Professional Documents
Culture Documents
A R T I C L E I N F O A B S T R A C T
Keywords: In this scholarly work, a novel mathematical model for single-phase, single-well simulation in hydrocarbon
Single-well simulation reservoirs was established using the fractional calculus theory. Specifically, the existing fractional generalization
Fractional calculus of Darcy’s law is reconstructed by introducing an auxiliary parameter to ensure that the definition of the con
Radial diffusivity equation
ventional rock permeability with dimension L2 was preserved. Subsequently, the newly constructed fractional
Finite difference method
Fractional-order derivative
flux relationship was substituted as a momentum equation in the mass conservation expression in the radial-
Grünwald–Letnikov cylindrical coordinate system. The resulting nonlocal time-radial diffusivity equation was solved numerically
using the block-centered finite-difference approximation and adopting the Grünwald–Letnikov formula for the
definition of the fractional-order derivative. The developed numerical scheme was verified by comparing the
approximate pressure solution against the semi-analytical solution of the simplified nonlocal time-radial diffu
sivity equation and the classic radial flow diffusivity equation. Furthermore, incremental material balance checks
were performed and exhibited to ensure the conservation of mass at each time step. Finally, a detailed sensitivity
analysis was performed to evaluate the impact of the order of fractional differentiation on pressure distribution in
the reservoir and wellbore. The proposed nonlocal time-radial diffusivity equation proffers a useful model for
single-well simulation in geological media that exhibit distorted flow paths and near power-law behaviors.
https://doi.org/10.1016/j.petrol.2020.107162
Received 9 August 2019; Received in revised form 10 February 2020; Accepted 4 March 2020
Available online 10 March 2020
0920-4105/© 2020 Elsevier B.V. All rights reserved.
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
Camacho Velazquez et al., 2006; Chang and Yortsos, 1990; Hardy and proposed fractional diffusion model conveniently allows for the
Beier, 1994; Razminia et al., 2015, 2014; Sahimi and Yortsos, 1990). consideration of anisotropy, stress-sensitivity, and multiphase flow etc.
Besides, the lack of a clear scale separation in unconventional reservoirs while modeling flow through porous media.
has contributed to the shortcomings of the classical theory in forecasting The plan of this paper is as follows: Section 2 details the theoretical
hydrocarbon production in shale gas reservoirs (Albinali and Ozkan, background for the re-construction of the constitutive fractional flux
2016; Chen and Raghavan, 2015; Holy and Ozkan, 2016; Obembe et al., expression and the derivation of the proposed nonlinear nonlocal time-
2017b, 2017c; Ozcan et al., 2014; Raghavan and Chen, 2018, 2016; Ren radial diffusivity equation. The implicit finite difference discretization of
et al., 2018; Ren and Guo, 2015). the proposed mathematical model including wellbore method and so
In general, the seepage equations developed using the fractional lution methodology are presented in Section 3. The semi-analytical so
calculus theory have been successfully applied to address the deviations lution (SAS) to the simplified version of the mathematical mode (i.e.
observed from the classical theory of fluid flow in porous media. This is uniform rock and fluid properties) is derived in Section 4. The SAS is
because such fractional diffusion models (i.e. using fractional derivative utilized later in Section 5 to verify the developed numerical model so
operators) provide an excellent tool for the description of memory and lution (i.e. comparison of the SAS to the numerical solution). Further
other hereditary properties of various materials and processes (Pod more, findings from synthetic examples of the single-well simulation are
lubny, 1998). In addition, fractional diffusion models imply the notion presented and discussed in the remainder of Section 5. The paper ends
that diffusion is irregular (i.e. anomalous diffusion) and the well re with some conclusions and remarks in Section 6.
sponses in such media follow a power-law behavior. A wealth of frac
tional diffusion models have been presented in the literature using the 2. Theoretical formulation
theory of fractional calculus to exhibit the subdiffusive nature of flow in
heterogeneous porous media (Al-Rbeawi and Owayed, 2019; Amir and 2.1. The classic representation of subdiffusive flux
Sun, 2018; Awotunde et al., 2016; Chen et al., 2014; Liang et al., 2019;
Obembe et al., 2018b, 2018a; 2017d, 2017e; Razminia et al., 2019; Sun The simplest constitutive relationship describing fluid flow in porous
et al., 2009) to name a few of the numerous citations. Subdiffusion de media is the classical Darcy’s law expressed as
notes a transport process where the mean square displacement of the � �
βc K
diffusing particle follows a non-linear function of time (i.e. power law ⇀ ⇀
uðx ; tÞ ¼
⇀
rΦðx ; tÞ (1)
μo
behavior). Besides, subdiffusion implies a memory formalism where the
history of the flow process during its journey in the porous media im
where βc ¼ 1:127 � 10 3 is a conversion factor equal, K[mD] refers to
pacts the macroscopic transport behavior. In general, subdiffusive
rock permeability and in its complete form is a tensor, μo [cp] denotes
transport has been reported to be more appropriate to describe fluid
the crude-oil viscosity, r is the gradient operator, and Φ[psia] is the
flow in a naturally fractured and disordered nano-porous media (Albi
potential.
nali and Ozkan, 2016; Obembe et al., 2017a; Raghavan and Chen, 2019,
The potential (Φ) in Eq. (1) is related to the pressure (p) through the
2018).
expression:
Typically, these fractional diffusion models are arrived at by modi
fying Darcy’s flux relationship via the introduction of fractional deriv
� �
ative operators resulting in variants of non-local constitutive temporal Φ Φref ¼ p pref γ Z Zref (2)
and/or spatial flux expressions. So far, numerous analytical and nu
merical techniques have been successfully applied to solve the resulting where the fluid gravity γ ¼ γc ρg, ρ denotes the fluid density (lbm/ft3),
fractional diffusion equations and many interesting findings have been γ c refers to the gravity conversion factor (0:21584 � 10 3), g is the ac
reported. Some of the key differences in the literature on fractional celeration due to gravity (32:174 ft/sec 2), p is the pore pressure (psia),
diffusion models comprise of the following; the interpretation of the pref is the pressure at reference conditions, and Z is the elevation from
fractional derivative operator (e.g. Caputo, Riemann-Liouville, Grün datum (reference), with positive values downwards.
wald-Letnikov, etc.) and the physical interpretation of the terms asso In recent years, fractional constitutive laws have been shown to be
ciated with the resulting modified flux expressions. Specifically, the adequate for representing the internal architecture of rock fabric and
physical interpretation of the pseudo-permeability coefficient (Kα ) where the presence of topological, geometrical and spatial influences
associated with current expressions is ambiguous and is a major draw result in distorted flow paths and loss of connectivity. Specifically, for
back to the widespread adoption of these models in the oil and gas subdiffusive flow, Eq. (1) can be generalized into the fractional Darcy
industry. law given by (Caputo, 1999; Raghavan, 2011):
The goal of this study is as follows; First, to propose a fractional
diffusion model for single-phase, single-well simulation in geological � � α
media utilizing a modified fractional Darcy flux expression as the gov βc Kα ∂1
(3)
⇀ ⇀ ⇀
uðx ; tÞ ¼ ½rΦðx ; tÞ�
erning momentum equation in the radial-cylindrical mass conservation μo ∂t 1 α
2
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
∂1 α t=Δt
X ð 1Þk Γð2 αÞ relevant physics becomes problematic.
α
f ðtÞ ¼ lim Δtα 1
f ðt kΔtÞ (4) Conversely, Eq. (7) is praise-worthy in that it allows the reservoir
∂t 1 Δt→0 k!Γð2 α kÞ
k¼0
engineer to separate distinct phenomena observed in porous media.
Specifically, this flux expression clearly separates the parameters that
where Γð:Þ denotes the Gamma function and 0 < α < 1 is the order of
control sub-diffusion (i.e. α and σ) from the conventional rock perme
fractional differentiation.
ability parameter which addresses petrophysical heterogeneity
Subdiffusive flux is characterized by 0 < α < 1 in Eq. (3) and inter
(permeability variation with space), multiphase flow, and stress-
estingly Eq. (3) reverts to the classical Darcy’s law when α ¼ 1. Of
sensitivity etc. This praiseworthy development is illustrated through
significance, the flux expression given by Eq. (3) incorporates the
the synthetic single-well simulation examples presented in Section 5 that
memory effects of fluid transport which were neglected in Eq. (1).
account for heterogeneity, anisotropy and stress-sensitive permeability
Several scholars in petroleum engineering literature have adopted Eq.
which will be impossible to adopt using Eq. (3).
(3) as an appropriate expression for the fluid flux for subdiffusive flow
(Albinali and Ozkan, 2016; Caputo, 1999; Lemehaute and Crepy, 1983;
2.4. Derivation of nonlocal time-radial diffusivity model
Obembe et al., 2017e; Raghavan and Chen, 2019) to name a few of the
many citations. However, the main drawback to this approach is that
Fluid flow in porous media is described mathematically by
physical interpretation of Kα in Eq. (3) is not straightforward. It is argued
combining the mass conservation equation, a representative constitutive
that Kα is a dynamic property with dimension ½L2 T1 α � and consequently
equation (momentum equation), and an appropriate equation of state
is different from the conventional Darcy permeability (K) with dimen
(Ertekin et al., 2001; Jamal et al., 2006). If we consider the
sion L2 . In addition, the literature indicates that static measurements are
two-dimensional (2D) radial-cylindrical grid geometry shown in Figure
not suitable to determine this parameter and that the only practical
(Fig. 1), the mass conservation equation is derived by acknowledging
method to estimate Kα is to match the transient pressure or flow rate
that over a given time interval (Δt ¼ t nþ1 t n ), the fluid coming from
data with an appropriate model (Ozcan, 2014). In the next sub-section,
neighboring blocks enters block (i; j) through block boundaries
the theory that constitutes the alternative approach to describe sub � �
diffusion in porous media is presented. (i 12;j) and (i;j 12) and leaves through its boundary blocks i þ12; j and
� �
2.2. Modified subdiffusive flux i; j þ 12 , including that fluid may enter or leave grid block ði; jÞ
through the action of an injection or a production well.
This section details the dimensionality of fractional and ordinary The statement of mass conservation for the radial-cylindrical coor
differential operators. To begin, if the ordinary time derivative operator dinate system for 2D flow is expressed in engineering form (Jamal et al.,
is replaced by the fractional representation as follows: 2006) on page 33 as follows:
d dα
→ (5)
dt dtα
Ztnþ1 � �� Ztnþ1 � �� Ztnþ1 � ��
ur Ar �� ur Ar �� uz Az ��
Go�mez-Aguilar et al. (2012) noted the dimensional inconsistency dt dt þ dt
Bo �r Bo �r Bo � z
associated with Eq. (5) by physical inspection of the expression since the tn i 1; j tn iþ1; j tn i;j 1
(8)
2 2 2
ometry. Accordingly, the following relationship was derived: where the porosity (φ) and oil formation volume factor (Bo ) are pressure
d 1 dα dependent, qsc is the volumetric rate at standard conditions of pressure
→ 0<α<1 (6) and temperature, Vb is the bulk volume of the grid block, αc is a volu
dt σ1 α dtα
metric conversion factor equal to 5.615, and Ar and Az represents the
where the parameter 0 < σ < 1 has a dimension of time (day) cross-sectional area normal to the flow in r and z directions.
(Go
�mez-Aguilar et al., 2012; Go�mez et al., 2015; Kang et al., 2019). The fluid volumetric velocities in the r and z directions can be re-
Therefore, Eq. (3) can be re-written as follows: written by the algebraic expressions as follows:
� � α 0 1
βc K σ 1 α
∂1
(7)
⇀ ⇀ ⇀
uðx ; tÞ ¼ ½rΦðx ; tÞ� � �� BΦ
μo ∂t1 α
β Kr σ1 α �� ∂1 α
B i 1;j Φi;j C
C
ur jr 1 ¼ c � α B C (9)
i ; j
2 μ o ∂t1 @ Δri 1; j A
where σ and α are the phenomenological parameters that can be ob
ri 1; j 2
2
tained by fitting the mathematical model with field data (i.e. wellbore
pressure). In addition, K is the conventional rock permeability which 0 1
may be a constant value or pressure dependent. � ��
β Kr σ1 α �� ∂ B
1 α C
BΦi;j Φiþ1;j C
ur jr 1 ¼ c � α B C (10)
iþ ; j
2 μ o riþ1; ∂t @ Δriþ12; j A
1
0 1
Until now, Eq. (3) has been widely adopted to model subdiffusive
flow through porous media. Specifically, this flux representation in � ��
βc Kz σ1 α �� ∂1 α BΦ C
B i;j 1 Φi;j C
troduces Kα and α as the model parameters that govern the subdiffusive uz jz ¼ � α B C (11)
i;j 1
2 μ o ∂t1 @ Δzi;j 12 A
transport through porous media. However, this approach fails to
zi;j 1
2
3
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
0 1
pressure.
� �� α BΦ C
β c Kz σ 1 α � ∂1 B i;j Φi;jþ1 C
u z jz ¼ �
� αB C (12) 3. Numerical formulation
i;jþ1
2 μo zi;jþ 1
∂t1 @ Δzi;jþ12 A
2
2 3 numerical estimate of the exact value of pðr; z; tÞ at the center of the grid
blocks ðri ; zj ; tn Þ.
Ztnþ1 6� ��
6 β c K z σ 1 α Az � ∂1 α � 7
7 (13) First, define a function ωαk such that:
6 � Φi;j Φi;j dt7
6 μ Bo Δz � ∂t1 α 1 7þ
4 o zi;j 1 5
tn ð 1Þk Γð2 αÞ
(14)
2
ωαk ¼
k!Γð2 α kÞ
2 3
MacDonald et al. (2015) showed that the coefficients ωαk can be
Ztnþ1 6� �� 7
6 β c K z σ 1 α Az �
6 � ∂1 α � 7 evaluated by means of the recursive formula:
6 μ Bo Δz � α Φi;jþ1 Φi;j dt7
7þ
4 o z i;jþ1
∂t1 5
tn 2
� �
2 α k
ωα0 ¼ 1; ωαk ¼ ωαk 1 (15)
k
Ztnþ1 �� �nþ1 � �n �
Vb φ φ
qsci;j dt ¼ i;j Therefore, the G-L fractional derivative operator defined in Eq. (4)
αc Bo i;j Bo i;j
tn may be re-written in abridged form as:
Equation (13) is the nonlocal time-radial diffusivity equation for h i
tn
single-well simulation in hydrocarbon-bearing rocks. The model is ∂1 α X
Δt
h i
tn tn
where Δt is the integer part of Δt .
Define fluid transmissibility terms by the expression:
� �
� 1
Tri�1; j ¼ Gri�1; j σ1 α
(17)
2 2
ri�1;
2
j μo B o ri�1; j
2
� �
� 1
Tzi;j�1 ¼ Gzi;j�1; σ 1 α
(18)
2 2
zi;j�1;
2
μo Bo zi;j�1;
2
where Tri�1; j and Tzi;j�1 represent the transmissibility in r and z flow di
2 2
rections, and Gri�1; j and Gzi;j�1 are the geometric factors between two
2 2
4
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
β Δθ
Gri ¼ 0 1 c j 0 1 (19) ri2 ri2 1
, , (29)
1;j
2 ri2 1 ¼ � �
2 B C B C3 2
loge @ri rL 1 A loge @rL 1 ri 1 A loge ri2= 2
i
2
i
2 ri 1
6 7
6 7
6 þ 7
where r1L ¼ rw , rNL r þ1 ¼ re correspond to the inner and external boundary
4 Δzi;j Kri;j Δzi 1;j Kri 1;j 5
2 2
2 3 2 3
Ztnþ1 1 α Ztnþ1 α
0 1 6 ∂ �7 6 ∂1 �7
� � 4Tzi;j 1 1 α Φi;j 1 Φi;j 5dt þ 4Tzi;jþ1 1 α Φi;jþ1 Φi;j 5dtþ
2 ∂t 2 ∂t
1 2 Δθ B 2 C tn tn
2βc =
j @riþ1 ri2 1A
2 2
0 1 (30)
� �
B 2 C The bulk volume for a grid block is given by (Aziz, 1979):
2βc 1 2 Δθ ri2 1 A
@riþ1 Reservoir inner reservoir intern
=
j
2 2
� �1=
r e Nr
αlg ¼ (23) To develop the numerical scheme, the time integrals on both sides of
rw Z tnþ1
� �� Eq. (30) are approximated by the expression FðtÞdt ¼ Fnþ1 Δt ,
αlg loge αlg tn
r1 ¼ rw (24) where Δt ¼ t nþ1 t n .
αlg 1
Therefore, Eq. (30) reduces to:
For i ¼ 1; 2; …:Nr 1 � 1 α�
∂ �� � 1
∂ α � ��
riþ1 ¼ αlg ðri Þ (25) Trnþ11 1 α Φnþ1
i 1;j Φnþ1
i;j þ Trnþ11 α Φnþ1
iþ1;j Φnþ1
i;j þ
i ;j ∂t iþ ;j ∂t1
2 2
riþ1 ri � 1 α� �� � 1 α � ��
L
riþ 1 ¼ � � (26) ∂
Tznþ11 1 α Φnþ1 Φnþ1 þ Tznþ11 1
∂
Φnþ1 Φnþ1 þ qnþ1
2 i;j 1 i;j α i;jþ1 i;j sci;j ¼
2 ∂t 2 ∂t
i;j i;jþ
loge riþ1=r
i
�� �nþ1 � �n �
Vbi;j φ φ
2
riþ1 ri2 αc Δt Bo i;j Bo i;j
2
riþ 1 ¼ � � (27)
2
2
loge riþ1 (33)
=ri 2
For slightly compressible fluids the right-hand side (RHS) of Eq. (33)
For i ¼ 2; 3; …:Nr can be represented in conservative form as follows (Jamal et al., 2006):
ri ri
(28)
1
riL 1 ¼ � � !
2 �� �nþ1 � �n � � � � �
loge ri=r Vbi;j φ φ Vbi;j φref
i 1 � ðco þ cφ Þ Pnþ1
i;j Pni;j
αc Δt Bo i;j Bo i;j αc Δt Boref
i;j
(34)
5
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
where Boref is formation volume factor (FVF) at reference conditions (i.e. Re-arranging the terms in Eq. (37) results in the following:
oil-bubble point pressure and reservoir temperature), φref is porosity at Tznþ1 Pnþ1 nþ1 nþ1
m Nr þ Trm 1;m Pm 1
reference conditions, co is the oil compressibility factor, and cφ is the
m Nr;m
h i
rock compressibility factor. Tznþ1 þ Trnþ1 þ Trnþ1 þ Tznþ1 þ Cm Pnþ1
m þ
Furthermore, for convenience, the index of the terms in Eq. (30) will
m Nr;m m 1;m mþ1;m mþNr;m
be identified with a single index ðmÞ rather than (i;j). This is achieved by Trnþ1 Pnþ1 nþ1 nþ1
Cm Pnm qnþ1
mþ1 þ TzmþNr;m PmþNr ¼
counting the grid blocks along the r direction first (horizontal sweep)
mþ1;m scm
n o n o
with i index, followed by the z direction with the j index, then the r
Tznþ1 γn Nr;m ðZm Zm Nr Þ Trnþ1 γnm 1;m ðZm Zm 1 Þ
direction again and so forth. This can be described mathematically by: m Nr;m m m 1;m
n o n o
m ¼ i þ ðj 1ÞNr (35) Trnþ1 γn ðZm Zmþ1 Þ Tznþ1 γn ðZm ZmþNr Þ
mþ1;m mþ1;m mþNr;m mþNr;m
h � i tion, Eq. (38) for an interior grid block, m reduces to Eq. (39).
Trnþ1 Pnþ1
m 1 Pm nþ1
γ nm 1;m ðZm 1 Zm Þ þ
m 1;m � �
h � i Trnþ1
m 1;m
Pnþ1
m 1 Trnþ1
m 1;m
þ Trnþ1
mþ1;m
þ Cm Pnþ1 nþ1 nþ1
m þ Trmþ1;m Pmþ1 ¼
Trnþ1
mþ1;m
Pnþ1
mþ1 Pnþ1
m γ nmþ1;m ðZmþ1 Zm Þ þ 2 ht i 3
n
h � i 6X
Δt �7
6 7
Tznþ1 Pnþ1
mþNr Pnþ1
m γnmþNr;m ðZmþNr Zm Þ þ Cm Pnm qnþ1
scm Trnþ1 6 ωαk Pnþ1
m 1
k
Pnþ1
m
k
7
mþNr; m m 1;m
4 k¼1
5
8 ht i 9
> n >
>
> > 2 ht i 3
<X Δt h � i>
= n
α
Tznþ1 ωk Pnþ1
m Nr
k
Pnþ1
m
k
γ nm kNr;m ðZm Nr Zm Þ þ 6XΔt �7
6 7
> > (39)
m Nr;m
>
>
: k¼1 >
>
; Trnþ1 6 ωα Pnþ1 k
Pnþ1 k
7
mþ1;m
4 k¼1 k mþ1 m
5
8 ht i 9
> n >
>
>
<X Δt h
>
i>
= (37)
�
Trnþ1 ωαk Pnþ1
m 1
k
Pnþ1
m
k
γnm k
1;m ðZm 1 Zm Þ þ 3.2. Modified wellbore model
m 1;m >
> >
>
>
: k¼1 >
;
In reservoir simulation, a well is modelled as a line source/sink term.
8 ht i 9 Considering one-dimensional single-well simulation and a block-
> n >
> >
>X
< Δt h �
>
i= centered grid enclosed between the external radius r1 and the well
Trnþ1 ωαk Pnþ1 radius rw , the production rate at the wellbore can be expressed by
k
mþ1 Pnþ1
m
k
γnmþ1;m
k
ðZmþ1 Zm Þ þ
mþ1;m >
> >
>
>
: k¼1 >
; Darcy’s law for radial flow as follows:
G w1 �
8 ht i
n
9 qsc1 ¼ P1 Pwf (40)
>
> >
> μo1 Bo1
>
<X Δt h i>
=
�
Tznþ1 ωαk Pnþ1 k
Pnþ1 k
γnmþNr;m
k
ðZmþNr Zm Þ þ
mþNr; m
>
>
mþNr m
>
> where Pwf is the wellbore pressure and Gw1 is the well geometric factor
>
: k¼1 >
;
expressed as:
� 2π βc kr1 Δz
qnþ1 (41)
nþ1
scm ¼ Cm Pm Pnm G w1 ¼ � �
! ln rrw1
Vbm
where Cm ¼ ðco þ cφ Þ
φref
αc ðΔtÞα Boref However, for a vertical well penetrating several blocks (i.e. different
m
layers), the well production rate must be allocated among the different
6
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
sure.
δwfm ¼ ωαk Pnþ1
m
k
Pnþ1
wfm
k
k¼1
Thus, for the case when the vertical well penetrates several blocks,
the contribution of well block m to the total well production rate can be Therefore, the wellbore pressure in wellblock m is computed from
obtained by substituting Eq. (43) into Eq. (42). Eq. (53).
� �
G wm h �i qscm μom Bom nþ1
qscm ¼ Pm Pwfref γ wb Zm Zref (44) Pnþ1 nþ1
wfm ¼ Pm þ Δt
1 α
þ δwfm (53)
μom Bom GWm σ 1 α
Given that the sum of the production rates of all wellblocks must add
up to the specified-rate-condition given by Eq. (45).
X 3.3. Solution methodology
qspc ¼ qscm (45)
mεψ w The successive Over Relaxation (SOR) iterative method was imple
mented to handle the non-linearity of the developed nonlinear algebraic
where ψ w is the set whose elements are the blocks penetrated by the equations derived above. Specifically, for each individual grid block we
vertical well. proceed as follows:
Combining Eqs. (44) and (45), Jamal et al. (2006) derived an
expression for the flowing bottomhole pressure as follows:
I. Assume guess pressure values (Pnþ1 ) at the unknown time (i.e. n þ
ðvÞ
m
�� �h �
P Gw m
�i 1).
P P γ Z Z þ qspc
II. With Pnþ1 , estimate all fluid properties (i.e. Bo , μo etc.) using the
mεψ w μom Bom m wfref wb m ref ðvÞ
Pwfref ¼ � � (46) m
P Gw m correlations presented in Appendix B.
mεψ w μom Bom
III. Solve Eq. (38) or Eq. (39) iteratively until Pnþ1 and the obtained
ðvÞ
Besides, for the special case of equal pressure drop ðΔP ¼ Pm Pwfm Þ in m
vertically stacked wellblocks, Jamal et al. (2006) showed that the pro pressure (Pnþ1 ) satisfy the convergence criterion. In this study,
ðvþ1Þ
m
duction rate of each wellblock can be prorated according to the the convergence criterion is set as:
expression: �
�Pnþ1ðvþ1Þ Pnþ1ðvÞ �
�
� �
Gw m max � m ðvÞ
m
� � 10 6 (54)
1<m<nr � Pnþ1 �
(47)
μom Bom m
qscm ¼ � �qspsc
P Gw m
mεψ w μom Bom
7
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
4. Semi-analytical solution to the homogenous single-well where the coefficients (β; ϱ and ζ) and variables associated with Eq.
simulation model (62) are defined appropriately in Appendix A. Furthermore, the tran
sient pressure distribution is obtained in the real time domain by uti
In the following, the semi-analytical solution (SAS) for the nonlocal lizing the Stehfest (1970) numerical inversion algorithm. Eq. (62) will
time-radial diffusivity equation considering a homogenous one- be used in Section 5 to verify the accuracy of the numerical solution.
dimensional (1D) radial-cylindrical closed reservoir (Fig. 2) is derived
employing the Caputo definition for the time fractional-order derivative. 5. Result and discussion
Derivation of the SAS comprises the following assumptions:
In this section, three example applications are provided to illustrate
1. Homogeneous porous medium. the applications of the developed single-well simulation model. Example
2. Constant fluid and rock properties. 1 (Section 5.1) presents the single-well simulation of a homogeneous 1D
3. No-flow/closed boundary at reservoir external radius. radial-cylindrical reservoir. Model validation was established by
4. Specified production rate at the inner boundary. comparing the reservoir pressure distributions obtained from both the
SAS and the numerical model. Further model verification was performed
Mass conservation for horizontal (negligible gravity effects) fluid by comparing the pressure solutions from the developed numerical
flow in the radial coordinate system can be expressed as follows model when α ¼ σ ¼ 1 and the pressure solution obtained from the
(Muskat, 1938): classic solution of the radial diffusivity equation (i.e. based on Darcy’s
1 ∂ ∂ flux). In the second example (Section 5.2), a single-well simulation of an
ð ρur Þ ¼ ðρφÞ (56) inclined heterogeneous 1D radial-cylindrical reservoir was presented.
r ∂r ∂t
Sensitivity analysis was performed to understand the influence of key
Recall the flux relation defined in Eq. (3) and substitution in Eq. (56) model parameters (i.e. α and σ ) on pressure distribution in the reservoir
leads to (i.e. grid block pressure). Finally, single-well simulation for a 2D
� � �� �
1 ∂ βc Kr σ1r α 2πrh ∂1 α ∂p
�
2πhφct ∂p anisotropic and heterogenous stress-sensitive reservoir is studied in
¼ (57) Example 3 (Section 5.3). Sensitivity analysis is also conducted to
r ∂r μo Bo ∂t1 α ∂r αc Bo ∂t
investigate the effect of α on wellbore pressure only. Review of the
α 1
Introduce the operator ∂∂tα 1 , to both sides of Eq. (57) and simplifying literature indicates that for geological media exhibiting subdiffusion,
results to: 0:47 < α < 0:94 and 0 < σ � 1 (Go �mez-Aguilar et al., 2012; Go �mez
� � � � α et al., 2015; Raghavan and Chen, 2019). Moreover, in Example 2 and
1 ∂ βc Kr σ1r α ∂p φct ∂ p Example 3, the reservoir fluid is slightly compressible, and the fluid PVT
r ¼ (58)
r ∂r μo ∂r αc ∂tα properties were calculated using the correlations presented in
Appendix B. All computations in this paper were performed using
Initial condition : pðr; 0Þ ¼ pi (59) MATLAB scientific computing language.
α � �
∂1 ∂p
No flow boundary at external radius:
∂t1 α
r
∂r r¼re
¼ 0; 0 < γ<1 5.1. Example 1 model validation
(60)
Consider a 0:5 feet (ft) (0:1524 m) diameter oil well located in 60
Specified production rate at wellbore boundary condition is acres (242811 m2) spacing operating with a specified rate and the
expressed mathematically by the expression: reservoir external radius characterized as a no-flow boundary. This
� � � � synthetic radial-cylindrical reservoir is a single layer with initial pres
2πhβc Kr σ1r α ∂1 α ∂p
qsc ¼ r ; 0<γ<1 (61) sure, thickness, horizontal permeability and porosity of 4000 psia
μo B o ∂t1 α ∂r r¼rw
(2:76 � 107 Pa), 30 ft (9:144 m), 80 mD (7:90 � 10 14 m2), and 0:23
Detailed development of the SAS of the transient pressure distribu respectively. The fluid in the reservoir has a density, formation volume
tion is presented in Appendix A with the resulting expression given by: factor, and viscosity of 48 lb/ft3 (768:89 kg/m3), 1:2 bbl/STB (1:2 m3/
pffiffiffiffiffiffi � pffiffiffiffiffiffi �� pi m3), and 0:45 cp (45 � 10 5 Pa-s), respectively. Other input data
�
b sÞ ¼ β ϱ I0
Pðr; ζsα r þ K0 ζsα r þ (62) employed for computation are listed in Table 1.
s2 α s
The reservoir spatial domain is discretized into 100 grid blocks
which are spaced logarithmically according to Eqs. (24) and (25). Then
the pressure distribution in the reservoir is obtained by solving Eq. (43)
for the numerical model (NUM) and Eq. (67) for the SAS. The profiles for
reservoir pressure versus radial distance obtained from the SAS and
NUM are shown in Figs. 3 and 4 for comparison.
The figures show the curves of pressure versus radial distance when
the production time t ¼ 1 and 3 days for a well producing at a rate of 500
STB/day (Fig. 3) and 1000 STB/day (Fig. 4), respectively. The solid and
Table 1
Input parameters for Example 1.
Parameter Value
t 6 days
α 0:61
ct 3:5 � 10 6
psia 1
(5:08 � 10 8
Pa 1
)
φ 0:23
Δt 0:25 days
σr 0:8
Fig. 2. Discretized 1D radial-cylindrical reservoir.
8
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
7:08 � 10 3 Kh
PD ¼ ðpi pÞ (64)
qsc μo Bo
6:3 � 10 3 Kt
tD ¼ (65)
φμo ct rw2
r
rD ¼ (66)
rw
re
reD ¼ (67)
rw
Fig. 6 presents the pressure distribution in a closed reservoir ob
tained from the numerical model (NUM) and the classic RDE at t ¼ 1 day
dotted lines denote the solutions from the SAS and NUM, respectively.
Examination of Figs. 3 and 4 show that there is an excellent agreement
between all the curves when t ¼ 1 day and 3 days respectively. Fig. 5
shows the pressure versus elapsed time at r ¼ 0:2604 ft (0:079 m) and
14:4897 ft (4:416 m) for qsc ¼ 500 STB/day and 1000 STB/day,
respectively.
Cleary, the plot indicates that there is an excellent overlap of the
pressure curves from the NUM and SAS.
Further verification of the developed numerical scheme was estab
lished by comparing the solutions obtained by setting α ¼ σ ¼ 1 in the
proposed nonlocal time-radial diffusivity equation and the classic radial
flow diffusivity equation (RDE). The solution of the radial flow diffu
sivity equation for a bounded circular reservoir with no-flow at the outer
boundary in dimensionless terms of dimensionless variables in Laplace
domain takes the form (Carslaw and Jaeger, 1940):
Fig. 6. Comparisons of classic radial diffusivity equation (RDE) and present
numerical solution (NUM) at an elapsed time of (a) 1 day and (b) 3 days
(qsc ¼ 1000 STB/day).
9
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
10
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
Fig. 12. Pressure variation with elapsed time in (a) Block 1, and (b) Block 20.
11
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
Table 3
Block pressure at Block 1 and Block 20 setting σ ¼ 0:8.
Time Block 1 Block 20
12
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
Table 5
Index of wellblocks in the reservoir.
Block location in r direction (i) Block location z direction (j) mEq. (37)
1 2 31
1 3 61
13
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
2. Fitting well test data with the proposed model provides the only similar works by (Holy, 2016; Ozcan, 2014; Raghavan and Chen,
realistic method of estimating the newly introduced phenomeno 2019) for further details.
logical parameters (i.e. α and σ ). Therefore, instead of Kh and skin
estimated from the classic radial diffusivity model from well test CRediT authorship contribution statement
data; Kh, α; and σ are all simultaneously estimated by fitting field
data or measured data with the numerical model using suitable Abiola D. Obembe: Conceptualization, Formal analysis, Methodol
optimization algorithms. The fitting process of the proposed frac ogy, Validation, Visualization, Writing - original draft, Writing - review
tional diffusion model with well test data is a topic of ongoing & editing.
research by the author, although interested readers can review
Nomenclature
rw Wellbore radius, ft
Rs Solution gas oil ratio (GOR), SCF/STB
SVgr Specific surface area of the grain, μm 1
t Time, day
tD Dimensionless time
Ti Initial temperature, ℉
u Fluid flux, ft/day
⇀
14
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
Greek Symbols
α Order of fractional differentiation, dimensionless
αc ¼ 5:615 Volumetric conversion factor, dimensionless
αlg Defined in Eq. (23)
βc ¼ 1:127 � 10 3 Conversion factor, dimensionless
∂α
∂t α Fractional-order derivative
r Gradient operator
γ Fluid gravity, psi/ft
γ API Oil API gravity
γ c ¼ 0:21584 � 10 3 Gravity conversion factor, dimensionless
γg Gas specific gravity
γK Permeability modulus, psia 1
γo Crude-oil specific gravity
ρo Crude-oil density, lbm/ft3
ρo Crude-oil density, lbm/ft3
Δr Size of grid block in r direction, ft
Δt Timestep, day
Δz Size of grid block in z direction, ft
ΔP Reservoir pressure drop, psia
φ Porosity, fraction
φref Porosity at reference conditions, fraction
φi Initial porosity, fraction
Φ Fluid potential, psia
Φref Reference fluid potential, psia
σ Auxiliary parameter, day
γo Oil specific gravity
μo Crude-oil dynamic viscosity, cp
μob Bubblepoint oil viscosity, cp
μod Dead oil viscosity, cp
ωαk Defined in Eq. (14)
Conversion factors
1 foot 0.3048 m
1 psia 6.894757 kPa
1cp 0.001 Pa s
1bbl/day 0.1589873 std m3/day
1 lbm/ft3 16.01846 kg/m3
1 mD 0.9869233E-6 m2
1 bbl/stb 1 m3/std m3
Herein, the Laplace domain solutions for the nonlocal time-radial fractional diffusivity equation employing the Caputo definition of the fractional-
order derivative is derived in detail.
Recall Eqs. (58)–(61)
� � � � α
1 ∂ βc Kr σ1r α ∂p φct ∂ p
r ¼ (A.1)
r ∂r μo ∂r αc ∂tα
15
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
Boundary conditions:
� � � �
2πhβc Kr σ1r α ∂1 α ∂p
qsc ¼ r 0<α<1 (A.3)
μo B o ∂t1 α ∂r r¼rw
α � �
∂1 ∂p
α
r ¼0 0 < α<1 (A.4)
∂t 1 ∂r r¼re
The Laplace transform of the Caputo time fractional derivative is defined below (Ishteva, 2005):
� �
∂α
L f ðtÞ; s ¼ sα FðsÞ sα 1 f ð0Þ; 0 < α < 1: (A.5)
∂tα
Taking the Laplace Transform of Eq. (A.3) with respect to time leads to
2b � �
∂ P 1 ∂Pb � �
þ b sα 1 pðr; 0Þ
¼ ζ sα P (A.6)
∂r2 r ∂r
where ζ ¼ μo φct
αc βc Kr σ1r α .
Substitute Eq. (A.2) into Eq. (A.7) results in
2b � �
∂ P 1 ∂P b
þ b ¼ ζsα 1 pi
ζsα P (A.7)
∂r2 r ∂r
Careful observation of Eq. (A.8) indicates a second order non-homogeneous differential equation, hence the solution would consist both com
plimentary and the particular solution.
First, the complimentary solution is of the form
pffiffiffiffiffiffi � pffiffiffiffiffiffi �
b
P ¼ A1 I0 ζsα r þ A2 K0 ζsα r (A.8)
where A1 , and A2 are coefficients to be obtained from the specified boundary conditions, I0 ð:Þ, and K0 ð:Þ are the modified Bessel functions of the first
and second kind with order zero.
Noting that the right-hand side (RHS) of Eq. (A.8) is independent of the space variable r, so P b p ¼ pi forms a particular-solution of Eq. (A.8).
s
Therefore, Eq. (A.10) is a general solution of Eq. (A.8).
pffiffiffiffiffiffi � pffiffiffiffiffiffi � pi
b ¼ A1 I0
P ζsα r þ A2 K0 ζsα r þ (A.9)
s
Applying the Laplace transform on the boundary conditions given by Eqs. (A.3)–(A.5) and imposing the initial condition leads to:
� � � �
b
∂P qsc μo Bo 1
¼ α
(A.10)
∂r r¼rw 2πrw hβc Kr σ r
1 s2 α
� �
b
∂P
¼0 (A.11)
∂r r¼re
Considering the above boundary conditions, the transient pressure distribution in the reservoir in Laplace space is governed by the expression:
� pffiffiffiffiffiffi � pffiffiffiffiffiffi �� pi
b ¼ β ϱ I0
P ζsα r þ K0 ζsα r þ (A.12)
s 2 α s
In this section, the correlations for calculating the PVT properties for the crude-oil are presented.
The oil specific gravity and molecular weight are computed using:
141:5
γo ¼ (B.1)
γ API þ 131:5
16
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
� �6:58848
Kw γ0:84573
Mo ¼ o
(B.2)
4:5579
Isothermal compressibility
Calculate isothermal oil compressibility using the Frashad et al. (1996) correlation.
co ¼ 10ð Þ (B.3)
5:4531þ5:03�10 4X 3:5�10 8 X2
where
X ¼ R0:1982
s T 0:6685 γg 0:21435 γ1:0116
API P
0:1616
(B.4)
Bubblepoint pressure
The bubble point pressure is estimated with the (Lasater, 1958) correlation
pf ðT þ 459:67Þ
Pb ¼ (B.5)
γg
where the gas mole fraction in the oil and Lasater bubble factor as calculated from:
� � 1
γo
xg ¼ 1 þ (B.6)
7:521 � 10 6 Rs Mo
� �
xg 0:15649
pf ¼ exp 0:59162 (B.7)
0:33705
Finally, the Jacobson (1967) equation is adopted to correct the calculated bubblepoint pressure for the effects of nitrogen in the surface gas as
follows:
Pb N2
¼ 1:585 þ 2:86yN2 1:07 � 10 3 T (B.8)
Pb
Al-Shammasi (2001) correlation is applied to calculate the Bubblepoint oil formation volume factor.
1:410 � 10 4 Rs 4:49 � 10 4 ðT 60Þ
Bob ¼ 1 þ 5:53 � 10 7 Rs ðT 60Þ þ þ þ
γo γo
(B.9)
2:06 � 10 4 Rs γg
γo
With the obtained values of isothermal compressibility, bubblepoint oil formation volume factor and bubblepoint pressure, undersaturated oil
formation volume factor is given by the expression:
h � �i
Bo ðPÞ ¼ Bob exp co P PbN2 (B.10)
Estimate the dead oil viscosity and bubblepoint oil viscosity applying the Glaso (1980) correlation and Chew and Connally (1959) correlation,
respectively.
17
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
� �
3:141 � 1010
μod ¼ 3:444
logðγAPI Þ½10:313 logðTÞ 36:447�
(B.12)
T
� �
� � 0:43þ 0:57
0:80 10ð0:00072Rs Þ
μob ¼ 0:20 þ μ (B.13)
10ð0:00081Rs Þ od
Calculate the undersaturated oil viscosity by using the Vazquez and Beggs (1977) correlation.
� �
! 2:6 P1:187 10ð 3:9�10 5 P 5Þ
P
μo ¼ μob (B.14)
Pb N2
References Giuseppe, E. Di, Moroni, M., Caputo, M., 2010. Flux in porous media with memory :
models and experiments. Transp. Porous Media 83, 479–500. https://doi.org/
10.1007/s11242-009-9456-4.
Acuna, J.A., Yortsos, Y.C., 1995. Application of fractal geometry to the study of networks
Glaso, O., 1980. Generalized pressure-volume-temperature correlations. J. Petrol.
of fractures and their pressure transient. Water Resour. Res. 31, 527–540.
Technol. 32, 785–795. https://doi.org/10.2118/8016-PA.
Al-Rbeawi, S., Owayed, J., 2019. Analytical and numerical analysis for temporal
G�omez-Aguilar, J.F., Rosales-García, J.J., Bernal-Alvarado, J.J., C�ordova-Fraga, T.,
anomalous diffusion flow in unconventional fractured reservoirs. In: SPE Middle East
Guzm� an-Cabrera, R., 2012. Fractional mechanical oscillators. Rev. Mex. física 58,
Oil and Gas Show and Conference. Society of Petroleum Engineers.
348–352.
Al-Shammasi, A.A., 2001. A review of bubblepoint pressure and oil formation volume
G�omez, F., Morales, L., Gonz� alez, M., Alvarado, V., L�
opez, G., 2015. Fractional thermal
factor correlations. SPE Reserv. Eval. Eng. 4, 146–160. https://doi.org/10.2118/
diffusion and the heat equation. Open Phys. 13, 170–176.
71302-PA.
Hardy, H.H., Beier, R.A., 1994. Fractals in Reservoir Engineering. World Scientific.
Albinali, A., Ozkan, E., 2016. Analytical modeling of flow in highly disordered, fractured
Holy, R.W., 2016. Numerical Investigation of 1D Anomalous Diffusion in Fractured
nano-porous reservoirs. In: SPE Western Regional Meeting, 23-26 May, Anchorage,
Nanoporous Reservoirs (DISS). Colorado School of Mines. Arthur Lakes Library.
Alaska, USA. SPE-180440-MS. https://doi.org/10.2118/180440-MS.
Holy, R.W., Ozkan, E., 2016. A practical and rigorous approach for production data
Amir, S.Z., Sun, S., 2018. Physics-preserving averaging scheme based on Grunwald-
analysis in unconventional wells. In: SPE Low Perm Symposium, 5-6 May, Denver,
Letnikov formula for gas flow in fractured media. J. Petrol. Sci. Eng. (in press).
Colorado, USA. SPE-180240-MS. https://doi.org/10.2118/180240-MS.
Awotunde, A.A., Ghanam, R.A., Al-Homidan, S.S., Nasser-eddine, T., 2016. Numerical
Hossain, M.E., Abu-khamsin, S.A., 2012. Utilization of memory concept to develop heat
schemes for anomalous diffusion of single-phase fluids in porous media. Commun.
transfer dimensionless numbers for porous media undergoing thermal flooding with
Nonlinear Sci. Numer. Simul. 39, 381–395. https://doi.org/10.1016/j.
equal rock and fluid temperatures. J. Porous Media 15, 937–953. https://doi.org/
cnsns.2016.03.006.
10.1615/JPorMedia.v15.i10.50.
Aziz, K., 1979. Petroleum Reservoir Simulation. Applied Science Publishers, London.
Iaffaldano, G., Caputo, M., Martino, S., 2005. Experimental and theoretical memory
Bell, M.L., Nur, A., 1978. Strength changes due to reservoir-induced pore pressure and
diffusion of water in sand. Hydrol. Earth Syst. Sci. Discuss. 2, 1329–1357. https://
stresses and application to Lake Oroville. J. Geophys. Res. 83, 4469. https://doi.org/
doi.org/10.5194/hessd-2-1329-2005.
10.1029/JB083iB09p04469.
Ishteva, M., 2005. Properties and Applications of the Caputo Fractional Operator (Msc.
Brinkman, H.C., 1949. On the permeability of media consisting of closely packed porous
Thesis). Dept. of Math., Universit€ at Karlsruhe (TH), Sofia, Bulgaria.
particles. Appl. Sci. Res. 1, 81–86. https://doi.org/10.1007/BF02120318.
Jacobson, H.A., 1967. The effect of nitrogen on reservoir fluid saturation pressure.
Camacho-Velazquez, R., de Swaan-Oliva, A., Vasquez-Cruz, M., 2011. Interference Tests
J. Can. Pet. Technol. 6, 101–105. https://doi.org/10.2118/67-03-04.
Analysis in Fractured Formations with a Time Fractional Equation.
Jamal, H., Sm, F.A., M Rafiq, I., 2006. Petroleum Reservoir Simulation: A Basic
Camacho Velazquez, R., Fuentes-Cruz, G., Vasquez-Cruz, M.A., 2006. Decline curve
Approach. Gulf Publishing Company.
analysis of fractured reservoirs with fractal geometry. In: International Oil
Kang, J., Zhang, D., Zhou, F., Li, H., Xia, T., 2019. Numerical modeling and experimental
Conference and Exhibition in Mexico, 31 August-2 September, Cancun, Mexico. SPE-
validation of fractional heat transfer induced by gas adsorption in heterogeneous
104009-MS. https://doi.org/10.2118/104009-MS.
coal matrix. Int. J. Heat Mass Transf. 128, 492–503.
Caputo, M., 2000. Models of flux in porous media with memory. Water Resour. Res. 36,
Krogstad, S., Lie, K., Møyner, O., Møll Nilsen, H., Raynaud, X., Skaflestad, A., Ict, S.,
693–705. https://doi.org/10.1029/1999WR900299.
2015. MRST-AD – an open-source framework for rapid prototyping and evaluation of
Caputo, M., 1999. Diffusion of fluids in porous media with memory. Geothermics 28,
reservoir simulation problems:SPE-173317-MS. In: The SPE Reservoir Simulation
2113–2130. https://doi.org/10.1016/S0375-6505(98)00047-9.
Symposium. Society of Petroleum Engineers, pp. 23–25 doi:SPE-173317-MS.
Caputo, M., Plastino, W., 2003. Diffusion with space memory. In: Geodesy-The Challenge
Lasater, J.A., 1958. Bubble point pressure correlation. J. Petrol. Technol. 10, 65–67.
of the 3rd Millennium. Springer, pp. 429–435.
https://doi.org/10.2118/957-G.
Carslaw, H.S., Jaeger, J.C., 1940. Some two-dimensional problems in conduction of heat
Lemehaute, A., Crepy, G., 1983. Introduction to transfer and motion in fractal media: the
with circular symmetry. Proc. Lond. Math. Soc. 2, 361–388.
geometry of kinetics. Solid State Ionics 9 (10), 17–30. https://doi.org/10.1016/
Chang, A., Sun, H., Zhang, Y., Zheng, C., Min, F., 2019. Spatial fractional Darcy’s law to
0167-2738(83)90207-2.
quantify fluid flow in natural reservoirs. Phys. A Stat. Mech. Appl. 519, 119–126.
Liang, Y., Chen, W., Xu, W., Sun, H., 2019. Distributed order Hausdorff derivative
https://doi.org/10.1016/j.physa.2018.11.040.
diffusion model to characterize non-Fickian diffusion in porous media. Commun.
Chang, J., Yortsos, Y.C., 1990. Pressure transient analysis of fractal reservoirs. SPE Form.
Nonlinear Sci. Numer. Simulat. 70, 384–393.
Eval. 5, 31–38.
MacDonald, C.L., Bhattacharya, N., Sprouse, B.P., Silva, G.A., 2015. Efficient
Chen, C., Raghavan, R., 2015. Transient flow in a linear reservoir for space–time
computation of the Grünwald-Letnikov fractional diffusion derivative using adaptive
fractional diffusion. J. Petrol. Sci. Eng. 128, 194–202.
time step memory. J. Comput. Phys. https://doi.org/10.1016/j.jcp.2015.04.048.
Chen, S., Liu, F., Burrage, K., 2014. Numerical simulation of a new two-dimensional
Muskat, M., 1938. The flow of homogeneous fluids through porous media. Soil Sci. 46,
variable-order fractional percolation equation in non-homogeneous porous media.
169.
Comput. Math. Appl. 68, 2133–2141.
Nigmatullin, R.R., 1986. The realization of the generalized transfer equation in a medium
Chew, J.-N., Connally Jr., C.A., 1959. A viscosity correlation for gas-saturated crude oils.
with fractal geometry. Phys. Status Solidi 133, 425–430.
Trans. Am. Inst. Min. Metall. Pet. Eng. 216, 23–25.
Nooruddin, H.A., Hossain, M.E., 2011. Modified Kozeny–Carmen correlation for
Darcy, H., 1856. Les fontaines publiques de la ville de Dijon: exposition et application.
enhanced hydraulic flow unit characterization. J. Petrol. Sci. Eng. 80, 107–115.
(Google eBook), Victor Dalmont.
Nur, A., Yilmaz, O., 1985. Pore Pressure in Fronts in Fractured Rock Systems. Dep. Of
Ertekin, T., Abou-Kassem, J.H., King, G.R., 2001. Basic Applied Reservoir Simulation.
Geophysics. Stanford U., Stanford, CA.
Society of Petroleum Engineers Richardson, TX.
Obembe, A.D., Abu-khamsin, S.A., Hossain, M.E., 2018a. Anomalous effects during
Farouq Ali, S.M., 1986. Elements of Reservoir Modeling and Selected Papers.
thermal displacement in porous media under nonlocal thermal equilibrium.
Forchheimer, 1901. Wasserbewegung Durch Boden. Zeitschrift des Vereines Dtsch.
J. Porous Media 21, 161–196. https://doi.org/10.1615/JPorMedia.v21.i2.40.
Ingenieur 45, 1781–1788.
Obembe, A.D., Abu-Khamsin, S.A., Hossain, M.E., Mustapha, K., 2018b. Analysis of
Frashad, F., LeBlanc, J.L., Garber, J.D., Osorio, J.G., 1996. Empirical PVT correlations for
subdiffusion in disordered and fractured media using a Grünwald-Letnikov fractional
Colombian crude oils. In: SPE Latin America/Caribbean Petroleum Engineering
Conference. SPE-36105-MS. https://doi.org/10.2118/36105-MS.
18
A.D. Obembe Journal of Petroleum Science and Engineering 191 (2020) 107162
calculus model. Comput. Geosci. 22, 1231–1250. https://doi.org/10.1007/s10596- Raghavan, R., Chen, C., 2019. The Theis solution for subdiffusive flow in rocks. Oil Gas
018-9749-1. Sci. Technol. d’IFP Energies Nouv. 74, 6.
Obembe, A.D., Al-Yousef, H.Y., Hossain, M.E., Abu-Khamsin, S.A., 2017a. Fractional Raghavan, R., Chen, C., 2018. Time and space fractional diffusion in finite systems.
derivatives and their applications in reservoir engineering problems: a review. Transp. Porous Media 123, 173–193.
J. Petrol. Sci. Eng. 157, 312–327. https://doi.org/10.1016/j.petrol.2017.07.035. Raghavan, R., Chen, C., 2016. Rate decline, power laws, and subdiffusion in fractured
Obembe, A.D., Hasan, M., Fraim, M., 2017b. A mathematical model for transient testing rocks. In: SPE Low Perm Symposium, 5-6 May, Denver, Colorado, USA. SPE-180223-
of naturally fractured shale gas reservoirs. In: SPE Kingdom of Saudi Arabia Annual MS. https://doi.org/10.2118/180223-MS.
Technical Symposium and Exhibition, 24-27 April, Dammam, Saudi Arabia. SPE- Razminia, K., Razminia, A., Baleanu, D., 2019. Fractal-fractional modelling of partially
188058-MS. https://doi.org/10.2118/188058-MS. penetrating wells. Chaos, Solit. Fractals 119, 135–142.
Obembe, A.D., Hasan, M., Fraim, M., 2017c. An anomalous productivity model for Razminia, K., Razminia, A., Baleanu, D., 2015. Investigation of the fractional diffusion
naturally fractured shale gas reservoirs. In: SPE Kingdom of Saudi Arabia Annual equation based on generalized integral quadrature technique. Appl. Math. Model.
Technical Symposium and Exhibition. SPE-188033-MS. https://doi.org/10.2118/ 39, 86–98.
188033-MS. Razminia, K., Razminia, A., Machado, J.A.T., 2014. Analysis of diffusion process in
Obembe, A.D., Hossain, M.E., Abu-Khamsin, S.A., 2018c. Evaluation of non-fourier heat fractured reservoirs using fractional derivative approach. Commun. Nonlinear Sci.
transfer on temperature evolution in an aquifer thermal energy storage system. Numer. Simulat. 19, 3161–3170.
Transp. Porous Media 124, 825–860. https://doi.org/10.1007/s11242-018-1100-8. Ren, J., Guo, P., 2015. Anomalous diffusion performance of multiple fractured horizontal
Obembe, A.D., Hossain, M.E., Abu-Khamsin, S.A., 2017d. Variable-order derivative time wells in shale gas reservoirs. J. Nat. Gas Sci. Eng. 26, 642–651.
fractional diffusion model for heterogeneous porous media. J. Petrol. Sci. Eng. 152, Ren, J., Guo, P., Peng, S., Ma, Z., 2018. Performance of multi-stage fractured horizontal
391–405. https://doi.org/10.1016/j.petrol.2017.03.015. wells with stimulated reservoir volume in tight gas reservoirs considering anomalous
Obembe, A.D., Hossain, M.E., Mustapha, K., Abu-Khamsin, S.A., 2017e. A modified diffusion. Environ. Earth Sci. 77, 768.
memory-based mathematical model describing fluid flow in porous media. Comput. Roeloffs, E.A., 1988. Fault stability changes induced beneath a reservoir with cyclic
Math. Appl. 73, 1385–1402. https://doi.org/10.1016/j.camwa.2016.11.022. variations in water level. J. Geophys. Res. Solid Earth 93, 2107–2124.
Ochoa-Tapia, J.A., Valdes-Parada, F.J., Alvarez-Ramirez, J., 2007. A fractional-order Sahimi, M., Yortsos, Y.C., 1990. Applications of fractal geometry to porous media: a
Darcy’s law. Phys. A Stat. Mech. Appl. 374, 1–14. review. In: Annual Fall Meeting of the Society of Petroleum Engineers, New Orleans,
Oldham, K.B., Spanier, J., 1974. The Fractional Calculus: Theory and Applications of LA.
Differentiation and Integration to Arbitrary Order. Academic Press. Stehfest, H., 1970. Algorithm 368: numerical inversion of Laplace transforms [D5].
Ozcan, O., 2014. Fractional Diffusion in Naturally Fractured Unconventional Reservoirs. Commun. ACM 13, 47–49.
ProQuest Dissertations Publishing. Sun, H., Chen, W., Chen, Y., 2009. Variable-order fractional differential operators in
Ozcan, O., Sarak, H., Ozkan, E., Raghavan, R.S., 2014. A trilinear flow model for a anomalous diffusion modeling. Phys. A Stat. Mech. Appl. 388, 4586–4592.
fractured horizontal well in a fractal unconventional reservoir. In: SPE Annual Thomas, O.O., Raghavan, R.S., Dixon, T.N., 2005. Effect of scaleup and aggregation on
Technical Conference and Exhibition, 27-29 October, Amsterdam, The Netherlands. the use of well tests to identify geological properties. SPE Reserv. Eval. Eng. 8,
SPE-170971-MS. https://doi.org/10.2118/170971-MS. 248–254. https://doi.org/10.2118/77452-PA.
Podlubny, I., 1998. Fractional Differential Equations: an Introduction to Fractional Vazquez, M., Beggs, H.D., 1977. Correlations for fluid physical property prediction. In:
Derivatives, Fractional Differential Equations, to Methods of Their Solution and SPE Annual Fall Technical Conference and Exhibition. SPE-6719-MS. https://doi.
Some of Their Applications, Vol. 198 of, Mathematics in Science and Engineering. org/10.2118/6719-MS.
Academic press. Zhou, H.W., Yang, S., Zhang, S.Q., 2019. Modeling non-Darcian flow and solute transport
Raghavan, R., 2011. Fractional derivatives: application to transient flow. J. Petrol. Sci. in porous media with the Caputo–Fabrizio derivative. Appl. Math. Model. 68,
Eng. 80, 7–13. 603–615.
Raghavan, R., 2004. A review of applications to constrain pumping test responses to
improve on geological description and uncertainty. Rev. Geophys. 42.
19