................5..............................1.......... 74 Boundary and Initial Conditions .5......................................................1 Steady State Filtration of Low Compressibility Fluid ................................ 127 3......................... 90 3......... 84 3...............2 Low Compressibility Fluids .....1 The WelgeMethod ...3 4.....................................................................................................................1...........2................................................1 The First Law of Superposition ..................................................................... 114 The Principle of Superposition .............................1...........1 Constant Production Rate ..... 76 2............ 95 3..................................2 4.1 Pressure Buildup Test Near No Flow Boundary .3.................... 77 2.................................1................. 139 The BUCKLEYLEVERETT Solution ....... 84 3.....................................3 Pressure Drop in Space and Time ..1 Steady State Filtration .. 114 3......1...................... 145 Influence of Gravity and Capillary Force ..2 Initial Conditions ......................2 Constant Pressure at the Interior Boundary and Closed Exterior Boundary 113 NonSteady State Filtration in Linear System ........................3 Calculation of MultiWell Problems ..............7...........4 The Spatial Distribution of Flow ... 123 3................2 Properties of the EiFunction .....2 Boundary with Constant Pressure ................... 125 3....4 The Equation of TwoPhase Filtration .............................1 Radial Systems with Constant Production Rate ............................5 2.............................. 78 Schematic of the Filtration Equations ......................................... 99 Dimensionless Variables ................................................2..... 110 3.6 Method of Image ................................4 2......................................................................1 Boundary Conditions .........3 Discontinuities in Porous Media .............................................................................. 105 NonSteady State Filtration in a Finite System ............... 120 3.................. 151 4.................................1 Elastic Porous Media..........................................2..........2...7.......... 90 3...........................4...........3........... 151 ...............3.............2 The Second Law of Superposition .....................................................................................................7 3...............................................................................................2 3................Table of Contents 4 2................................ 71 2.....2 Constant Pressure Boundary ..... 120 3........................................................... 94 3........... 142 4........7. 132 References .. 128 3..........................................................................................2...............6 3..............
...................................2.....5 4..........2...2 Part 2 ......4 Relative Permeability ........ 185 6.................................................2 The Mobility Ratio .....................1..1..........1......1............................1.......... Liquid and Gas Permeability...............................................3 Equation of Filtration for Real Gas ............ 187 6................. 179 5.................... 185 6...............................1................2 .. 183 6.. Equation of State ........ 182 6....................................1............................. 186 B .........................2.................Questions ....................1 Linear Displacement ..........................2........ 183 6................ Anisotropy........1 Porosity Measurements . 187 6..1 Part 1 ................. 183 6........1.... 185 6........... 184 6........ 185 6...............2 Displacement in an Inclined Layer .........2......................... 182 6.1......Table of Contents 5 4..........2 Equation of Filtration for Low Compressible Fluids.........................2............ 166 5.............1 Continuity Equation....................................................................7 Calculation of the Cumulative Influx ...................................................... 163 5.........2....................................1................................................................. 183 6...............1........................2.. 185 6...4.................... Capillary Pressure ..................... 168 5............................ 184 6.........................1.................................. Saturation............................................2....... 184 6...........2 Evaluation of Capillary Pressure Curves................................ 155 References ........ 187 6. Boundary Conditions..10PressureBuild Up of a ShutIn Well..........3 Permeabilty Measurements ................................. 186 6.1................3...3....... 187 6...........1.....3 Darcy Equation.....8 Finite Radial Systems ....1.........................................................1....3................ 187 6.4 Displacement in Inclined Layers ...................1 BuckleyLeverett Theory.............2................................................9 Physical Meaning of Superposition..............................................3 PistonLike Displacement in Linear Systems ..................................6 NonSteadyState Filtration with Constant Production Rate through Infinite Radial Systems ..2........2...............................1...................................................1.......3 The Capillary EndEffect ....................... 175 References .2 Effect of Gravity and Viscosity on the fCurve.................3 Supercritical Displacement ..3 6 Examination Outline ......................4 Evaluation of Relative Permeability Functions.5 SteadyState Filration through Radial Systems.1.....................2 Wettability.........1............1 Porous Material .....1 A ...................................4 SteadyState Filtration through Linear Systems for Liquids and Real Gases .......3 Part 3 ....................................................................................... 182 6....................1..........4..2......................Questions .... 163 Propagation of a Displacement Front ....................................... 152 4.....1......................................2..........1 5... 182 6..........................................................1 Part 1 ....................................2............................................ 183 6...............1...................................................................3.................................................................4 Imbibition ............... 184 6..................... 159 5 PistonLike Displacement ........................................1.1.2..................................................4..................................2 Influence of the Capillary Force .............. 154 4.1...................................................2........................... 164 5........2.........1...... 183 6................................... 181 6.... 187 6......................................... Influence of Capillary Pressure on the Saturation Profile ........ 182 6....................................... Darcy´s Law.................................1....................................................1.................................................................
....................2........2...............................2...................2.......................3.... 188 6............ 189 6.....4 Filtration of Low Compressible Fluids through Radial Systems with Closed Boundaries.....3 NonSteadyState Filtration through Infinite Systems ............2.........2.................2........2 Part 2 ................................5 Second Law of Superposition...............................................2.........1 Derivation of the BuckleyLeverett Theory ...... 189 6........3 Part 3 .......................................3.. 188 6.......2.1 Derivation of Special Forms of the Equation of Filtration 1.......................3 PistonLike Displacement in Linear Systems ...................................... 189 ....... 188 6.........6 Methods of Image........................2........... 189 6...................... 188 6........2............................ 187 6........................3....................2................. 187 6............................3...............2 Derivation of Special Forms of the Equation of Filtration 2...2...4 Supercritical Displacement.............2.Table of Contents 6 6......................... 189 6.............2.....................2 Welge Method . 188 6........................2.2.....2..........
................................................2: Figure 1.6: Figure 1....K.......................10: Figure 1........29: Figure 1.................28: Figure 1........27: Figure 1................................................................. 53 Drainage and imbibition relative permeability characteristic (after Craig) ............12: Figure 1...................................List of Figures 7 List of Figures Figure 1..................... 30 Non wetting fluid saturation versus the effective pore size distribution .... 22 Capillary pressure versus saturation of the wetting phase for the model in Figure 1........................................................................................................................................................22: Figure 1................. 51 Schematic diagram of water invasion into porous media permeabilities of the wetting phase (after Craig).....................................................................................15: Figure 1..................Hall) .......................................... and Rassel) .................................................. 40 Variation in gas permeability with mean pressure and type of gas (from Klinkenberg) ...............21: Figure 1....26: Figure 1........................ 32 Schematic diagram of DARCY’s experiment ......9: Figure 1..........................................................Link) ..........16: Figure 1.............................................. 37 Schema of permeability measurement for unconsolidated media (from Monicard) ......................................................................... and Yuster). Terwilliger.............1: Figure 1............................................................ 13 Packing of spheres and porosity (after P.N................................................................ 51 Schematic diagram of the PENNSTATE device for relative permeability measurements (after Morse.......... 39 Hassler type core holder (from Monicard)................................................30: Figure 1........... 54 Schematic diagram of a field segment.......... 48 Schematic diagram of a device for measuring relative permeabilities of the wetting phase (after Rappoport and Leas)....... 16 Pore compressibilities of rocks (after H......................... 25 Hysteresis of the capillary pressure curve ............................................ 18 Comparison of wetting to nonwetting fluid .......11: Figure 1....1: Definition of representative control volume for porosity measurements .... 50 Schematic diagram of a device for relative permeability measurements (after Hafford)......31: Figure 2.......................8: Figure 1.......19: Figure 1................. 26 Drainage and imbibition in capillary tubes....................................... 20 Modeling the porous medium as a bundle of cylindrical rods ...........20: Figure 1...................... 54 Typical water/oil relative permeability characteristic (after Craig) ......................................................... 19 Definition of the contact angle by Young ............ Jenks....................25: Figure 1....... 31 Bivariant pore radii distribution (from Dullien and Mehta) .....23: Figure 1.....24: Figure 1....................14: Figure 1................... 17 Pore volume compressibility factor in terms of overburden pressure (after I................. 27 The dimensionless capillary J function curve (after Leverett) ....................7: Figure 1............ 27 Equilibrium between gravity and capillary forces.....................9...................... 59 ..................5: Figure 1............... 34 Air permeameter: Schematic Flow diagram (after Monicard) .3: Figure 1. 20 Illustration of the principal radii of the curvatures ........... 24 Assembly with mercury pump for capillary pressure measurement (Purcellmethod)...................................................... 23 Schematic diagram of a diaphragm device for capillary pressure determination by drainage (after Welge and Bruce)...............................................................Fatt) ..................................4: Figure 1.......17: Figure 1.......................................................18: Figure 1.. 15 Sediment compaction and porosity (from Krumberlain and Sloss) ...................................13: Figure 1.................. 49 Schematic diagram of a device for relative permeability measurements (after Osoba).......... 41 Schematic diagram of a device for measuring relative permeabilities of the nonwetting phase (after Leas.............................................
.10: Figure 3.....2: Figure 4... 86 Illustration of steadystate filtration in a radial system ......12: Transformation of the coordinate system .............................List of Figures 8 Figure 2............................ 155 Capillary pressure and relative permeability functions used in the ..................................4: Figure 4.......... compressible fluid) . 109 Dimensionless pressure for single fractured well in an infinite acting system (after Gringarten............... 124 Superposition of several wells in a infinite reservoir ........................................ 117 Variable production rate in case of a ideal reservoir (after Hurst) ...............................10: Figure 4......... 146 Cumulative production by linear displacement (after Marle) ..........3: Figure 2.......................20: Figure 3...........3: Figure 4.......................13: Figure 3................... 96 Plots of pressure drop in the vicinity of a well (infinite reservoir.......9: Figure 3.......6: Figure 3............................12: Figure 3...................................8: Figure 3..... 132 Calculation of fractional curve (after Marle) .......................................................... 129 Pressure buildup curve near a discontinuity......17: Figure 3......................15: Figure 3........... 108 Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder). 152 Influence of the velocity of displacement on the distribution of saturation regarding the capillary force (by Douglas et al 1958)... 145 Determination of average saturation of the wetting phase after breakthrough (after Welge)....................function (after Chaumet)...........................7: Figure 4...............11: Figure 3..... 128 Production from a well near impermeable boundary (after Bear)...5: Figure 3....8: Figure 4.........................2: Figure 3............................................. 131 Production in the vicinity of a boundary with a constant potential (after Bear) ..... 155 Displacement in countercurrent.................21: Figure 4....... 144 The displacement front as discontinuity of saturation (after Marle) .............19: Figure 3..... compressible fluid) . 153 The displacing efficiency as a function of velocity (by Kyte......................7: Figure 3......................16: Figure 3......................................................................9: Figure 4........ 63 Volume element in a cartesian coordinate system......... 142 Propagation of saturation profile (after Marle) ..................................... 99 Solution for the infinitive and finite radial filtration problem with closed boundary and constant pressure drop (after Van Everdingen and Hurst) .................................... with two production wells.............................................. and Ragavan) .............................. 122 Pressure change at point R in infinite reservoir........... Ramey.................................................................. 126 Pressure buildup analysis plot (after Horner) ........ 68 Illustration of the boundary conditions.................................. 108 Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder)... 94 Plots of pressure drop in the vicinity of a well (infinite reservoir.........3: Figure 3.........4: Figure 3..................................................... 96 The flow rate in function of the dimensionless variable Kt/r2 (after Chaumet)...................... 107 Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder)......14: Figure 3............... 153 "Endeffect” in case of a wetting displacing phase (after Marle)......... 89 Plots of the Ei(z).............................................5: Figure 4............................................... Rappoport 1958)........................................ 154 Endeffect” in case of a nonwetting displacing phase (after Marle) ............2: Figure 2........1: Figure 3..4: Figure 3................................ 126 Application of the second law of superposition on a well with a variable production .. 87 Plots of production equation for gas wells ...................................................1: Figure 4..... 77 The Radial Coordinate System .....11: Figure 4...... 102 Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder).................................... 148 The influence of gravity on the fractional curve (after Marle).18: Figure 3..........................6: Figure 4.................
.................. ..................................................List of Figures 9 Figure 4...5: Figure 5.................................................................... 166 Influence of the mobility ratio on front propagations in case of a linear displacement ........... 170 Forces acting on the displacing front....................... (by Graham and Richardson) .....................7: Figure 5..........................2: Figure 5.......................................................6: Figure 5....................................... 158 Comparison of saturation profiles according to different mathematical models........................3: Figure 5.... 173 Supercritical displacement in inclined layer (after Marle) ........ 175 ............... 164 Schematic diagram of pistonlike displacement .......................4: Figure 5..... 171 Position of the displacing front by favorite mobility ratio (after Marle).............. 169 Possible positions of the displacing front in inclined layer.1: Figure 5...........................................................................8: calculation by Blair ................................... 172 Position of the displacing front by unfavorable mobility ratio................ 157 Recovery in case of linear counterflowing imbibition and the experimental determination of the influence of a certain in corelength..13: Figure 4...................14: Figure 5.................................... 157 Distribution of pressure and saturation in case of linear (counterflowing) imbibition (by Blair)..............
List of Figures 10 .
These so called pores may contain a variety of fluids such as air. a complex network can be formed which is able to carry fluids. Only these permeable and porous media are taken into consideration in this volume. • In Chemical Engineering: Porous medium is applied as filter or catalyst bed. dispersed within it in either a regular or random manner. oil etc. • In Hydrology: The porous medium is a water bearing and sealing layer. If the pores represent a certain portion of the bulk volume. water.1 Fundamental Properties of Porous Media A porous medium is a solid containing void spaces (pores). Various examples can be named where porous media play an important role or where the technology requires them as a tool. • In Petroleum Engineering: Porous medium (reservoir rock) stores crude oil and natural gas 111 . either connected or unconnected. • In Soil Science: The porous medium (soil) contains and transports water and nutrients to plants.
one should distinguish between: • consolidated • unconsolidated porous media.1 Porosity 1.1.1) where Vp Vs is the void volume (pore volume) and is the volume of the solid material. Porosity is defined as the ratio of pore volume to total volume.1 General Aspects and Definition Basically one must distinguish between two groups of porous media: • intergranular • fractured. Definition of Porosity The porosity of porous media is defined as the ratio of the volume of the pores to the total bulk volume of the media (usually expressed as fraction or percent). where: VT = Vp + Vs . which can be expressed as: V p VT – V s φ = . (1. Materials having both fractured and intergranular porosity are called dual (double) porous media. VT VT (1. concerning the mechanical properties.2) . In a consolidated porous medium the particles (grains) are held together by a cementing material.= . On the other hand. in the other type the grains are loose. Let us select any point of the porous media and its environment with a sufficiently large volume VT. A typical characteristic of a consolidated medium is the possibility to form shaperetaining samples.112 1: Fundamental Properties of Porous Media 1..
If the errors in measuring VT and VP are ∆VT and ∆VP then Eq.2 Determination of Porosity The determination of the porosity with sufficient accuracy is not a trivial issue.1: Definition of representative control volume for porosity measurements Porosity is a statistical quantity which depends on the magnitude of the total volume taken into consideration (see Figure 1.: VT = 109m3) the calculated porosity can deviate greatly from the true value. (1. since it contains the reservoir fluids.+ φ Vp VT where ∆φ is the error in calculating the porosity.1.1). 1.= .g. especially for small samples. On the other side if the volume is too large the porosity may deviate from the real value due to the influence of heterogeneity. If the selected volume is too small (e. therefore the volume VT should be large enough in order to obtain the statistical average of porosity. The storage capacity of a reservoir rock always depends on the effective porosity. 1.1: Fundamental Properties of Porous Media 113 Basically one must distinguish between two kinds of porosities: • Total porosity (isolated pores are considered also) and • Effective porosity φ eff (effective in the sense of fluid transport).2 will lead to: ∆V p ∆V T ∆φ .3) . Figure 1.
26 0.1: Relative errors in measuring porosity ( ∆V T = ± 10 [ m ] ) φ[%] VT[m3] 2. 1.10 0.37 0.1 ×10 –3 = 5. φ VT φ (1.70 0.09 1 4 12 20 –9 3 The following quantities are necessary in order to calculate the porosity based on Eq.114 1: Fundamental Properties of Porous Media Assuming that ∆V p = ∆V T = ∆V then Eq.4 103 kg when it was saturated with brine.25 2.1 103 kg.20 12.13 3. and 10.1.4) According to Eq. Assuming an error ∆V=109 m3 for the values of V.80 0. Example 1.109 50. It is completely saturated with brine having a density of 1200 kg/m3. 1. .18 0.1: A core plug has a radius of 1.50 0. • the solid volume (Vs). Solution: Weight of brine in the plug w = 10.60 6.4.40 4.30 0.50 3. • the void volume (Vp).50 13.30 0.2: • the total volume (VT).50 0. and 5.60 0.00 4 8 16 32 64 25. Calculate the effective porosity of the core plug.3 can be written as: ∆φ 1 .16 6. the relative error of the porosity measurement depends on VT and φ.0 102 m in length. The dry core plug weighted 5. 1.25 102 m.3 ×10 kg –3 .10 1. this error can be up to 50% as shown in Table 1.00 1. Table 1.25 1.4 ×10 –3 – 5.= ∆V 1 + .
2: Packing of spheres and porosity (after P. Compaction . Figure 1.3 Compaction Figure 1.2 ×10 –3 Bulk Volume of plug VT = r πh = 24.18 or 18% VT 2 –6 Porosity of plug 1.2 shows porous media built with spheres of equal size. The compaction .1: Fundamental Properties of Porous Media 115 Volume of brine (pore volume) –6 w 5.5 ×10 Vp φ = .= 4.is irreversible.42 ×10 3 ρw 1. no characteristic factor has been introduced yet to describe the compaction as a property. However.3 ×10 V p = .K.= .Link) .of a sediment depends on the greatest depth a rock reached during its genesis. The spheres are ordered in three different ways to illustrate the effect of compaction on the porosity of a pack.in contrast to the compressibility .1.= 0. Figure 1.3 shows the porosity of clay and sandstone as a function of depth.and thus porosity .
elastic and thus compressible medium.only minor though .116 1: Fundamental Properties of Porous Media 1.6) where φ0 is the porosity at the pressure p0.3: Sediment compaction and porosity (from Krumberlain and Sloss) However. Figure 1.1. cφ is small and normally regarded as a constant. The pore volume alteration during the pressure drop in the reservoir has its source in the elasticity of the solid. .c φ = . Therefore cφ will be a function of porosity. . Change of pressure inside the pore space during production also affects the porosity. (1. .4 Compressibility of Porous Media Reservoir rock is not considered to be a rigid system but as a .4 illustrates this relation. Figure 1.5) ≈ φ 0 [ 1 + cφ ( p – p0 ) ] . The isothermal compressibility of porosity is defined as: 1 ∂φ . φ ∂p T Integration of the preceding equation leads to: φ = φ0 e cφ ( p – p0 ) (1.
which is illustrated in Figure 1.1: Fundamental Properties of Porous Media 117 Figure 1.5.4: Pore compressibilities of rocks (after H. .Hall) The compressibility of the pore space is influenced by overburden pressure too.N.
5: Pore volume compressibility factor in terms of overburden pressure (after I.Fatt) .118 1: Fundamental Properties of Porous Media Figure 1.
2 Capillary Properties 1.8) 1. The drop of liquid will take a certain shape due to the interfacial tensions acting on it. i (1.6: Comparison of wetting to nonwetting fluid The contact angle is used as a measure of Wettability. which are: the interfacial tension between fluid 1 and 2. If the contact angle is larger than 90°.2 Wettability A fluid drop on a plane solid surface can take various shapes.1: Fundamental Properties of Porous Media 119 1.2. .7) ∑ Si = 1 .. Young. The respective shape (either flat or shaped like a pearl) depends on the wettability of the considered solid. 150 years ago. Effective pore volume of the porous media Summing the saturations results in: (1. between a drop of liquid and a plane of a solid surface. In the case of a wetting fluid. then the fluid is referred to as nonwetting. the contact angle is smaller than 90°.1 Saturation Basically pore space may contain several phases. defined the contact angle as a consequence of the static equilibrium. Figure 1.2. Figure 1. for air and mercury the air. In case of air and water the water is the wetting phase. The saturation of a certain phase is defined as: Volume of phase i in the porous media Si = .6 illustrates that property. σ12 σs1 and σs2 the interfacial tensions between solid and fluids.
11) Figure 1.9 and Eq.g.7.03 N/m.073 N/m and between oil and water about 0.2. 1. can be calculate by the Laplace equation: 1.1. At room temperature the interfacial tension between water and air is 0. 1. the pressures at both sides of the fluid interface are not equal. called capillary pressure (Pc).7: Definition of the contact angle by Young From Figure 1.120 1: Fundamental Properties of Porous Media Fluid 2 σ12 σS1 Solid σS2 Fluid 1 Figure 1. = p nw – p w r 1 r2 (1. P c = σ 12 . This pressure difference.10) Interfacial tensions and thus θ are regarded as temperaturedependent.9) (1.+ . Eq.8: Illustration of the principal radii of the curvatures . oil and water).10 will result: σ s2 – σ s1 cos θ = σ 12 σ s1 + σ 12 cos θ = σ s2 (1. 1.3 Capillary Pressure For any two immiscible fluids (e. When regarding oil reservoirs it is necessary to consider the specific rock and fluid properties in order to determine whether the reservoir rock is water or oil wet.
then Fluid 1 spontaneously intrudes into the pore space. it is easy to calculate the porosity and water saturation of the wetting fluid as follows: π φ = 1 – 4 (1. In order to illustrate this capillary equilibrium in a simplified manner. . – . asin . until equilibrium is achieved.9). one could imagine the porous medium as a bundle of infinitely long glass rods with uniform radius R (see Figure 1.13) Based on the geometry of the cylindrical rods Figure 1. 1. This is called imbibition.9 the capillary pressure can be calculated from Eq. Fluid 1 will occupy the smaller pore spaces first.11 as: σ 12 P c = r where r = r 1 which is the radius of the water air interface. – acos . Equilibrium is achieved when Fluid 1 uniformly occupies pore spaces with greatest possible interfacial curvature. 2 + 2 .14) . Fluid 1 will displace Fluid 2.8 illustrates the principle curvatures radii r1 and r2 which are elements of the Laplace equation. (1. If a porous medium is completely (100%) saturated with the nonwetting Fluid 2 and contacted by the wetting Fluid 1. If air is the nonwetting and water is the wetting fluid. Thus in the case of imbibition.12) and the water saturation can be calculated from: 4 S 1 = 4–π r R R r r 2 . Since one of the two principle radii of curvature (r2) is infinite. R R r + R r + R R (1. then σ s2 = 0 and thus σ s1 = σ 12 and cos θ = 1 .1: Fundamental Properties of Porous Media 121 Figure 1.
10. Let pw be the pressure in the wetting phase and pnw in the nonwetting phase. then the capillary pressure is defined as follows: p nw – p w = P c ( S w ) (1.16) .9: Modeling the porous medium as a bundle of cylindrical rods Since S1 and Pc are functions of r.122 1: Fundamental Properties of Porous Media Figure 1. the following is also valid: Pc = Pc ( S1 ) (1.15) For a bundle of rods having a radius of R = 7.3 105 m and σ12 = 0.037 N/m the relationship between the capillary pressure and S1 is shown in Figure 1.
1: Fundamental Properties of Porous Media 123 Figure 1. The sample (core) and the diaphragm are both saturated with the wetting fluid. The saturation can be calculated with the help of material balance.10: Capillary pressure versus saturation of the wetting phase for the model in Figure 1. Consequently this applied pressure equals capillary pressure which corresponds to the given saturations. One of the devices used for measurement of the capillary pressure curve is illustrated in Figure 1. The upper and lower cells are separated by a diaphragm.g. air. e.9 1. until capillary equilibrium has generated a constant saturation. The nonwetting fluid surrounding the core.11.4 Measurements of Capillary Pressure in a Porous Medium The various kinds of measurement methods are all based on the same principle: A constant pressure is exerted on the porous medium. The porous medium is placed into the upper chamber. which is impermeable to the nonwetting fluid.2. . is then set under constant pressure and is thus pressed into the porespace of the core.
The original devise used by Purcell is shown in Figure 1. Nitrogen is used for applying a constant pressure to force Hg into the core.124 1: Fundamental Properties of Porous Media Figure 1. This procedure enables the capillary pressure to be determined as a function of the Hgsaturation. then the volume indicator of the pump is set to zero. After completing the described drainage and measurement procedure. It is essential to pay attention to the fact that capillary equilibrium should be achieved at every pressure step so the displaced volume is no longer a function of time. it will never come to a complete displacement of the wetting fluid. However. . displaced by the non wetting fluid. then the cell is filled with Hg up to the level indicator (nonwetting fluid Hg will not be imbibed into the pore space of the core). The so called connate water saturation always remains. A constant Hg level is held by the pump. This procedure is then repeated several times at higher pressures.11: Schematic diagram of a diaphragm device for capillary pressure determination by drainage (after Welge and Bruce) The wetting fluid. It is easier to measure the capillary pressure using mercury as a nonwetting fluid. flows through the diaphragm into a graduated pipette to be measured. it is possible to establish the capillary pressure curve for imbibition by reducing the pressure stepwise inside the upper chamber. At each pressure step the Hgvolume is measured which has to be pumped into the cell in order to maintain the level.12 with the following procedure: The core is evacuated inside the cell.
then a certain pressure must be applied to enable the nonwetting phase to intrude the pore space. Another method uses a centrifuge to determine the capillary pressure curve. Due to the necessity of capillary equilibrium for a correct reading of the injected volume.12: Assembly with mercury pump for capillary pressure measurement (Purcellmethod) . Figure 1.g. If the core is saturated completely with the wetting phase (e. Final pressures may range from 1 to 100 MPa. Initially the core is saturated with the wetting fluid and the volume extruded is dependent on the rotational speed (up to 20 000 rpm).1: Fundamental Properties of Porous Media 125 Calibration of the device is accomplished with the empty cell.13 shows a typical capillary pressure curve. Figure 1. applying a wide range of pressure steps. This pressure is the threshold pressure which depends on the largest pore diameter. water) at the beginning of the measurement. a measurement may take from hours to several days.
empty capillary tubes are placed into a tub filled with a wetting fluid.13: Hysteresis of the capillary pressure curve The capillary pressure curve will be asymptotic to a certain limit value of saturation . It makes no difference in the measurement of Pc if (i) the capillary tube is filled first with the wetting fluid and placed in the tub (drainage). the menisci will rise against gravity until equilibrium between gravity and capillary force is reached.13). . This appearance (hysteresis) may also be explained using capillary tubes. then the height of the meniscus will depend on the saturation process (see Figure 1. in case of imbibition. This means that the porous medium is imbibing the wetting fluid immediately.15 . In the case of an uniform radius of the capillary tube (see Figure 1. the height of the meniscus of a wetting phase above the contact level will be independent of the displacement process. Thus the capillary pressure curve appears as a hysteresis (see Figure 1. This procedure of a wetting fluid being displaced by a nonwetting fluid is called drainage. the nonwetting fluid with which the core is initially saturated will be displaced spontaneously by the wetting fluid.the so called connate water saturation. If .as graphically illustrated in Figure 1. or (ii) an empty tube is placed in the tub and the wetting fluid enters the tube (imbibition).15). On the other hand.126 1: Fundamental Properties of Porous Media Figure 1.14). If this procedure is repeated with capillary tubes having sequentially different diameters.
P c ( S w ) = h ( S w ) ( ρ w – ρ nw )g ..1: Fundamental Properties of Porous Media 127 Figure 1.. Figure 1. r (1.17) In addition.19) .15: Equilibrium between gravity and capillary forces Thus at equilibrium the following is valid (for a single capillary tube): 2σ cos θ P c = h ( ρ w – ρ nw )g = .18) (1. Correspondingly the capillary pressure function may be expressed in terms of h = h(Sw).15 illustrates that Sw is a function of height h if the porous medium is regarded as a bundle of capillary tubes with equal length. but different diameters.14: Drainage and imbibition in capillary tubes Figure 1. ( ρ w – ρ nw )g (1. Therefore the saturation may be determined as a function of the vertical distance from the Pc = 0 plane. Pc ( Sw ) h ( S w ) = .
⋅ PcL cos θL σ L where PcR PcL σR σL θR θL (1. Table 1.028N/m θw/o = 33° to 55° σHg/a = 0. is the contact angle measured under reservoir conditions. 1. is the interfacial tension under reservoir conditions.2: Interfacial tension and contact analyses In the Laboratory σw/a θw/a = 0. The conversion is based on Eq.2 shows some interfacial tension values measured in the laboratory and estimated values in reservoir.2. 1.20) is the capillary pressure under reservoir conditions.5 Conversion of Laboratory Data In Order to use capillary pressure data measured in the laboratory for capillary pressure determination under reservoir conditions. a proper conversion of this data should be done at first. Pc(Sw) is the capillary pressure of the wetting phase (e.20 cos θ R σ R P cR = . Table 1. is the interfacial tension measured under laboratory conditions.48 θHg/a = 140° . is the contact angle measured under laboratory conditions.128 1: Fundamental Properties of Porous Media where ρw ρnw g is the density of the wetting phase. is the capillary pressure measured under laboratory conditions. is the gravitational constant.g.07 N/m = 0 In the Reservoir σw/o = 0. is the density of the nonwetting phase. water).
07 cos θ L σ L P cR 5850 h = .= .⋅ 20600 = 5850Nm 0.1: Fundamental Properties of Porous Media 129 Example 1.= 2m ( ρ w – ρ o )g ( 1100 – 800 ) ( 9.71 ⋅ 0. if the laboratory measurement of Pc using air/water system is 20600 N/m2 at Sw = 0.7. Use Table 1.81 ) .028 P cR = .2 and the following data: ρo ρw θR = 800 [kgm3] = 1100 [kgm3] = 45° Solution: cos θ R σ R 2 0.7 and above the level of Pc = 0.2: Calculate the height of a transition zone at Sw = 0.⋅ PcL = .
is the contact angle. is the interfacial tension between fluid 1 and 2.k J ( S w ) = .16: The dimensionless capillary J function curve (after Leverett) .130 1: Fundamental Properties of Porous Media 1. is the porosity. σ 12 cos θ φ where Pc σ12 θ k φ (1.2..6 The Leverett Function According to Leverett the capillary pressure curves of core plugs . Figure 1.obtained from samples of the same formation . is the permeability..21) is the capillary pressure. These investigations resulted in the dimensionless Jfunction (see Figure 1.16) which is given by: Pc .may be correlated with other properties.
The extent to which the distribution curves of the two methods diverge is illustrated in Figure 1. This function is a distinctive property of reservoir rocks and is therefore very often used to characterize a porous medium. 1. The function shown is similar to the function of cumulative frequency distribution in statistics. Figure 1. Characteristic for this method is the fact that the pore radius is measured directly. and the other one determined by photography of thinsections.17 shows the measured non wetting fluid saturation Se(Pc) versus the effective pore size re(Pc) calculated from Eq. Again consider a bundle of equally long capillary tubes. which corresponds to a specific saturation as indicated by Eq.1: Fundamental Properties of Porous Media 131 1.17: Non wetting fluid saturation versus the effective pore size distribution Another way to determine the poreradii distribution is by means of statistical methods (e.22) Where r is defined as the smallest radius which has to be filled with the nonwetting fluid in order to reach a certain part of the pore space.18. It is customary to use the terms re (pore entry radius) and Se (effective saturation).17 indicates that each capillary pressure value can be related to a certain radius.15. 1. from thin ground sections). . Figure 1. then it is possible to set up a Vi = Vi(re)function using the Purcellmethod: 2σ cos θ P c ( r ) = .7 Pore Size Distribution Eq.2.22. 1.g. which shows the pore volume distribution function determined by the centrifuge method (the curve with sharp peak).= P c ( S Hg ) r (1. so the circular crosssections of the capillary tubes correspond to a certain partition function.
132 1: Fundamental Properties of Porous Media Figure 1.18: Bivariant pore radii distribution (from Dullien and Mehta) .
only a portion of the mass particles flows and the remaining part forms the flooded framework.1: Fundamental Properties of Porous Media 133 1. First investigations regarding filtration date back to the year 1825. Results of this experiment were unfortunately lost. Henry Darcy proved this hypothesis in the year 1856 using the equipment illustrated in Figure 1. Muskat and Reed suggested to give the unit of permeability the name Darcy. is the crosssection of the packing. He measured the time elapsed until the water level had resumed a certain height.⋅ h .3. L where K S L h (1. From the results he deduced that the pressure drop caused by the filter was proportional to the velocity of filtration.. Botset. is the length of the packing.⋅ .24) where: A is the cross section of the porous medium perpendicular to the direction of flow . Darcy found that the flow through the sand filter corresponded to the following formula: S Q = K ⋅ . 1.19. whereas in the case of filtration. Velocity of filtration had been 5 [m3m2] per day. Investigations were made with various sand gravel packs. In the same year Wyckoff. At the first World Oil Congress in 1933 the permeability was defined by Fancher.1 Darcy’s Law In the year 1854 Dupuit made experiments with urban water filters in London. From that time on the equation is called Darcy’s Law: k ∆p q = A ⋅ .23) is the permeability coefficient. and Barnes. Lewis.3 Permeability It is important to distinguish between mass flow and filtration: In the case of mass flow all particles in the field of flow are in motion. is the difference in piezometric head. µ L (1. Chaumont had the idea of diging a trench parallel to the river Garonne (in Southern France) and then to dewater this trench using an Archimedian screw.
[ atm ] k = . if at a 1 cm2 crosssection a fluid with 1 cP viscosity flowing with a rate of 1 cm3/s will cause a pressure drop of 1 atm/cm: ] µq ∆P [ cp ] [ cm / s .17 is valid for a laminar and steadystate onphase flow through a porous medium only. 2 A L [ cm ] [ cm ] By using the SIunits 3 (1. Figure 1. Moreover.= ./ ./ . the fluid has to be largely incompressible. 1.= 1Darcy .25) .2 Definition and Units of Permeability Even today it is customary to use DARCY [D] as a unit of permeability.19: Schematic diagram of DARCY’s experiment 1.3. The permeability of a porous media will be 1 Darcy.134 1: Fundamental Properties of Porous Media k L ∆p µ is the permeability as a material property of the porous medium is the length of the porous media in direction of flow is the pressure difference along the porous medium is the viscosity of the flowing fluid The Darcy’slaw in Eq.
3: A cylindrical core having a radius 2.48 ×10 Exercise: Calculate the permeability of a core plug from the following test: • Flow rate • Inlet pressure = 2.104 [m2] = 0.= [ m ] 2 [m] [m ] The relationship between the two units is: 1 Darcy = 0.s] .3 [m]. Solution: From Darcy’s law k ∆p .106 [m3s1] = 5 [bar] – 13 –6 –2 3 –1 –2 3 [ m ] = 0. the differential pressure across the core was 10 [bar].s].q = A ./ .002 [Pa. was flooded with brine at a steady rate of 1.148 [ D ] 2 • Outlet pressure = 1 [bar] • Length of core • Area • Viscosity = 0.0.001 [Pa. Assume brine viscosity 0.001 [ Pas ] .1: Fundamental Properties of Porous Media 135 [ Nsm ] [ m s ] [ Nm ] 2 . Calculate the absolute permeability of the core.3 ×10 [ m ] 10 [ bar ] k = 1.µ L µ Lk = q .1 [m] = 1.106 [m3s1].54 102 [m] and length of 0.A ∆p 1 ×10 [ m ] 0.× –4 2 s 20.× .987 x 1012 [m2] Examples 1.3 [ m ]k = .
Since gas is the nonwetting fluid it does not alter the original state of the core and the measurements can be repeated. (1.27) . the gas commences to flow through the sample.mostly formed cylindrically (core) . the ideal gas behavior is assumed. Usually gas is used to measure permeability of core samples instead of liquids. is the outlet pressure.( p 1 – p2 ) . and thus permeability may be calculated using Darcy’s law for ideal gas as follows: kA q = – .3.26) is the inlet pressure.02 to 0. It can also be used for low permeability cores where a higher pressure difference is required.05 [m] • cubes with 0. Samples from a consolidated media can be shaped as regular geometrical forms: • cylinders with a diameter ranging from 0. then: qp = q a p a . 2 is the dynamic viscosity of gas at the average pressure p To convert the flow rate q to a flow rate measured under atmospheric pressure (pa). is the permeability of the core. µL where: p1 p2 k q µ (1.. The rock sample . p1 + p2 is the flow rate of gas at average pressure p = .136 1: Fundamental Properties of Porous Media 1.02 [m] length of the side.is fixed in the device with a sealing rubber gaiter. By applying constant pressure.3 Measurements of Permeability The method depends on the following factors: • consolidation of the medium • core size • fluid properties • the applied pressure.
29) (1. A (p 2 – p 2) 1 2 (1.28) (1. 1. From Eq. 2 µLp a From Eq.29 the permeability ka can be written as: – 2q a µL pa k a = . 1. Figure 1.28 into yields: –k a A p1 + p2 q a = .20: Air permeameter: Schematic Flow diagram (after Monicard) ..21. The device may be described as a cylinder inside which a porous medium is positioned between two lattices see Figure 1.19).27 the flow rate q can be written as: q a pa q = ..( p 1 – p 2 )  . 1.30) where qa is the flow rate at atmospheric pressure pa.1: Fundamental Properties of Porous Media 137 where: qa is the flow rate of gas at atmospheric pressure pa. p Substituting Eq.. The prearrangement for unconsolidated media is very similar to the equipment of Darcy (see Figure 1.
• compatibility of large pressure gradients. • permeability measurement in two directions. Permeability measurement in a vertical direction presupposes the casing of the core by an impermeable rubber gaiter. In doing so air. water or oil can be pressed through the core in an axial direction. The Hassler core holder is a common device. The fluid is then pressed horizontally through the lattice and the core. This enables the core to remain permeable in these areas when cased by the rubber gaiter. Advantages of the Hassler core holder are: • excellent sealing. Hereby the permeability measurement is possible both in horizontal and vertical direction (see Figure 1. • optimal selection of core length. The liquid is either injected with constant gas pressure or by use of a micropump.138 1: Fundamental Properties of Porous Media In order to measure the permeability of very compact media using a liquid. high pressure gradients must be applied.22). • independency from the fluid used. Then the area of the cylindrical surface is covered both at the inflow and outflow opening to one quarter with a lattice. . Horizontal measurement makes a sealing of the top surfaces of the core necessary.
1: Fundamental Properties of Porous Media 139 Figure 1.21: Schema of permeability measurement for unconsolidated media (from Monicard) .
22: Hassler type core holder (from Monicard) .140 1: Fundamental Properties of Porous Media Figure 1.
1: Fundamental Properties of Porous Media
141
1.3.4
Klinkenberg Effect
Klinkenberg noticed that the permeability of gas is not the same as for liquids and in addition the gas permeability depends on pressure. The correlation between gas permeability ka, liquid permeability k1 and mean pressure inside the core, pm is given by Eq. 1.31. The parameter b depends on the gas used. b k a = k l 1 +  p
m
(1.31) (1.32)
Figure 1.23: Variation in gas permeability with mean pressure and type of gas (from Klinkenberg) Physically the Klinkenbergeffect may be explained by the phenomenon of surface slipping of gases caused by the Brown motion. This surface slipping decreases with increasing pressure. However, at low pressure this effect is responsible for the deviation of the gas permeabilities.
142
1: Fundamental Properties of Porous Media
1.3.5
Analogies between the Laws of Darcy, Ohm and Fourier
The following is a comparison of different laws which have a similar form to DARCY’s law: Darcy’s law k ∆p  q = A  µ L Ohm’s law of electrical current: 1U  J = A  ρL Fourier’s law of heat conduction:
∆T Q = Aλ  , L
(1.33)
(1.34)
(1.35)
where: A J k L Q q U ∆p ∆Τ µ λ ρ is the cross section [m2] is the amperage [A] is the permeability [m2] is the length [m] is the rate of heat [Js1] is the flow rate [m3s1] is the voltage [V] is the difference in pressure [Pa] is the difference in temperature [°C] is the dynamic viscosity [Pa.s] is the thermal conductivity [Wm1 °K] is the electrical resistance [Ωm]
The form of Eq. 1.33, Eq. 1.34 and Eq. 1.35 is similar. These analogies enable a simulation of filtration processes with the help of electrical models and the adoption of mathematical solutions obtained from heat flow problems.
1: Fundamental Properties of Porous Media
143
Table 1.3: Comparison between the laws of Darcy, Ohm, and Fourier Darcy Flow Rate u[m s ] Permeability coefficent
k K = µ
3 –1
Ohm Amperage J[ A] Electrical conductivity 1 γ = ρ 1Ωm Voltage
U[V]
Fourier Heat conduction rate Q [ Js ] Thermal conductivity λ WmK Temperature difference
∆T [ K ]
–1
[ m Pa s ] Pressure difference
∆p [ Pa ]
2
–1 –1
1.3.6
Filtration Velocity
The velocity of filtration is defined as a fluid volume q flowing through the surface A of a porous medium within unit time:
–1 q u =  [ ms ] . A
(1.36)
u is an arithmetical quantity. In comparison to the actual velocity of flow in pore channels, great differences can be recognized. However a statistical average is easily calculated as follows: qu v =  =  , Aφ φ where φ is the porosity and v the displacement velocity. If a fluid at a velocity of u = 1 [mday1] is injected into a porous medium with a porosity of 0.1, a specific fluid particle will be transported within a distance of 10 [m] in one day. (1.37)
Therefore the correction factor may often be significant.43) (1. However.= αu + bu . Further the velocity of mass filtration Qm may be introduced as the product of filtration velocity and density: ρu = Q m ⁄ A . 1. where Qm is not too large..38) If bu is small . b can be defined as: βρ b = – b’ρ = – .144 1: Fundamental Properties of Porous Media 1.39) where b..41) becomes negligible in oil reservoirs. .38 will be reduced to the Darcy formula and α can be calculated by: µ α = – .then Eq. Then Eq.38 can be written as: βQ m µQ m ρdp = – . The unit of ß is [m] and is considered as a characteristic length of the porous medium.3.in comparison to the term of first order . 1 +  dx Ak µA The order of magnitude of ß ranges at 105[m]. in gas reservoirs Qm is large in the vincinity of wells. dx 2 (1. where A is the crosssection. Therefore the correction factor: βQ m µA (1. respectively ß. k In consequence of experiments.7 Quadratic Equation of Filtration Darcy’s law is only an approximation of the general equation of filtration given by: 2 dp .40) (1. 1. k (1.42) (1. is independent of the fluid properties.
he observed that at a constant rate the pressure gradient is larger in cases of gas liberation.06 [m] in diameter. The so called Hassler core holder was already described in Figure 1. in which he simulated an oil reservoir with gas drive by using a tank filled with sand and gassaturated oil at a pressure of several bars. From the results of this experiment. Thus a two phase flow causes an additional pressure drop for the flowing system. In doing so. the slogan “more wells. Though this theory was greatly contested . An overestimation of the JAMINeffect led Herold to the opinion that a given pressure gradient can only set a certain number of oilgas interfaces (gas bubbles) in motion. The retention of the wetting phase at the outlet is called end effect and was also discovered by Hassler.20 [m] long and 0. In 1926 Becker and Parkhurst had the idea to consider oil and gas in a reservoir as liquid and gaseous phases of the hydrocarbon system being under thermodynamic equilibrium. Hassler observed that it is essential to determine the permeability not only for dry cores but for cores with any saturation. In the year 1927 Uren conducted experiments.25. Uren recognized that on one side the detached gas reduces the permeability. It lasted until the year 1945 when this phenomenon was understood. oil and gas in a hydrocarbon reservoir. First the cores were saturated with oil derivates which were then displaced by air. Brunner and Deahl were convicted that a gas bubble was not able to plug a porous medium due to the existence of various cross linkings between the pore channels and the possibility of the gas molecules diffusing through the oil. In the same year Power made experiments with a sandfilled pipe 2. He injected saturated oil slightly above bubble point pressure into the pipe.especially by Versluy’s and Lewis. The model used by Wyckoff and Botset was 3 [m] long and 0.22. The knowledge that finally led to a systematic research of multiphase filtration in porous media originally referred to the gas/oil ratio. They . but on the other side provides the energy needed to mobilize the oil. In the year 1936 Hassler made experiments which contributed greatly to the solution of multiphase problems. However. more oil” was accepted in practice. Hassler.4 Relative Permeabilities It was not until 1917 that Lewis discovered the coexistence of water. Hassler measured the permeability of air and determined the saturation by weight control of the core. This was possible due to the advantage of a quick demount and mount of the core. Brine was used to ensure measurability of the electrical conductivity as shown in Figure 1.05 [m] in diameter. this would mean that the drainage area of a well was limited.1: Fundamental Properties of Porous Media 145 1. Wyckoff and Botset conducted research work with brine and carbon dioxide in the same year as Hassler made his investigations. Power concluded that the additional pressure loss was due to separated gas bubbles caused by the JAMINeffect.
This definition is valid for krnw. k 0 ≤ k rw ( Sw ) ≤ 1 . He introduced not only the capillary pressure into the equations of multiphase filtration. (1.44) (1.µ nw L where: krw is the relative permeability of the wetting phase krnw is the relative permeability of the nonwetting phase.45) The same indices are also valid for the flow rate q and the viscosity µ. all efforts were made to extend the validity of the Darcylaw to multiphase filtration. but together with Buckley he also set up the theory of frontal displacement which will be covered in Chapter 4.1 Definition of Relative Permeability As the historical review showed. (1. The relative permeability of the wetting phase is defined as: kw ( Sw ) k rw ( S w ) = . Finally. then the following formulas may be set up: kk rw ∆p .µw L kk rnw ∆p .146 1: Fundamental Properties of Porous Media measured the pressure at 10 equidistant points determining the permeability both of the wetting (kw) and nonwetting (knw) phase as a function of saturation. is determined by the gas/oil ratio. The variation of viscosity and saturation of the samples. Leverett had conducted several experiments concerning the relative permeability in case of a fluid/fluid system and then extended these to threephase wateroilgassystems. If this is possible. . They made the observation that saturation may be independent of the absolute permeability at a given gas/oil ratio.q nw = – A . at steadystate flow.q w = – A . 1.. also.4.46) where kw is the effective permeability of the wetting phase.
1: Fundamental Properties of Porous Media
147
1.4.2
Relative Permeability Measurements
The following three methods can be applied: • The Hasslermethod: The principle behind this method is the ability to adjust the phase pressures and the filtration velocity independently by the use of one or two diaphragms. • The PENNSTATE method: Two fluids are injected simultaneously at constant rates. • The Welge method: The principle is the calculation of the relative permeability using the results received from the displacement experiments.
1.4.2.1
The HASSLER method
Leas, Jenks and Russel used the Hassler device in their instrumentation as shown in Figure 1.24. Hasslercore holder contains a core (C) placed on a diaphragm (D). The diaphragm is an artificial porous medium exhibiting a relatively large capillary threshold pressure. It is fully saturated with the wetting phase to prevent the flow of the non wetting phase downstream. Since the diaphragm is completely impermeable to the non wetting phase, it is perforated in order to inject gas into the system. The wetting phase can be considered as static under equilibrium conditions. This equilibrium will be achieved if the pressure drop along the core keeps balance to gravity. The pressure can be regulated, so the saturation distribution can be regarded as homogeneous. The pressure drop is measured along the core, and the wetting phase pressure is regulated by a mercury flask. The saturations are directly marked by the meniscus m. In doing so the relative permeability of gases is determined. The wetting phase remains immobile being in capillary equilibrium with the flowing nonwetting phase. Using the device by Rappoport and Leas (see Figure 1.25) it is possible to measure the relative permeability of the wetting phase. A significant distinction from the method described above is the use of two diaphragms between which the core is positioned. The desired pressure of the nonwetting phase can be fixed. The wetting phase circulates from top to bottom. In order to obtain a homogeneous saturation distribution in the core, it is essential to regulate the flow velocity so that the pressure drop approximately becomes equal to the hydrostatic pressure difference.
148
1: Fundamental Properties of Porous Media
Figure 1.24: Schematic diagram of a device for measuring relative permeabilities of the nonwetting phase (after Leas, Jenks, and Rassel)
1: Fundamental Properties of Porous Media
149
Figure 1.25: Schematic diagram of a device for measuring relative permeabilities of the wetting phase (after Rappoport and Leas) The Hassler method is considered theoretically adequate and efficiently reproducible. Summarized, the essence of this method is that only the relative permeability of the wetting phase will be measured, whereas the nonwetting phase is immobile. As an example of the efforts of many scientific groups to develop a modification of the Hasslermethod  in order to measure the relative permeability of both phases some efforts were made by Osoba as shown in Figure 1.26. The method used by Hafford as shown in Figure 1.27 which can be considered as a simplification of the instrumentation of Osoba.
150 1: Fundamental Properties of Porous Media Figure 1. Although Hafford measured the saturation by weighing. since demounting the core for weighing (being under high pressure during the measurement of relative permeability) causes the expansion of the fluids. Therefore it may be possible that certain fluid quantities will be displaced from the core. . the accuracy of this experiment is questionable. Loomis. it is possible to run the experiment without a diaphragm on the outlet side. Crowell and Richardson investigated this problem and verified the reproducibility of the Hassler method by neglecting the amount of fluids displaced from the core during demounting.26: Schematic diagram of a device for relative permeability measurements (after Osoba) If the flow rate is high enough to prevent capillaryendeffects (explained in Chapter 4).
In many cases it is referred to as the dynamic steadystate displacement method. Figure 1. therefore the name PENNSTATE.2 PENNSTATEMethod This method was developed by Yuster and colleagues at Pennsylvania State University.2.28 presents the instrumentation used by Morse. Terwilliger.4.28: Schematic diagram of the PENNSTATE device for relative permeability measurements (after Morse.27: Schematic diagram of a device for relative permeability measurements (after Hafford) 1. and Yuster) .1: Fundamental Properties of Porous Media 151 Figure 1. The basic principle behind this method is the following: Figure 1. Terwilliger and Yuster to measure relative permeability.
The water and oil filtrated simultaneously through this artificial medium. since conductivity not only depends on saturation but also on the fluid distribution.free of turbulences. Likewise Yuster determined the saturation by weighing. The filtration processes will only coincide with the results of the Hasslermethod. 1.2. little spheres were packed to simulate a porous medium. Between two transparent slices. The core placed inside the holder has been fitted between two porous packings which should provide both satisfactory mixing of fluids and the prevention of capillary end effects at the inlet and outlet of the core. Problems arise concerning the application of this method in connection with the development of capillary contact between the core and the two packings. . if each phase individually forms some channel system of its own and maintains this. through the porous medium. The film verified that water and oil form their own flow channel systems in which the wetting fluid occupies the smaller pores. 1.3 Saturation Distribution and Relative Permeability The main difference between the Hassler method and the PENNSTATE principle is the fact that the Hassler method considers the movement of only one fluid.152 1: Fundamental Properties of Porous Media At first the core is embedded in a plastic tube in order to achieve the indispensable sealing. Determination in terms of measuring the electrical conductivity has not been very successful.4. Further it was noticed that these channels maintain their position. But in case of dynamical methods both phases are in motion. The flow has been laminar and . Increasing oil saturation affected a growing number of oil channels and a decrease of water channels. Therefore it is evident that the mobile phase forms continuous paths.in spite of great tortuosity . The motion of the fluids had been filmed and magnified. In consequence the fluids should not flow alternately through the same channels as small drops or larger filaments. The so called APIresearch project 47B was carried out at the University of Oklahoma. called channels. It is also difficult to guarantee the homogeneous distribution of the phases.3 WelgeMethod This method was at first developed by Welge and it is based on the evaluation a continuous two phase displacement. The fluids are injected at a constant rate by two micropumps.4. two experiments were conducted in order to analyse the nature of phase distribution in porous media. In 1949 and 1950.
but also depends on the saturation distribution (in consequence of the structure of the pore channels and the wettability of the porous medium). The nonwetting fluid was colored synthetic resin. Figure 1. Basically it must be distinguished between two kinds of displacement: • the wetting phase is the displacing fluid (imbibition) • the wetting phase is the displaced fluid (drainage) The phase whose saturation has been increased after displacement is always considered the displacing phase. Just as supposed the relative permeability is not only a function of saturation.29. The method also proved the theory of separated flow through the pore channels. As the wetting fluid. The preparation of the sample by grinding off progressively. After transforming the fluids into a solid state. the saturation distribution due to the relative permeabilities was made observable. AMOCO also made experiments on sand packing as a porous medium.1: Fundamental Properties of Porous Media 153 Simultaneously the distribution of the residual oil saturation after water displacement has been investigated. It was noticed that the residual oil is distributed in few oil filaments occupying the relatively larger pores. Differences between imbibition and drainage are illustrated in Figure 1. a great amount of photographs. enable a journey through the porous medium. Woodmetal was used.29: Schematic diagram of water invasion into porous media permeabilities of the wetting phase (after Craig) .
30 illustrates this aspect. Figure 1. The deviation of the curve in the direction of displacement is referred to as the hysteresis of relative permeability. Figure 1.31 shows two oil/water permeability function pairs to indicate the role of wettability.154 1: Fundamental Properties of Porous Media The distribution of the nonwetting phase at drainage differs from the one at imbibition.30: Drainage and imbibition relative permeability characteristic (after Craig) Figure 1.31: Typical water/oil relative permeability characteristic (after Craig) . In addition Figure 1.
4 1.S. Doherty series (1971) Fatt. and Davis. D.: 5. J. 1953) Krumbein. W.7 1..: "Relative Permeability to Liquid in Liquid Gas System" Trans AIME 192 (1952) 1. AIME (1952) p. Jr. Technip.1: Fundamental Properties of Porous Media 155 1.: "Laboratory Measurements of Relative Permeability" Trans. Paris (1980) Morse. and Russel.S. W.T.D. 3 of the Henry L.N.: "Reduction in Permeability with overburden pressure" Trans.A.H. Crofts Inc. Ch.3 1. Jenks. P. S.: "Relative Permeability Measurements on small Core Samples" Producers Monthly II (1947) Rappaport.J.8 1. Petroleum Technology (Jan.L.: "The Reservoir Engineering Aspects of waterflooding" Monograph Vol.A.L. and Leas.5 References 1.K. R. Yuster.2 1.5 1. et al.329 Hall. W.: "Basic Petroleum Geology" OGCI Publications. F.1 1.C. Terwilliger.: "Properties of Reservoir Rocics: Core Analysis" Institut Francais du Petrole Publications Ed.9 Craig. New York.: "Relative permeability to Liquid in Liquid Gas System" Trans. H. Leas.P.: "Strategically and Sedimentation" Appleton century publication.J. R. W.P.6 1.F. and Sloss. L. I. L.. AIME 192 (1951) Link. Tulsa Oklahoma USA (1983) Monicard.10 Osoba. AIME 192 (1951) .
156 1: Fundamental Properties of Porous Media .
perhaps fortified by experimental evidence. The conceptional assumption is more comprehensive than the equation. However. Fick’s law relating flow of matter of a component to its concentration gradient in a multicomponent system. They usually take the form of relationships between fluxes and driving forces. Such a simple relationship does not always hold. There are more or less powerful approaches for abstracting and simplifying natural phenomena. 257 . Ohm’s law relating electrical current to electrical potential gradient etc. For example: Newton’s law relating shearing force to velocity gradient. In each case the simplification is carried to the point where the model is still amenable to mathematical treatment. There exist a large number of constitutive equations describing relationships between fluxes and driving forces. They are commonly referred to as field equations. insufficient to yield specific answers unless further equations are supplied. They form an under defined mathematical model. We do so to emphasize that these equations define the assumed behaviour of ideal continua. yet is not so simple as to miss those features of the studied phenomena it is intended to describe. either the fundamental nor the field equations contain information regarding the properties of the particular continua under consideration. but they will always refer only to the most important aspects of the process. Fourier’s law relating heat flow to temperature gradient. These mathematics are summarized by the term mathematical model. The resulting equations of continuity are generally written in the form of partial differential equations. momentum and energy of various kinds. A more general law of the continuum theory is the law of conservation of the extensive properties as mass. Any system properties not included in the mathematical model can’t be taken into consideration in further calculations. These mathematics aim at approximating these processes in a more or less sufficient way. These equations are definitions extracted from physical experiences. although outwardly the assumptions take the form of equations. linear dependence of a flow on some conjugated force. While these equations determine the features of a mathematical model we refer to this as fundamental equations. In all these cases we see a simple. We mention as examples of fluxes those of mass. The description of motion of a continuum are based on the so called constitutive assumptions or constitutive equations.2 Equations of SinglePhase Filtration A quantitative description of a physical process always requires a mathematical formulation. momentum and energy. The constitutive equations are often referred to as phenomenological equations because of their dependence on experimental evidence.
The field equation express the conservation of the fluid mass. In order to solve this set of equations. In case of filtration only a certain part of the particles are moving. further equations are required to determine the initial state and the boundary conditions. The mechanical properties of the fluids are formulated through the equations of state. Basically one must distinguish between mass flow and filtration. all other mass particles form a solid matrix. During mass flow all mass particles of the system are in motion.258 2: Equations of SinglePhase Filtration Mathematical models of filtration are generally based on Darcy’s law. .
the fluid body will accelerate as long as the resultant acceleration is not zero.1 Differential Form of the DarcyLaw Consider a certain volumetric element in space with fixed boundaries. but it will keep its velocity. Its length is δs and has a cross section δΑ.1 Fundamental Equation of Filtration.1: Schematic diagram of a field segment . In filtration process the most important forces are listed in the following: • force of compression (acting on surface) • force of gravity • forces of inertia • frictional forces • capillary forces (acting on body) (acting on body) (acting on surfaces) (acting on surfaces) Figure 2.2: Equations of SinglePhase Filtration 259 2. Two types of forces should be distinguished: body forces and surface forces. To simplify but without losing the generality we take a cylindrical element. 2. Regarding Newton’s law of motion. Equal forces are acting on the bulk of fluid within the control volume.1. The surface of this element is build from stream lines and the base surfaces are perpendicular to it. we seek a value for velocity such that the resultant acceleration equals the null vector. The size of the volume must be selected in a manner so that random effects may be statistically eliminated.
3) . A very complicated labyrinthic structure of the pore channels makes accurate calculations impossible. in reference to the crosssections of the pore channels. The second term refers to the turbulence and is proportional to the squared velocity.260 2: Equations of SinglePhase Filtration Convective and local acceleration are mostly so small in case of filtration. Capillary forces are surface forces between two immiscible fluids which are separated by an interface inside the pore channels. (2. Such cases are known to exist in hydrology and are called unsaturated filtrations or unconfined flow.2) where φδA is the free surface (pore surface) of the element and ∇pδs the pressure difference along δs. are important. which are proportional to the velocity of filtration. Calculation of these forces is only possible in sum. The frictional force is actually to be regarded as a surface force acting on the surfaces inside the pore channels.1) In reference to a defined elementary volumetric element in Figure 2. (2. Therefore only an average value referring to one volume unit is considered. and the frictional force F µ .since it results only from a mathematical derivation and not from observation and measurement. Frictional forces are generally composed from two terms.just as is the velocity of filtration . If the flow is laminar the second term disappears. The forces of inertia and the forces of turbulent friction can be neglected. We assume that only the force of compression F p . The following equation for the equilibrium of forces may be set up: Fµ + Fp + Fg = 0 (2. The capillary force is therefore regarded as a ficitous surface force. capillary forces are considered only if the waterbearing layer is partially empty. Therefore in case of filtration frictional forces become volumetric forces. but these are only fictious . The first represent the adhesive forces and is proportional to the velocity inside the pore channels. the force of gravity F g . • The force of gravity: F g = – ρ gφδAi 3 δs . forces of inertia may be neglected for steady state and nonsteady state filtration as well.1 those forces may be mathematically specified as follows: • The force of compression: F p = –φδA∇pδs . In the case of a onephase filtration.
po ρ (2.2 .7) ρ = ρ(p) (2..8. φ (2.∇ψ .6 the differential form of the law of Darcy is obtained: k u = – .u + i 3 ρg = 0 .4) In addition.7 becomes: kρ u = – .( ∇p + i 3 ρg ) µ We introduce a so called potential function instead of pressure: p dp ψ = gx 3 + ∫ . • The laminar frictional force: F µ = – B µδAuδs . yields: ρ∇ψ = ∇p + ρgi 3 If Eq. 2.4 into Eq.6) (2. then Eq. 2. µ Differentiation of Eq.10 is substituted into Eq. φ or after reducing: B ∇p + µ .Eq. 2. where B is a coefficient fort a given porous medium.10) (2. the term φ ⁄ B is defined as k and interpreted as the permeability. (2.2: Equations of SinglePhase Filtration 261 where the unit vector i3 is directed upwards. 2.7 is obtained. 2.1 leads to: B – ∇p + µ . 2. Substituting Eq.u + i 3 ρg ⋅ φδAδs = 0 .5) (2. 2.9) .8) where p o is a reference pressure at x 3 .9. After transforming Eq. Then Eq. 2. 2.
k∇ψ µ in detailed form: ∂ψ ∂x 1 (2.2 Anisotropic Porous Media The Eq.k 21 µ u3 k 31 u1 k 12 k 13 ∂ψ k 22 k 23 . Thus ρ u = – . the Darcy’s law becomes the following: . In an isotropic porous medium the permeability k is a scalar. ∂x 2 k 32 k 33 ∂ψ ∂x 3 ij ji (2. but in an anisotropic medium is a tensor. 2..1. 2.12 for a coordinate system with axis parallel to the principal directions of the porous medium.11) k 11 ρ u 2 = – .9 is a linear vectorvector equation. Writing Eq.262 2: Equations of SinglePhase Filtration 2. These principal directions are orthogonal to each other. Directions parallel to these coordinate axes are called principal directions (axis) of the porous medium.12) If the matrix is symmetrical (k = k ) it is possible to transform the coordinate system so that all values apart from the main diagonal become zero.
In this coordinate system Eq. 2.2: Transformation of the coordinate system ∂ψ’ ∂x’1 (2.13 becomes a little bit more complicated: .13) u’1 ρ ∂ψ’ u’2 = – .0 k 2 0 . µ ∂x’2 0 0 k3 u’3 ∂ψ’ ∂x’3 where: 3 k1 0 0 ψ’ = g i=1 ∑ x’i cos αi + ∫po ) ρ( p p dp (2..14) α i are defined as the angles between the coordinate axes and vector of gravity.2: Equations of SinglePhase Filtration 263 x3 x’3 x’1 α3 osα1 gi1c α1 3 sα co gi 3 α2 x1 g x2 x’2 Figure 2.
. x’3 ) is different to the one in the coordinate system ( x 1 x 2 x 3 ) since gravity is not parallel to the x 3 axis.k∇ψ’ = – . According to Eq. the proportions are not the same in different coordinate directions. The potential function ψ in the coordinate system ( x’1. x’2.2 Transformation of the coordinate system ( x 1 x 2 x 3 ) into ( x’1.15 the velocity of filtration is proportional to the potential gradient. ∇p + ρg ∑ i i cos α i µ µ i=1 (2.15) Figure 2. However.264 2: Equations of SinglePhase Filtration 3 ρ k u = – . x’3 ) . x’2. 2.
. .17) 2.1 Incompressible Fluids The potential described by Eq. If this is not the case the fluid is considered incompressible: dρ . (2.20) (2..21) . dp T Vf By substituting: Vf = m ⁄ ρ .2: Equations of SinglePhase Filtration 265 2.21.. This constant is called the coefficient of isothermal compressibility which is defined by: 1. 2. (2.dVf c = – .2 Low Compressibility Fluids For low compressible fluids one may assume that the fractional change of volume of the fluid as pressure changes at constant temperature is constant. a constant.18) (2.2. where m is the mass of fluid.8 includes the density of the fluid.18 we obtain: m d .19) (2. c = – m dp ρ dp ρ After integration of Eq.22) where: ρ o is the density of fluid at any reference pressure ( p 0 ) .2 Equation of State 2.2. into Eq. 2. ρ 1 dρ 1 . yields: ρ = ρo e c ( p – po ) (2.16) (2. 2.= 0 dp or after integration: ρ = constant . . It is a function of pressure. .= .
Rs )q os (2.23) 2. Rs ) = B o ( p. T. This conversion factor can be defined as: volume of fluid under reservoir conditions B = volume of fluid at standard conditions Formation Volume Factor for oil is the volume of 1 standard m3 oil (1 m3 tank oil) under reservoir conditions p. Tthe rate of fluid under reservoir conditions can be calculated by: q o ( p. Since these fluids are compressible a change of phase also can take place (gas evolving from oil). This change of volume is mainly due to changes in pressure and temperature. but in this text only one phase is considered.266 2: Equations of SinglePhase Filtration Applying TAYLOR’s rule and neglecting the terms of higher order we get the following approximation: ρ ≈ ρ0 ⋅ [ 1 + c ( p – po ) ] (2.2. This change of volume should be accounted for by using a conversion factor (B) which is called Formation Volume Factor.T and solution gasoil ratio Rs. T.3 Formation Volume Factor The volume of produced fluid at surface conditions is usually different than the volume of fluid entering the wellbore at reservoir conditions.24) .
28) (2. yields: Mp s ⁄ RT s B g = Mp ⁄ RTZ ( p ) (2..p ρ = ..33) (2. 2.27 we obtain: 1 dρ 1 . RT Z ( p ) (2.30) (2.30 into Eq.32) (2.26.∂Z .4 Ideal and Real Gases For ideal gases according to the law of BoyleMariott: m pV = . V Substituting Eq. 2. 2.– . M (2. 2. 2.26) where Z(p) is the real gas compressibility factor. yields: Mρ = . 2.25) where m is the mass of gas.27) (2..27 yields: dρ M.c g = .= .= ρ dp p For real gases Eq. 2. M the molecular weight and V the volume.32 .31) ..p RT Taking the derivative of Eq.RT ..25 into Eq.c g = .= dp RT Dividing Eq.29) (2.27 becomes: M. Since the density of a gas is defined as: m ρ = .28 by Eq.ρ dp p Z ( p ) ∂p The formation volume factor for gas is defined by: ρs B g = ρg Substituting Eq. Thus the compressibility of a real gas is defined as: 1 dρ 1 1 . 2.2: Equations of SinglePhase Filtration 267 2.2.
T ) = V s Bg where: Vs is the volume of gas at standard conditions. The fluid mass content in the control volume is: ρφdx 1 dx 2 dx 3 .= C ⋅ p pT s Where Ts.5 Equation of continuity The equation of continuity describes the law of mass conservation.35) (2. We use a rectangular coordinate system and consider a parallelepiped as a control volume (see Figure 2.ps are the standard temperature and pressure. The volume of gas under any pressure and temperature can also be calculated by: V ( p. Figure 2.34) 2. (2.3) with a porosity φ.2.33 becomes: p s TZ ( p ) TZ B g = .268 2: Equations of SinglePhase Filtration After simplification Eq.36) .37) (2.3: Volume element in a cartesian coordinate system The pore volume is then: φdx 1 dx 2 dx 3 . (2. 2.
∂t (2.43) .41) Also considering the filtration in the direction of other coordinates.43 yields : k ∂ ∇ . Eq. 2. .38 and Eq. (2.ρ 2 .ρ 2 ∇ψ = . 2. k 3 ∂ψ ∂ .2: Equations of SinglePhase Filtration 269 The change of quantity during a time interval dt is: ∂ ( ρφ ) .ρ 2 .45) (2. ∂x 3 ∂ (2.ρ 2 .40) The change of flowing quantity results in: ∂– . 2. + . + . = .dx 1 dx 2 dx 3 dt.( φρ ) µ ∂t Or in detailed form: ∂ .15 into Eq. k 2 ∂ψ ∂ .44) (2.39) dx 2 dx 3 dt .38) On the other side.( φρ ) = ∂t i=1 ∑ i ( ρui ).( φρ ) = ∇ ( ρu ) ∂t Substituting Eq.42) where the term dx 1 dx 2 dx 3 were already cancelled. k 1 ∂ψ ∂ .42 becomes: ∂ – . 2.( ρu 1 )dx 1 dx 2 dx 3 dt . After writing it in vector form Eq.( φρ ) ∂x 1 µ ∂x 1 ∂x 2 µ ∂x 2 ∂x 3 µ ∂x 3 ∂t (2. . . ∂x 1 (2. the quantity of fluid flowing through the surface at x 1 and x 1 + dx 1 is: ( ρu 1 ) and ( ρu 1 ) x 1 + dx 1 x1 dx 2 dx 3 dt (2.41 result in: ∂ – . 2.
2.48) . µ ∂t (2.46) If the 3rd coordinate direction is vertical then Eq.( φρ ) . 2.44 yields: k k ∂ ∇ .47The second term on the left hand side is often very small compared to the first. 2. For these cases the following equation is sufficient: k ∂ ∇ .47) If the 3rd coordinate direction is vertical then Eq.gi 3 = . 2.ρ∇p + ∇ ρ 2 . 2.14 yields: 3 ρ 2 ∇ψ = ρ∇p + ρ 2 g ∑ ii cos i=1 αi .ρ∇p = .( φρ ) .270 2: Equations of SinglePhase Filtration Differentiation of Eq. µ µ ∂t (2. (2.46 into Eq.46 become more simple: ρ 2 ∇ψ = ρ∇p + ρ 2 gi 3 Substituting Eq.
∇ρ = .53) (2. 2. φ and µ are constant and Eq. then ρ. 2.50 becomes the Laplace equation: ∂ 2 p. 3 ki Then Eq.+ .54) Substituting Eq.. 2. 2.51) (2. 2.50) 2.49) In the case of a homogenous porous media and k = const then Eq. (2.= 0 2 2 2 ∂x 1 ∂x 2 ∂x 3 (2. 2.+ .3 Special Forms of the Equation of Filtration 2.55) .22 the following transformation can be made: ρ∇p = ρ 0 e c ( p – p0 ) c ( p – p0 ) 1 = 1 ∇ρ ∇p = .( φρ ) cµ ∂t (2.49 becomes: ∂ 2 p.∂ 2 p. 2.∇ ρ 0 e c c (2.3. 2.∂ 2 p. i = 1.54 into Eq.∂ 2 p.48 becomes: ∇ [ k∇p ] = 0 .2 Low Compressibility Fluids Based on Eq.+ k 2 .2: Equations of SinglePhase Filtration 271 2.= 0 2 2 2 ∂η 1 ∂η 2 ∂η 3 For isotropic porous media Eq.1 Incompressible Fluids Assuming an incompressible fluid and porous media.+ k 3 .+ .∂ 2 p∇ [ k∇p ] = ∇ 2 p = .49 becomes: ∂ 2 p∂ 2 p∂ 2 p∇ [ k∇p ] = k 1 .52) (2.48 becomes: k∂ ∇ .3.= 0 .50 can be simplified by introducing a new independent variable: k1 η i = x i .+ . 2 ∂x 2 ∂x 2 2 ∂x3 1 Eq. 2.
Then these functions in Eq.22 into Eq.1 Elastic Porous Media The porosity and the permeability of an elastic porous medium are both functions of pressure: k = k ( p ).57) (2. 2.55 results in: k ∂ ∇ . 2. Eq. φ = φ(p) (2. 2.61) φ = φo e Substituting Eq. 2.ρ o ∇p = .59) K [ m s ] is defined as the piezometric conductivity.58) (2. 2.48 leads to: .. Without serious restictions one may assume that the interdependencies are small and the alteration of porosity and permeability are proportional to pressure changes.{ φρ o [ 1 + c ( p – po ) ] } µ ∂t (2.56 into Eq.23 and Eq.60 are similar to Eq. Eq.55 becomes: 1 ∂p . 2. 2. 2. K ∂t where: k K = µcφ 2 –1 (2. The viscosity µ can be regarded as constant as well hence the fluid compressibility is small.60) The actual functions have to be determined by measurements.61 together with Eq. 2. For a homogeneous isotropic and incompressible porous media and constant fluid viscosity..56) In Eq.22 and may be written as follows: k = ko e ck ( p – po ) cφ ( p – po ) (2.58 is identical to the Fourierequation of heat conductivity. 2.23 gives: ∇ρ = ρ o c∇p .∇ 2 p = . c and p 0 are constant.3. 2. 2. Substituting Eq. 2.2.272 2: Equations of SinglePhase Filtration Differentiation of Eq.57 ρ 0 .
2.e µ ∂t o o Eq.66 into Eq...62) (2. 2. 2. After simplification: ∇2 [ e ( c + ck ) ( p – po ) φo µ ( c + cφ ) ( c – c ) ( p – p ) ∂ ( c + c ) ( p – p ) o .68) .67) Eq.65 and Eq.e φ k c + ck ∂t (2.[ e k o ] ] = .65) and c + c k = c + c φ In addition: e ( c + ck ) ( p – po ) ≅ 1 + ( c + ck ) ( p – po ) (2.= . 2. 2.∇ 2 p = . ko ∂t k o ∂t (2.63 and Eq.[ φ ρ e ( c + c φ ) ( p – p o ) ] ∇ ..64 leads to: φ o µ ( c + c φ ) ∂p φo µc t ∂p ..64) Since c φ and c k are very small it can be assumed: e ( cφ – ck ) ( p – po ) = 1 (2.2: Equations of SinglePhase Filtration 273 ko ρo ( c + c ) ( p – p ) ∂ k o ∇p = . 2.63) After differentiation it is quite evident that Eq.62 can be written as: ko ρo . (2.64 are identical. 2.67 shows that the compressibilities of the fluid and the porous medium are added in the piezometric conductivity term: ct = c + cφ where ct is the total or ultimate compressibility of the system.66) Substituting Eq.∇ [ ∇ ( e ( c + c k ) ( p – po ) ) ]= µ ( c + ck ) φo ρ o ( c + cφ ) ( c – c ) ( p – p ) ∂ ( c + c ) ( p – p ) o .e φ k ∂t ko (2.[ e φ o ] .
72.∇p = φc g µ . 2.( φρ ) = φ ∂ρ = φ ..74) which is called the real gas pseudo pressure. 2.48 becomes: k Mp M2p ∂p ..71) (2. µ RT Z 2RT µZ ∂t After simplification of Eq.φc g µ . p b µZ (2..= .∇ .= .76) (2.75) . µZ µZ ∂t AlHussainy..69) Substituting Eq. This function enabled the following derivations: dm ( p ) 2p ∇m ( p ) = . 2.⋅ ρ dp (2..φc g µ .. 2..∂t ∂t ρ ∂p ∂t 2p ∂p Mp ∂p M.=φ c g .∇ k .∇p = . yields: 2p 2p ∂p . µZ ∂t RT Z ∂t 2RT Thus. Eq.... RT Z and the compressibilty is 1 dρ c g = .48 and taking the porosity as constant since the rock compressibility is several orders of magnitude less than the gas compressibility.48 can be developed in the following way: ∂ 1 .70) (2.∇p = .30 the real gas density is: Mp ρ = ..4 Real and Ideal Gases From Eq. the right side of the Eq. 2.∂t dp ∂t µZ ∂t (2.69 into Eq.72) (2. Crawford then introduced the following function: p pdp m ( p ) = 2 ∫ .274 2: Equations of SinglePhase Filtration 2. Ramey..73) (2.∂ρ ρ ∂p = .= .. 2.∇p. dp µZ and ∂m ( p ) 2p ∂p dm ( p ) ∂p .
. Ideal gases are characterized by the following terms: 1 Z = 1. For ideal gas Eq..and µ = constant. µ Substitution of this equation into Eq.76 into Eq. 2.73 results in: ∂m ( p ) ∇ [ k∇m ( p ) ] = φc g µ .kp ∂t (2.2: Equations of SinglePhase Filtration 275 Substitution of Eq. 2. 2.77) Assuming that the porous medium is isotropic and homogeneous. The differences that cg and µ are both functions of pressure and hence Eq.58 for low compressibility fluids.78) This equation is identical to Eq. 2.. ∂t (2.75 and Eq. 2.78 leads to: φµ ∂p 2 ∇ 2 p 2 = . p where p is defined as the average pressure of the considered volume and time interval.77 becomes: φc g µ ∂m ( p ) ∇ 2 m ( p ) = . c g = . 2.78 will not be linear.74 becomes: p2 m ( p ) = . 2.80) (2. Eq. k ∂t (2..79) . 2.
gas. . (2. This leads to a general equation for the boundary conditions: ∂ψ a . u.82 is valid in the case of water intruding from the surface into the reservoir. In case of an open boundary it is either the potential or the gradient of the potential given at the boundary as a function of time: ψ Γ = ψ ( x. t ). 2.. Infinity is more a mathematical fiction than a fact but it is very useful for the solution of the equations.81) In both cases the velocity across the boundary is calculated by: kρ ∂ψ . ∂n x ∈ Γ. b and c are functions of time.84) It is possible to combine the boundary conditions Eq. or ∂ψ ( x.83) x∈Γ (2.83. t ) ψ′Γ = . In addition it is also valid at faults and pinch outs.= 0.1 Boundary Conditions The space where filtration takes place may be limited or infinite.or waterbearing layer. The condition in Eq. 2. 2.µ µ ∂n Γ ∂ψ .+ bψ = c. A boundary is considered open if the fluid is able to pass through and closed if it is not.u n = .276 2: Equations of SinglePhase Filtration 2. In the case of a closed boundary the following mathematical formulations may be made: kρ kρ ∂ψ . ∂n Where n is a unit vector normal to the boundary Γ .5 Boundary and Initial Conditions 2.82) (2.µ ∂n (2. The condition in Eq. x ∈ Γ. 2.n = – .81 is valid at the top or bottom boundaries of an oil.= 0.85) a.82 and Eq.5.∇ψn = – . (2. ∂n x ∈ Γ.
86) By solving problems such as filtration of incompressible fluids or a steady state filtration ( t → ∞ ) no initial conditions are required. and p ( x. 2. The state of the system at t = 0 is called initial condition: ψ ( x.83 must be applied at the surface of a well (inner boundary) producing (or injecting) at a given rate. Figure 2. The expression initial conditions refers to this practice.4: Illustration of the boundary conditions . t 0 ) = p 0 ( x ) (2.5. 2.2: Equations of SinglePhase Filtration 277 Eq. t 0 ) = ψ 0 ( x ) .2 Initial Conditions In order to solve filtration problems which have time dependent solutions it is necessary to know the state of the system at a certain date. This is usually the temporal starting point t = 0.87) (2.
278 2: Equations of SinglePhase Filtration 2.5.89) (2.88) Both ψ 1 and ψ 2 are potential functions of the two regions separated by the discontinuity surface...= ρ . .ρ ..3 Discontinuities in Porous Media Permeability may change by leaps and bounds at the contact surface of two regions of the porous medium.. µ ∂n µ ∂n (2. In such a case the potential and the normal component of the velocity at the boundary must be continuous: ψ1 = ψ2 and k 2 ∂ψ 2 k 1 ∂ψ 1 .
kp ∂t 2 .( φρ ) ∂t Equation of Flow DarcyLaw ρ u = – .6 Schematic of the Filtration Equations Equation of State ρ = ρ(p) φ = φ( p) Equation of Mass Conservation ∂ ∇ ( ρu ) = – . c g = . µ.( p ) K = K ∂t φc t µ p 2 Real Gas: Mpρ = RTZ pm ( p ) = 2 ∫ .k ∂t 1 pIdeal Gas: Z = 1 ..2: Equations of SinglePhase Filtration 279 2..dp Zµ p0 ∂m ( p ) ∇ ( k∇m ( p ) ) = φc g µ ∂t Homogenous and isotropic Porous Medium φc g µ ∂m ( p ) 2 ∇ m ( p ) = . µ g = const .( φρ ) µ ∂t Incompressible Fluid and Porous Medium: ρ. m ( p ) = p µ 2 2 φµ ∂p 2 ∇ p = .k∇ψ µ ρ∇ψ = ∇p + ρgi 3 ρ ∂ ∇ . φ = const ∇ ( k∇p ) = 0 Homogenous and isotropic Porous Medium ∇ ( ∇p ) = ∇ p = 0 Low compressible Fluid and elastic Porous Medium ρ = ρ 0 ( 1 + cf ( p – p0 ) ) φ = φ0 ( 1 + cφ ( p – p0 ) ) ct = cf + cφ ∂p ∇ ( k∇p ) = φ0 c t µ ∂t Homogenous and isotropic Porous Medium 2 1∂ k ∇ p = .k∇p = .
= .3 yields following results: ∂p ∂p .2 and Eq. A.= cos θ. y ) .1) To convert Eq.∂r ∂x ∂r ∂y ∂r ∂p ∂x ∂p ∂y ∂p ( x.4 and Eq.4) (A.cos θ + ..= .cos θ sin θ ∂y ∂r ∂x (A. A.cos θ + ∂p sin θ ∂r ∂x ∂r ∂y ∂r ∂x ∂y ∂p ∂x ∂p ∂y ∂p ∂p ∂p .= .∂θ ∂x ∂θ ∂y ∂θ Since: ∂x ∂y .+ . A.8) .1 to a radial coordinate system the independent variables r and θ have to be considered.2) (A.= – .5) (A. A.∂x + ∂p .3) ∂p ( x..⋅ cos θ = .= ..∂y ∂p .= .. A.5 in Eq.280 2: Equations of SinglePhase Filtration Appendix A Derivation of Laplace Equation in Radial Coordinates The Laplace equation in the Cartesian coordinate system is: ∂2 p ∂2 p ...= ..6 with cos θ leads to: 2 ∂p ∂p ∂p .= ∂p ∂x + ∂p ∂y .= 0 ∂x 2 ∂y 2 (A. ..7) (A.cos θ ∂θ ∂x ∂y Inserting Eq..= sin θ ∂r ∂r ∂y ∂y . y ) . A. . The Cartesian coordinates can be expressed: : x = r cos θ y = r sin θ (A.6) (A.= –r sin θ.r cos θ ∂θ ∂x ∂θ ∂y ∂θ ∂x ∂y Multiplying Eq.+ .r sin θ + .= r cos θ ∂θ ∂θ and ∂p ∂p ..+ .cos θ + ∂p sin θ ∂r ∂x ∂y ∂p ∂p ∂p .( –r sin θ ) + .
⋅ sin θ + 1 ⋅ .+ 1 cos θ ∂p . ∂p .∂p ( r. ∂θ = ∂y r ∂y Then Eq..13 becomes: ∂2 p ∂∂p .∂.= cos θ .11) (A.∂.9) Adding Eq..⋅ ( – sin θ ) = ∂p ⋅ ( cos θ + sin θ ) = ∂p ∂r r ∂θ ∂x ∂x (A.sin θ r ∂θ ∂r r ∂θ Also Eq.2 ∂r ∂θ r ∂r ∂y cos θ ∂∂p 1 ∂p ..8 and Eq. A.∂x ∂x ∂x ∂r ∂x ∂x ∂θ ∂x and ∂ ∂r ∂.= sin θ .6 is multiplyed with sin θ and Eq.⋅ .– .sin θ .∂p ( r.cos θ sin θ ∂y r ∂θ ∂x (A.2 ∂r ∂r r ∂θ ∂x sin θ ∂ ∂p 1 ∂p ..12) (A.15) (A. A.cos θ ∂p – 1 sin θ .= x = cos θ.= y = sin θ. ∂p + ∂θ .⋅ cos θ + . ∂θ = ∂x ∂x r ∂r . A. A.– .⋅ cos θ = ∂p . ∂p .+ . ∂y ∂x ∂θ ∂y ∂y ∂y ∂y ∂r and ∂r . A. A.7 with sin θ and some rearrangements lead to: 2 ∂p 1 .sin θ . ∂p + ∂θ . θ ) = ...cos θ r ∂θ ∂θ r ∂r – sin θ r cos θ r (A..9 yield: 2 2 ∂p 1 ∂p .cos θ .∂p ∂r r ∂θ ∂y Since ∂ ∂r ∂. A.+ .7 with cos θ the same calculation leads to: ∂p .10) If Eq.⋅ ∂p ⋅ ( – sin θ ) = ∂p sin θ – .2: Equations of SinglePhase Filtration 281 Multiplying Eq. θ ) = .13) (A.14) (A.16) .12 becomes: ∂∂p ∂2 p ..
282
2: Equations of SinglePhase Filtration
Substitute Eq. A.15 and Eq. A.16 into Eq. A.1, with simplifications, yields :
2 ∂ 2 p 1 ∂p   +   + 1 ∂ p = 0   ∂r 2 r ∂r r 2 ∂θ 2
(A.17)
Eq. A.17 can be written as:
2 1  ∂p   ∂ r  + 1 ∂ p = 0 r ∂r ∂r r 2 ∂θ 2
(A.18)
3
Solutions of the SinglePhase Equation of Filtration
In order to find a suitable solution to the filtration equation, one should know what type of condition can be used to describe a flow regime. Generally there are two conditions: NonSteady State and Steady State. The NonSteady State may be subdivided into the following conditions: A  Transient condition: This condition is valid at an early, relatively short time, where the pressure response in the reservoir is not affected by the presence of an outer boundary, thus the reservoir appears infinite acting. In well testing this condition is applicable when the production rate is deliberately changed for a short time, the pressure response is measured for a few hours and the boundary effects will not be felt and therefore the reservoir is mathematically infinite. B  Late Transient condition: This condition exists in the period when the boundary effects start to show up in the pressure response. This will occur when the well test period takes a longer time, or the reservoir is smaller than expected. C  PseudoSteady State condition: The Pseudo (or Semi) Steady Sate condition occurs after the late transient condition and is valid for a reservoir which has been producing for a sufficient period of time so that the boundary effect has been felt by the pressure response. The outer boundary could be impermeable to fluids (no flow boundary) or a constant pressure boundary. Steady State condition: The Steady State condition occurs also after the late transient period. This condition is applicable when the production rate is constant and fluid withdrawal will be exactly balanced by fluid entry across the open boundary, so that there is no change in pressure with time in the whole reservoir.
383
384
3: Solutions of the SinglePhase Equation of Filtration
3.1
Steady State Filtration
3.1.1 Steady State Filtration of Low Compressibility Fluid
Filtration is steady state if the potential (and so the flow) at any point of the system is independent of time ( ∂p/∂t = 0 ). For constant fluid viscosity, Eq. 2.57 will be reduced to: ∇( k ∇p ) = 0 Eq. 3.1 is similar to Eq. 2.49 for incompressible fluids. For homogeneous and isotropic porous media (k = k = constant) follows: ∇2p = 0 (3.2) (3.1)
A linear one dimensional model is used for the sake of simplicity. Mathematically it can be described by: ∂ p = 0 ∂x2 and its solution after integrating Eq. 3.3 twice is given by: p = ax + b (3.4)
2
(3.3)
where: a and b are constants of integration. The actual value of the pressure is then determined by boundary conditions. A production rate (q) is assumed negative when flowing out of the porous medium and positive when flowing into it. Thus q becomes negative when produced and positive when injected. The boundary conditions are used to determine a and b as follows: At x = 0 is p = p i . Then b is determined from Eq. 3.4 as: b = pi kA dp  At x = L is   = qB µ dx (3.5)
6) and by substituting into the second boundary condition (B. 3.1.4 then dp = a.). and ∂ p ∂θ 2 2 = 0 .C. Chapter 2. Taking the derivative of Eq. 3. 3. which will leave Eq.2 Steady State Filtration in a Radial System The Cartesian coordinates can be transformed into radial coordinates. then the filtration is radially symmetrical.7) The value of p .2 in the following form: 1 ∂ ∂p .∂ p 1 r + = 0 ∂ r 2 2 r ∂r r ∂θ 2 (3. as shown in Appendix A. a is determined as: µqB a = kA Substituting Eq.3: Solutions of the SinglePhase Equation of Filtration 385 where B is the formation volume factor and q is negative. 3.x kA (3. 3. 3.4 yields: µqB p – p i = .5 and Eq.9) If the pressure (or potential) distribution is independent of θ .8) (3. dx (3.7 into Eq.pi becomes negative in case of flow taking place in direction of the xaxis (q is negative) and positive in the inverse case.
3.10 twice leads to: p = a ⋅ lnr + b We set a constant pressure at the inner radius rw: p = p wf at r = r w and a constant rate q.15) µqBa = – .14 yields: (3. 3. 2πhk b can be calculated from Eq. 3. From Darcy equation: kA dp = –q B at r = r w . 3. 3.12 with respect to r (dp/dr =a/r) and substituting it into Eq.. 3.1: The Radial Coordinate System This simplifies Eq. Taking the derivative of Eq.17) (3.13) (3.9 to: 1 ∂ ∂p r = 0 r ∂ r ∂ r Eq.14) where A = 2πr w h is the inner surface and h the thickness of the layer.10 can be expressed in the following form also: ∂ p 2 (3.12) (3.10) 1 ∂p + = 0 r ∂r ∂r Integrating Eq. Note: outfow rate is negative while inflow rate is positive.12 as: b = p wf – a lnr w (3.11) (3. µ dr 2 (3.386 3: Solutions of the SinglePhase Equation of Filtration Figure 3.16) .
19) (3.19 leads to: 2πhk p e – p wf q = – . (3.20) .r p wf – p = . If the inner radius rw represent a well then pwf is called as bottom hole flowing pressure. 3..12 leads to: µqB. µB ln ( r e ⁄ r w ) This equation is known as Dupuit equation.2: Illustration of steadystate filtration in a radial system Considering an outer boundary ( r = r e ) with constant pressure pe Eq.lnr w 2πhk Substituting the values of a and b into Eq..ln 2πhk r w (3.18) Figure 3.3: Solutions of the SinglePhase Equation of Filtration 387 and µqBb = p wf + . 3.
1. 2.30 one can set up the following equation: µQ m βQ m Mp .–  πhMk r w 2πhµ r w r (3.+  dr Z ( p )µ MAk r 2πr 2 hµ pi rw p r (3.21 becomes: 2Qm RT 1 βQ m p 2 ∫ . 2. 3. x MAk µA (3. 1 + .dp = – . 3.3 Steady State Gas Filtration From Eq. At a given exterior radius re the pressure is pi.25 becomes: µQ m RTZ r e βQ m r w 2 2 p i – p wf = – .ln . then Z(p) is substituted by Z. 3. 1 –  πhMk r w 2πhµr w re (3. Then Eq.41 and substituting of ρ by Eq.23) In the case of a radial symmetric filtration (see Figure 3.26 may be set up as follows: p i – p wf = – AQ m – BQ m where: 2 2 2 (3.388 3: Solutions of the SinglePhase Equation of Filtration 3.2) the surface area is A = 2rπh and Eq.26) The fraction r w ⁄ r e is generally considered small and so the last term ( 1 – r w ⁄ r e ) ~ 1. 3.74 into Eq.24) or Q m RT βQ m 1 1 r m ( p ) – m ( p wf ) = – .27) . for p = pi and Eq.dp = .dp = – . 1 +  x MAk Z ( p )µ µA pi p (3.21) (3.ln .+ . 3.21 is integrated along a linear segment: 2Qm RT βQm p 2 ∫ .+ .22) Substituting Eq.22 we obtain: βQ m 2Q m RT m ( p ) – m ( p i ) = – . 1 +  dx RTZ ( p ) Ak µA Then Eq.∫ .25) Assuming that the pressure difference between the two ends of a linear porous body is not very large. . 1.
ln πhMk r w βRTZ B = 2 2 2π h kMr w Qm is negative for producing wells and positive for injection wells..27 instead calculating them with Eq. 3. 3. Eq.= A + B Q m Qm 2 2 (3. 3. 3.28.3: Plots of production equation for gas wells Q m = C ( p i – p wf ) 2 2 n (3.28) This equation is useful to calculate gas well production rates as a function of pwf. Eq. 3.3.3: Solutions of the SinglePhase Equation of Filtration 389 µRTZ r e A = .29) The constants C and n have to be determined by fitting of Eq. . In practice the following equation is also used: Figure 3.27 can be transformed into the following form: p i – p wf .29 to the measured Qm and pwf values.3a is equal to B and the intercept is A. In the same way we can evaluate the constants in Eq.28 indicates that the slope of the line in Figure 3. The graphical illustrations are shown in Figure 3.
Further it is also applied for the evaluation of pressure buildup curves and for the water inflow into hydrocarbon reservoirs. 2.2.31) The well should produce continuously at a rate q.t) for r w < r < ∞ t > 0 (3.390 3: Solutions of the SinglePhase Equation of Filtration 3.1 Radial Systems with Constant Production Rate Eq. According to Darcy’s law we obtaine the following equation: 2πr w hk ∂p . The radius of the well is rw. homogeneous. It is a base for calculations of production rates of hydrocarbon wells and deep water wells. The fluid has a viscosity µ and a compressibility c.2 NonSteady State Filtration in Infinite Acting Systems 3. It is assumed that the fluid bearing layer has a thickness h.33) (3.58 transformed into the radial yields: 1 ∂p 1 ∂p + = r ∂r K ∂t ∂r ∂ p 2 2 k K = µφc (3. a permeability k and is horizontal. a porosity φ. At t = 0 the pressure is the same everywhere: p = pi r > rw t = 0 (3.34) . = – qB µ ∂ r r = rw The boundary condition in infinity is: p = pi r = ∞ t>0 (3.30) This equation is of essential importance when solving technical problems.32) We seek for a solution in the form: p = p (r. isotropic and infinite.
1 dp 2z dp = = = . the first B. 3. in Eq. 3.dp = – .39) (3. It is possible to reduce the number of independent variables by introducing a new one in the form of: w = x +y 2 2 (3. y ) (3.35) where x and y are the side lengths and z is a function of two variables.dp 2 2dz K∂t K ∂z ∂t K t Kt d z ∂p dp ∂z 2r dp 2r .+ 2 Kt d z K 2 t 2 d z 2 Kt d z Kt d z 2 ∂r Substituting Eq.dp 4z d p = .= dz ∂r Kt d z Kt r d z r dz ∂r ∂ p 2.41 into Eq. 3. in Eq.+ = . To calculate the hypotenuse of a right triangle we are looking for a solution of: z = z ( x. 3.42) To solve Eq. x and y.40.C.Eq.37) In a similar manner we assume that p is a function of only one variable z and we introduce the following relationship: r z = K⋅t Based on this assumption we get: 1 1 r z.32 can be transformed as well to: z dp µqB= – dz 4πhk (3.43) Using Eq.38) (3.41) + (4 + z) dp = 0 dz (3.40) (3. 3. 3.∂p ∂z = .dp 4r d p 2.38 the B.30 yields: 4z d p dz 2 2 2 2 2 2 2 2 2 (3. Using Eq.3: Solutions of the SinglePhase Equation of Filtration 391 To explain the solution method used below we take Pythagoras theorem as an example.42 the boundary conditions must also be transformed.35 can be replaced by z = z(w) (3. 3.36) So the variables x and y in Eq. 3.38 and Eq.44) .C.39 . 3.33 can be transferred into: p = pi z = ∞ t>0 (3. 3.∂p = .
392 3: Solutions of the SinglePhase Equation of Filtration Eq.45) Let y = z then Eq. Eq. 3.z d z dz 4 dz dp dz (3.dξ ξ –ξ (3.49) (3. 3.51) and the second substitution with the notation ξ = z ⁄ 4 and by separation of variables. taking the exponential of Eq.42 can be written as: z d p dz or: d dp 1 dp .z .46) 2 2 + dp z dp = – dz 4 dz (3. 3.52) .50) (3.48) (3.49 y = A1 e or z –z ⁄ 4 dp = A1 e dz –z ⁄ 4 (3.51 becomes: e dp = A 1 . 3.47) (3. 3. = – .48 yields: z lny – lnA 1 = – 4 where A1 is a constant of integration.46 becomes: d y = – y/4 dz By separation of variables: d y y = – dz 4 By integrating Eq.
This integral is defined as the so called exponential integral and its numerical solution can be found in any mathematical handbook: ∞ e Ei ( – x ) = – ∫ .53) Using Eq.56) The calculation of the pressure drop at the well bottom is made by substituting r = r w and p (r. 3.t) = pwf ( t ) into Eq.54) 2 The integral in Eq.Ei – .53 becomes: µqBr p i – p ( r.dξ ξ x –ξ (3. t ) = .57) . r w  p i – p wf = .51 and the boundary condition in Eq.52 yields: ∞ pi – p = A1 z⁄4 ∫ e . 3.Ei – . 4Kt 4πhk 2 (3.44 and assuming that rw is very small rw so that the condition . 3.56: µqB. 3. 4πhk 4Kt 2 (3.55) Finally Eq. 3. 3.53 cannot be solved in a closed form. A 1 can be determined as: 4Kt µqBA 1 = – 4πhk (3.dξ ξ –ξ (3.3: Solutions of the SinglePhase Equation of Filtration 393 Integrating Eq.≈ 0 is valid.
If z « 1 the series in Eq. = – 0. In the vicinity of z = 0 the TAYLORSeries of – Ei ( – z ) is defined as: z– Ei ( – z ) = – γ – ln z + z – .function (after Chaumet) 3. 57722 – ln .58) (3.. 57722 – ln z then: r 4Kt – Ei – .59) Kt 2 r (3.58 will have very small values in terms higher than the third term. 80907 + ln 4Kt 2 r 2 2 (3.60) .+ .= 0. 3.2 Properties of the EiFunction The function – Ei ( – z ) is illustrated in Figure 3. 57722 is the EULERConstant. 4 where γ = 0..4: Plots of the Ei(z).2.394 3: Solutions of the SinglePhase Equation of Filtration Figure 3. which makes the following approximation valid: – Ei ( – z ) = – 0..4.
0. 80907 + ln 4πhk µqB p i – p e = – . 3.61) The error of this approximation becomes less than 1 % if the following condition is satisfied: Kt . A graphical illustration of these values indicates a parallelism of the pressure drop curves.56 leads to: µqBp i – p ( r. 3. 3. 0. This fact can be verified mathematically with Eq. . 3.ln 4πhk re µqB. The pressure is evaluated at the the two radii seperately using Eq.63 we get: µqBp e – p wf = – .ln 2 2πhk rw 2 re rw (3. 3.64 from Eq.> 10 2 r (3. 3.62) Eq.64) By substracting Eq.> 10 . 0.= – . This means that the pressure difference between two radii becomes a constant for a given production time interval.2.65) The pressure difference is therefore independent of time but Eq.61 where rw and re are the two considered radii: µqB.60 into Eq.61 enables the calculation of the pressure drop between two selected radii. p i – p wf = – . 2 r Further it is possible to illustrate the spatial distribution of pressure according to Eq.63) Kt  2 re (3. 3. 3. 3.3: Solutions of the SinglePhase Equation of Filtration 395 3.65 is only valid if Kt . 80907 + ln 4πhk Kt 2 r (3.3 Pressure Drop in Space and Time Substitution of Eq.5. t ) = – .61 as shown in Figure 3. 80907 + ln 4πhk Kt  2 rw (3.61 and the difference gives the pressure drop between the two points.
61.6: Plots of pressure drop in the vicinity of a well (infinite reservoir.396 3: Solutions of the SinglePhase Equation of Filtration Figure 3. compressible fluid) Figure 3. the staggered line calculations by Eq. 3. compressible fluid) Figure 3. The continuous line represents calculations by Eq.6 shows the pressure distribution as a function of log r for various t. . 3.5: Plots of pressure drop in the vicinity of a well (infinite reservoir.56. The linear section of the plot is shifted parallel in time.
= 1 [sec] K 0.1015 [m2] r w = 0. 6 [days] t2 > K 0. 1 2 At a 100 m distance from the well the time limit for application of Eq. 3.61 in the following cases: a) For calculation of the bottom hole flowing pressure b) For calculation of the pressure at 100 [m] distance from the well Solution: At the well radius 3.1: The following data of an oil bearing layer are known: Permeability: Porosity: φ =0.= 10 [sec] = 11. 3.= . 1 2 .2[] Thickness: Well Radius: Oil Viscosity: h = 5 [m] k = 10 [mD] = 20.s] Oil Compressibility: c = 109 [Pa1 (104 bar1)] Production rate: q = 10 [m3/day] (Bo=1) For how long should the well be produced to allow the use the Eq.61 is valid if : –2 10r w 10 ⋅ 10 t 1 > .1 [m] µ = 103 [Pa.= 10 ⋅ 10 .61 instead of Eq. 3.57 will be much higher: 4 10r w 6 .3: Solutions of the SinglePhase Equation of Filtration 397 Example 3.
92 ⋅ 10 0. 2 – 12 Time is large enough to apply Eq.= – 0. 0. Solution: Since: – 3 – 10 10 5 86400 µqB. 0. 80907 + ln 4πhk Kt  2 r w 6 5 0. 35 [MPa] 2 0.2: The task is to determine the pressure at the well radius and at r = 100 [m] after 30 days of production. The bottom hole flowing pressure is calculated as follows: µqB. p ( r = 100 ) = p i + .= . 36 [MPa] 2 100 . 92 ⋅ 10 [Pa] – 12 4πhk 4 ⋅ π ⋅ 5 ⋅ 0. 02 ⋅ 10 = .1. The reservoir data are to be taken from Example 3. 92 ⋅ 10 0. 1 ⋅ 30 ⋅ 86400 = 10 ⋅ 10 – 0. pwf = p i + . 1 [m sec ] –3 –9 µcφ 10 ⋅ 10 ⋅ 0. 80907 + ln  = 9. 1 µqB. 80907 + ln 4πhk Kt  2 r w 6 5 0.= .61. 1 ⋅ 30 ⋅ 86400 = 10 ⋅ 10 – 0.= 0. 3. 02 ⋅ 10 2 –1 k 0.398 3: Solutions of the SinglePhase Equation of Filtration Example 3. 80907 + ln  = 8.
3. centered by the well inside an infinite.40.  . 3. 3.7: The flow rate in function of the dimensionless variable Kt/r (after Chaumet) 2 . t ) = – Bµ ∂ r Eq. t ) = –qe (3. 3. 3.3: Solutions of the SinglePhase Equation of Filtration 399 3. 3.66) z – 4 (3.t)/q is illustrated as a function of Kt/r2.A 1 Kt d z Kt z ∂r respectively from Eq.68) Substituting Eq.4 The Spatial Distribution of Flow We assume a cylinder with radius r.67) µqB = .51 yield: ∂p 2r dp 2r e = = . 3.e 2πrhk r – 4Kt 2 (3. homogenous and isotropic layer. Eq.67 and Eq.Kt = .68 into Eq.38 ∂p 2r µqB. The quantity of fluid flowing through the cylindrical surface is a function of time and may be calculated from Darcy’s law: 2πrhk ∂p q ( r.2. Figure 3.7 the relation q(r.38 and Eq.e ∂r Kt 4πhk r 2 r – 4Kt 2 (3.66 gives: r – 4Kt 2 q ( r.69) In Figure 3.
Eq. K 2 ∂ 2 (3.75) (3.72 into Eq.71) (3.30 can be done and as rw.70) (3. 3.30 and Eq.73) PD C + pi = pi rD r w > rw tD rw . 3.33 leads to the following formulas: ∂ ∂ 1 1 [ P D C + p i ] + [ P C + p i ] = [ PD C + p i ] 2 rD rw ∂ ( rD rw ) D K ∂ ( rD r w ) t D r w ∂ . c.≥ 0 K 2 (3. Thus Eq.3 Dimensionless Variables Simplifications of the calculation may be achieved by introducing the so called dimensionless variables. 3.76) A similar transformation of the variables in Eq.72) Substitution of the variables given in Eq.76 become: . 3.74) 2πr w hk ∂ [ P C + pi ] = –q Bµ ∂ ( r D r w ) D PD C + pi = pi rD r w = ∞ t D rw . K are constants and therefore can be cancelled. 3.Eq.Eq. 3.3100 3: Solutions of the SinglePhase Equation of Filtration 3. pi. r radius: r → r D = rw Kt kt time: t → t D = . 3.31.= 2 2 φµcr w rw 2πhk pressure: p → PD = ( p – p i ) qBµ Thus r = r D rw rw t = t D K qbµp = P D  + p i = P D C + p i 2πhk 2 (3.70 .73 . 3.= 0 K 2 (3.
The logaritmic approximation (for tD/rD2>10) of Eq.80) Solutions of this boundary value problem are obtained by using Eq.∂PD + = 2 r D ∂ rD ∂ tD ∂ rD PD = 0 rD > 1 tD = 0 tD > 0 tD > 0 2 (3.Ei – .[ 0 80907 + ln ( t D /r D ) ] 2 (3.[ 0. 2 4 ( t ⁄ r 2 ) D D 2 (3.85) .84 becomes more simple: 1 P Dw = .. 3.Ei –  2 4t D 2 Thus PD may be regarded as a function of t D ⁄ r D : 2 1  1 1 P D ( t D ⁄ r D ) = – .79) (3.78) ∂P D = –1 ∂ rD rD = 1 PD = 0 rD = ∞ (3.82 becomes: 2 1 · P D = .83) Figure 3. 3.84) At the wellbore r D = 1 .Ei – 2 4Kt ⁄ r 2 Bqµ w Substituting the dimensionless variables into Eq.82) (3. 3.( p i – p ) = . Rearranging of this equation leads to: 2 2 1 r ⁄ rw 2πhk . 3.56.81 yields: 1 rD P D = – .82 with r e ⁄ r w = ∞ . so Eq.77) (3.3: Solutions of the SinglePhase Equation of Filtration 3101 ∂ PD ∂P D 1. 3.8 shows the function PD according to Eq. 80907 + ln t D ] 2 (3.81) (3.
8: Solution for the infinitive and finite radial filtration problem with closed boundary and constant pressure drop (after Van Everdingen and Hurst) .3102 3: Solutions of the SinglePhase Equation of Filtration Figure 3.
Solution: r r D = .= . 3.3: Convert the following variables to dimensionless variables using the data in Example 3.4: Calculate the pressure at a distance of 100 [m] from the well after 5 days of production.63[MPa] .15. Use the data given in Example 3.= 4.= 1000 rw 7 Kt 0.83 or the graphical illustration of the Ei function in Figure 3. 84 ⋅ 10 [Pa] 2πhk From the ( r e ⁄ r w = ∞ ) curve in Figure 3. 32 2 r2 r 2 r2 100 w w K = 0. Solution: At this time the approximations used in Eq.61 and Eq.= D = ( p – p i ) qµ = ( 9.= – 1. The first task is to calculate the dimensionless variables: Kt r 2 Kt 0. PD =1.= .1. so one must use Eq. 3. 1 ⋅ 5 ⋅ 86400 2 ( t D ⁄ r D ) =   = . r = 100[m].3. 59 ⋅ 10 2 2 rw 0.8. 1 [m sec 2 –1 ] 5 Bqµ. .8.3: Solutions of the SinglePhase Equation of Filtration 3103 Example 3. 54286 ⋅ 10 6 –5 = – 2. 02 ⋅ 10 2πhk . 01 Example 3. 3.= 2. 1 6 6 2 ⋅ π ⋅ 5 ⋅ 0.1 .Example 3.84 are not applicable. 1 ⋅ 30 ⋅ 86400 t D = . 37 ⋅ 10 ⋅ 0. 63 ⋅ 10 – 10 ⋅ 10 ) – 3 – 10 10 86400 – 12 t = 30[days] p = 9.
3. 79 [MPa] 2πhk 2 (3.72 follows: 6 5 qBµp = p i + p D .86) .Ei –  = 1.3104 3: Solutions of the SinglePhase Equation of Filtration Since: 1 rD P D = – .= 10 ⋅ 10 – 1. 15 = 9. 15 2 4t D From Eq. 84 ⋅ 10 ⋅ 1.
90) (3. 3.88) (3. 3. rather than under a constant production rate. These two variables are both regarded as functions of time. 3. 3. The main interest though appears to lie in the determination of the flow rate q or the cumulative production Q at the well radius rw. Because the bottom hole pressure pwf and the pressure in infinity are both given and thus pressure at r will range between these two values. only PD is different.31 and Eq. .∫ r = r dt . 3.t 2 2 rw µφcr w pi – p pi – p = .= . The function p(r.92) rD and tD are similar to those used in Eq.87) The flow rate is calculated by Darcylow (see Eq.91) (3. for example to control water conning problems. The boundary condition in Eq.32): 2πr w hk ∂p q ( t ) = – .30 and the conditions in Eq.33 are still valid. Also note that Eq.t) is no longer of interest for us in this case.56 is not applicable in this case since the production q is not constant. Bµ ∂ r r = r w and the cumulative outflow by integrating Eq.71.32 though is substituted by: p = p wf r = rw t>0 (3.88: 2πr w hk ∂p Q ( t ) = ∫ q ( t ) dt = – . 3. ∂ r Bµ w 0 0 t t (3.= D p i – pwf ∆pwf (3.4 The Infinite Radial System with Constant Pressure at the Interior Boundary Sometimes the well is produced under a constant bottom hole flowing pressure (pwf).70 and Eq. 3.3: Solutions of the SinglePhase Equation of Filtration 3105 3. 3. Eq.89) Again dimensionless variables are applied: r r D = rw Kt k t D = . 3.
V.31 and Eq.Eq.97) ∂P D 2 2πhφr w c∆p wf ∫ dt D ∂ rD rD = 1 0 Let: tD QD ( tD ) = ∫ ∂ rD r 0 ∂P D D =1 dt D (3.12 (for r e ⁄ r w = ∞ ) as a function of tD then: ( t ) = 2πhφr w c∆pwf Q D ( t D ) Where ∆p wf = p i – p wf 2 (3.9 to Figure 3.99) . 3.) described by Eq. Eq. 3. 3.95.Eq.92 are then substituted into Eq.3106 3: Solutions of the SinglePhase Equation of Filtration Eq.C.C.92. 3.79 is replaced by the B. 3. Since QD(tD) can be obtained from Table 3. 3. in Eq. 3.96) The above boundary value problem (B. except that the B.9 Figure 3. 3. described by Eq. 3.90 . in Eq. 3.∂P D + = 2 rD ∂ rD ∂ tD ∂ rD PD = 0 PD = 1 PD = 0 rD > 1 rD = 1 rD = ∞ tD = 0 tD > 0 tD > 0 2 (3.P.30.3 at the end of Chapter 3 or from Figure 3. 3.33.C.93 .93) (3.89. 3. 3. 3. Substituting the dimensionless variables in Eq. The following equations are obtained: ∂ PD ∂P D 1.96 is similar to the B.V. leads to: 2πr w hk Q ( t ) = – . using Eq. in Eq.dtD = k D 0 2 (3.98. The B.C.94) (3.Eq.98) Figure 3.12 show the graphical solution of Eq.95) (3. Bµ tD tD Bµφcr w ∂ ∫ ∂ ( rD rw ) [ pi – ∆pwf PD ] r = 1 .80.87 can be converted to dimensionless B.77 .P. 3.
Q D ( t D ) = 8. 895 ⋅ 10 6 The cumulative oil production after 100 days will be: = 2πhφr w c∆p wf Q D ( t D ) = 2 –9 6 6 3 ⋅ π ⋅ 5 ⋅ 0. Solution: The dimensionless time results in: 7 Kt 0.100 ⋅ 86400 = 8. 1t D = .= .3).1.3: Solutions of the SinglePhase Equation of Filtration 3107 Example 3. 1 From the Van Everdinger and Hurst solution (Table 3. 1 ⋅ 10 ⋅ 2 ⋅ 10 ⋅ 8. 64 ⋅ 10 2 2 rw 0. 2 ⋅ 0.9: Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder) . 2 [m ] 2 Figure 3.5: Consider the well data from Example 3. 895 ⋅ 10 = 1117. The bottom hole flowing pressure is fixed at: p wf = 8 [MPa] Calculate the cumulative production after 100 days.
.3108 3: Solutions of the SinglePhase Equation of Filtration Figure 3.11: Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder).10: Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder) Figure 3.
.3: Solutions of the SinglePhase Equation of Filtration 3109 Figure 3.12: Solution for the infinitive and finite radial filtration problem with closed boundary and constant bottom hole pressure (after Silder).
77. This operation is called LAPLACE transformation and is a little bit complicated.3110 3: Solutions of the SinglePhase Equation of Filtration 3.∂P D + = 2 ∂ tD rD ∂ rD ∂ rD ∂ PD PD = 0 rD > 1 tD = 0 tD > 0 2 (3. Only the results are given here: . t D ) (3.104) This solution can be achieved by transforming Eq. 3.72 are again applied. 3.1.101) ∂P D = –1 ∂ rD rD = 1 ∂P D ∂ rD = 0 r D = r eD (3. the initial condition in Eq. The differential equation given by Eq.5 NonSteady State Filtration in a Finite System 3. 3. followed by integration and retransformation.1 3.100) (3.100 .103 in the form of: D = P D ( r D. 3.103 into the complex plane.5. 3.79 are further valid. 3. In this case the boundary is not considered to be located in infinity but at a finite distance with radius reD: ∂P D 1.103) We now seek a solution for the boundary value problem described by Eq.70 . 3.5.Eq. 3.102) tD > 0 (3. The only difference can be found at the exterior boundary.1 Constant Production Rate Closed Exterior Boundary The dimensionless variables of Eq.Eq. 3.78 and the boundary condition in Eq.100 .Eq.
+ t D 2 r eD – 1 4 r eD ln r D 3r eD – 4r eD ln r eD – 2r eD – 1 – . (3.3: Solutions of the SinglePhase Equation of Filtration 3111 2 rD .107) The roots increase monotonous.P D = . . if n increases. This means: e 2 –α1 tD >e 2 – α2 tD >e 2 –α3 tD This is the cause for a monotonous decrease of the exponential factor. If r w « r e it is possible to write the well flowing pressure formula in a more simple form: 2t D 3 P Dw = . 3.[ ln t D + 0.109) . The approximation for large tD is: 2t D 3 P Dw = .103 will have an identical solution which is: 1 P Dw = .Eq.+ ln r eD – 2 4 r eD (3.80 and Eq. 3.105) 2 2 2 2 J 1 ( α n r eD )Y 1 ( α n ) – J 1 ( α n )Y 1 ( α n r eD ) = 0 where Ji and Yi are the Bessel functions. From this one may conclude that if tD is large enough this sum will become very small and thus can be neglected. 3.– 2 2 2 r eD – 1 4 ( r eD – 1 ) J 1 ( α n r eD ) [ J 1 ( α n )Y 0 ( α n r D ) – Y 1 ( α n )J 0 ( α n r D ) ] e + π ∑ 2 2 α n [ J 1 ( αn r eD ) – J 1 ( α n ) ] n=1 α n are the roots of the equation: ∞ –αn tD 2 2 2 (3.108) If the pressure disturbance has not yet reached the exterior boundary.+ ln r eD – 2 4 r eD –αn tD 2 e J 1 ( α n r eD ) +2 2 2 2 α n [ J 1 ( α n r eD ) – J 1 ( α n ) ] n=1 ∞ 2 ∑ (3. Eq. 3. 80907 + Y D ( t D ) ] 2 (3.Eq.77 .100 .8 for different values of reD.105 is expressed graphically in Figure 3.106) Eq. 3.
100 . then Eq. 3.1.102 and Eq. if the exterior boundary is not yet reached. but the boundary condition in Eq. 3.100 . 3.Eq.85 and Eq. 80907 ) + . 80907 + ln 4πhk Kt . Eq. 3.+ Y ( t ) 2 rw (3.Eq.112) The solution of the boundary value problem is described by Eq. 3.111) 3.110 is the combination of the equations Eq. 3. 3.110) ∞ 2 ∑ Eq.113 becomes: . It is evident that the function YD(tD) will become zero. With dimensioned variables Eq. 3.102 remain the same. 3. 0. but a constant pressure at the exterior boundary.113. p wf = pi + .+ 2 ln r eD –  + 2 4 r eD – αn tD 2 J 1 ( α n r eD ) e 4 2 2 2 α [ J ( α r ) –J 1 ( αn ) ] n = 1 n 1 n eD (3.2 Boundary with Constant Pressure If the system has an open boundary.5.107.112 considering again that r w « r e is given by: – βn tD 2 e J 0 ( β n r eD ) 2 2 2 2 β [ J ( β ) – J 0 ( β n r eD ) ] n=1 n 1 n ∞ 2 P D = ln r eD – ∑ (3.109 becomes: µqB. 3. The solutions of the finite and the infinite systems are therefore identical for a given time t . 3.3112 3: Solutions of the SinglePhase Equation of Filtration where: 4t D 3 Y D ( t D ) = – ( ln t D + 0.113) β n are defined as the roots of this equation: J 1 ( β n )Y 0 ( β n r eD ) – Y 1 ( β n )J 1 ( β n r eD ) = 0 It tD is large enough it is possible to neglect the summation term in Eq. 3.103 is substituted by: PD = 0 tD > 0 r D = r eD (3.
V.12 for different ratios of r eD = r e ⁄ r w .118) The solution of the above B.P. Again the dimensionless pressure form is used.4: The rate at the interior boundary (e.Figure 3. 3. but the dimensionless cumulative inflow QD(tD). Example 3.5 and the exterior radius (re) of the reservoir was estimated as 320 [m]. tD is 8.6: Consider the data given in example 3. 3.64 107 From Figure 3.9a: QD= 3.3: Solutions of the SinglePhase Equation of Filtration 3113 re P D = ln rw which is identical with Eq. The differential equation and the boundary conditions may then be written as follows: ∂ PD ∂P D 1. is given by Eq.5.2 Constant Pressure at the Interior Boundary and Closed Exterior Boundary Just as in Section 3.∂PD + = 2 r D ∂ rD ∂ tD ∂ rD PD = 0 PD = 1 ∂P D ∂ rD = 0 1 ≤ r D ≤ r eD rD = 1 r D = r eD tD = 0 tD > 0 tD > 0 2 (3.5 106 2 3 The cumulative production will be: Q = 2πhr w ∆p wf Q D ( t D ) = 440 [m ] .117) (3. is determined from Figure 3. but the pressure pwf is set at a certain value.99. at the well radius) is not fixed.20. Calculate the cumulative production after 100 days.g.116) (3.115) (3.5.9 .114) 3. (3. Solution: From the solution of example 3.
Ramey and Raghavan used a flow model "Uniform Flux Fracture" which is the first approximation to the behavior of a vertically fractured well. where the fluids flow linearly to the fracture or the wellbore of the horizontal well. 3.erf . so that there is a pressure drop created by the fracture.3114 3: Solutions of the SinglePhase Equation of Filtration 3. .123) .119) ∂p Bµ = – q ∂x Ak p ( ∞. Fluid enter the fracture at a uniform rate per unit surface area of the fracture.6.121) (3. .1 Linear Flow with Constant Production Rate For a vertically fractured wells. the fracture intersects the wellbore perpendicularly.6 NonSteady State Filtration in Linear System Linear flow systems are applicable in the case of a massive fracture or a horizontal well. 2 Kt Ak x f 4Kt where: is half length of the vertical fracture and xf A is the surface area of the flow 2 (3. Consider a one dimensional flow problem described by: ∂ p ∂x 2 2 1 ∂p = K∂t x = 0 x = ∞ x>0 t>0 t>0 t = 0 (3.120) (3. t ) = p i p ( x. or horizontal wells in an infinite acting reservoir. – Ei . the fluids will flow linearly to the vertical fracture or the horizontal well especially at the beginning of the flow and after a relatively short time period.122) Gringarten. 0 ) = p i (3. This pressure difference can be calculated from the following equation: xf –x f qBµ 2πKt p – p i = .
= 0. 3. Example 3. 2 ⋅ 10 ⋅ 10 – 12 A) After 5. 1 ⋅ .2 and 15. 1 . Figure 3. 3.125 can be approximated by 1 pD = .6 min. 2 ⋅ 60 t D1 = K .1 Eq.1 after 15.125 becomes: pD = πt Dx f (3. t1 3 5. 3.13 shows the relation between pD and tDxf. Solution: 2 –1 k 0. f 2 t Dx 2 4t Dx f f (3.Ei .1 a half slope straight line is obtained which indicates linear flow behaviour.7: A) Calculate the pressure drop created by a vertical fracture having a length of 5 [m] after 5.1. For t Dxf > 0. Use the data in example 3. B) Compare the pressure drop in part A with the pressure drop.= 3.6 min.123 to a dimensionless equation the following dimensionless variable is introduced: r w t Dx = t D  f xf 2 (3.3: Solutions of the SinglePhase Equation of Filtration 3115 In order to convert Eq.127) which indicates that during a short time the flow into the fracture is linear.126) (3.2 min.124) using the definition of pD and tD the dimensionless pressure is obtained by: pD = 1 1 –1 πt Dx erf  – .= . 1 [m sec ] –3 –9 φµc 0. 02 ⋅ 10 = . created by the unfractured well in example 3.125) For t Dxf > 10 Eq.80907 ] f 2 For t Dxf > 0. 12 ⋅ 10 2 2 rw 0.[ ln t Dx + 2.= 0.
12 ⋅ 10 ⋅  2.: tD .6 min.25 The pressure drop after 5.6 min. 6 ⋅ 60 t D2 = K . 36 ⋅ 10 2 2 rw 0. 1 r w = t D2  xf 2 –3 t Dx f2 = 15 Since tDxf > 10 .= 0.( 2. 80907 ] = 2. is: µBq∆p = . 14 ⋅ 5 ⋅ 0. then pD2 can be calculated from: 1 pD2 = . is: µBq( – 10 ⁄ 184300 )10 ∆p1 = . 25 ) = 415 [kPa] – 12 2πhk 2 ⋅ 3.2 min. 02 ⋅ 10 After 15. 1 ⋅ .13. 76 f 2 Then the pressure drop after 15. 1 = 3.> 10 .1 and after 15.p D1 = .p D2 = 509 [kPa] 2πhk B) For the unfractured well in example 3.3116 3: Solutions of the SinglePhase Equation of Filtration t Dx f1 r w = t D1  xf 2 3 0.: t2 3 15. PD1 = 2.[ ln t Dx + 0.= 9.6 min. 5 2 = 5. 0 From Figure 3. then pD can be calculated from: 2 rw 1 pDw = .[ ln t D + 0. 43 2 . 80907 ] = 4.
Figure 3. 76 ∆p f p Df The pressure drop created by the unfractured well.= . 61 2.= 4.13: Dimensionless pressure for single fractured well in an infinite acting system (after Gringarten. is 1. 43 = 1. and Ragavan) .6 min.3: Solutions of the SinglePhase Equation of Filtration 3117 p unf ∆p unf .6 times higher than the one created by the same well with a single vertical fracture. Ramey.. after 15.
1: Solution of Transient Filtration in the Case of Infinite Radial System Mathematical Formulation Distance Dimensioned Dimensionless r r D = rw Kt t D = 2 rw 2πhk P D = ( p – p i ) qBµ 2 r Time t Pressure p Differential Equation 1 ∂p 1 ∂p + = 2 r ∂r K ∂t ∂r ∂ p 2 ∂P D 1. rD = ∞ 2.∂P D + = 2 r D ∂ rD ∂ tD ∂ rD ∂ PD Initial Condition Boundary Conditions 1. r ≥ rw .Ei – . 80907 + ln 2 D 2 r tD 2 rD 2 Solution in Case of Kt . = – qB µ ∂ r r = r w p = pi . t ) = .Ei –  2 4t D r2 D Kt  P = 1 0.3118 3: Solutions of the SinglePhase Equation of Filtration Table 3. r = ∞ ∂P D = –1 ∂ rD rD = 1 PD = 0 .) Well Radius p = pi . 80907 + ln 4πhk . 0. rD ≥ 1 .) Outer Boundary Solution µqBr p i – p ( r. t ) = – . t = 0 PD = 0 . tD = 0 2πr w hk ∂p .≥ 10 2 r µqBp i – p ( r. 4πhk 4Kt 2 tD 1 rD P D  = – .
t D = 0 Initial Condition Boundary Conditions 1. Trivial [ P D = 1 . rD = ∞ . 80907 + ln t D ] 2 PD = 1 . rD = 1 .2: Solution of Transient Filtration in the Case of Infinite Radial System Conditions Dimensionless Variables: Distance r r D = rw Kt t D = 2 rw 2πhk P D = ( p – p i ) qBµ r r D = rw Kt t D = 2 rw pi – p P D = p i – p wf 2 Constant Production Constant Borehole Pressure Time Pressure Differential Equation ∂P D 1. tD > 0 Borehole Pressure rD = 1 . rD = ∞ . t D > 0 1 P D = .[ 0. r D = 1 ] tD Cumulative Influx QD ( tD ) = 0 ∂P D ∫ ∂ rD r = 1 dtD D 2 Trivial [ Q ( t ) = qt ] Q ( t ) = 2πhφr w c∆p wf Q D ( t D ) . tD > 0 ∂ rD rD = 1 PD = 0 . tD = 0 2 ∂P D 1. tD > 0 . tD > 0 . PD = 0 .) Well Radius 2.∂P D + = 2 r D ∂ rD ∂ tD ∂ rD ∂ PD PD = 0 . rD ≥ 1 .) Outer Boundary ∂P D = –1 . rD ≥ 1 .3: Solutions of the SinglePhase Equation of Filtration 3119 Table 3.∂PD + = 2 rD ∂ rD ∂ tD ∂ rD ∂ PD PD = 0 .
3. Eq.128) where C is a constant.7 The Principle of Superposition In mathematics the superposition theorem states that any sum of individual solutions of a linear differential equation is also a solution of this differential equation. 3. For all single pressure changes the cummulative production (influx) can be calculated by Eq. 3.128 can be written in the following form too: n We ( t ) = C ∑ j=0 ∆p wf . According to the theorem of superposition the overall influx caused by the n subsequent finite pressure drops is the sum of the elementary solutions: n We ( t ) = C ∑ ∆pwf QD ( tD –tDj ) j=0 (3. The pressure derivative can be approximated piecewise by finite differences: dp wf dτ pj – pj – 1 = t Dj – t Dj – 1 t Dj – 1 ≤ τ ≤ t Dj (3. 3.3120 3: Solutions of the SinglePhase Equation of Filtration 3. The pressure can then be approximated by a step function.1 The First Law of Superposition Consider a well producing with a continuously changing bottom hole pressure as it is shown in Figure 3.Q D ( t D – t Dj )∆t D ∆t D (3.131) .7.128 is the original form of the van EverdingenHurst solution.99. Eq. the superposition theorem is considered to be one of the most powerful tools to get the solutions of complex flow problems without solving the differential equation for different boundary conditions over and over again.14. In practice.129) or after replacing the summation by integral: tD We ( t ) = C ∫ 0 dp wf dτ Q D( t D – τ ) dτ (3. Vogt and Wang improved this model assuming piecewise linear pressure change instead of a stepping one. but for different boundary conditions.130) τ is the integration variable.
3..+ t Dn – t Dn – 1 Q D ( t Dn – τ ) dτ ∫ t Dn – 1 t Dn Let u = t Dn – τ .132 can be rewritten as follows: t Dj t Dn – t Dj ∫ – Q D ( t Dn – τ ) dτ = ∫ ∫ 0 Q D ( u ) du = (3..– .3: Solutions of the SinglePhase Equation of Filtration 3121 and then the integral in Eq.– .∫ Q D ( t Dn – τ ) dτ + t D2 – t D1 t D1 0 t D1 (3.132) pn – 1 – pn . Q D∗ ( t Dn – t Dn – 1 ) t Dn – 1 – t Dn t Dn – 1 – t Dn – 2 . 3..∫ Q D ( t Dn – τ ) dτ + .135) t D1 t D1 t D2 – t D1 pn – 2 – pn – 1 pn – 1 – p n .134) Eq.. and the integrals in Eq.Q D∗ ( t Dn ) + ...132 yields: p 1 – p 2 p 0 – p 1 p0 – p1 W e ( t n ) = – C .. + . 3. then du = – dτ .133) t Dj – 1 t Dn – t Dj t Dn – t Dj – 1 t Dn – t Dj – 1 ∫ 0 Q D ( u ) du + Q D ( u ) du Introducing the integral of the QD(tD) function tD Q D∗ ( t D ) = ∫ QD ( u ) du 0 (3..131 can be splitted into n terms: t D1 t D2 p 1 – p2 p0 – p1 W e ( t n ) = – C .. Q D∗ ( t Dn – t D1 ) + (3.
.137) ∑ j=0 ∆p j Q D ( t Dn – t Dj ) = ∑ j=0 ∆p j ∆  Q D∗ ( t Dn – t Dj ) if n → ∞ ∆t Dj (3.138) Figure 3. + .135 becomes: ∆p 2 ∆p 1 ∆p 1 W e ( t n ) = – C .– ..14: Variable production rate in case of a ideal reservoir (after Hurst) . 3.136) = –C ∑ j=0 ∆p j ∆  Q D∗ ( t Dn – t Dj ) ∆t Dj where ∆pj ∆p j – 1 ∆p j ∆ . 3.– ∆t Dj ∆t Dj – 1 ∆t Dj Comparing Eq.128 it is evident that: n n–1 (3. Q D∗ ( t Dn – t D1 ) + t D1 ∆t D2 t D1 ∆p n ∆pn – 1 . 3.. = .136 with Eq.– ..3122 3: Solutions of the SinglePhase Equation of Filtration Let be ∆p j + 1 = p j – p j + 1 and ∆t j + 1 = t j – t j + 1 then Eq. Q D∗ ( t Dn – t Dn – 1 ) ∆t Dn ∆t Dn – 1 n–1 (3.Q D∗ ( t Dn ) + .
144) (3.139) If the rate of well 1 is set to q 1 = 1 then the pressure drop is defined as ∆p 1∗ .56 yields again: r1 µB .146) . 3.15 illustrates the pressure change of two wells inside an infinite reservoir. 3.∆p 2∗ = . Analogous the formulas for well 2 producing with a rate q 2 = 1 and starting at a time t 2 may be set up: r2 µB .2 The Second Law of Superposition Figure 3.3: Solutions of the SinglePhase Equation of Filtration 3123 3. at a time t1.139 and Eq. The pressure change at any time t > t1 at point R can be calculated using Eq.141) Using Eq.140) (3.Ei – . 4πhk 4K ( t – t 1 ) 2 (3.Ei – . Eq.Ei – . 3.∆p 1∗ = .7.Ei –  µB 2 4 ( t D – t D1 ) 2 2 (3. 4πhk 4K ( t – t 1 ) or dimensionless: r D1 2πhk 1 P D1∗ = . 3. At first well 1 is put on production.145) (3. 4πhk 4K ( t – t 2 ) r D2 2πhk 1 P D2∗ = . ∆p 2∗ = . ∆p 1∗ = .140 it is trivial to state that if well 1 producec at a constant rate q 1 ≠ 1 the pressue change at point R will be proportinal to q 1 : ∆p 1 = q 1 ∆p 1∗ and ˆ P D1 = q 1 P D1∗ (3.Ei –  µB 2 4 ( t D – t D2 ) ∆p 2 = q 2 ∆p 2∗ 2 2 (3. with rate q1.56: q 1 µB r1 ∆p 1 = .142) ˆ where q 1 is equal to q1 but it is dimensionless.143) (3.
4πhk 4K ( t – t 1 ) 4K ( t – t 2 ) 2 2 (3.147) (3.15: Pressure change at point R in infinite reservoir.148) (3.3124 3: Solutions of the SinglePhase Equation of Filtration and ˆ P D2 = q 2 P D2∗ If both wells are under production the pressure changes at point R may be added: ˆ ˆ P D = P D1 + P D2 = q 1 P D1∗ + q 2 P D2∗ ∆p = p i – p r ( t ) = q 1 ∆p 1∗ + q 2 ∆p 2∗ = r1 r2 µB . q 1 Ei – . + q 2 Ei – . with two production wells .149) Figure 3. .
148 . 3. As shown in Figure 3. Bottom hole flowing pressure of well 1 is determined by the second law of superposition as follows: n P D1 = ˆ ∑ qj j=1 1 P Dj∗ (3.t3. Wells started with production successively at the times t1.152) The value of the bottom hole pressure is the result of this equation where the qj represent the rate changes at the times tj.61.3 Calculation of MultiWell Problems Fig.149: µB p i – p wf1 ( t ) = 4πhk n ∑ j=1 rj q j E i –  4K ( t – t j ) 2 (3. Thus Eq. 3. 3.17 the wells 2 → n are projected imaginatively into well 1.16 shows various wells put on production with constant rates inside an infinite acting reservoir. 3.4 Single Well with Variable Production Rates When applying the second law of superposition the distances of the wells 2 → n from well 1 are not considered. Summation is only made for t > t j .Eq.7. 3.7. The diagram includes the overall production rate. For the calculation of the flowing pressure it is permitted to use the logarithmic approximation formula Eq. 3.151) r1 is the radius of well 1.t2.3: Solutions of the SinglePhase Equation of Filtration 3125 3.152 becomes: . Then every rj is substituted by rw: µB p i – p wf1 ( t ) = 4πhk n ∑ j=1 rw q j E i –  4K ( t – t j ) 2 (3.150) or from Eq.
16: Superposition of several wells in a infinite reservoir Figure 3. – . 80907 + ln 4πhk j = 1 Figure 3. ∑ q j ln ( t – t j ) + 0. 80907 + ln  = 2 rw K  2 r w ∑ qj j=1 n (3.3126 3: Solutions of the SinglePhase Equation of Filtration µB p i – p wf1 ( t ) = – 4πhk n ∑ j=1 K ( t – t j ) q j 0.153) n µB .17: Application of the second law of superposition on a well with a variable production .
Eq. but the well was shut in during the time ∆t . Then at the end of t1 the well is shut in ( q 2 = 0 ) for the time ∆t . This curve is of great importance since the determination of permeability and static reservoir pressure is made possible.155 to: 0.18 illustrates a pressure buildup curve.3: Solutions of the SinglePhase Equation of Filtration 3127 3.154 is simplified to: Bq 1 µ ( t 1 + ∆t ) p ws ( t 1 + ∆t ) = p i + .5 Pressure Buildup of ShutIn Well At first the well is put on production during time t1 with a constant rate q1.156 should give a straight line portion and the slope of this portion is m and according to Eq. Figure 3. When the shutin pressure (pws) is plotted versus Horner time in a Semilog plot.156) This equation calculates the pressure buildup at the well bottom at time t1 until the end of the buildup test. 3.157) . To make up for this discrepancy. 1832Bq 1 µ ( t 1 + ∆t ) p ws = p i + . 3.155) Instead of pwf (flowing pressure) pws (shut in pressure) was used which states that the well is shut in.ln ∆t 4πhk (3. we assume that the well is producing with a rate q 2 = – q 1 (as injector) during the shutin time ∆t . 3.154) Eq. 3. 3.154 considers the well is on production during the time t = t1 + ∆t . Then: q 1 + q2 = 0 So Eq. The pressure drop after the time t = t 1 + ∆t may then be calculated by Eq.153 as follows: µB p i – p wf1 ( t 1 + ∆t ) = – 4πhk q 1 ln ( t 1 + ∆t ) + q 2 ln ( ∆t ) + 0.log hk ∆t (3.7.156: 01832µq 1 B hk = –m (3. Usage of the common logarithm changes Eq. 3. 80907 + ln K  ( q 1 – q 2 ) 2 r w (3.
In the cartesian coordinate system the distance between point B(x1.6 Method of Image The second law of superposition may also be useful in the case boundaries exist inside or at the edges of the reservoir. It can easily be taken for granted that for symmetry reasons no flow will occur through the axis of symmetry (fault).Ei – .149 the pressure drop is given by: r1 r2 Bµqp ( x. Such a case is illustrated in Figure 3.7. the problem is reduced to a dualwell problem in an infinite reservoir as discussed before. Therefore the potential distribution inside the real region will be identical in both cases.18: Pressure buildup analysis plot (after Horner) .0): r 2 = ( x 1 + x 10 ) + x 2 2 2 2 (3. will be produced with the same rate. If an image of well 1.x2) and the well (x10. + Ei – .160) Figure 3.0) is: r 1 = ( x 1 – x 10 ) + x 2 2 2 2 (3. t ) – p i = .158) and from the image well (x10. This boundary may be a fault as well as a pinch out.3128 3: Solutions of the SinglePhase Equation of Filtration 3.159) According to Eq. 4πhk 4Kt 4Kt 2 2 (3.19. mirrored to the other side of the fault. 3.
161) (3. p ( x. t ) – p i = . + Ei – .162) 2 2 2 2 2πhk ( x 1 – x 10 ) + x 2 2 ( x 1 + x 10 ) + x 2 The pressure drop is calculated by setting r 1 = r w and r 2 = 2x 10 .ln (3. Kt 1 . 80907 + ln 2πhk or with cartesian coordinates 2 2 ( x 1 – x 10 ) + x 2 µqB.3: Solutions of the SinglePhase Equation of Filtration 3129 Figure 3.p ( x. r w p ( x. t ) – p i = . t ) – p i = .61: µqB.+ ln 2 r1 r1 . 4πhk 4Kt Kt 2 2 Kt . . 0. 80907 + ln . r2 (3. Eq. Ei – . 0.163) . 3.19: Production from a well near impermeable boundary (after Bear) By applying the approximation Eq.160 yields: x 10 µqB.+ . 3.
p ws = p i + .1 Pressure Buildup Test Near No Flow Boundary The well is put on production with a constant rate q.+ Ei – . ≈ Ei – . = b K ( t 1 + ∆t ) Kt 1 Then 3. ln .7.+ ln . 2 2 2 2 2 2 (3. – Ei – . then: x 10 x 10 Ei – . + 4πhk 4Kt 4K ( t – t 1 ) x 10 x 10 Ei – .3130 3: Solutions of the SinglePhase Equation of Filtration 3. ln . For the first two Eifunctions it is possible to use the logarithmic approximation formula without any restriction. rw µqB. 3. Ei – .168) .166) If ∆t is large it is possible to apply the logarithmic approximation to 3.167) 2 2 (3.– ln 2 4πhk ∆t x 10 and K∆t . Kt K ( t – t 1 ) where the shutin time is ∆t = t – t 1 .165 for all terms and thus: K ( t 1 + ∆t ) µqB. Thus Eq. 3.165 becomes: ( t 1 + ∆t ) Bqµ.6. – Ei – . ln .164) (3. The pressure buildup curve can be calculated in the same manner as demonstrated before (by using the second law of superposition).154 may be written as follows: x 10 x 10 µqB. until the time t1 and then it is shut in.– b 4πhk ∆t (3. .164 according to Eq. ( t 1 + ∆t ) p ws – p i = . 4πhk ∆t K ( t 1 + ∆t ) K∆t since: ∆t = t – t 1 . – Ei – .165) If ∆t is small the last Eiterm becomes zero and the preceding one is practically constant. 2 x 10 (3. ( t 1 + ∆t ) p ws – p i = . r w p ws – p i = .
(m2=2m1).20 shows the pressure buildup curve.169) Figure 3. 3. The first section of the curve is described by Eq. 3. The slope of the second straight line (displayed in the semilogarithmic coordinate sheet) is exactly the double of the slope of the first. Figure 3.3: Solutions of the SinglePhase Equation of Filtration 3131 ( t 1 + ∆t ) Bqµp ws = p i + .167 (after a short time no boundary effects) and the second by Eq.20: Pressure buildup curve near a discontinuity .169 (after the pressure disturbance reached the boundary).ln 2πhk ∆t (3.
7. The specific boundary condition given by Eq. If the real region is again mirrored.21: Production in the vicinity of a boundary with a constant potential (after Bear) The pressure change at well bottom can be calculated by a more simple equation.ln 4πhk r2 µqB.ln 2 4πhk ( x – x ) 2 + x 2 r1 1 10 2 2 2 2 2 2 (3.172) and x + x 10 ≅ 2x 10 . but we consider the image injection well.81 is used. It is evident that the distribution of the potential in the real region will again be the same for both cases. r 1 p i – p ( x.21. Ei – .= – .170) (3.3132 3: Solutions of the SinglePhase Equation of Filtration 3. t ) = . – E i – . 2. 4πhk 4Kt 4Kt Using the logarithmic approximation and cartesian coordinates yields: µqBp i – p ( x. 3.6. The corresponding equation is given by Eq.( x 1 + x 10 ) + x 2 .151. For this case: x – x 10 = r w therefore: µqB. t ) = – .2x 10 p i – p wf = – .2 Constant Pressure Boundary Now let us consider the case of a constant potential at the outer boundary as shown in Figure 3.171) Figure 3.ln rw 2πhk (3. We ignore the boundary. The rate for an injection well is q 2 = – q 1 : r2 µqB. the image well will now be regarded as an injection well.
Trans. Publications. P.. and Hurst. Ramey.P.: "The application of the Laplace transformations to flow problems in Reservoir" Trans. Van Everdingen.1 3.C. H. "Applied pressure analysis for fractured wells" J. Slider.8 References 3. AIME (1949) Gringarten.4 3.: "Practical Pertroleum Reservoir Engineering Methods".: "Dynamics of fluids in Porous Media" Elsevier New York (1972) Chaumet.(1976) .F.3: Solutions of the SinglePhase Equation of Filtration 3133 3.S. Paris (1964). W.C. U. Tech. A.F. J. R. A.A. Pet.5 Bear..K.2 3. Petroleum Publishing Company Tulsa O. Editions Technip. (July 1975) 887892.: "Cours de production" Thome III Econlement monophasique de Fluides dans les millieux poreux. and Raghavan. H. I. AIME 259.3 3. Jr..
020 26.126 130.129 44.386 11.3134 3: Solutions of the SinglePhase Equation of Filtration Table 3.661 74.785 33.897 23.488 20.313 16.136 37.265 59.855 25.742 17.30 0.417 219.207 38.691 29.249 151.70 0.Van Everdingen: Constant Pressure QtD Functions for Infinite Acting Radial Reservoirs tD 0.743 6.697 36.979 40.517 33.048 32.247 34.966 123.533 195.574 48.145 213.0 6.778 13.571 100.940 8.606 0.968 107.0 8.758 0.859 143.852 192.081 98.904 110.869 7.517 63.808 136.914 173.0 5.058 36.338 78.076 24.894 .270 127.068 29.406 26.208 196.457 8.855 12.550 131.045 140.223 137.461 9.124 184.639 176.825 211.416 155.638 119.308 31.928 69.539 5.202 4.549 170.50 0.101 22.05 0.447 3.784 216.403 124.259 19.60 0.153 5.526 tD 355 360 365 370 375 380 385 390 395 400 405 410 415 420 425 430 435 440 450 455 460 465 470 475 480 485 490 495 500 510 520 525 530 540 550 560 570 575 580 590 600 610 620 625 630 640 650 660 670 675 680 690 700 710 QtD 121.777 37.858 46.418 36.611 34.635 139.738 116.625 58.497 85.494 37.10 0.469 tD 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 QtD 21.151 33.818 30.520 0.935 28.322 162.314 6.679 32.317 54.664 146.500 22.00 0.785 77.555 27.167 17.184 158.189 117.895 61.450 15.968 51.444 168.284 114.244 25.226 75.563 38.540 103.433 42.298 21.411 7.429 18.701 22.336 35.698 165.590 18.034 41.781 43.019 104.0 2.633 26.013 15.542 203.589 94.40 0.251 1.023 86.883 16.20 0.272 39.512 71.972 133.937 31.671 20.544 197.919 39.01 0.057 101.689 0.684 14.502 209.0 7.434 10.473 187.569 2.083 120.192 30.737 67.000 0.15 0.898 1.684 24.064 148.367 111.774 183.166 189.062 89.495 105.883 34.112 0.385 tD 96 97 98 99 100 105 110 115 120 125 130 135 140 145 150 155 160 165 175 180 185 190 195 200 205 210 215 220 225 230 235 240 245 250 255 260 265 270 275 280 285 290 295 300 305 310 315 320 325 330 335 340 345 350 QtD 41.791 27.357 179.90 QtD 0.945 161.069 181.183 169.278 0.174 27.640 153.391 134.886 80.417 32.466 24.090 72.291 23.565 159.626 39.575 91.573 15.878 200.453 141.090 95.011 18.428 81.131 14.233 13.851 38.201 205.020 1.545 88.277 49.949 10.827 113.25 0.0 3.3: Hurst .437 108.0 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 1.262 144.131 64.838 126.964 9.801 157.046 1.80 0.084 92.0 9.359 1.404 0.029 154.699 129.565 30.319 12.854 208.974 35.856 150.913 11.845 19.965 83.331 40.588 97.648 53.140 1.080 20.976 56.443 29.084 42.735 42.684 41.
107.830 574.3: Solutions of the SinglePhase Equation of Filtration 3135 Table 3.781 660.128 446.795 263.867.373 401.912 1.330 1.669 481.214 411.100 4.100 5.847.024 1.283 389.425 2.013 410.520 825.900 8.257 1.800 5.280 727.002 1.065 308.514 232.000 5.139.200 4.850 3.486 1.700 6.000 8.318 239.353 280.144 355.088 838.807.318 1.260 1.250 1.460 350.290 1.922 782.490 1.000 7.125 1.500 4.595 1.171.267 372.379 682.200 6.437 643.338.872 1.409 1.815 278.560 306.350 4.379.325 1.276.344 4.475 1.629 268.545 321.668 248.308 345.500 2.672 620.800 7.593 1.410 1.975 2.594 914.441.360 1.220 422.950 3.220 1.100 7.966 957.543 396.437 493.100 3.433 335.300 6.942 365.566 979.421.168 440.160.903 881.300 4.800 4.270 1.210 1.200 1.900 1.033.666.310 1.625.066 312.332 586.459 925.010 1.198 1.729 1.775 1.222 1.877 705.203 1.959 398.376 340.077 1.090 716.265 1.012 1.572 380.875 337.000 510.460 1.075.275 2.440 1.340 1.300 8.850 4.118.000 1.011 1.120 1.350 1.747.500 1.900 3.000 328.850 1.806 597.090 1.386 367.750 4.200 3.585 580.150 1.420 1.737 551.200 2.522 557.523.093 671.225 2.350 3.574 1.439 1.048 360.503.800 6.900 4.600 7.050 3.160 1.278 1.418 413.726.900 2.300 5.725 1.400 1.770 348.000.408 1.100 8.482.243 .712 903.950 4.965 614.567 311.028 859.470 1.925 1.927 1.650 2.325 2.957 250.950 1.600 4.309 936.025 1.025 2.450 3.724 1.080 1.600 5.700 3.854 1.141 1.462.275 1.200 5.120 1.053 303.003 382.400 8.150 4.900 6.130 1.400 4.300 7.828 370.562 316.225 1.700 4.028 1.721 236.000 6.688 1.274 274.250 4.200 8.390 1.750 1.030 298.307 499.200 7.859 386.704 373.075 2.950 5.211 288.887.755 637.475 2.859 1.974 870.670 222.544.432 384.743 1.750 3.640 693.675 1.700 7.359.380 1.706 1.150 3.296.893 1.022.299 563.490 1.370 1.167 399.255.128.230 1.705 392.250 3.500 6.075 1.922 375.145 539.980 224.418 1.300 2.150.532 603.400 5.314 313.020 237.170 221.825 1.598 749.337 534.500 5.888 283.043.180 1.372 626.992 794.800 1.995 293.543 299.646.907.011.820 416.383 1.800 3.545 275.425 1.709 1.075 816.090 827.280 1.948 287.550 3.086.525 1.073 265.167 505.816 892.843 343.650 1.019 1.190 1.706.214 428.300 3.108 1.786 404.904 229.725 760.030 1.329 1.773 968.000 3.055 318.100 2.065.760 326.150 2.020 1.400 2.575 1.767.100 6.320 1.515 261.139 377.120 234.042 805.016 457.212 1.350 2.558 487.067 849.800 2.700 5.819 255.336 1.625 1.650 3.234.500 990.550 2.480 1.125 394.450 1.450 4.305 1.799 301.945 463.060 1.400 3.110 1.400 7.300 1.086 247.600 6.650 4.400.666 1.906 1.600 2.501 245.585.192.430 1.686.4: Continuation 720 725 730 740 750 760 770 775 780 790 800 810 820 825 830 840 850 860 870 875 880 890 900 910 920 925 930 940 950 960 970 975 980 990 1.181.700 1.720 362.912 242.000 2.688 353.514 296.197 406.388 257.390 1.517 324.400 6.945 545.181 270.858 1.827.140 1.550 1.317.144 946.375 1.113 648.125 1.771 475.700 2.142 338.097.550 4.082 1.450 2.072 591.861 516.319 1.600 3.077 452.605.695 522.032 323.125 2.040 1.066 631.068 568.833 771.050 2.900 7.100 1.750 2.480 330.600 1.449 738.953 260.496 361.230 349.054.500 3.050 4.850 2.245 252.608 409.250 2.070 1.597 358.252 608.196 434.863 469.175 1.213.473 290.289 226.787.240 1.420 285.375 2.050 1.820 1.298 1.564.070 387.225 1.037 1.958 333.646 1.224 1.175 2.500 7.729 273.
01011 1.51011 3.0107 9.01010 6.0105 7.5108 3.828106 2.125.975.900 10.252106 1.700 8.697109 2.538104 3.699104 9.328108 2.708.996 2.427105 4.51012 2.095.000 17.065.000 35.0107 4.016106 1.01012 1.438 2.51010 3.984107 6.368108 6.0105 4.878 2.645108 4.689106 6.780.01011 9.006.5109 3.108107 2.555 2.01010 5.891010 6.051105 1.0106 4.610106 5.5106 2.5109 2.01010 2.189105 1.0108 1.000 30.966.961106 3.0108 1.560.927.500 9.291010 3.01010 9.0108 6.000 50.780 3.01010 4.508 6.0108 4.184 2.0109 9.0105 5.0106 7.200 9.761105 8.0107 5.0105 2.747108 1.500 15.358 13.643109 6.01011 7.751010 4.134106 1.5: Continuation 8.000 80.421011 .5105 2.0105 8.000 25.921010 2.0105 9.457 14.0109 4.0107 6.0109 6.247 7.363.861 2.01010 1.600 9.866106 9.967 3.313105 6.0109 1.473 2.0107 8.414109 7.471010 5.121 15.066104 4.100 9.868109 5.0106 9.0106 1.01010 1.817104 6.096 9.0108 2.732 2.01010 7.220108 7.510108 5.0106 1.429108 1.085.105.000 1.000 90.099 11.047.203.0106 8.505109 3.400 9.900 9.0109 7.066108 7.5106 3.650.942 1.581010 7.880108 2.164.726 5.144.725107 1.000 40.081011 1.544105 7.100107 4.800 5.000 70.0107 1.216 2.800 8.398106 2.0109 8.01011 2.0108 8.000 12.865107 8.0106 5.607107 3.0109 5.800 9.462105 2.01011 1.628 2.3136 3: Solutions of the SinglePhase Equation of Filtration Table 3.758106 7.183109 7.816106 8.0109 1.368 4.965105 1.095107 1.947.909108 8.0108 9.164.032107 5.000 9.797107 9.51010 2.184.500 20.600 8.166 1.700 9.986.5108 2.350.01011 8.389 17.126105 2.005.064105 5.284 21.928107 7.5105 3.688.326105 1.01012 2.0107 1.899.5107 2.103109 2.0108 7.299109 4.0105 6.01010 8.000 125.0107 1.267104 7.300 9.171010 1.087109 4.071107 5.771108 3.026.065 1.299 12.0109 2.633.517106 4.000 75.308104 4.604107 2.000 60.288109 1.0106 2.796 2.911106 1.281010 1.227 2.01011 4.586.000 100.0108 5.551010 1.5107 3.01011 6.0105 1.113104 1.01011 5.0106 6.046.744 2.191010 5.781105 3.021010 3.0107 7.948109 1.51011 2.531.
2) The relation between density and pressure is given by Eq. The displacing phase can either be the wetting phase or the nonwetting phase.23.1) (4. Assuming that there is no mass transfer between these two phases at the interface separating them and that phase equilibrium has been achieved between the two phases. The concept of relative permeability as described in Chapter 1 actually based on the mathematical description of this phenomenon.4) (4.5) 4137 . As in a onephase filtration it is assumed that the process is isothermal and thus density and viscosity will be functions of pressure only: ρ 1 = ρ1 ( p1 ) µ1 = µ 1 ( p1 ) ρ 2 = ρ2 ( p2 ) µ2 = µ2 ( p2 ) (4. 2. 2. Using the same assumption which has been previously discussed.22 and Eq. 2.3) (4.7.( ∇p 2 + ρ 2 gi 3 ) µ2 Index 1 refers to the displacing phase index 2 to the displaced phase. the fundamental equations of a twophase filtration may be set up in the same form as in Eq.4 4. In this chapter a general mathematical approach was developed to describe the movement of these two fluids.( ∇p 1 + ρ 1 gi 3 ) µ1 kk r2 u2 = – . The difference between the two phase pressures is specified as the capillary pressure which is a function of saturation: p2 – p1 = Pc ( S1 ) (4. kk r1 u1 = – .1 TwoPhase Filtration The Equation of TwoPhase Filtration This chapter deals with the physical phenomenon of two immiscible fluids flowing simultaneously through porous media.
8) .6) (4. 2. It is evident that S1 + S 2 = 1 and so ∂ ( φS 1 ρ 1 ) ∇( ρ 1 u 1 ) = – ∂t ∂ ( φS 2 ρ 2 ) ∇( ρ 2 u 2 ) = – ∂t (4.43 in the fact that fluid 1 only takes the portion S1 of the pore space φ and fluid 2 the portion S2. S1 and S2 are defined as the saturations of the two phases.4138 4: TwoPhase Filtration The equation of continuity only differs from Eq.7) (4.
9) (4. 4.13 and Eq.+ φ .11) (4.16) (4.14: ( S1 + S2 ) ∂ ( u1 + u 2 ) .12) (4. .+ φ .= 0 ∂t ∂x it is evident that u is independent of x and therefore: ∂u .14) Instead of coordinate x3 simple x was written.= 0 ∂x ∂t ∂S 2 ∂u 2 .17) .+ ρ 1 g µ 1 ∂x kk r2 ∂p2 u 2 = – .= 0 ∂x ∂t (4. 4.2 Vertical TwoPhase Filtration of Incompressible Fluids For this case we assume a vertical one dimensional filtration of incompressible fluids in a porous medium.15) (4.= 0 ∂x Then a new function is introduced: u1 F 1 = u and (4.Eq.13) (4.+ ρ 2 g µ 2 ∂x p2 – p1 = Pc ( S1 ) S 1 + S2 = 1 ∂S 1 ∂u 1 .8 may be written as follows: kk r1 ∂p1 u 1 = – .1 . Now it is useful to introduce the total velocity of filtration as a new variable: u = u1 + u2 When adding Eq.18) (4. 4. 4. Therefore Eq.10) (4.4: TwoPhase Filtration 4139 4.+ φ . .
22 we obtain: dP c ∂S 1 µ 1 F 1 u µ 2 ( 1 – F1 )u .19) where F1 is the portion of fluid 1 in reference to the total flow.= .= ( ρ 1 – ρ 2 )g + . The usual symbol of fractional flow value or function is f1. We use both F1 and f1 to distinguish between two cases.⋅ µ 1 µ 2 dS 1 ∂x µ1 µ2 .– k r2 u u .9 and Eq.⋅ .24) (4.19 into Eq. 4.= .⋅ ∂x ∂x dS 1 ∂x ∂x (4.20) (4.– .⋅ .25) .⋅ .23) (4.– dS 1 ∂x kkr1 kk r2 F1 is then obtained in the following form: µ 2 k ( ρ 1 – ρ 2 )g k . 4. F1 and F2 are called fractional flow functions.11 yields: ∂P c ( S 1 ) ∂p 2 ∂p 1 dP c ∂S 1 . 4.10 leads to: ∂p 1 µ1 F1 u .= .dP c ∂S 1 F 1 = .= 1 – F 1 u u (4. .21) (4. 4. 4.= – ρ 1 g – ∂x kk r1 ∂p 2 µ 2 ( 1 – F 1 )u .22) Then Eq.+ .+ k r1 k r2 k r1 k r2 Also F1 can be expressed as: kkr2 dP c ∂S 1 1 + . Substitute Eq.– ∆ρg uµ 2 dS 1 ∂x F 1 = µ 1 k r2 1 + µ 2 k r1 (4.20 and equating the result with Eq.= – ρ 2 g – kk r2 ∂x Taking the derivative of Eq. 4.4140 4: TwoPhase Filtration u2 u – u1 F 2 = . 4.18 and Eq. 4. We use f1 if the capillary pressure is neglegted and F1 if not.21 is subtracted from Eq.+ .
+ .⋅ .+ k r1 k r2 then ∂S1 F 1 = f 1 + ψ 1 ⋅ ∂x where f1 and ψ1 are functions of S1 and u.= 0 φ dS 1 ∂x ∂x ∂x ∂t (4. Substituting u 1 = u ⋅ F 1 into Eq.16 leads to: u ∂F 1 ∂S 1 .28: ∂S 1 ∂S 1 u df 1 ∂S 1 ∂ .29) (4.13 and Eq.⋅ ψ1 = µ 1 µ 2 dS 1 ..30) (4.+ k r1 k r2 and k u . 4.26) (4.27) .– k r2 u f 1 = µ1 µ2 . ψ 1 ⋅  + .⋅ .= 0 φ ∂x ∂t or after using Eq.28) (4.+ .4: TwoPhase Filtration 4141 If µ 2 k ( ρ 1 – ρ 2 )g . 4. 4.dP c .
u Eq.27 are no longer required.= f 1 .= 0 φ dS 1 ∂x ∂t (4.28: u1 F 1 = . 4.30 is not linear. 4. In the year 1942 Buckley and Leverett published their theory which enabled great progress on behalf of multiphase filtration.1: Calculation of fractional curve (after Marle) The task is to compute the velocity w of a given saturation at any given point i.⋅ . In this chapter though the problem is discussed in a simplified manner. 4.+ . thus a solution can only be achieved numerically. Eq.11 and Eq. According to Eq. to ”track” .32) 1 df1 dS1 f1 f1 0 0 S1m S1M 1 S1 Figure 4.4142 4: TwoPhase Filtration 4. 4.31) (4.3 The BUCKLEYLEVERETT Solution Eq.30 becomes: u df 1 ∂S 1 ∂S 1 . • Immiscible fluids. This theory neglects the capillary force also considers the following assumptions: • Incompressible fluids and porous media. 4. • Darcy’s law of two phase filtration is valid and • Linear displacement.e.⋅ .
.⋅ .= . If f1 is plotted versus S1 as shown in Figure 4. On one side at x = 0 the velocity u1 = u is constant. S1m and S1M are the possible minimum and maximum values of S1. This term (w) is only a function of S1 and u.dx + .4: TwoPhase Filtration 4143 the displacing front. Mathematically.1 which illustrates the so called Sshaped curve which is characteristical for most porous media. 4.37) is the injection rate of phase 1 and is the cross section area of the porous media. this means that for every constant S1 value there is a function: S 1 ( x.33: ∂S 1 ∂S1 .dt = 0 ∂x ∂t and since: dx = wdt this leads to: ∂S 1 ∂S 1 w . This means that the displacing phase is injected at a constant rate.+ . 4. t ) = constant which gives the location x as a function of time. Therefore it is sufficient to know the distribution of initial saturation and the velocity of displacement in order to calculate the saturation distribution.= 0 ∂x ∂t Comparing Eq.2.32 leads to: q 1 df 1 u df 1 w = .36 with Eq. 4. This curve has an inflection point at which the differential of f1 is a maximum.33) (4.34) (4.⋅ Aφ dS 1 φ dS 1 where q1 A (4.35) (4. It is assumed that the initial distribution of saturation in a vertical porous medium at t = 0 corresponds to the curve shown in Figure 4. Any given value of the saturation S1 is valid for a specified point x and time t.36) (4. Taking the derivative of Eq.
4144
4: TwoPhase Filtration
If porosity φ is also regarded as a constant then the speed of propagation for every value of saturation will, as a result of Eq. 4.37, be proportional to: df 1 dS 1
(4.38)
Figure 4.2: Propagation of saturation profile (after Marle) As time elapses the distance travelled for all saturations is plotted. As expected the points with small or large saturation values progress at a lower speed than areas with a middle saturation value as shown in Figure 4.2. It means that after a certain time period we will have a saturation profile without physical sense since several xvalues correspond to two different saturation values. This difficulty only appears after the displacement has travelled a certain distance since the initial saturation distribution was assumed to be continuously decreasing. If at t = 0 the saturation S1 equals S1m everywhere then these problems will arise for every time t > 0. These difficulties are found also in other fields of physics for example in case of supersonic streaming of gas. The solution achieved for the saturation distribution may only be interpreted if the profile is not continuous. (Figure 4.3 displays this discontinuity) As a consequence of the conservation of mass the location of the discontinuity must be fixed in a way so that the areas on both sides of the discontinuity are equal in size. (In Figure 4.3 the crosshatched area).
4: TwoPhase Filtration
4145
Figure 4.3: The displacement front as discontinuity of saturation (after Marle)
4.3.1
The WelgeMethod
Let the initial saturation in a porous medium be S1a which can either be smaller or larger than S1m. Then the profile of saturation is plotted at sequential dates t1<t2<t3... The discontinuity in saturation and the front saturation S1f should be determined for every profile. It will be proven that this value is the same for all t. As already mentioned it is essential that the discontinuity may not contradict the law of conservation of mass. This means that the area below the analytical curve must be equal in value to the area below the profile corrected by the discontinuity as shown in Figure 4.3. The area below the analytical curve is calculated as follows:
S 1M S 1M
∫
S 1a
ut w ( S 1 )t dS1 = φ
∫
S 1a
df 1  dS 1 = ut [ f 1 ( S 1M ) – f 1 ( S 1a ) ] φ dS1 (4.39)
4146
4: TwoPhase Filtration
The area below the corrected profile (Welge approximation) will be:
S 1M
( S 1f )t ( S 1f – S 1a ) +
∫
S 1f
w ( S1 )t dS 1
ut df 1 =   ⋅ ( S 1f – S1a ) + f 1 ( S 1M ) – f 1 ( S 1f ) φ dS 1 S = S 1 1f Eq. 4.39 and Eq. 4.40 though must result in: f 1 ( S 1f ) – f 1 ( S 1a ) df 1 =   S 1f – S 1a dS 1 S 1 = S 1f
(4.40)
(4.41)
There is only one point that can satisfy Eq. 4.41. This point is the tangency point of the line drawn from point S1a to the curve f1, which is independent of time. The meaning of Eq. 4.41 is illustrated in Figure 4.4. The tangency point also gives the value of the saturation at the front (S1f).
Figure 4.4: Determination of average saturation of the wetting phase after breakthrough (after Welge)
47 into Eq.+ f 1f dS 1 (4. 4.48) .( 1 – f 1f ) φAL (4. S 1 = S 1bf and Eq.+ f 1f dS 1 1 From Eq.= . 4. 4.⋅ dS 1 u u dt q dt After integrating.42 becomes: df 1 C = f 1f –  S 1f dS1 Substituting Eq..42 yields: df 1 f 1 = ( S 1 – S 1f ) ⋅ .37.4) and Eq. Eq.can be expresses as: (4. S1 = S1f (as shown in Figure 4.w = .= φAL dS 1 Q (4.⋅ . 4.43) (4. . one can write the equation of the tangent of the fractional flow curve as follows: df 1 f 1 =  S 1 + C dS 1 1 where .46) (4.44 becomes: df 1 1 = ( S 1bf – S 1f ) ⋅ .43 into Eq..is the slope of the tangent and C is the interception of f1 axis. 4. Substituting Eq.0 .45) df dS 1 df 1 φ φ dx φA dx . 4. 4.42) df dS 1 At f1 = f1f . 4. (4.= .4: TwoPhase Filtration 4147 In order to calculate the average saturation of phase 1 in the swept zone behind the front ( S 1bf ) .47) where Q is the cumulative amount of injected displacing phase (1).45 and solving for S1bf yields: Q S 1bf = S 1f + .44) At f1 = 1.46 can be written as: df 1 .
Figure 4. It illustrates that after breakthrough the cumulative production of the displacing fluid (Q1) will increase rapidly and the cumulative production of the displaced fluid (Q2) will decrease by the same amount so the total cumulative production (Q) is a linear function. 4.3. based on material balance. Figure 4.5: Cumulative production by linear displacement (after Marle) . is given in Reference 4.5 shows the cumulative production by a linear displacement.48.4148 4: TwoPhase Filtration There also is an alternative derivation of Eq.
) Estimate the time of the water breakthrough. From Eq.) The water saturation at the water front (S1f). d.60 0.65 = 0.40 0.70 0. Solution: a.99 Calculate the following: a.) The point of tangency represents the water saturation at the front is S1f = 0.) Extension of the tangent to fw=1.50 0.) The average oil saturation in the swept area of the core (behind the water front).1: A cylindrical sandstone core having diameter of 0. the average water saturation behind the front is Swbf = 0.96 0. Saturated with oil (35 API) and irreducable water saturation (S1m = 0.) The velocity of the water front.2). Water was injected into the core with a steady rate of 0.85 0. the cumulative water injected can be determined from: .0.92 0. The fractional flow was calculated and listed below: Sw: Sw 0.35.00 – 0.62 0.48. length of 1 [m] and porosity of 20%.20 0. e. c.) From the plot of the fractional flow curve (fw verses Sw): b.) The cumulative water injected to reach the average water saturation behind the front determined in part (B).55.30 0.75 0.65 0.1 [cm3s1] to displace the oil. b.05 [m].75 fw 0. c. Then the average oil saturation behind the front = 1.4: TwoPhase Filtration 4149 Example 4.00 0.55 0.65. 4.
055 .5 ) ⋅ 0.⋅ .= 0.⋅ 2.2 ⋅ .1 w = .= 30 min wf 0.055 cm ⁄ s 2 Aφ dS 1 π ( 2.775 4 1 – f 1f d.55) = 2.17 Then the velocity of the front calculated from: q 1 df 1 0.4150 4: TwoPhase Filtration S 1bf – S 1f 3 3.17 = 0.⋅ 100 ⋅  = 175 cm 1 – 0.= .14 ⋅ 25 0.) The slope of fw curve at (Sw = 0.55 Q = φAL .) The time of water breakthrough is: L100 t = .= .2 e.65 – 0.
49) Displacement proceeds in vertical direction from bottom to top. The relation: kk r2 ( ρ 1 – ρ 2 )g . f1(S1i)] is not so steep but the saturation at the front becomes larger and displacement more effective. 4.1 Influence of Gravity Eq.26 may be written in the following form: kk r2 ( ρ 1 – ρ 2 )g 1 f 1 = .⋅ 1 – . A stable frontal displacement is only possible if the heavier fluid is maintained below the lighter fluid.⋅ µ2 u (4. Because this state is not stable and the fluids will exchange their positions in counterflow.4 Influence of Gravity and Capillary Force 4.4. If the velocity of filtration increases the fraction will become smaller and the fractional curve shifts back to the case ρ1 = ρ2. The theoretical case of ρ1 < ρ2 can’t be handled by the theory previously discussed. The second term of Eq.1 + .⋅ µ2 u k r2 µ 1 .50) is simply the relation between the frictional force and gravity.6).49 vanishes and gravity has no more influence.⋅ k r1 µ 2 (4. The tangent drawn from the initial point [S1i. The limiting points of the curve f1 remain unchanged since they only depend on the krfunctions (see Figure 4. If ρ1 > ρ2 the term inside the parenthesis becomes smaller than 1 which means that the fractional curve shifts to the right.4: TwoPhase Filtration 4151 4. At first the curve f1 for ρ1 = ρ2 is drawn. 4. .
6: The influence of gravity on the fractional curve (after Marle) 4. by the method of finite differences. . The solution can only be achieved with help of numerical methods. In case of a slow displacement the capillary force is larger than the viscous forces. This is expressed in a rather flat saturation profile.7 which has been discussed previously.30 for various rates of filtration with the help of the method of finite differences. 4. In the case of a fast displacement the profile becomes steeper and tends to the BuckleyLeverett solution.4. It can be observed that when the displacing phase reaches the end of the medium the displacing efficiency is larger at a fast displacement than at a slow displacement. for example. The BuckleyLeverett solution neglecting the capillary force is illustrated in Figure 4.2 Influence of the Capillary Force When considering the capillary force one must regard the fact that Eq.30 is not linear. 4.4152 4: TwoPhase Filtration Figure 4. The other profiles were calculated by applying Eq. Discussion of such methods of solution would surpass the objective of this textbook.
Rappoport 1958) Figure 4.8: The displacing efficiency as a function of velocity (by Kyte. Two things are of importance: First the efficiency of displacement is at certain values independent of velocity.4: TwoPhase Filtration 4153 Figure 4.7: Influence of the velocity of displacement on the distribution of saturation regarding the capillary force (by Douglas et al 1958) Figure 4.8 shows the oil recovery versus the rate factor for different core lengths and for a strong waterwet system. Second the time period between .
On the other side the displacing phase may be nonwetting.9: "Endeffect” in case of a wetting displacing phase (after Marle) . Breakthrough takes place at S1M and in the following the saturation profile will tend towards the line of the saturation S1M. 4. The point of breakthrough is defined as the moment of first outflow of the displacing phase.3 The Capillary EndEffect The endeffect is a phenomenon at which the wetting phase is held back by the capillary force at the boundary of the medium. The deviation is effected by the capillary endeffect.4154 4: TwoPhase Filtration the arrival and breakthrough of the displacing phase at a small displacing speed is large.9 the fluid arriving at the boundary of the medium accumulates and causes a peculiar deformation of the saturation profile. Until saturation at the boundary has not yet reached the value S1M the capillary force will be larger than zero and the capillary gradient is infinitely large.10.4. Figure 4. Then the endeffect must be considered in opposite. In Figure 4. The saturation of the displacing phase remains S1m and the profile developes as shown in Figure 4.
4 Imbibition Let us consider a porous medium contacted with the wetting phase at its bottom surface and all other sides are covered by a impermeable layer. Due to capillary forces the wetting phase tends to intrude at the bottom side and thus displaces the nonwetting phase in counter flow. Figure 4. It is assumed that the fluids are incompressible.11: Displacement in countercurrent Therefore: .4: TwoPhase Filtration 4155 Figure 4.4. At the initial time t = 0 the saturation of the wetting phase is S1m.10: Endeffect” in case of a nonwetting displacing phase (after Marle) 4.
⋅ k r1 k r2 + k . Eq.+ k r1 k r2 (4. Eq.4156 4: TwoPhase Filtration u 1 + u2 = 0 (4. Afterwards they are subtracted one from another yields: dP c ∂S 1 µ1 µ2 u 1 . 4..+ .12. 4. 4. 4.⋅ dS 1 ∂x k r1 k r2 From Eq.+ .51 and Eq.= 0 ∂x ∂x ∂t or ∂S 1 + ∂S 1 1 ∂ϕ 1 ∂S 1 ∂ .51 along with Eq. 4.14 remain further valid.52) (4. the capillary pressure is zero.56) (4.53) where: k ( ρ 1 – ρ 2 )g + ϕ 1 = µ1 µ2 .⋅ .= 0 φ dS 1 ∂x ∂x ∂x ∂t Eq.10 by kr2/µ2. 4. 4.54) (4.dP c ψ 1 = .ϕ 1 + ψ 1 . thus: . Therefore: + + ∂S 1 u 1 = ϕ 1 + ψ 1 ⋅ ∂x (4.58) At the inlet x = 0.57 is in its form identical with Eq. 4.9 – Eq.51) Eq. 4. 4.15.+ φ . 4.13 leads to: ∂S 1 + ∂S 1 ∂ + .39 into Eq.57) (4. The boundary conditions are: At the outlet x = L: + + ∂S 1 u 1 = ϕ 1 + ψ 1 = 0 ∂x x = L + (4.⋅ µ 1 µ 2 dS 1 . = k ( ρ 1 – ρ 2 )g + k . ψ 1  + .2. Then Eq.30. 4. 4.55) Substituting of Eq.9 is divided by kr1/µ1 and Eq.
57 – Eq. 4. The result is shown in Figure 4.59 can only be solved numerically. Figure 4.13: Distribution of pressure and saturation in case of linear (counterflowing) imbibition (by Blair) .12 and Figure 4.12: Capillary pressure and relative permeability functions used in the calculation by Blair Figure 4.59) The boundary value problem Eq.4: TwoPhase Filtration 4157 ( S1 ) x=0 = S 1M (4. It is appropriate to mention that the pressure gradients of the phases are corresponding to the counterflow opposed to one another.14. 4.13. To illustrate this it is of use to regard the calculations by Blair (1960). These numerical results were verified by experiments of Graham and Richardson shown in Figure 4.
4158 4: TwoPhase Filtration Figure 4.14: Recovery in case of linear counterflowing imbibition and the experimental determination of the influence of a certain in corelength. (by Graham and Richardson) .
423 (1958).: Trans. AIME 215.. AIME.1 4... H. 96 (1958).. Doherty Series N.R. J. P. AIME 146. J.C. 107 (1942). and Wangner.7 .R. Trans.2 4.S.Jr. L. (May 1960) Buckley. Jr. Marle.A. paper no.A. U. AIME secondary Rec.6 4. Wichita Falls.. P. 3 of the Henry L.195. AIME 215.M. Symp.. Blair. R.5 4./Dallas (1971). J.M.J..F. Trans. C. Welge. 4.5 References 4. 91 (1952). and Rappoport. Gulf Publishing Company (1981).: "The Reservoir Engineering Aspects of Waterflooding" Monograph Vol.: "A Simplified Method for Computing Oil Recovery by Gas or Water Drive" Trans.Y.4: TwoPhase Filtration 4159 4. F.: "Calculation of Linear Waterflooding Behavior Including the Effects of Capillary Pressure" Trans.4 Blair. and Leverett. Douglas.3 4. Kyte. M. 1475G. Craig.: "Multiphase Flow in Porous Media" Institut Francais du Petrole.M. J. Texas.
( ∇p 2 + ρ 2 gi 3 ) µ2 p 2 – p1 = P c ( S1 ) Equation of State ρ = ρ(p) µ = µ( p) Equation of Continuity ρ1 = ρ1 ( p1 ) µ1 = µ1 ( p1 ) ρ2 = ρ 2 ( p2 ) µ2 = µ2 ( p2 ) S1 + S2 = 1 ∂ ( φS 1 ρ 1 ) ∇( ρ 1 u 1 ) = – ∂t ∂ ( φS 2 ρ 2 ) ∇( ρ 2 u 2 ) = – ∂t ∂ ( φρ ) ∇( ρu ) = – ∂t .( ∇p + ρgi 3 ) µ kk r1 u 1 = – . One Phase Basic Equation of Motion Two Phase k u = – .1: Summary of the Equations of One Phase and Two Phase Filtration.4160 4: TwoPhase Filtration Table 4.( ∇p 1 + ρ 1 gi 3 ) µ1 kk r2 u 2 = – .
1 + .+ ρ 1 g µ 1 ∂x kk r2 ∂p 2 u 2 = – . µ 1 = constant.1 + .– ∆ρg uµ 2 ∂S 1 ∂x f 1 = µ 1 k r2 ..4: TwoPhase Filtration 4161 Table 4.= – ρ 1 g – ∂x kk r1 ∂p 2 µ 2 ( 1 – f 1 )u .2: Summary of the Equations of Two Phase Filtration. ρ 2 = constant µ 2 = constant kk r2 ∂P c ∂S 1 .= 0 ..µ 2 k r1 u ∂f 1 ∂S 1 ∂S 1 .= .∂S 1 ∂x ∂x ∂x ∂x ρ 1 = constant. Two Phase 1Dimensional kk r1 ∂p 1 u 1 = – .+ φ .= 0 ∂x ∂t . .= .– .+ . .+ ρ 2 g µ 2 ∂x p2 – p1 = Pc ( S1 ) ∂p 1 µ1 f1 u . ..= – ρ 2 g – kk r2 ∂x ∂P c ∂S 1 ∂P c ( S 1 ) ∂p 2 ∂p 1 .∂t φ ∂S 1 ∂x S1 + S 2 = 1 u = u1 + u2 u1 f 1 = u ∂S1 ∂u 1 ...
4162 4: TwoPhase Filtration .
If the displacement velocity is not extremely small the BuckeleyLeverett discontinuity may be considered as a sufficient good approximation. It is defined as displacement front.4) 5163 . The meaning is illustrated in Figure 5. This part is called a transition zone. It is recommended to introduce further simplification by specifying the displacement as pistonlike.1) and behind the front: S 1 = S 1M and u2 = 0 (5. The displaced phase flows ahead of the front and the displacing phase is predominant behind the front.5 5. (5.3) (5.1 PistonLike Displacement The Mobility Ratio It was observed that in case of neglecting the capillary forces between displacing and displaced phases a discontinuity in the saturation will develop which may be presented by a sharp front.2) The mobilities of the displacing and displaced fluids are: k r1M λ 1 = µ1 and k r2M λ 2 = µ2 respectively.1: It is assumed that ahead of the front: S 1 = S 1m and u1 = 0 (5. If the capillary force is not neglectable then the saturation profile becomes continuous and a more or less smooth profile.
8) . ν 3 ) where ν is a function of x and t.6) The propagation velocity of this surface is ν = ( ν 1. x 3 + ν 3 ∂t. x 2.ν 3 = 0 ∂t ∂x 1 ∂x 2 ∂x 3 (5.+ .= . t ) = 0 (5. ν 2. At time t the position of the front may be described by the following function: F ( x 1.ν 2 + .2 Propagation of a Displacement Front It is assumed that the flow regimes behind the front just as ahead of the front are at steady state and the fluids are incompressible.⁄ λ2 µ1 µ2 (5.5) The mobility ratio is the ratio between the mobilities of the displaced phase ahead of the front and the displacing phase behind the front and it is a constant for a given rockfluid system Figure 5.7) (5. t + ∂t ) = 0 When differentiating with respect to t: ∂F ∂F ∂F∂F . x 3.ν1 + .1: Comparison of saturation profiles according to different mathematical models 5. The new position of the surface after a time period ∂t can then be described as follows: F ( x 1 + ν 1 ∂t.5164 5: PistonLike Displacement The mobility ratio (M) is defined as: k r1M k r2M λ1 M = . x 2 + ν 2 ∂t.
t ) : p1 = p2 Then the component of velocity normal to the front must be continuous. The following boundary conditions are valid: ψ1 = ψ1 (2) ψ2 = ψ2 (1) x∈Γ (1) (2) (5.– kρ ∇ψ ∇F = 0 ∂t φµ where the relation between u and ν is: u = φν The potential function for a constant density is: p ψ = gx 3 + ρ (5.5: PistonLike Displacement 5165 or in vector form: ∂F .= . 5.10) (5..9) (5.10 into Eq.11) (5. 5.16) .∇ψ µ Substituting Eq.µ 1 ∂n µ 2 ∂n (5.15) (5.9 we obtained: ∂F .+ ν ∇F = 0 ∂t The Darcy’s law in vector form is given by: kρ u = – . t ) = 0 and the (2 ) distribution of the potential ψ 2 between the surface F ( x.14) x∈Γ If neglecting the capillary force the pressure must be equal at the surface F ( x..12) (5. From Eq.13) The problem is now formulated according to Muskat (1934): Determine the distribution (1) of the potential ψ 1 between the border Γ and the surface F ( x. 2.89 the equation of motion for the front may be written as follows: k 1 ρ 1 ∂ψ 1 k 2 ρ 2 ∂ψ 2 . t ) = 0 and the border Γ ..
= 0 ∂x 2 ∂2 p2 . 3.11: ∂F kρ ∂F kρ φ .20) (5.17) 5. The front of displacement has proceeded to xf during the time period t as shown in Figure 5.1 Linear Displacement Let us look upon a horizontal displacement inside a linear medium with length L. ∇ψ 2 ∇F = 0 ∂t µ 1 ∂t µ 2 (5. Just as in region 1 the filtration in region 2 is also a single phase filtration and according to Eq.2.3: ∂2 p1 .5166 5: PistonLike Displacement or in consequence to Eq. and x = L are: p 1 = p 10 p 2 = p 2L x = 0 x = L (5. ∇ψ 1 ∇F = φ .– .= 0 ∂x 2 0 < x < xf (5.19) Figure 5.18) xf < x < L (5.21) At the displacement front: .– .2: Schematic diagram of pistonlike displacement The boundary conditions at x = 0. 5.2.
19 become: p1 = a1 x + b1 p2 = a2 x + b2 (5.25) (5.32) (5. 5.18 and Eq.30) The velocity of the front is: dx f u1 ν f = . 5.Eq. b1.= ∂x ∂x Then the initial condition is given by: xf = 0 at t = 0 (5. 5.23) (5.26) a1.31) . 5.28) (5.= λ 2 ∂x ∂x Since λ 1 ⁄ λ 2 = M .= dt φD where φ D is the displaced fraction of the pore volume which is determined by: φ D = φ ( S 1M – S 1m ) (5. 5. 5. then: ∂p 2 ∂p 1 M .29) (5.25 and Eq. a2 and b2 are all constants of integration which can be determined by the Substitution of Eq.23 into Eq.22) (5.24) After integrating twice Eq.26 leads to: p 2L – p 10 a 1 = xf + M ( L – xf ) b 1 = p 10 M ( p 2L – p 10 ) a 2 = xf + M ( L – xf ) M ( p 2L – p 10 ) b 2 = p 2L – x + M( L – x ) f f (5.20 .5: PistonLike Displacement 5167 p1 = p2 x = xf ∂p 1 ∂p 2 λ 1 .27) (5.
34) (5.34 is then integrated yields: φD 1 t = . The question arises now is: .35 becomes: 1 2 t D = Mx Df + .36) Then Eq.= – kλ 1 a 1 ∂x From Eq. 5. In case c) the displacement is evenly balanced over the whole thickness. 5. During the displacement the boundary of the phases become inclined. In order to simplify Eq.27 and Eq.33) (5.2 Displacement in an Inclined Layer Let us now regard a non horizontal layer in which both the displaced and displacing phase are in static equilibrium if ρ 1 > ρ 2 and the heavier displacing phase is below.37) Eq. 5. defined below. 5. The time t is proportional to xf if M = 1.( 1 – M )x f2 2 kλ 1 ( p 10 – p 2L ) (5. 5.MLx f + .37 is illustrated graphically in Fig. 5. Figure 5.2.= .5168 5: PistonLike Displacement and ∂p 1 u 1 = – kλ 1 .35 the dimensionless variables. 5. the Darcy velocity in Eq. In case b) the displacing phase pushes forwards at the bottom of the layer and affects an unfavorable efficiency of displacement. which means that the two phase mobilities are equal in value.3) which shows the relation between tD and xDf for different values of M. were used: xf x Df = L kλ 1 ( p 10 – p2L ) t D = φD L 2 (5.35) where t is the time period in which the front proceeds to xf.4 illustrates three cases: a) the Initial case at t = 0 a static condition is reached.31 can be expressed as: dx f kλ 1 ( p 10 – p 2L ) ν f = . (5.33.( 1 – M )x Df 2 (5. 5.dt φD xf + M ( L – xf ) Eq.
2 0 0 1 2 3 4 5 M=0.5 xDf 0.S2M.1 tD Figure 5. where S2M is the initial saturation of the displaced phase.5: PistonLike Displacement 5169 When does displacement b) and when does displacement c) takes place? Same simplifications which has been previously discussed are applied: • The phases are incompressible.6 M=10 M=5 M=1 M=0.0 0.8 0.4 0. 1.S2m and the relative permeability of phase 2 is zero. • Behind the front the saturation of the displacing phase is S1M = 1 .3: Influence of the mobility ratio on front propagations in case of a linear displacement . The relative permeability of phase 1 is zero ahead of the front. • Ahead of the front the saturation of the displacing phase is S1m = 1 .
( ∇p2 + ρ 2 gi 3 ) µ2 (5.4: Possible positions of the displacing front in inclined layer.38) (5.5170 5: PistonLike Displacement Figure 5. Behind the front: kk r1M u 1 = – .40) (5.( ∇p1 + ρ 1 gi 3 ) µ1 u2 = 0 ∇u 1 = 0 Ahead of the front: u1 = 0 kk r2M u 2 = – .39) (5.41) (5.42) .
In order to answer the original question it is not of importance to know the position and shape of the front at all times. At both points Eq.Eq. In the Figure 5.5: PistonLike Displacement 5171 ∇u 2 = 0 The boundary conditions are as follows: u1 = u u2 = u u1 ⋅ n = 0 u2 ⋅ n = 0 p1 = p2 u1 ⋅ n = u2 ⋅ n in infinity behind the front in infinity ahead of the front at the impermeable boundaries of the layer (5.5 two neighboring points at the front are considered.38 .46 must be valid: p1 ( r ) = p2 ( r ) p 1 ( r + dr ) = p 2 ( r + dr ) (5. 5. This stable shape of the front then proceeds translatorial.46 have a semi steadystate solution and in addition which kind? The following idea was provided by Dietz (1953). 5.44) (5. The general solution was previously discussed.5: Forces acting on the displacing front. Those are r and ( r + dr ) . This though is so complicated that solutions are only achieved numerically.46) The conditions formulated in Eq. 5. Figure 5. Actually it is essential to know if the boundary value problem Eq.47) .43) (5. It is satisfactory to know if a stable shape is formed after a certain distance of displacement.45) at the front (5. 5.46 neglects the capillary forces and the continuity of filtration velocity at the front.
–  u dr = 0 kk r1M kk r2M (5.6: Position of the displacing front by favorite mobility ratio (after Marle) Figure 5. 5.38 and Eq.47: ( ∇p 1 – ∇p 2 )dr = 0 Substitution of ∇p 1 and ∇p 2 from Eq.. µ1 r2 ⁄  µ k 2 (5. 5. 5.5172 5: PistonLike Displacement Since: p 1 ( r + dr ) – p 1 ( r ) = ∇p 1 dr p 2 ( r + dr ) – p 2 ( r ) = ∇p 2 dr from Eq.42 into Eq.52) where M is the Mobility Ratio: k r1 M = .6 illustrates the position of the front.( 1 – M ) kk r1M (5. 5.48) (5. The mobility of the displaced fluid is superior.53) Figure 5.49) (5.51) This equation indicates that if a stationary solution for the fluid phase exists it must be perpendicular to the vector: u µ1 w = ( ρ 1 – ρ 2 )gi 3 + . In this case . The heavier fluid 1 displaces the lighter fluid 2 from bottom to top.50 leads to: µ2 µ1 ( ρ 1 – ρ 2 )gi 3 + .50) (5.
5.52. 5. The heavier but more mobile phase displaces the lighter and less mobile phase.7 states that again a stationary and stable front will exist. This critical velocity may be calculated using Eq.54) Since i 3 ⋅ u = u ⋅ sin α and u ⋅ u = u k ( ρ 1 – ρ 2 )g sin α u k = – µ1 . Both terms of the right side of Eq. Figure 5.7: Position of the displacing front by unfavorable mobility ratio.55) . Now let us see the case where M > 1 : α Figure 5. If the velocity of filtration is reduced to zero the front will turn to the horizontal position and maintain this position.54 becomes: (5. The displacement proceeds from bottom to top. At the critical velocity the vector w is perpendicular to the axis of symmetry of the layer. This front though becomes more and more flattered as front velocity increases. With increasing displacement velocity the vector w turns to the direction of u . 5..5: PistonLike Displacement 5173 M < 1 . This position is semi steadystate and stable. In consequence the scalar multiplication with u becomes zero: u µ1 w ⋅ u = ( ρ 1 – ρ 2 )gi 3 + . If the velocity exceeds a certain critical value a stationary position of the front becomes impossible to maintain.( 1 – M ) u = 0 kk r1M (5.( 1 – M ) k r1M 2 Eq.52 become positive.
A case of ρ 1 < ρ 2 will not be discussed.5174 5: PistonLike Displacement uk is the critical velocity of filtrations. A filtration with velocity higher then uk is called supercritical. . In this case though the heavier fluid is on top of the lighter fluid which makes the front unstable.
57) Figure 5.56) (5.[ h 1 u 1 + h2 u 2 ] h (5. The front has travelled a considerably far distance and the state is to be referred to as practically stable. t ) and 1 u = . The theories of Le Fur and Sourieau are used for explanation purposes in this work. That means. In this case u1 is the same at every point of the layer which contains the fluid 1 (section h1(x. In an abitrary but fixed point xa the value h1(x.8: Supercritical displacement in inclined layer (after Marle) If ρ 1 > ρ 2 the displacing fluid takes the lower part of the layer and h1(x.t)). t ) + h 2 ( x.t) will monotonously increase with time. The same is valid for u2 in the layer section h2(x. the rate of changes is everywhere moderate.5: PistonLike Displacement 5175 5.3 Supercritical Displacement If the velocity of displacement is larger than the critical velocity the interface will become more and more extended Figure 5. In this case the Dupuitassumption is valid which states that the equipotentials are perpendicular to the layer.t). At first it is assumed that displacement is at an advanced stage. The relation between h1 and h2 is: h = h 1 ( x.t) is referred to as the distance between the bottom of the layer and the front.8. This displacement is called supercritical. .2.
0.u2 h p 1 ( x.60) (5.69) . t ) = p 2 ( x.5176 5: PistonLike Displacement The velocity of filtration in direction of the xaxis is: kk r1M ∂p 1 u 1 = – .+ ρ 2 g sin α µ 2 ∂x (5. t ) γ = g sin α * * (5.+ ρ 2 g cos α = 0 ∂y The equation of continuity is set up for the whole layer as: ∂h 1 ∂ ( u1 h1 ) .61) (5.+ ρ 1 g cos α = 0 ∂y ∂p 2 . t ) = p 1 ( x.66) (5.67) (5.+ φ ( S 1M – S 1m ) . .64) (5.= 1 – S 1 h h1 w 1 = .58) (5. t ) p 2 ( x.= 0 ∂x ∂x and the following average values are defined as: h1 S 1M + h 2 S 1m S 1 = h h1 S 2m + h 2 S 2M S 2 = .59) Since filtration is parallel to the axis of the layer at every point (a consequence of the Dupuitassumption) the following equations are valid: ∂p 1 .62) (5. 0.65) (5.+ ρ 1 g sin α µ 1 ∂x kk r2M ∂p 2 u 2 = – .63) (5.u1 h h2 w 2 = . .68) (5.
79) .78) (5. 5.5: PistonLike Displacement 5177 S 1 – S 1m * k r1 ( S 1 ) = k r1M S –S 1M 1m (5. 5. 5.77) (5.75 into Eq. t ) + h 2 ρ g cos α 1 2 Since: S 1 – S 1m h 1 = h S 1M – S 1m and from Eq.71) As before the capillary force is neglected and the pressures of the phases must be equal at the front. h 1.68: p 1 ( x.+ φ .73) (5. t ) = p 2 ( x. t ) + h 1 ρ g cos α = p2 ( x. 5. t ) and from Eq.63 . . h 2.58.75) (5.73: P c ( S 1 ) = p 2 ( x.72) (5. 5. t ) – p 1 ( x.70) S 1 – S 1m * k r2 ( S 2 ) = k r2M S 1M – S 1m (5.76) (5.= 0 ∂t ∂x * * * * * * * * * (5.+ ρ 2 γ µ 2 ∂x S1 + S2 = 1 ∂S 1 ∂w 1 .74) (5.+ ρ 1 γ µ 1 ∂x kk r2 ∂p 2 w 2 = – . 5.Eq. t ) = S 1 – S 1m h . 5.59 and Eq.67 and Eq.62 leads to: kk r1 ∂p 1 w 1 = – .( ρ 1 – ρ 2 )g sin α S 1M – S1m Substitution of Eq. 5. . then: p 1 ( x. Eq.
Only the function P c ( S 1 ) has nothing to do with capillary pressure.80) (5. The two dimensional approximation of the supercritical displacement corresponds with a one dimensional displacement where γ S 1.5178 5: PistonLike Displacement w1 + w 2 = 0 p2 – p1 = Pc ( S1 ) * * * (5. S 1 p 1. 4. * .14. etc. because the capillary forces were neglected due to precondition Eq.76 .9 . are physically corresponds to the actual factors.Leverett Eq. Therefore the fictive saturation is the average saturation the fictive pressure is the pressure at any select point. k r2 ( S 2 ) Pc ( S1 ) * * * * * is the fictitious gravity. All factors.Eq.81) Eq. except Pc. w 2 k r1 ( S 1 ).72. are the pseudo relative permeabilities is the pseudo capillary pressure. p 2 w 1. 5. the average saturations. are the fictitious phase pressures are the fictitious fractional velocity. 5.Eq.81 may be considerated identical with the equations of Buckley . 5. 4.
Y..3 Dietz. Wetenschap.: "Multiphase Flow in Porous Media" Institut du Petrole..1 5.N.M.83 (1953). . Akad. Proc. M.: "A theoretical approach to the problem of encroaching and bypassing edge water" Koninkl.. (1937).: "Flow of Homogeneous Fluids through Porous Media". Marle. Gulf Publishing Company (1981). McGraw Hill Book Co. D.2 5. C. B56. Ned. N. Muskat.3 References 5.5: PistonLike Displacement 5179 5.
5180 5: PistonLike Displacement .
Correct and sovereign answers attest that the candidate is familiar with the subject and understands the topics very well. A . the most important formulae and figures supported by short explanations are asked. With the help of his/her notes. The final mark will range between excellent (1) and fair (3). . Every colour marks a question referring to a distinct part of the subject. As far as the Aquestions are concerned. In case of optimum answering. In doing so. At first all Aquestions have to be answered orally by adducing short and clear definitions without any derivations. the mark fair (3) is offered. The candidate gets 2030 minutes to prepare the answers to the Bquestions. the candidate has to present the answers orally. the candidate is examined if he/she is skilled to use his/her knowledge in solving problems which had not been discussed in the lectures.Fundamental concepts.6: Examination Outline 6181 6 Examination Outline The candidate draws three cards of different colours. the candidate is failing in case of only one bad mistake or one lack of knowledge.The theoretical background is asked. Moreover. Every card comprises two questions: A and B. the demonstration of his/her understanding of physical meanings is more important than the exactness of derivation. The candidate must be able to derivate formulae. However. the most important formulae have to be written and figures have to be sketched. B . the candidate can endeavour to get a better mark by approaching to answer the Bquestions. In doing so. Five minutes are the optimum period of time for answering the three Aquestions.
1. • Darcy´s law • definition of permeability • compare liquid and gas permeability • explain anisotropy .Questions 6.1. Saturation. Anisotropy Liquid and Gas Permeability. Capillary Pressure • definition of saturation • explanation of the term wettability • explain the terms wetting and nonwetting • What is capillary pressure? • illustrate the capillary pressure curve 6.1.1.2 Wettability.1.3 Darcy Equation.1 6.1.1.1 Part 1 Porous Material • definition of porous medium and porosity • classification of porous material • What is compressibility? • What is compaction? • What is the difference between compaction and compressibility? 6.1 A .6182 6: Examination Outline 6.
Boundary Conditions • compressibleincompressible fluid • equation of filtration for low compressible fluid.1.2.2 Equation of Filtration for Low Compressible Fluids.1.3 Equation of Filtration for Real Gas • real gas pseudo pressure • equation of filtration for real gas • ideal gas 6.1.1.2.2.2 6.4 SteadyState Filtration through Linear Systems for Liquids and Real Gases .1 Part 2 Continuity Equation.2.1.4 Relative Permeability • define and explain relative permeability 6.1. Darcy´s Law. Equation of State • equation of continuity • main forces in the filtration process • differential form of Darcy´s law • equation of state 6.6: Examination Outline 6183 6. piezometrical conductivity • boundary conditions: closedopen boundary 6.1.
2.7 Calculation of the Cumulative Influx • infinite radial systems with constant pressure at the inner boundary • dimensionless variables 6.2.5 • equation SteadyState Filration through Radial Systems • explain the term radialsymmetric with respect to potential distribution • Dupuit equation 6.1.6184 6: Examination Outline • explain the term steadystate • steadystate filtration for liquids • steadystate filtration for real gases • What is the major difference between the two equations? 6.2.2.6 NonSteadyState Filtration with Constant Production Rate through Infinite Radial Systems • differential equation • inital and boundary conditions • solution with Eifunction • solution with lgfunction 6. constant pressure) • plot the pressure changes and the cumulative influx for constant boundary conditions • asymptotic solutions for large periods of time .1.8 Finite Radial Systems • differential equation • constant boundary conditions (constant production rate.1.1.
2 Effect of Gravity and Viscosity on the fCurve.3 6.6: Examination Outline 6185 6.1 Part 3 BuckleyLeverett Theory • assumptions • ffunction • saturation profile • front saturation 6.1.2.1.1.2.9 Physical Meaning of Superposition • explain the principle of superposition • first law of superposition • second law of superposition 6.1.3.3.10 PressureBuild Up of a ShutIn Well • illustrate the idea of a pressurebuild up • Horner plot: What do you plot? What can you determine from the plot? 6. Influence of Capillary Pressure on the Saturation Profile • illustrate the effect of gravity on the fcurve • explain the effect of viscosity on the fcurve • influence of capillary pressure on the saturation profile • illustrate the influence of displacement velocity on the saturation profile • What is the capillary end effect? .1.
1.4 Displacement in Inclined Layers • illustrate possible positions of the displacing front in an inclined layer and the forces acting at the front • Dietz stability analysis • What is supercritical displacement? .3.1.3.6186 6: Examination Outline 6.3 PistonLike Displacement in Linear Systems • define the mobility ratio • illustrate the idea of pistonlike displacement • front velocity • explain the effect of different mobility ratios on the pressure distribution 6.
2.1 6.3 Permeabilty Measurements • Which factors influence the applied method? • illustrate common permeabilty measurements • analogies to the Darcy equation 6.4 Evaluation of Relative Permeability Functions • How can relative permeability functions be measured? 6.1.1.1 Part 1 Porosity Measurements 6.1.2 B .1.2.2.2.2.2.2.2.2 6.6: Examination Outline 6187 6.Questions 6.1 Part 2 Derivation of Special Forms of the Equation of Filtration 1 • derive the equation of filtration for an incompressible fluid .2 Evaluation of Capillary Pressure Curves • How can the capillary pressure of a porous medium be measured? • conversion of laboratory data • Leverett function 6.
2.2.2.2.2.2.2.2.5 Second Law of Superposition • calculation of multiwell problems • wells with variable production rate 6.2.2.4 Filtration of Low Compressible Fluids through Radial Systems with Closed Boundaries • illustrate the principle of dimensionless variables • derive equations for boundaries with constant production rate and constant pressure • give solutions for the early and the late period 6.2 Derivation of Special Forms of the Equation of Filtration 2 • derive the equation of filtration for an ideal gas • derive the equation of filtration for a real gas 6.6 Methods of Image • illustrate the principle of the method • pressurebuild up test near a no flow boundary for early and late period • pressurebuild up test near a constant pressure boundary .6188 6: Examination Outline • derive the equation of filtration for a low compressible fluid 6.3 NonSteadyState Filtration through Infinite Systems • derive the equation of filtration for a radial system with constant production rate • describe the properties of the Eifunction 6.
1 Part 3 Derivation of the BuckleyLeverett Theory • assumptions • derive the BuckleyLeverett solution 6.2.2.4 Supercritical Displacement .3 PistonLike Displacement in Linear Systems • assumptions • derive the velocity of the front • derive the time period in which the front proceeds to xf 6.2.2.2 Welge Method • illustrate the principle of the Welge method • How do you determine the front saturation? • How do you determine the average saturation? 6.3.3.2.3.3.3 6.6: Examination Outline 6189 6.
6190 6: Examination Outline .
This action might not be possible to undo. Are you sure you want to continue?
Use one of your book credits to continue reading from where you left off, or restart the preview.